Regarding skill MM

Hello everyone,

some time ago, I opened up a poll with the question, whether you want skill MM in random battles or not. To explain, the thing generally referred to as “skill MM” is a principle where battles are put together not only with players with suitable tanks, but also with players of the same skill level. The result was as follows:

42 percent of people do not want this at all, 58 percent do. Of the 58 percent, 41 percent of players wants to have this based on your personal rating and only 17 percent wants it based on winrate.

So much for the poll (7954 votes cast). Well… it was a bit pointless poll, because the cold hard truth is that it is almost certain skill MM will not appear in random battles (and by almost certain I mean 99,9 percent chance). Both Storm and SerB were clear about this matter and as far as we know, their position has not changed even a little. Personally, I was always against skill MM and I there are numerous reasons for that (for example, mixing noobs with noobs does not make anyone learn anything), but the thing that occured to me a day ago when I was watching my own stats is: how exactly do you define a good player?

There are cases of obviously good or obviously bad player. A person with 38 percent overall winrate is a bad player (or a troll, but that’s another matter). A person with 65 percent overall winrate – well, here we can safely assume it’s a good player. Or… can we?

I’ll use myself as an example. I think we all can agree that I am not an excellent player, but I am also not too bad, XVM shows me as “green” and I guess that’s fine. Ever since the Christmas event started, I was using the KV-1S to slowly grind the 100k XP for 3 days of free premium. In my battles with it, I reached an overall winrate of 64,74 percent (156 battles). So, uhhhh, does that mean I am a really, really good player?

No. It means the tank is just broken. I am not a bad player, but my overall winrate is around 53 percent (only like 50 battles of which were played in a platoon and like 20 in a company, so no, I didn’t get carried). But if I wanted however to APPEAR as a good player, I could always do this:

- make a new account
- buy gold, credits, premium account time
- buy a premium tank I am good at, free-exp over the Soviet line using this tank to the KV-1S
- start driving around with the KV-1S

Premium tank I am good with? No problem: my Dicker Max has 60,93 percent winrate after 190 battles, my T-25 has 68,37 percent after 98 battles. Pick one. Again, this does NOT mean I am a great player, it means I found a vehicle that sits with me really well (just as there are tanks I am absolutely hopeless with).

This is called “statpadding”. Using broken, overpowered or simply only well-fitting tanks, you appear much better than you really are. However, this would also probably mean that using skill MM in random battles, these players would appear in high-skill battles, obviously ruining them for players who are REALLY skilled. I mean – sure, you can train hard and become better and better, but really good players (or “true unicums” if you like, who didn’t participate in statpadding) have something you can’t learn: this killer instinct. It’s the same with any skill really, you can become a good violin player, but without that “something”, you will never be truly great. Game skills work like that – players like Quickybaby are simply great, they adapt, they react, a lot of it lies in training and effort, but there is also that instinct that tells you what to do and that’s something I for example will never have, no matter how hard I train.

What I am basically saying, right now, people who are skilled are swearing at tomatoes (please note that there are 48 percent winrate scrubbies who actually think they are good and blame their teams for their performance, I am not talking about those), but if really good people recieved what they wanted (skill MM), “statpadder” would be the new “tomato”.

Another issue with this is the amount of unhappy people it would create. You see, I know I am not a great player and I realize it. But there are TONS of people out there who think they are good for whatever reason. These people, who think of themselves more than they are would end up playing with other noobs and they wouldn’t be happy. What do you think would be their reaction?

“Damn, I got kicked into the noob league, I have to get better”



I think we all know which one would that be. This is yet another issue by the way, changing such a fundamental rule two years into the game would only create confusion and unhappy feelings (“I’m playing for two years now, why am I playing with noobs?”). Make no mistake: there are a LOT of people with delusions about their skill level. That’s simply human nature: when confronted with obstacles, rather than accept the fact it was our fault, we look for reasons elsewhere. Some people get over this actually and realize it, but a lot of people don’t and skill MM would make this entire crowd very, very unhappy.


Personally, I partially blame XVM for the current situation. Lately I’ve seen yet again several cases where players go “OMG tomato team I quit” – strangely enough, it’s usually not even unicums or good players saying that, but people around 48-50 percent winrate. If you see the “skill” of your players in advance, it can either boost your confidence, or it can demoralize you, and even that would be fine, IF the system was set right.

A lot of players complain lately that their teams are full of red players – as in, red is bad. Well, if the majority of the teams consist of red players, it doesn’t mean all the red players are brainless idiots. It means the system, that decides whether a player recieves color “red” or “orange” doesn’t work properly. If red is bad, how come the majority of the players are red, when (out of definition), the majority of the players is AVERAGE? Part of the answer probably lies in the fact that it wasn’t most likely an average player, who felt the need to write that part of the XVM mod.

This color issue actually influences our behavior on a subconscious level. I believe it’s a form of the Halo effect. Our “first impression” is not based on that player’s performance, but on what color was he assigned to by a mod. In this sense, the same action of one player can be judged differently based only on that color: a “red” tank hides behind a building? He must be a dumb noob, camping here! a “purple” player hides behind a building? Ah, he’s waiting in ambush, surely because he expects the enemy, smart! – and so on.

Personally, I think two (mutually exclusive) things should be done with XVM:

- rework the coloring scheme to actually reflect the average winrate and display it as “average” in a different color than red or orange. Both colors mean “warning” on most basic level and players subconsciously associate them with negative feelings even if the average player is performing fairly well in that battle. Please note that the average winrate in WoT is 49 percent, not 50 or 52. Some dark green would be nice, with red reserved only for truly bad noobs.

- ban that part of XVM mod altogether

Anyway, this is just my personal line of thought. As long as there will be ratings and winrates, there will be statpadding. It just is like that. I am sure all this was written before, but I just felt like writing it again, to sort out my own thoughts too.

222 thoughts on “Regarding skill MM

    • Yep, seconded.

      By the way, I get sick and tired of statpadders with a 60% WR bitching about my 53% WR as if that is bad. My favourite comment (from people playing like brainless zombie chickens): “you are not good enough to talk to me!” Well, excuse me, but FU and LTP in something else than your T1/T18/KV-1S.

      • Like someone on wot forums said few months ago – we need noobs in our teams for us to be good. Being good in wot is relative, it depends on how good other players are. If you’re unicum now and suddenly you get thrown into battles with other unicums only, you’re not a unicum anymore, you’re plain average. My whole playstyle is focused around bashing noobs, not superior situational awareness or map and game mechanics knowledge, just being plain better player than most (considering the 48% WR average). Im fine with it, Im good only because others are bad, but I bet there would be alot of hurt egos when people who now are considered very good to unicum would see they’re not as good as they thought they are. And there would be some tears.

        “This is called “statpadding”. Using simply only well-fitting tanks, you appear much better than you really are.”

        Other thing, SS I think you overused term “stat padding” here. KV1S, hellcat – yeah those are OP. Playing tiers 5 and lower – yeah, thats statpadding too. But playing a tank that you like and you do good in, why is that statpadding? I mean, fuck am I suppose to play tanks I dont like just to make my stats more objective? Is good arty player supposed to play meds just to make his stats “non stat padded”? I dont mean like t18 or old su26, but if you, lets say, do great in e100 but alsolutely suck balls in IS7, so you obviously stopped playing is7 and played e100 alot (because who the fuck wants to play tank he cant play), why would that be statpadding?

        • ” lets say, do great in e100 but alsolutely suck balls in IS7, so you obviously stopped playing is7 and played e100 alot (because who the fuck wants to play tank he cant play), why would that be statpadding?”

          I dont think so,
          Neither the E100 nor the IS7 are your typical statpadding tanks. They both have their strengths in your typical Tier X game, but not so overpowering in such a way that other tier X tanks are outmatched by their performance. Far from it, actually – Both the E100 and IS7 can be fairly well countered.

          Lets be honest here.

          Winning is fun. Its rewarding. I would totally understand if the player in your example was playing the E100 instead of the IS7. Not per se so he keeps his stats higher, but simply because he has more fun in the E100. He wins more. He gets to kick some butt in his E100, while in his IS7 he gets owned alot. Naturally he would rather play his E100.

          The same goes for many players, if not all. We all keep tanks we like to play, because they suit us. They are fun. They make the game fun. I keep fun tanks like the Hellcat, T-34-85, Type59, the Jagdpanther and many, many more. I dont think of these as statpadding tanks, these are simply the vehicles I prefer and have a good time with. Getting decent stats with these is more a side-effect of having a good time.

          I dont think stats shouldnt affect your choice in the tanks you want to play. While stats are important, they are not THAT important. I would rather keep playing the game for the game and having a good time, not for the stats.

          • Its weird, your post looks like you’re not agreeing with me, but you actually are agreeing with me, whats up with that.

            • Ahh,.. isnt there an edit button somewhere? :P

              I think you refer to the last part. I write “I dont think stats shouldnt affect your choice” where I simply wanted to write “should”.

              One typo made my post contradictory. Maybe I should work for WG.

      • statpadders , especially low tier one, are nothing but fking bully of newbies. Like a big baby who refuse to grow up. If they mocking you with your 53% WR, it’s because their are jealous. Personally i have 52% WR , all by doing solo random battle(i don’t have much time to play clan war, and only platoon with my friends when we have free time at weekends,in 5~10 battles). But i never use XVM mode for see other player WR. Even with bad player team, doesn’t that mean you will lose. What make a good player is they action in a battle, not they statistic.

    • I think SS failed massively with this “skill mm” poll. The options to choose from were badly laid out because you can’t create skill based mm based on win ratio or wg rating. It is mathematically, physically, statistically and intelligently impossible. It needs a different kind of rating like elo. That’s the first mistake.

      Then there was the question does not even ask what SS actually wanted it to ask. He wanted to know do we want skill based mm or not. Not whether we want skill based mm based on principle that can never work. The poll was as badly laid out as this:
      Do you like meat?
      a) yes I like chicken
      b) yes I like fish
      c) no I hate all meat
      So according to your own poll only two kinds of meat exist. We know that is not true.

      The rest of the post clearly shows to me that SS is hard against skill based mm but doesn’t know anything about what kind of _different_ (yes, there are different versions of skill based mm!) skill based systems can be created or how skill based mm is done in other games. The poll asked the wrong questions and as such the answers are meaningless. I voted for not skill based mm in the poll because I don’t want skill based mm based on win ratio or wg fail rating. But I want skill based mm in wot based on elo! It is a proven concept, it works very well in many many games and no game has done skill based mm based on win ratio or wg rating…

      I tried sending SS messages so he could correct his poll but he either missed my posts, didn’t read them or ignored them. I hope next time you put a poll on your website you SS at least spend 2 minutes trying to understand what is the question, what do yo want to know? Next time you want to know about skill based mm and do people want it here is a poll you can use:

      Do you want skill based mm?
      a) no
      b) yes

      As for your article it is full of all kinds of mistakes and common misunderstandings about skill based mm. Noobs fighting each others, skill not being verifiable at all or statpadding.


  1. ” tomato team i quit ” well I personally prefer fight to the end and in 40% cases tomato team wins because of me …

    • So I guess that makes u a unicum, no?
      Stupid me for having doubts about SS first argument…

        • Well 50% win rate plus 40% of the otherwise 50% lost games makes 70% WR. U r either too modest or u exaggerate your carry powers a bit :)

  2. I would like to jump right into this and say that skill-based MM of any kind is bad, because it pushes everyone to being average. Also i would like to say, having a tank with 100 battles, or even 150 battles is by no means A RELIABLE STATISTIC. Best compared to one’s whole stats, but up to 100 battles a lot of RNG causes play in the actual stats of vehicles, simple example: I once had a 12-losing streak in my VK3001H for a double (i stopped trying to do effort back then), on 100 battles that already represents over 10% of my overall statistics in pure losses!

    And indeed, about the XVM thing hwere people say “OMG i quit”, they will never become unicums or good players. Unicums know that a chance is just that – a chance – if you have 30% chance, you have a 30% chance, it is your choise to take it or leave it, but if you leave it you know that chance will drop dramatically. I’ve won games with 30-40% chances, and lost plenty with 60-70% chances. People should take a less black & white look at things where stats are concerned and look at a broad picture.

    XVM additionally has configs already available, you can take whatever colours you like. Personally i’d start putting bright red on from anything below 51% WR.

    • Also with lack of editing majority of ppl according to XVM aren’t red but yellow.

      • a uncium will never quit because they want to make damage in the battle becaue its always very annoying to do 0dmg and lose

  3. XVM is a good tool if used be good players or players who want to learn to be better. Unfortunately, there are too many baddies with it and “tomato team i quit:…

    • I mainly used XVM when I got into a tier 10 with my WZ-132 (or any low-tier in a hi-tier game situation) so I’d know which of their larger-than-mine tier players are bad so I could flank them too. So for instance in my Pershing if a Tiger II has like 56% I won’t fight it, but I will go and flank a 49% E-75 (if the opportunity is given, I won’d suicide just to get to it of course). So I used XVM just as a tool, and I think people are naive to believe it’s predictions.

      • That doesn’t make sense. XVM remains completly useless except for some additional info. No knowledge to be found there and if you think XVM improves your gameplay, you are badly mistaken.

        • XVM tells you if you can be aggressive in your gameplay or not. What alex said makes sense. He’s taking advantage of the fact that the E-75 is an average player, and so he is targetting it first for an ‘easy’ kill. But, he doesn’t want to attack the other two tanks because they’re not average players and will retaliate at his actions.

          But, killing above average players is even better because you want to stop them from dealing damage to your team. You’ll know that they’re capable of even carrying the game, so it’s best that you eliminate the most threatening player to your team first.

          • You guys are putting too much importance on whos blue, purple or red while its mostly irrelevant to the outcome of the game. You need tactical knowledge to play this game effectively and when you learn that, it won’t matter whos red or purple. Noobs focus their gameplay based on XVM calculation which get them even more distracted which leads to bad gameplay.

            It is very much impossible to accurately calculate someones skill therefor XVM remain useless in determining player skill and strategy for gameplay …

  4. You point out a lot of XVM’s downsides without mentioning the upside value: Can I trust player X to help win a game? That is the primary use that Blue/Purple random players get out of it, IMHO. I feel like I can trust most yellows, greens and above to play as a team to win, but I cannot say the same thing for oranges and reds.

    This means that if I see a bunch of reds and oranges defending the flag I will stay behind with them because I can’t trust them to help each other. It also means that if I see 3 tanks going hill on Campinovka and they are yellow or above, I will support them, but if it is reds and oranges, well…I’ll let them find out what happens when they overextend.

    Skill matchmaker will never happen, but I sure as hell use “skill battlegroup maker” when I play in a random and select what people to help using XVM, and with a 58% win rate, I must be doing something right.

    • But can you trust a purple or a blue player? For example, if you yourself are red, yet see a blue player holding the other flank, can you really be sure he does the reasonable thing (holding out waiting for the possibility of reinforcements), or is it possible he will go like “Fuck this, there are reds on the other side, they won’t push anyway, I’m outta here”…

      • I would say that since, and the statistic escapes me at the moment, but I want to say only 5% of WOT players use XVM, that it would be unlikely that red/orange and even most yellow players even have XVM installed.

      • Haha, nicely said, Im blue, and I often screw up in a match where I die first.
        And I got réally demotivated by xvm, turned it of after 10 battles.

      • Had a game last night where a so-called “good” player felt it was more important to farm damage and pad his stats than go back to reset the base to allow us to cap and get a win. Half the team was begging for the reset, but he simply refused to go back. His excuse? “I only have 168 hp and I’m not going to die so bad players can win a match.”

      • If that purple/blue player has a solid ammount of games played (which can be also seen ingame via XVM) then, with the above 90% probability, I would say yes, you can trust him.

    • I’ve seen players with high win rate act damn stupid, and ppl with low win rate (around 46-47%) carry the game. The best thing is just not to trust those that aren’t in your platoon, even a unicum player could be distracted from gaming, Relying trully on any unknown random player never turned out to be a good idea for me at least :)

      • Heck I don’t even rely fully on my platoon mates. XD

        Funny thing is, apparently I am “blue” on several tanks, indicating I should be pretty darn good. Hellcat for instance: 70% winrate over 299 battles. But then there is my Object 268: 38% winrate over 80 battles… ROFL! Pretty sure that looks beyond tomato-like to the XVM users amongst my team mates, and so I don’t get a lot of assistance when I ask for it. Which is rather annoying!

        • Yup but 80 battles are withing the RNG and bad luck with bad teams :D The funny that I have like 48% or less with my T54 and I have more than 280 battles… and I love the tank… On the other hand I hated the tier 8 VK ausf.a, and I have 60% win ratio with around 250 battles… :D So yeah, sometimes it is just too damn strange

        • Don’t worry, I have like 60 or so battles on my T110E5 and only 43% WR (overall I have 52%), but that’s just because I went a few battles with it, lost 80%, and then built up WR with clanwars, I don’t even play E5 in pubs anymore.

  5. You are damn right! I hate this part of XVM because I saw plenty of guys giving up, suiciding, just because he saw a lot of “reds” in the team… Some times we have even won.

    The other thing is, that for example even an super-über unicum can get unlucky, and for example ammoracked, or badly crippled by a lucky arty player. (because I think unicum guys aren’t the ones that stay behind one rock all game) and than there are those bad players (like my little brother) that can have an epic game, just by luck! When he can penentrate, even set tanks on fire, when he was speeding at 40km/h and barely aimed at all.

    So all in all, I think the good thing would be if the game would try to balance the team after he choose the 30 players! For example, when there are like 4 red players and 4 unicum players, and they are not platooned, the MM shouldn’t put them on one side, instead both teams should get 2 red and 2 unicum players (and the rest is the avrage color, I have no idea what is). Or just leave it as it is…

    • I have never seen true unicum who gave up and suicided at the beginning of battle because of ‘reds’ in team. But yes, I have seen some yellow/greens who did this, but it was very rare behaviour.

      Most of unicums value their precious stats too much to do it;) It is always better for your statistics to lose battle in which you played well and tried to win than to lose doing literally nothing by suicide.

      • On a more serious note I never suicide in this situation because winning with worse team against better opponent is always very rewarding experience and brings a lot of fun. Especially when your so-called tomatoes teammates positively surprises in this battle. It happens not so often, but happens. Maybe your enemies are statpadders this time and and your win chances are better than you though?

    • Very much this.
      I don’t mind the current system, but when you have blues and greens one one team and reds on another, when you are the only good player on a red team it’s soul destroying.

      No matter what you do holding a flank, the team lemmings another and all dies…

    • The problem with that idea is it gives every tomato a good player to carry them and a tomato on the other team. So tomatoes benefit most and learn least.

  6. Hello.

    The Matchmaker works most of the time very well. Sometimes there are not equal skilled teams.. but there is one big problem with the Matchmaker.. how it handles platoons.

    Platoons overrule everything else. The MM doesnt check if those platoons are top tiers or low tiers.
    IF there is a three man platoon of 3x T49 tanks on the one side, there can be a 3x T29 platoon on the other side.

    The point is, that the MM tries to balance the platoons of both teams, but it doesnt check which tanks does platoons konsists.

    • Well exactly, it’s giving everyone a fair chance. You have 5 tier 5s, 5 tier 6s and 5 tier 7s on every team and that’s balanced. It doesn’t matter who’s in a platoon – you can just as well comunicate via chat and Y commands.
      If you did have the platoon balance thingy you’d have a lot of mirror battles – a T29 platoon versus a Tiger platoonon tier 7, a KV-1S platoon versus a KV-2 platoon on tier 6 and a T49 vs SU-85 platoon on tier 5. Not too fun of you ask me, I personally like a challenge of three T29′s, mostly because if you’re winning you can overpower them, and if you’re losing you’re losing anyway, might as well YOLO into them.

  7. I think XVM should be banned. In a fight in my E75 our E100 types over chat that we are a team of noobs and suicides in to the lake on Lakeville. We did you loose but only just. If ou had just taken some shots from people we would have won, they only had 2 or 3 low HP tanks left. Checked his service record and he had a 44% WR. I am an average player at best (above average (just) according to noobmeter) but surely XVM is making this person a worse player. If He didn’t have XVM telling him that we no chance of winning he would have actually played and may get better instead of ruining what was a good fight and a possible win.

    • The baddies will always have lame excuses to not play properly.
      In your case, that E-100 may have no XVM installed, he may just be on crack or is metally challenged and came up with this excuse to suicide.

  8. My words!

    Remove the damn % win chance for games… Iam so sick with ppl yelling when they see me in the team “OMG WE WIN WE HAVE 60% CHANCE”, but in fact 20% of those 60% are only bcoz of me and lets be real… Even the best unicum is part of a 15 man team and cant win the game entirely alone.

    • People seem to struggle with statistics anyways. Somehow, lots of players (generally not unicorns, those should be able to understand that battles are won because of their own effort) appear to think “OMG 51%.. it’s a certain win. I’ll just relax in base.”

    • Even the best unicum is part of a 15 man team and cant win the game entirely alone.

      I ve seen many cases: OMG we have player with 60+ WR (or something like that), or player with this or that particular tank, we win. = lessening the chances to win thru less than usual effort from all who read and understand that statement. This allso goes for win chance from XVM.
      XVM – never tried, never will.

  9. I am much happier since I disabled win chance on xvm. I tend to judge players by colour, but I use to congratulate them on great game if they surprise me by good play, I don’t lack that. But I must agree, lately I’ve seen a lot of OMG TOMATO TEAM I DROWN which irritates me a lot.

    After all I use xvm only to evaluate enemies. Those purples and blues must die first lolz.

    • Exactly. As I already said, how I play mediums. When I’m in Pershing, I’ll gladly engage a shit E-75 but will stay the fuck away from a good KT.
      When I still had XVM, I constantly searched for the names of the enemy unicums, and if they were killed I got a morale boost, but that was hardly important to the outcome, since 3 dead average teammates don’t make up for a dead enemy unicum.

  10. XVM can be bad for good players too.

    Nothing more annoying than an arty that focuses you and only you because you have a higher WN7/8 than the rest of the team.

    • Right, that would explain a lot. Been wondering why arty seems to pick on me quite a bit.

    • So darn true, 15 tanks on my team and three arties on their team focus their fire on me.

  11. What I still would like to see is limited matchmaking for example +-1 tier point battles.. I really hate being in a battle in my tier 6 against tier 8s and that I know that even if i do a good match. More than often my effort wont be enough to effect the final outcome of the battle..

    Now there is some tanks that have less of a problem meeting +2 tanks but for some tanks it is just don’t work.

    For example if i go out to battle in my tier 3 tank and meet tier 5s i cant do anything against them sometimes even from the side/rear and with gold. This I don’t feel is a challenge of the +/-2 matchmaking as more mission impossible..

    This is one of the things i know a lot of my Clanmates feel that is very annoying and some have even stopped playing because of it.

    What WG really need to do is look of the the balance of the game and how MM works because how it is now it is just broken and annoying..

    • If you can’t penetrate 2 tiers higher tank from sides/back with gold, then you’re doing something wrong. Using HE? Aiming for bad spots? Even KV-1 is easily penetrable from the back and that thing is a beast, same with AT-2 or Matilda.

      • try to pen a kv1 with the pz1c , or pen the matilda with the 1c…. there are only 2 places… the top part of the tank and the top part of the turret… and since they are taller than u its kinda hard to do it.

    • I remember when I played wayyy back with +-3 tiers. You don’t know how it felt going against tier 10s in my T29, no gold ammo, 198 pen is what I had. And I got Ace Tanker out of that, shooting Mauses in the ass!! You should all stop whining about the +-2 tier rule, besides, there are plenty of other low tier tanks to harass. If you’re attacking KVs alone in your Pz.I.C you are doing something wrong in the first place.

    • I remember even +4/-4 MM (beta it was I think) when I was driving KV-1 and StuG against IS-4 and Jagdtigers That was challenging:)
      For someone who learnt to play with +4/-4 and later +3/-3 tier MM, current +2/-2 tier MM looks like “easy mode”.

  12. quote: “numerous reasons for that (for example, mixing noobs with noobs does not make anyone learn anything)”

    that a non-issue, and not a valid reason
    why the fuck should I care if noobs don’t want to learn? people who want to learn will learn and get better at the game, morons will not
    as for “stats padding”, again, a non-issue since they will eventually hit a brick wall when their skill is outmatched
    you say that you have 70% WR in your E-25 .. so?! are you gonna play that tank forever – the moment you play other tanks and show your real skill, it only goes downhill from there

      • if you are thinking about MM system (based on leagues like SS was talking about) Player with unreal(better) stats will play with “big boys”, and if you play only that tank you will continue to play with pro players if you change the tank you will lose (because you don’t have general skill) and you will get down where you belong after some time. And you will create new account and you will start to “stats padding” and…over and over again. If that is fun for them let them do it :).

        B.t.w. if skilled MM is based on equal teams (equal mix from unicums, noobs and other) if you are not really unicum your team will lose because your unreal score and other team will win and your stats will start to go downhill).

      • brag about what stats, in a skill MM stats padding does not matter
        he will play his E-25 forever and will e good with that tank ONLY until WG decides, or not, to rebalance it

        you have a twisted and eroneus view on how exactly skill based MMs actually work

    • good point mate,
      I will add only two things:

      1. quote: “numerous reasons for that (for example, mixing noobs with noobs does not make anyone learn anything)”
      – why should you build some leagues of noobs or unicums or average players (like football leagues), there is no need from that. You could just slit players in two equal teams base on their skill (example: 2 unicums + 3 above average + 5 average + 3 under average + 2 total noobs in each of the teams). And about waiting time in the queue if in that moment unicum players are rare you could take 1 for each team and give more slots in the team for other players with other score range.

      2. qute: “rework the coloring scheme to actually reflect the average winrate and display it as “average” in a different color than red or orange.”
      - for example now noobs are yellow, and after some time (need some time to ppl learn new color schema) you will start reading in the chat “omg lemon team => xxxx has drowned”. this wont change anything (bad player is bad player no matter if he is purple or red or whatever color).
      I use my own color schema for players skills (I made some adjustments with consultation with my friends we were watching how players are playing). You can change his color for more optimistic feeling in the game but when I see some1 is chasing some light tank (how is ruining for his life) and not capping – passing trow the cap zone to make a kill and after that we lose the game. He is bad player no matter what color he has now red, tomorrow yellow…

      And you need to use xmv with some reserve, no1 is 100% what will happen in the battle. And I never give up until the last ammo shell or my last 1 hp. :)

    • @zMeul:
      so what’s the point of having teachers if kids don’t want to learn?
      People who want to learn WoT are sometimes afraid to even ask clearly superior player about what to do simple because of this “shut up 800WR brainless noob…” attitude of some players… so if you don’t care about the noobs, don’t at least whine about them…

      • the teachers are there for the kids who want to learn, to get better at understanding the world
        those who don’t, end up in whatever hell-hole they dig with their own hands

        man sole existence is based on survival and adaptation, you don’t want to .. you die – simple as that

        quote: “so if you don’t care about the noobs, don’t at least whine about them…”
        not that I don’t care about the noobs, I’m dragged down by the sheer number of utter morons (not to be confused with noob = new player) that plague the battles in WoT since it’s release
        I stopped playing a 3rd time in 3 years because of this. I’m done wasting my own time and money on asswipes and bots who leech battles
        players who actively contribute to the battles should be the ones rewarded, those who sit at the red line should get shit

  13. Thats why most of the good “skill based” proposals were not elitist “hurr durr, unicums fight each other, tomatoes bot against each other” crap, but skill/stat distribution among both teams.

    In other words, you would get one purple/blue player for each tier (if possible, ofc), evenly spread out in both teams.
    So instead of “ZOMG, enemy team has unicum tier X, while we have siemka E100, fuck this, I am out”, we would get “Fuck yeah, their tier X unicum has an equal opposing tier X unicum, now thats more fair”.

    See where I am going?
    There would be no need for XVM predicted win rate, as battles would be “fair” and much more skill dependent – if your (or yourself) top tier unicom failed to kill theirs and instead padded his damage ratio by hitting lower tier tanks, blame him as much as tomatoes.
    If he did, great, you can now win easier and commend him for doing his job, right?

    I know the poor MM hamster is not up for that task, but lately, randoms are just sad and it changes are needed asap.

    • So you want:

      Same numbers of unicums/normal players/bad players
      Both unicums/n/b having the same tier (because T8 vs T10 unicum is unfair)
      Both unicums/n/b having the same class, I guess? (because arty T10 vs heavy T10 is unfair)
      In case of unicum/n/b platoons have those balanced

      And do that for all 15 players to have balanced chances.

      Good luck with that and not increasing the waiting times tenfold. Or more. Most likely more.

      • That idea came to my mind also. It is not mandatory, that every team draw has all the skill group. If one team has platoon of unicums, I think, it’s fair that other team should has something like this too (Maybe even equal number of great players (blue)). I’ve experienced situations unicum platoons vs team of tomatos and every time it’s slaughter. That battles just need some variation of counterweight

    • But your system, even if it would be possible to implement it with Nya-chan improvements) would create always balanced battles. As a result in long run EVERYONE would get 50% wr (49% because of draws). I do not understand how it is fair and right to have bots and unicums winning both half of their battles.

      Where is incentive to improve your skill in this system? No matter how bad you are, you would get balanced teams and 49%wr.

      Bad player should either improve his skill level or suffer;)

      • Wrong. On average (in the long run) everyone plays in balanced teams, but obviously in individual games this often doesn’t happen. Forcibly balancing the teams would have little to no influence on win rate as it is now: the point is to reduce cheap wins and hopeless losses to have more consistent challenge/fun.

        • “Wrong. On average (in the long run) everyone plays in balanced teams, but obviously in individual games this often doesn’t happen. Forcibly balancing the teams would have little to no influence on win rate as it is now: ”

          That is simply not true. Now we are not playing in balanced teams in long run (with exception of average player), because we has also our skill level which influence outcome. Imagine a unicum A. What is average number of unicums if we count his teams from 10000 battles? More than 1, because A IS a unicum.What about enemies? It is less (if unicums are 0,1% of population). Subsequently there would be less tomatoes/greens/blues in ‘A’ teams.

          Same with tomatoes and so on. And this is the reason that better players win more in long run.

          After introduction dUG system in battles in long run Unicum A will be always by balanced out by unicum B in enemy team. It is huge change and I would like to see some argument that it doesn’t influence win ratio

          • ” Subsequently there would be less tomatoes/greens/blues in ‘A’ teams.. ” I meant less than in enemies teams (because of one spot would be “reserved” for unicum’)

  14. Nice post, I don’t also like skill MM at all. Because I would be one of those players who would go to play with tomatoes. And I think many players think I’m a tomato too. I have only 49% WR (and it’s not really rising, slowly, very slowly only), and I think there are couple of things why it’s not rising. First, I play mostly with friends over Skype, and we are having fun, not trying to get our stats better. And second thing, I only have laptop and wireless 3G connection. So I always have something like 24fps and minimum of 100ms ping. Ping usually is 120-140ms. Sometimes even higher. And if these guys with 60fps, 20ms ping and 51% WR would try to play with these settings, I think their WR wouldn’t be so good. Or if it is, they really are good. And if I got system that has 60fps and 20ms ping, my WR might even rise, but that’s just guessing. Probably I would only look at cool new graphics and drive over cliff…

    But also system can affect in WR, everyone doesn’t have money or motivation to buy high end computer to play some online game.

    • Yes, fps has a major influence on play. For the first year and a half, I played on a laptop so my fps was averaging around 15 to 18. Something like 11,000 battles with this. My win rate was I believe 50. I got a new computer and my fps went to 50+ and my win rate is now 52 and slowly climbing.

      For the longest time, I rarely played above tier V due to the slow fps. I knew that my computer was holding me back, so I played the low tiers to be able to enjoy the game. I am sure that this has affected my XVM score. So, now people probably see my stats and say I play mostly low tiers because I am seal clubbing. The funny thing is now I enjoy the upper tiers more and play the low tiers mainly to work up a tree. Unfortunately, my OCD causes me to play a tank until I have some multiple of 50 or 100 games played on it before I move on. (Note: I really hated having to play 50 battles in the T7 Combat Car.)

  15. I voted against skillbased MM, because the poll was not accurate enough. There are two different ideas of how that MM should look like:
    a) Make both teams have the same number of bobs and pros (to simplify it) (this one is the shitty version)
    b) Match only good or only bad players on both teams of a battle (this one’s the way to go)

    About your post, to be perfectly honest, there are lots of different ways to handle such issues (rerolls, statpadders, TC/platoon whores, sealclubbers) when having a skillbased MM working with WR (definitely the best way).
    1) WG could take solo win ratios only
    2) WG could divide the solo WR into 3 (or more?) different groups when rating players in the MM. Tier 1 – 4, 5 – 8, 9 – 10. So, if you padded your WR in KV-1S or T18 and then jump into your BatChat, you will not get matched like a pro, but like a noob instead, because you’re most likely terrible at high tiers.
    3) Put your tanks’ WR in relation to the server average of players who play that tank and have the same overall win ratio as you. That way, a 65% sealclubber with most his game in KV-1S and 65% wins in it, will be rated the same as someone who plays without padding his stats and has like 55% wins in some random tier 6 tank and 50% overall. (random figures, too lazy to look things up).

    It’s not like there’s no reliable way to tell who is good and who isn’t etc, like you make it look like in your blog entry. It’s actually quite easy.

    • Also, contrary to common (and your) belief, if you match noobs with noobs, they will still learn a lot. If they are better than the people they meet, their WR will increase and as a result they will meet better enemies. Once they start losing too much again they will meet worse players, etc. A players’ MM weight and skill rating would constantly change and as a result people would actually learn faster than now, because the difficulty level would increase gradually. In contrast to that, noobs get roflstomped by sealclubbers the way it is now, because there’s at least 2 of them in every battle.

  16. I think this the even matching of people would create problems of their own, like “oh you’re a unicum? well there are no unicums in queue so screw you”.

    • it could simply match 38% – 45% people all the same and 57 – 70% (solo WR), could lower that limit every 10 or so sec in queue. only americans would have a problem, because they really lack players.

    • the general problem lies in the unequal distribution of skilled players where one side has 9 above 50% and the other has not even 3.

      you also have cases of equal numbers of skilled players but those are on different tiers. the ones on bottomtier will ahrdly be able to face the others on equal terms, thats what the other equally tiered players are for.

      all in all it cant remain as it is right now. to many games are almost settled before the battlecounter is down.

      btw SS: you can easily recognize Statpadders. they either have a very low battlecount with a suprisingly high eff and WR or you just take a look at what he is driving… KV1S or hellcat? better dont trust him. in the end you find out after the match when you check the stats ingame and see what tanks have been driven.

      *cough* Rainbowdash_GER *cough*

    • I don’t think it’s really necessary to balance a unicum with another unicum, maybe two blues or a bunch of greens maybe? Just to balance out that “Chance to Win” formula of XVM by combining all the player’s skills and W/R.

  17. Skill based MM actually would punish players that are skilled and help players that are bad.


    Becouse now there at top and they can use their skill to gain more damage/exp/credits.
    with skill MM they will have average skill so the benefits will be more average.
    Also generall WR spread will be more dense in the centre of distribution.

    So skill based MM is a wish created by bad players so they would not get in fights with good players.

    Good platooned players are happy to have tomatoes in both teams.

    • imo this would affect blue/purple players if they were as many as red and yellow. The fact is there r only “few” of them compared to the rest, so they would end up being top players among a team with a more even spread of green/yellow/red players.
      So if the matter is “to learn smthg” why not learn it from the opposing unicum player and NOT from your unicum teammate.
      The thing is I find it very dispiriting when one team is full of purple/blue/greens and the other just red with few yellow. Imo this is a way of seal clubbing too – an argument I would accept in this case is the number of battles, deep red after 15 k battles, well smthg is not right.

      So my point is a kind of loose skill matching WITH number of battles calculated in somehow

    • I think (basing on stats) that I’m above average. And I want skill based MM.
      I don’t care about that I would have less exp/credit from battles.
      I care that I would have much less frustration during battles watching like tomatoes just stand and watch and lose.
      I care that I would lost a battle not because there were bigger morons in my team, but because other team was better. Of course it means the same, but there is a thin border between a tomato team sealclubbed and a team of skilled players that faced better ones.

      If you have read WG patent claim, you would notice phrases like “challenge battle” – I don’t want battles that are 0 challenge and battles that are an unbearable challenge, but simply battles that are a challenge.

    • Then that means they’ll get better than the other reds and rise up to be yellow or green. It’s gonna be an easier learning curve for the players, instead of battling unicums when they’re still trying out the game (baby sealclubbers)

  18. Ofc, no skill matchmaking will ever come to WoT cause it will cost $$ to develop, and all WG cares about is $$. Why develop something that would hurt your coffers if there’s “nothing” wrong with the system…

    Now if only WarThunder would make a skill based MM, they would stomp WoT flatly on the ground. Don’t get me wrong, I dun like WT at all, but if WT is smart enough to care more about the Players than WG does, they might sink a company that’s already well established.

    • How to implement skill based matchamker cheap:
      1. Take any formula, effi, wnX, count it for player when he logs in – simple.
      2. According to that formula put player in one of few (like 3 or 4) corresponding bin with players with similar skill value, each bin should have similar number of players – (proper quantile).
      3. For each bin run separate instance of already implemented matchmaker (with all it’s hidden agendas).


      It could even work faster than current matchmaker because of smaller numbers of players than current matchmaker has to handle.

      • They can always use the XVM Win Chance formula and simply mix-match players in queue to balance it out.

  19. WN6-7-8 is a very complex way of evaluating one’s performance. It takes a lot of different things into consideration, and values them differently.
    For example: What should be kill/death or damage dealt/damage taken ratio?
    In my opinion, over 1. But most players are well below this. Are they average? No, they are bad players. If they can’t on average pull their weight it means that they play badly. Vehicle argument doesn’t count in bigger scheme of things. This is why most players are valued by XVM as bad and not average.

  20. Good article.
    I can agree here:
    “mixing noobs with noobs does not make anyone learn anything”

    Thats absolutely true, I remember that I for myself learned a lot of good positions by watching good players (although at that time, I had no clue about XVM) and then going at these positions myself.

    Where I can not agree is this:
    ” These people, who think of themselves more than they are would end up playing with other noobs and they wouldn’t be happy. What do you think would be their reaction?”

    Why shouldnt they be happy? They think of themselfes as the total pwners, and in games with noobs they can pwn. And if they pwn enough, they climb up the ladder and get better enemies.

    All in all I am against skill MM. In my eyes the game would be way more static and boring, much camping of good players overall. Lets be honest, the good players rush to good positions in the first minutes and wait for the enemy to make their push. In 8 times out of 10, the enemy guys pushing are the “red noobs”. But that is nothing bad. It makes every game unpredictable. If you have only good players in a game, some maps will be a lot more boring than they are already.

    Skill MM: predictable, boring camping most of the time with only good players
    without SKill MM: games are unpredictable, you just dont know what will happen, and often the action starts when a “red noob” makes a push at a certain position

  21. I really wanted the option to hide your statistics implemented, it was planned but later SerB cancelled it.
    Not having stats displayed would be much better than what we have now. Winratio and XVM coloring set expectations, which are often wrong. XVM brings much more bad things than it helps, at least for me.

  22. skill mm will only unveil that their are 60% noobs and idiots online meanwhile and THAT is the average. So if you make them all green it doesn’t make them better players. The fact that average skill goes down does mean we have more and more noobs, accept this – the wot playerbase is growing fast. Color changing doesn’t change that or improof their skill or brain.
    I don’t care if noobs decide to stop playing but i care the good players stop playing because they are sick of always beeing top damage dealer and top xp and still losing. Thats why i play less time than i did before, it makes me sick (i myself have 1660wn7 and about 8800 wg rating).
    But still xvm is important, as i play more aggressive if i am the only good player (to kill much enemies and make it easier for the noobs) and play more defensive if the whole team is good (and we have more time until they get killed).
    I often play 7/7 battles now and the skill based mm is more fun than random. And i don’t think the noobs learn anything from others. Even if i tell them what to do and what not, noone cares shit. They play their style, get killed and leave battle … when should they learn someting? As you say, you have the instinct or not. If not, stay with the noobs as you can’t get it and you won’t annoy others this way that have it.

  23. “If red is bad, how come the majority of the players are red, when (out of definition), the majority of the players is AVERAGE?”
    Someone on the NA forums mentioned this about WN8′s scale when it was finally being “released” as the new standard, that it’s “average” is above the actual playerbase’s “average”

  24. You are absolutely right. I like how some people can’t admit to, if there would be skillMM, then not just means, that you have better teammate, but better enemies too. So it could be fun if you pwn some noobs, but if the opponent is smarter it could ruin your stat too. So, at the end, everyone would be average. If you use the winrate for skill MM.

  25. Noobs gonna noob, nothing new about it. And I think you got the Skill MM wrong. It cannot be about putting noobs and pros in separate games, but placing the same amount of noobs and pros in each team against EACH OTHER. Everyone will play on the same server against everyone else, but that would eliminate the ragequits after someone sees 30 or 40% win chance game.

    As far as XVM goes, it can be useful and can cause harm, all depending on who uses it.

  26. I’ve been avoiding XVM stat like the plague for a very long time, then tried it once it was merged with the rest of the XVM (no dokan, standalone exe etc.). To be honest I disabled it pretty fast. Why?

    It just spams too much information that (as mentioned in the article) may not be accurate, yet it is presented as such and it doesn’t account for random factors (bad day, somebody else playing, stock tank, statpadder…).

    False hopes for easy victory or grim desperation over potato team WILL influence your performance – but almost always in a negative way. The worst kind of idiots are the “hurr durr this is lost noob team I quit suicide!!!” at the start of the battle – those should be punished (ideally by breaking a keyboard on their backs :P).

    Many times after the battle ended while looking at the damage/kills of some player it seemd like he’s a damned killer machine unicum – guess what, below 50 % WR.

    I have a quite nice WR (61%, 16.5k battles, trying to play all nations/classes; only arty isn’t my favorite) and I can fuck up just like the worst noob, it just happens – so don’t you ever dare to rely on me! :D

  27. Well, I started to use XVM just recently, sometimes it’s handy when deciding whom to shoot first :P But what would I appreciate the most from it would be a function to show said players gamestyle (camper/balanced/aggressive) based on their survivability, avg spots etc – vbaddict does something like this IIRC. I’ve had way too many situations when we were 4 tanks against 3 in some classical “corner of death” situation so I rushed to distract the opponents to make them move and show their sides etc. just to realise the other 3 tanks from my team had zero balls altogether. Hmm, maybe I’m gonna submit this idea to XVM crew :P

  28. The only form i would like to see Skill mm is when they balance out skill like they do with tank weight.

    For example each player gets a formula when entering MM like: [WG stuff for MM]*[Tank MM weight]*[Player skill weight]

    Then the sum of this formula is taken and balances each team with skill/tank combo with a maximum difference of 10%.

    The players we see in a average battle today would still be in that same battle. Only difference is that players are balanced by skill two instead of only tanks.

    It would cause a lot of battles lost by MM law of Murphy derp to be avoided.

    • true was thinking the same
      at least they will be even teams on both sides and not the well known issue where you have
      a platoon of unicoms and other great player vs a team with majority noobs and some green and yellow in them giving the other team a seal clubbing paradise action

  29. Maybe I am a scumbag, but I often use “red retards in my team” as a meat shield or as a bait (like, “go omg…”, he goes, takes some hits, I profit from that) but in my defence, he can’t realibly do anything in this battle and thanks to these tacticts I can carry the game and use normally useless team mates, better players wont obviously do what I want…

    • Do u think the next time this “meat shield” hears from a blue/purple player (as u seem to suggest u r) >Go do this or that< he will? I think not, and this could easily be the reason many "noobs" just don't follow legit advices, not even when u tell them to cap with u and it's perfectly safe for them.

      What I want to say is it's been a long time since I seen in battle msgs from top players about how to attack and what strategy to follow or even where the LT should scout.

      It's not right for the team playing for u just to have ur stats increased. It's sad that some times unicums just hide in a0 or whatever corner and just emerge shortly before end and instead of resetting cap (no I'm not talking about lost battles) they prefer to take two or three kills of almost dead enemy tanks and then go hide to survive until end or drown.

      Red or noobs can learn from playing with each other. After all when this game started it was noobs against noobs, right? Unless u had some kind of divine intervention…
      BUT it's more achieving and fulfilling and constructive to fight WITH and win AGAINST slightly better players than u, so that u strive to get better and not feel u r the dumb kid that evry1 makes fun of in the party ;)

      • “meat shields” – i used to hate this expression. but once i got better i understood why it was being used.

        WoT is a team game and every tanks plays a role. Since I am only one out of 15 I have no control over 14 of them. I can’t force them to play in a specific way. I have to accept the way they play and make the most out of it.

        It they play like meat shields, this is how I must use them. If I wouldn’t, they would die for nothing. Using baddies as meat shields basically means honouring their play style.

        • well it doesnt take many games to convert a “meatshield” into an avid fan of “@#$@# camping $h!@# tomato noob” (<–taken from real ingame msg! :P) who just hides in battle and isnt "usefull" for anything.

          This tactic in my opinion just reproduces the problems with new players who their only fault is they have under 1k battles.

          In the end U have created a selfish good for nothing noob whose attitude is going to last longer than if u took some time to really advice him in battle (not talking about trolls or mentally challenged <–number of battles is indicator )

  30. SS, I agree that skill matching teams is counter productive but what I and a lot of other players want is skill balancing. The diff obviously is the MM would ensure the teams have similar ratios of bad, avg, good, very good, and unicum players ( based on who is in the queue). All WG has to do is pick a metric — WR, WN7, or personal rating and test it out. I realize many won’t agree on the metric used but no reason WG cant test it on the test server. It doesn’t have to be perfect, prevent matches where one team has a 3 Uni platoon and the other has one green and 14 reds.

  31. 1 “This is called “statpadding”. Using broken, overpowered or simply only well-fitting tanks, you appear much better than you really are”

    2 “What I am basically saying, right now, people who are skilled are swearing at tomatoes, but if really good people recieved what they wanted (skill MM), “statpadder” would be the new “tomato”.”

    3 “Personally, I partially blame XVM for the current situation”

    - 3 things i consider most important details!

  32. WR MM is still better option imho. it would change the differences between padders. and maybe even change the overall view on tank balance

  33. I think not the color red is the problem but attitude against these palyers.
    Many of them are really bad player and will never learn anything but there are some people who just have less fights (1k guys) to get skill and also want to get better . You should consider that. If i come in a 45% fight my first reaction is oh crap noob team again but then I try to find a way how victory is possible. Maybe try to kill enemy pro tanker or try to give tips or just try to deal as much damage as you can. of course there are some fights which are impossible to win but if you never win a fight below 50% win chance your win rate will be never better as 50%.
    My conclusion: xvm great program if you know how to use!

  34. “These people, who think of themselves more than they are would end up playing with other noobs and they wouldn’t be happy”

    Thanks you mention them, I think those people are the biggest issue in WoT at the Moment. Why? Well, i did 6k dmg on E-50M, then died and watched an IS-7 (Not a tomatoe but “below average”) That was ramming an E-100 and fired HE using autoaim on him.
    “Why the fuck am I playing my ass off if such players Ruin the game” I asked myself…
    Unfortunately those players don’t recognize what they did wrong, they apso mostly don’t get financial punishments because of using premiumaccounts or just changing Gold into money (a “tomato” that’s also a reallife friend of Mine recently told me he needs 15000Gold in 2 month to buy Tanks and change xp O.o).
    There are 2 solutions for this Problem:
    1. Skill mm puting noobs together with noobs would Not let them learn anything but it seems like they enjoying the game no mater if they pwn or fail without doing a Single point of damage, so where’s the Problem?
    2. There are tons of data collected or collectable in a Single battle, why not writing a algorythm that Shows a Player hints like “aim Full before you fire” or “don’t stay without moving while you are reloading”. I’m missing an Advanced tutorial that Shows how to use autoloader, hulldown, sidescraping etc…

    • Yeah, the hints could be a bit better. I am missing something as basic as: “Try to not show the opponents your sides”. I have seen “dozens” of battles where people go into open and when they spot the enemy, they just stop and start shooting, oblivious to the fact that the enemy has their front to them and they are showing him their side.

  35. Not many ppl know but in WT there actually IS a SKILL MM present and operating. It works the easiest to implement, but player harsh way – If you are good you are treated as if you you were half a tier higher(So you will be “Top tier” 2 times less). Once you try it you clearly feel that it is not a good solution. Which is why I will always wote against skill mm.
    The even worse thing is that if you are bad you are treated as if you were half a tier lower. Which means that your top tiers are constantly a band of morons.
    Actually most of the good players have left arcade mode and play historical battles where there is not enough players to allow the skill mm to work.
    All that is publicly know is that WT skill MM depends on win ratio and K/D ratio. As this blog is not about WT, but many ppl whine “we want skill MM”. Well go and try it. A really refreshing experience.

  36. For me personally, I only use XVM for the WN7 ratings on players.. to me, I think WN7 tells more about a player then the win%, or color.. I also know that if I play a tier 6 tanks, and have a full team of reds/low WN7 players, I asked for it.

    I remember how I was when I first started playing, and I can guarantee, I was red all the way to my first tier 10 tank.. then I started trolling the forums, and seen what WN7 was, and started to pay attention to it, and that in itself changed my play style to an extent. I also started to use wot statistics, and paying attention to what my WN7 was after each match, and what I was doing in matches where it was high, and compared it to what I did when it was low.

    another factor I take into account when using XVM is, battle count.. I could have a ton of the so called “red” players… but if they have a decent battle count, I can assume they know sort of what they are doing, and how to play the game. so I need not have to worry to much about them.. if I see anybody with a WN7 of “1″, i can rest assured that about 99.9% of the time, they are a bot.. and I haven’t been let down yet with that thinking. seen a platoon yesterday in fact, where there was a platoon of 3.. one had a WN7 of 800 something, the other 2 were a “1″.. I asked if they were bots, and the guy with the WN7 of 800 something, said ” lol, they are logged in accounts on my laptop, i do that so if I die, I can switch to another tank” I was like, really? your pathetic…

    • For me, the worst is “red” players with high battle count. Someone with 20-30k battles, 45% w/r and low performance mean he actually learn nothing.

      • yea, if i see a red player with 12k+ battles and a WN7 of something below 500, then I know he hasn’t “gotten the game” as of yet.. But I can still hopefully rely on him to at least, know how to function, and not yolo up the middle.. that is my hope anyways.. but the colors of my WN7 are based off of the WN7, not the win rat as the winrate can be influenced drastically by platoons / tank companies..not saying that is always the case. but you know what I mean.

  37. hey all delusional players come to WarThunder they calculate Skill into MM.

    PS very good article SS, you deserve a blog medal.

  38. Thing is you got to be good to statpad. I’ll take for example me as well. For the last 2 months i average 1900wn7 with various tanks. But my overall wn7 is only 1295 cause i was a huge noob when i started (suicide rushing my hellcat cause i thought it was the worst tank ever) . Now if I start a new account i’d be unicum material. But imagine a guy that has an average of 1250wn7 but his recent wn7 is much closer to his overall, if he starts a new account he’ll still be green. You know what i mean? You got to be good to statpad efficiently. If you’re a unicum with only 600 battles and the 550 are in kv1s no one is gonna respect you as a player anyway.

    I know a yellow guy that started a new account cause his stats didn’t represent how he plays anymore… His new account is still yellow. #FAIL. :P

  39. Some thoughts what WG could implement better:

    • Every new player has to accomplish the tutorial and participate in 100 battles on a seperate “newbie” server, before playing on the regular servers. Therefore frustration among new players would not be so high.

    • Launch a league server which will replace the as now known ‘public servers’. There would be still a pretty big mix of good and not so good players within each league. There would also be a pretty wide margin between bad, average, good and expert, due to the factors you mentioned .. favorable tanks, statpadders, winrate, WN7 etc.

    • People would swear a lot less, the battles would become much more intense, you could trust/rely your team a lot more, it’s more fun to shoot players which are actually harder to get and don’t let them shoot like lemmings. The game would simply adapt on your skill.

    • The prevention of statpadders (e.g. favorable tanks) and tank balance is WG task.

    • The “OMG tomato team” players would still exist even without XVM. The whine would just shift to the end of the battle when they die. When you have leagues and a player is accusing others to be ‘noobs’, he implicit calls himslef a noob too.

    What do you think about this approach? Do you see any flaws?

    • “People would swear a lot less”

      So naïve~

      Also 100 battles on “newb server” = 100 AFK battles to pass it faster and the newbie server is insta-unplayable. People would just cheat it, like that cheat any game tutorial they don’t want to go through.

      • Good point. I guess you are right with the rigged AFK battles.

        But there still is the idea of a league, which is implemented on many games and there it works well.

  40. I would suggest putting stats from last for example 500 battles. I am new player im not good at this game (only over 3k battles). I am orange i have about 800 points in this mod i don’t actually know whether it is efi or wn6. The point is that if it is efi I should be really close or even on the green level(according to the last 500 battles). That means that I am two levels higher and my winrate is not 49 precent but 50.

  41. Letme first start with a disclaimer: i’m a borderline “unicum”(i’d say entry level one, for some unknown reason as to how wn7/8/* is calculated my values differ wildly between those and my WR has taken a dive on these past months) and as such tend to be an elitist asshole when playing(i don’t tolerate stupidity in the game, i play to win and to compete at high level)

    So, having laid that out first letme continue: i don’t use XVM and i don’t like it.
    I rather remain ignorant of the rest of the team stats(you can still somewhat fathom the probable result seeing the lineups and the players/platoons whetever they belong to elite clans, etc) and give everyone the benefit of the doubt than be sour and pissed from the countdown.
    BUT, the people i platoon usually have XVM and -not surprisingly- it’s result is usually correct(there’s always overturning ofc).
    ¿And you call a 48% WR player average?, it might be by a numericalk standard, but XVM is right in calling it a “RED”, 48% WR player is terribad, it’s a BOT, a noobtard or worse, hell, i can get better WR by not moving from the base and just turning the turret and firing!.
    Worst is people with more then 10K matches and 48% WR, you need to be a special kind of baddie to have that WR after 10K battles, i mean, how and WHY are you still playing when you lose more games than you win?!?!, it wrnches my gut to see such tards in T10 matches…

    48% means you’re useless, that for the game it’s the same wheter you’re there or not, in fact it means that you you aren’t even likely to kill a single tank before dying (let’s expand on this: on 15v15, if everyone has one kill -or an equivalent damage output- then we can call it “average”).

    now 52/53%, that’s average, that’s someone that at least is doing something positive for the team but for whatever reasons falls short(low battle count, no platoon mates that teach him how to play properly, bad computer, or just not having the *spark* or initiative).
    An interesting measure of a good player is if he can “carry”, ask yourself that question “can this guy carry the game if he ends up alone or in disadvantage?, or he’ll cower in a hole and die like a dog?)

    For example, i have a friend who falls into that category, he’s been playing for quite sometime(but has few battles as he only plays weekends “casually”) but he can’t grasp basic concepts for high-performance gameplay: he doesn’t angle his tanks properly(in fact, he didn’t knew about angling and effective armor thickness until not long aso), doesn’t knows the best map positions depending on tank, has poor tactical insight(can’t read the flow of the game properly), is not familiar with all the other tanks and their capabilities, etc.
    Hence he can’t “carry” a game most of the time as he’ll be outwitted easily

    • A very large part of the players have close to zero knowledge about game mechanics, good and weak sides of different tanks, maps, etc…Even the minimap and fast message buttons are something they are not familiar with. Worse, they will most probably never bother to learn more.
      Still they have tier 10 tanks.

    • Your math doesn’t work there, homie.

      You’re saying, to be average,a player has to win more than they lose. That’d be cool except every game has a winner and a loser. Not everyone can win more than they lose, especially factoring in draws.

      Average, as it relates to winning a game, IS the equivalent of “useless.” It’s literally a coin flip. The “average”player has no impact statistically over a win or a loss…that’s what makes them average.

      A bad player brings the team down… more than simply having no impact.

      • a 48% player IS bad, because what they don’t lose they draw, hence, they don’t or can’t affect the outcome of a battle, they’re fillers.
        that is a definition of a bad/noob player.
        Then you have the terribads which appear to ACTIVELY make their team lose(as unfathomable as that appears), without resorting to team damage some people play so bad that they actually tilt the game against their tea, maybe it’s you top T10 going headless chicken, or a terrible terrible arty player.
        so yeah, for this game, being “average” means you’re bad as might as well not be there.

        And sadly, those 48% tards will reach T10 eventually and ruin the endgame for everyone.

        a match comes to my mind in a 3-man platoon in a T10 match, we did a combined 13K dmg, EIGHT TANKS of our side did ZERO DMG, EIGHT tier 9/10s…, E100s, WF E100, JPE100, T57..
        EIGHT, those people should uninstall as they have no sving or place in this game whatsoever

        • If all those average “baddies” uninstalled… 53% players would then become 48% players.

          I think you’re struggling with the concept of a statistical average here…

      • In general the comment is correct but u have to factor in the number of battles each player has… In a simplified model, One 10k 40% WR player could mean Ten 1k 60% WR players if they all played against him. (I said it’s a simplified model)

  42. Oh SS, you so silly

    How is XVM to blame for this?
    Blame the dumbasses who make the skill formulas and don’t compensate for sealclubbing and other statpadding jackassery.

    Also, people leave the game based on the calculated winchance, most of the time. What this means is that its not when they have reds on their team, but MORE reds than the opposing team, which is entirely justified. Often, it just ain’t worth the time.

  43. The overall skill level of a team measured against that on the other side I don’t consider much of an issue. With 15 tanks a side it usually evens out quite well.

    The problem comes when you throw platoons into the mix. If one side ends up with a platoon of 3 unicums sealclubbers driving around in T54′s and the other gets 3 muppets cocking about in tanks that demand expert handling or just aren’t very good, it doesn’t take a genius to guess who’ll come out on top.

    So I’d be wishing for a middle-ground, where the MM at least gives it a shot to match platoons against platoons of about the same level of competence, whereas the rest of the team would be made up not using skill matching.

  44. Xvm is broken because It dont takeś intonaccount spotted dmg for scouts etc. Proper effi rating should show performance of Player in that particular vehicle in current battle. 1300 effi is useless when in td he has 2200 and 750 in scout tanks.

    • Oh, really?

      Walking into a club and telling the best chick your winrate and eff is the best make-out tactic ever!

      “Hey beybe, I have 1800 eff…”

      And she gets wet.


  45. Personally, I use XVM for 2 reasons: 1.) damage counter, as an armature youtuber, it makes it 10x easier to see what games I would consider good (3000+ damage usually), and what ones I should maybe not put up :P and 2.)It lets me see MY progress more so than anyone else’s. When I first started using XVM I was borderline orange (48 I believe) since then I have tried and seen gains in my rating (53 currently) I would not consider myself a “good” player, but I am technically above average and I would tend to agree. I own some people, and fail miserably to others. What people fail tor realize is that XVM is merely a tool. It doesn’t help you or hurt you unless you let it, it is yours to mold, change and use at your own discretion. When I am in one of the “tomato fields” or what ever you call it, I typically call it a challenge, 35% chance to win? So you’re saying there’s a chance? lol I know, cheesy, but if you really want to do better, quitting every second or third game wont cut it, no matter how good you are. anyway, its your opinion, and that is mine, so have fun, and murk a unicum for me :P


  46. Skilled MM will never work because player skill is not a constant. It is HEAVILY influenced by tank choice, amount of equipment and crew qualifications, how tired the player is, etc. Which contradicts the exact purpose of Wot, to be able to try new tanks, new playstyles, even after a tiring day at work, without being punished for it.

    Skilled MM downsides are known and have been proven by other MMOs.

  47. Yes but I would still like the teams be even with stats, have noobs play with pros and that but just shuffle the average WR evenly. It’s annoying and not really fun to top score in every two games and still lose them all ;_;

  48. I have used XVM in the past and would again if I could face trying to get a config that doesn’t intesely annoy me. However the problem is that people are seeking to find one number that tells them all they need to know about a player.

    For this you really need about 4 numbers:
    Battle count,rating/WR , battles in current tank, rating/WR in that tank.

    Because it’s no good to know that the guy is 49% overall with 10000 battles if he has 65% in the current tank with 400 battles in. Conversely if you can see he is 65% overall but has 200 battles at 40% in his E-100 then perhaps he’s been paddering somewhere. and you aren’t so worried about his E-100.

    Really to assess him you want to see his overall and recent stats in all tanks and the current tank and doing that for all 29 players in the battle will take you about 15 mins, perhaps. Now you expect the MM to manage all this and throw you in a battle in 60s in a way that keeps everyone happy?

    Anyway lets see how it works out when they start to use skill-MM for Team Battles as see if everyone is happy with that.

  49. Personally by skilled MM I would like to see even teams.I have nothing against Unicums playing against Tomatoes as long as both teams have them.So a team with 10 avg players 2 unicums and 3 Tomatoes should face a team with somehow the same players skills.
    There`s no fun when a team wipes the floor with the other in the first 2 mins,and personally I don`t like knowing that wherever I play like a Uncium or if I am AFK,the outcome of the battle won`t change since the team is too good/bad.

    The MM should take in consideration how good the player is in that tank,and not overall,so stapadders won`t drag the teams down.

  50. you can change the colors yourself. is easy. just google it.

    the XWM is good for good players to see if the 2-3 tanks that are nearby will help him or not.
    i was in plenty of situations with a bigger tier tank and lost because he didn’t know how to help me.

    if i see a good player i know i got good chances (yes.. only chances) to do something usefull with him on that part of the map.

    XWM does not tell you who is good and bad or who will win or lose. i win 30% chance battles and lost with 90%. if you play in a team with a unicum platoon but theyr connection drops or they just want to trool (i did that) you will still lose the game, even if the XWM will tell you “95%” chances to win.

    ppl should learn that XWM is a indicator not a prediction.

  51. I only use xvm to know which players I meet that seem to be a real threat. But every unicum can be a tomato sometimes in a game, and every tomato can get a unicum game. I, myself, as “above average” just try my best to get a good rating

  52. How to get good stats:
    Option #1: Get OP tank.
    Option #2: Pubstomp with a platoon.
    Option #3: Spam gold rounds.
    Option #4: Seal-club.
    Ultimate option #5: all of the above. (Example: 3-men platoon of T18s shooting shooting HE, and HEAT against armored tanks)

    • none of those work anymore with WN8, which is one of the main reasons why WN8 is a much better metric than all the previous ones

      (note that XVM doesn’t – yet – show WN8, but noobmeter, vbaddict and wotlabs all do)

      the only way to statpad WN8 is play a bad tank well: which is not easy, because bad tanks are bad :)

      • You are wrong because a few days ago I met a re-roll account with +2000 WN8 and under 1k battles running on his E-25, which is an officially OP tank.

        • Perhaps this reroll is just a good player?

          But you are right of course. The existence of one good player totally disproves the viability of WN8…

          BTW: WN8 has a different scale. 2000 is not unicom anymore.

  53. I disagree with the article on multiple points…

    Anyone who is making intelligent steps to better his gameplay – even through various forms of statpadding, pedobearing, cherry picking OP tanks, platoon- and TC whoring – is still a LOT better player than the unwashed, braindead masses. AT LEAST HE IS DOING SOMETHING TO PERFORM BETTER. So YES, even pedobears with 5K battles in Loltractors are better players than 95% of the WoT playerbase. JUST DO SOMETHING. We all know the t10 players who still refuse to learn the basics after 10-20-30k battles.

    Skilled players are swearing at tomatoes because they ruin their game experience on the long run. It’s a fun for a while to kill them in the 100s, then it gets fucking boring. And they cannot just avoid random battles altogether, because the xp / crew / credit grinds. A lot of the skilled players want a challenge, not even better (but ultimately useless) stats. In a different league (with skill based matchmaker for example) the games would be less frustrating (less completely retarded team mates) and more challenging (better enemies).

    Saying that low skilled, inexperienced players (fighting in a “low skill league”) will not improve without the skilled players is pure bullshit, sorry. This game is not that complicated, anyone who has the drive to get better at playing WoT, can pick countless tools to better his gameplay (tutorials, blogs, replays, streams, whatever… or god forbid, common sense (gasp)), to rise over the ignorant hordes.

    • 100% agree, the league system works even for a retarded game as LOL, it worked at quake3 online, and it works in a lot of other games…its as easy as giving them the chance to playin their league or the next one, with a max of X battles with that +1 league… if he doesnt improve, bad luck, try again after u play 10x times the battles u played AND failed in that +1 league.

      u only give them that chance if they are “close” to the limit between leagues.

      it worksonly if u have a lot of leagues, not just 3 or 4 as i’ve read here… with 10-15 diferent leagues u can find people with your similar skill and at max u will have to wait what? 1m? i prefer to wait 5m and have a decent battle, with a decent team and a decent enemy, than wait 3s, and end up 15-2 or 2-15

      the problem? the farmers… as always, but if a tank is OP, nerf it to the ground, as if it were german. and they will have to find a new tank to drive…. another solution? max X% of battles with the same tank counts for the points of the league, so u cant overperform to a superior league using only 1 or 2 tanks.

  54. Q: “If red is bad, how come the majority of the players are red, when (out of definition), the majority of the players is AVERAGE?”
    A: Median =/= Average. It’s very simple:
    2 unicums with 2000 WN7 + 18 noobs with 0 WN7.
    Average WN7 for all these players is 200, so noobs (majority of the players!) are below average and unicums are a lot above average.

  55. You don’t rly understand how XVM and ratings work huh SS?

    First of all win ratio isn’t part of any sensible rating, it’s meaningless data.

    Playing “only well fitting” tanks is not stat padding, I can’t imagine anyone forcing himself to play tanks that he can’t play with or doesn’t like in the name of justice. Every one has keepers and tanks which he just wants to grind and forget…

    ” If red is bad, how come the majority of the players are red, when (out of definition), the majority of the players is AVERAGE?” – Wtf not they aren’t, where did you pull this bullshit from? Majority of players are retared. Even server average winratio is like you said is less than 50% (winratio is meaningless but w/e), so that clearly indicates that majority is bad.

    The XVM rating is already lifted up way too much and you want to make it easier lawl. Most of the “green” players are utterly brainless garbage (and even most of the “blue”, that’s just matter of having enough gold), but with so many red cripples and bots, it’s not hard to get good stats.

    You blame the XVM? Are you kidding me? It’s fault of how benefital this game is for idiots. Get a premium account, premium tank and you will get 20k per battle by just idling on cap. Noob penetrates once per battle in his 8 tier premium tank and he thinks, he must be awsome player cuz he just got 30k income from battle YAY! It’s fault of the game that most of the noobs can succesfuly live in delusion that they know anything about this game. Even without premium and gold, pure idiot can get to most of X tier tanks in ~4k battles. The economy and overwall system of this game is way too retard friendly. The massive RNG makes it even worse. Most of the tomatos have problems with hitting Maus from 50m not to mention finding a weakspot. But from time to time, once for 1000 battles even greatest tomato can get top gun, and he thinks he’s awsome YAY! Just watch one of those pathentic “RNG no comments” WG videos… 8 amo rack blow ups in a row? Why not! Blind shooting with troll gun on moving Kv-2 and killing someone from 500m? Why not! The amazing ablility of noobs to just turn turret and hit you perfectly without aiming, especialy with slow aiming russian heavy tanks? Happens every battle!

    And in the end about skill based matchmaker… Skill based matchmaker is utopian dream and it doesn’t exist. Yes, there are tons of games with semi skill based match maker (actually most of current games have it, no idea why not WoT, for example War Thunder, Hawken), it’s just solved by making the skill match maker with quite wide matching (like utter noob can’t play with pro, but average will play with good), or based on temporary efficenity of player in few last battles/days/weeks whatever. First of all you will never find “equal” players. Such thing does not exist. People are not equal. Secondly, with the match making system that takes into account only tiers for now, sometimes you have to wait for battle even for few minutes. Imagine what would happen if it also tried to fit players of similiar skill. Slightly longer waiting time for “average” players, but unicums would have to wait for HOURS or arrange the games to get to play…

  56. OH. SS, I invite you to the ASIA server. Being a 52%er and above for around 5k games makes you a good player. Anything above 55% is just godly…

  57. Please think what would happen to stat padder after applying his padded stats to skill based MM – he would be brought to his more or less real skill level pretty fast. No more stat padding. Is it wrong?

    About noob-team-quitters – one of biggest pains is false-positive – you have Red Army team, you quit, and they somehow manage to win – you get shit for your 2x battle. Really cool. The only thing is such false-positives happen very rare.
    Now looking from other side – what is the reason that Red Army team (without one tank) wins – most probably there are XVM whores on the other side, who just got a thought that if XVM shows 60% WR, it is enough to drive and shot generally in enemy direction.

    Apply skill based MM and a player would know, that he fights people with similar skill that he has – no need to identify tomatoes and unicums – all around are equal threats. That coloring part of XVM becomes somehow unnecessary. Unlike now, when XVM can tell you if a teammate is a moron, or a player that can help you, if treated properly.

    XVM was created because of players needs and is only a symptom of broken design.

    Personally I think that WoT designers are afraid of making a game that says “You suck” to a below average player. A lot of people don’t react good on such message. And being put in skill based mm in the lower part could mean just that.

  58. The way I use XVM:
    Look for the seriously bad teammates, and avoid them. With winrate that low, they may do things worse than afk ie: teamkill.
    When a baddie talk shit and would not stop, smack him on the face with his own WR.
    Accuse unicum platoon of sealclubbing, guilt them FTW

  59. Actually the easiest solution is the one that Mechwarrior Online has settled on – just don’t track stats at all. Do that, and nobody would worry about any of this crap, they would just shut up and play.

    • I completely agree!

      I was so dissapointed when they canceled the option to hide your stats. It would spare me a lot of the bullsh@t in random battles.

  60. First of all: Great read.

    1: I think you are right about the colour issue, because of the consequences you outlined, although i think the real issue has another component: I think only players that perform average or below tend to leave battles and suicide. If good players tended to do that quite often, their stats would reflect it, so they won’t be rated as good players. That means that, while the colour of a player itself is not conclusive, it depends entirely on the player how mods like XVM influence him. I myself deactivated the whole “win chance calculation” before or during battle, because i don’t see it being useful to me. Seeing low numbers might demotivate me (or motivate, depending on the day, but that wouldn’t be that often) and seeing high numbers might lower my attention. When i see a team of red players it might have the same effect but in general i know, that the performance of a player in a battle depends on the battle itself. Players have good battles and bad battles. On the other hand i managed to kill a T110E3 unicum because my brother and me (both in the Löwe) recognised the threat, positioned accordningly and focused on killing him. That took the best player AND one of the best tanks in the enemy team out – all thanks to the information provided by XVM.
    If you can’t use XVM to you advantage and you realise you get demoralized by it too often, the best idea is to uninstall it again, since that allows you to treat each tank according to the danger that tank poses to you and not what the player in it might or might not know.

    tl;dr: The information XVM provides might not be presented in the best colour scheme, but what a player makes of it is his choice. Some players can use XVM to their advantage, some can’t. XVM-stat cannot be recommended unconditionally.

  61. I generally agree with that part about XVM, but I must say, that atm I’m using it, and my experience is: if both teams are mostly yellow, with some green and red, the fight should be quite interesting. I hate purple players in any team. The propability that purple player fails is the same as for any other player. They can fight great, but when they fail, they fail epic. Plus they are able to retarded behaviour. Like killing people that do not leave spawn point fast enough. Like killing scouts, that do not go where they want them to go. Like killing people that “steal their kills”. Like fighting each other instead of enemy. Like klling people from “enemy” clean being their teammates. I’ve witnessed it all and more.

  62. wow, I wish I would have time reading all the comments.

    just a brief comment of mine… now playing without xvm since thousands of battles. I remember all the battles with 5% win chance, where me and my “tomato team” have beaten green enemy teams like a bunch of noobs. lol… personally, I find xvm totally useless and distracting. I use performance tracking services to follow my own progress to see where I need to improve.

    what defines a good WoT player? thousand of different things.

  63. Podle mě není winrate správný způsob pro posouzení kvality hráče, protože výsledek bitvy dokáže jeden hráč ovlivnit jen zřídka (cca v 10% případů) což ukazuje onen rozdíl 40% – 60% výher
    Dále je winratio závislé i na konkrétním tanku, s některým tankem je totiž výrazně jednodušší ovlivnit výsledek než s jiným.
    No a v konečném důsledku je výsledek ovlivněn kvalitou týmu.

    Rozdělení podle zkušeností není zřejmě ten nejlepší nápad, problém s rajčaty spíš je, že díky všelijakým akcím se i velmi podprůměrný hráč dostane k vysokým tierům a tam dokáže pak napáchat daleko víc škody než na tieru 1-5

    Zde ukážu jeden krásný příklad hovořící za vše.
    Takovýto hráč by podle mě neměl překročit tier 5

  64. I have XVM and it is useful in spotting who on the other team needs to die soon and who would be good to team up with in an informal platoon. I have been in teams with 38% CTW and won and have been on teams with 78% CTW and been trounced (The sneaky bastards on the other team were not paying attention to XVM and did not know that they were supposed to loose). Not much is certain except that if you don’t try you are certainly going to loose.

  65. Just a notion. There can be average players by skill, and by points. And those two do not necessarily meet.

    eg. Its correct that majority of players are yellow or below, when they are compared in xvm. But whatever color they are put, IT DOESNT MAKE THEM ANY BETTER. They are still shit and useful for nothing. Green would become new “pea” or something else, but they would be dissed until end of the world.

    Where I play, we call red players as “kurwas”, polacks, siemkakurwas, eastern european fucktards, and so on. And that trend is quite true, most the poor players are from the eastern europe. Thats just a fact. Now you just have to get over it, you and your countrymen are just shit, live with it or get better.

    • Generalisation level – OVER NINE THOUSAAAAAAAAAAAND!

      While the number of tards from Poland is high (which is not surprising really, since Poles are quite a large percentage of playerbase) you should also note, that there are some extraordinary players too (Loofah for example) and Polish clans are quite strong (both in cw and tournaments). Do I really need to mention some of those clans?…

      I wonder how do you call noobs from western europe (surprise, surprise – personal skill does not correlate with nationality)? Tea-tards? Frog-tards? Pasta noobs?


  66. I do agree that visible stats cause rage, and also that players are delusional about their skill – it’s a problem with humans in general, unfortunately.

    However, banning stats wont stop stat rage, because then players will judge by other visible criteria: which side has more ‘better’ tanks, or the ‘better’ side of the map, or whatever. At the end of the day, judging a likely outcome based upon previously known patterns is a thing we humans do, and usually pretty well – but if you want a list of all the situations where our pattern-judging skills fail, check :)

    As for the rating itself (as in, padding WR via KV-1S): check out the new WN8 rating, and the WN wiki:

    The main feature of WN8 is that it makes many forms of stat-padding useless – it doesn’t alter WR stat-padding, but nothing can, and it values WR only very slightly, so it doesn’t help your WN8 rating to pad your WR that much. Most importantly, it stops people playing OP tanks to get ‘better’ ratings than their natural level of talent suggest.

    And just finally: WN8′s colour boundaries are derived from population distributions – thus a red player really IS bad, and a yellow player (ie, me…) really IS average. Observations lie, data doesn’t :)

    • “The main feature of WN8 is that it makes many forms of stat-padding useless”

      While WN8 does a decent job at minimizing the effects of some forms of statpadding it is a far claim to say it makes them “useless”.

  67. there are some really strange twists in your argument.
    at first i dont know why xvm shows you as green because i dont use xvm and dont know what information bases this colour on, be it wr, eff or something else. i think most will agree that wr is one of the worst indicators for “skill” so the majority voted for a rating based mm.
    considering that there was recently a post on ftr about a new method to measure whether a tank is op or not and the newly developed wn8 i really dont know why u still use winrate as an indicator and conclude therefore that it would be senseless to introduce a skillbased mm. sure wn8 still has some flaws, mainly because the wot api doesnt show necessary information but it for sure is a lot better than winrate.

    secondly skill based matchmaking isnt a clear term, some mean what you described, others mean something totally different. i wouldnt like to see your version of a skilled mm. a matchmaker that balances the teams not only around battle value but skill indicator too would be favourable imho. cant be that hard to shuffle the teams around a bit so that not all tomatoes are in one team and all unicums in the other. yeah that would result in most winrates gravitating to 50% but as winrate is meh anyway no one should care about that. unicums will still carry and keep their high wn8 and tomatoes would still die in the first two minutes and keep their red rating. in the end we would just have more balanced and thrilling matches.

  68. Oh well, you finally made me register to comment.
    This post is one of your worst.

    Two points:
    1. Regarding your KV1S statistics
    You won 101 out of 156 battles. An average winrate of 49% should have given you 78 wins. So you won 23 battles more than expected.
    You admitted to padding your winrate by doing “like 20 [battles] in a company” and a few platooned (even platooning with two 53%-ers means a handful of games won more). This easily explains most of your winrate increase. Add to that the fact that 150 battles is a real small sample size and that the specific tank might just be one that fits your playstyle (so you win a couple battles more even playing averagely).

    SS, you are a great asset for the community and one of my personal Wot heroes. But like most players and forum posters, you are at war with statistics – and losing it.

    Most of your arguments are DIRECTLY IN FAVOR of skill based matchmaking!

    If matchmaking happens by winrate it means that padders will be put in matches with players better than them. There is no stronger incentive to NOT pad your stats. Skill based matchmaking would punish the padders!

    And a personal side note: Players don’t behave badly cause of XVM. Players behave badly because they lack character and XVM is just their tool. If there was no XVM they would use other tools. Instead of harrassing baddies, they would harras arties, or TDs. or light tanks. The other day I had two top tier heavies taking two minutes to push our arty into the water – “sorry lol this is for youtube”.

    What this game needs is an effective report system with hard bans.

    • One other things:
      Skill based MM ist basically a league system.
      There is a reason why all serious sports are organized in leagues.

      It is better for everyone involved when professionals at the height of their abilities dont play against the local seniors home’s team.
      It is better not mix good players with handicapped ones (Paralympics anyone?).
      It is better not to let 25 yr old compete with 10 yr olds.

      Only of course in WoT some people believe this mixing would somehow improve the experience.

  69. Skilled MM would be bad and i think everyone who has even a little bit of sense knows why. What i think they should however change is the distribution of noobs/unicums on both teams. One team shouldn’t get 2 unicums and all tomatoes. One of those unicums should be on the other side and that would be at least fair fight. No one wants to play with 99% players who are skilled as they are because that would mean much harder games to win and everyones WR would tend to go down to 50%, but no one also doesn’t want to be the literally only good player in his team while enemy team has like 50% of green,blue, purple players. XVM’s win chance tends to be accurate, but however it can be totally wrong and i managed to win 43% win chance battles and sometimes lost 70% + battles. How the battle will end depends on many factors including a lot of luck. And i seriously doubt any unicum will suicide just because he got into bad team. I for example wouldn’t suicide, but will actually try much harder to win the battle and do everything i can to help those “badies” to win it. Also the thing where you mentioned how average/bad players considering themselves good players is called Dunning-Kruger’s effect. It’s perfectly explained on wikipedia. Check it out.

    • So you are against skill based MM but find it is really important that MM considers skill when matchting the teams?!

      And this is brilliant:
      “XVM’s win chance tends to be accurate, but however it can be totally wrong and i managed to win 43% win chance battles and sometimes lost 70% + battles.”
      So you managed to win a 43% game lose a 70% one and that somehow shows the win chance can be “totally wrong”. If the win rate would be accurate you would be expected to win about 4 in 10 43% games and lose 3 in 10 70% games.
      Having said that: I deactivated the win chance because on every tier below maybe 9 win chance is not decided by some aggregate of skill but mostly by the fact which team’s imbeciles suicide first. But still: Statistics! I’ve lost 90% battles. So what. This should happen once every ten games hence: “win *chance*”.

  70. I know my stats suck hard:
    but I don’t want skillbased mm for one reason – for every failstreak, keyboard and mouse smashed out of frustration there’s a match where I shine – when it looks that it will be another loss I WAAAAAAAGHHHHH!!!!!!!!! my way to victory, do some good damage, make a “you shall not pass!” or win on my own against all odds (Kolobanov x6). I play this game for those epic moments, that’s all.
    And if skillbased mm was implemented such stunts would be even less possible. :/

    • I see what you did there. You are actually a decent player and fear that skillbased mm will take away from you the chance to pwn enemy teams consisting if potatoes.

      But fear not: That’s what team companies are made for.

      • Me, a decent player? Oh, stop it you… *blush*
        But yeah, getting a game in which I can pwn requires: enemy being dumb and my team being even dumber, but not completely retarded so, they can do at least some dmg.

        I don’t play companies. During my wot “career” I platooned for two matches. Solopubbing for the win! :D

        • I recently read the advice that solopubbing is the best way to become a stronger player. You are doing it right.

          • It’s possible that it’s the best way to get into some psychiatric facility, but who cares? SANITY IS FOR THE WEAK! :D

  71. You make 1 important mistake, the same that Storm and SerB is making every time, when they talk about skill based matchmaking.

    It’s not about putting noobs with noobs and skillozords against skillozords.

    It’s about making battles where number of unicums and number of tomatos will be equal in both teams.
    The point is to make teams more equal, so number of battles were your team wins 15:3 and you have no time to deal damage or number of battles where your team loses 3:15 and you are alone against 5 after 2-3 minutes of battle.

    It doesnt matter if my team is composed of 14 tomatoes and me or 14 unicums and me – as long as on the other side are 14 tomatoes and green or 14 unicums and green.

    Battles were I have only tomatoes in my team, while on the other side are 5 purple players are no fun at all.

    • One more thing about XVM – xvm is a tool, blame idiots who whine and quit. It’s their stupidity not xvm that makes them quit.

      Why most people are red? They dont. Most people are yellow and orange.
      However xvm has a bit inflated stats – WNx is based on results taken from pages like vba, which are used mostly by players better then average, that’s why average in XVM is like 30% from the top.

      Even without it, there is no reason why most popular color should be the one for average player. If bad players are playing more then good ones, they will be overrepresented in battles.
      Also you can compare skill distribution in tier 10 battles and tier 1-4 battles. The lower the tier, the more tomatoes in average battle, which is understandable

    • Let’s take a look at your proposal from perspective of afk/bot player. System would find another afk/bot and throw him in the enemy team, yes? Same for other players.

      So in long run our little afker would have balanced battles and thus achieve 49% WR. Same as green player, unicum player etc. So no matter how (un)skilled they are, after thousands of battles all of them would have similar wr.
      I would consider this a flaw, because the main goal during every particular battle is to win (and have fun). I think we agree on that? Kills, damage, spots are far less important. Clever tactic, teamwork and knowledge of game mechanics are just a tools.

      • I wish I could edit my posts (‘as flaw” …)
        Now we have incentive to improve skill level, if someone is bad he is crippling his teams chances, losing more and getting less fun and credits/exp. After changes this incentive is gone.
        He would be just carried over more often because of balanced/mirror teams, getting more credits/exp (bonus for winning). So we would have more terrible players in high tiers… There too many of them even now.

  72. Great article SS I personally don’t really care about winrate cause their are so many factors that can affect it, myself I have a 47% winrate and it may look like I’m not good but I am defiantly not bad I understand the game mechanics very well and have good combat strategy but terrible execution but sometimes I can execute them flawlessly.

    • The average overall WR on NA server is about 49%, if your overall WR is lower then your simply below average. Does that mean your bad? Sort of. The reasoning may be due to low number of games, you continue to blame bad teams after 5k battles if your WR isn’t increasing.

      The difference with “blue” and “purple” players is that we play to win even when the odds are stacked (tomato team) and occasionally we pull off wins in such cases which is why we have higher WRs.

  73. Anyone ever thought about not trying to persuade WG to bring “any”-rate MM into random battles and create new type of battle ?

  74. Perfect article. But, to be honest, XVM stats should be merely banned. For people who know how to use it, it is a cheat (focusing the able players of the other team and denying them the right to play); for the others … well … “OMG tomato team I quit”
    It adds nothing positive to the game experience. Except for the lammers and stat padders.

  75. As Ascender said the basic problem with a skill based MM is that everyone ends up being average!

    The first thing to understand is that rankings are relative. They measure your position amongst the players you play against.

    Simplistically, imagine a pool of players where some players win 60% of the time and some 40% of the time. In a simple win rate based MM the 60% WR players would only play against other 60% WR player and they’d all soon be 50% WR players.

    It’s important to understand this. As the players are all of similar skill the wins will be evenly distributed across the battles. Any given player is no better than the others and won’t be able to influence the battle greatly.

    The same is true of the 40% WR players playing each other. They too would be soon be 50% WR players and so all the players would end up playing against each other. As soon as your WR improved you’be be matched against other slightly better players and your WR would return to 50%!

    So matching better players against better players means you’re all average!

    If you base the MM on kills then as you get better you’ll only be playing other similarly skilled players and your kill rate will drop.

    If you want to be ranked as better than other players you have to play against those other players.

  76. I can’t blame xvm for this stuff happening around but the idiots that are missusing it.
    Win chances are there to give a perspective about the game.You can lose matches with 99% win chances as well as winning matches with 1% win chance.
    Those crying about low win chances are either retarded or to bad to worry about anyway.
    Yeah, maybe it demoralizes minds here and there, but no-one foces you to keep it on(i took win chances off for example because I don’t count on them and neither need them) in the first place.
    A “red” player is an universal term.Everyone used to xvm describes m as the players from average ratings to very bad ratings…A true “red” team doesn’t appear that often…there will always be average to good players in them.Also, matches with full red teams against full good teams happen 1 time in a million.
    Xvm musn’t be used to predict matches…it’s used to give you an idea of what tactics you can use, which players you might be able to count on, which enemies you should be careful with(not avoid, but not yolo either) and how to adapt trough the course of a battle…that’s all.
    And about unicums screaming about noobs…I agree with that only partially…I mean, it’s annoying sometimes to lose a match because a certain player does a foolish mistake which was the reason of losing a match which could been won.
    Yeah, I do mistakes too, often, but when I do those and lose I shut my mouth and admit I done something wrong because I know I have a part(or even full) of fault.

    XVM is a good mod that it’s been missused by most of the players, no matter if they are skilled or not

    • Oh, and I agree on your skill mm thoughts.
      Noobs will always play with noobs and will never improve.
      Skilled players will play with skilled players untill they’ll all reach an average win ratio which is a negative aspect of the game(You play to win and you have fun on winning, not on losing ffs…).
      And the statpadder stuff too…but those will turn to average skill level in no time.

  77. Why does it have to be tomatoes with tomatoes and unicorns (or disguised tomatoes as it were) with unicorns?

    Is it that hard to understand the concept of balancing the skill base tween the sides? Sure, stat padders are going to mess that up, but how bad a problem is that really? Is the game loaded with rampant stat padders?

    Just have the MM check what the players stats are and not pile all the tomatoes on one side.

    • Once the two teams are balances (same number of tomatoes, unicorns etc) then with evenly matched teams win rates will group a 50% across the teams. Win rate will cease to be a useful differentiator and we’ll have to pick something else.

    • Not hard, but balancing teams consisting tomatoes and unicums is pointless – tomatoes will get smashed by not tomatoes as they are now, and unicums will be frustrated as they are now.

      See battles with mixed teams – usually it looks like tomatoes die first, and battles are finished by alive players that are above average. Sucks for tomatoes.

      Separating kinds would result with tomatoes playing with tomatoes, and unicums playing with unicums.
      I think that there would be much more camping at tomato level (1/5th of bots and afkers) than in unicum one (after all – they want to maintain their stats, so waiting may not exactly work).

  78. Well intristing article
    i do agree on some parts and disagree on others
    i do disagree on the fact that all thing that 50% and bellow players would play in seperate server and the goood and above players play in another no this is just a stupid implacement
    what should be done is the MM should make the teams not only even in tanks but in skill
    now it is just a bit broken you have teams where there is sometimes a platoon of unicoms and then you see that the others players are good or great and then the enemy team is full of reads(mostly) some yellow and some green players in it
    also to clarifie one thing they are 2 class of noobs
    the newbies where they just start the game and they know nothing about the game at all and WG does nothing about that if you tell a new player that this tank has this tanks has scout MM he will ok and???
    which means these players get put into battles and you have the better players raging at the guy saying him noob scrub and many other good cosmetics things
    and then we have the PURE noobs
    theseeee are the idiots to say who have tones of battles on there belt and still play worse then newbies
    this is a product of the dickness of other players someone above mentioned using the reds as meat shields…. they are many giving the so called tips and you have them gotten killed then the player who got killed in favour of the pro get pissed and they start not hearing others even on the most logical thing which is cap
    now should skilled MM be implaced as i mentioned above putting the bad/new players in one server and leaving the so called unicoms and stuff on another is retarded

  79. I have 55% win ratio and based on xvm, noobmeter etc i’m green in most stats. However it is often that i just don’t have the feel of the game and do bad decisions, in addition im put into a lot of roflstomp battles, on the stomped side. Yet stats show me good cause half of my tanks are fun tanks im good with and are maxed out (crew, equipment).

    Stats are justs showing the numbers not the causes.

  80. Am I stuck with yellow/green stats because I play with a.) a majority of stock tanks b.) many tanks that aren’t considered very “good” (popular) [I like underdog vehicles because it feels nice when I can make them work for me] and c.) a majority of sub 100% crews?


    Very good read btw, SS!


  81. >>>>>>a “red” tank hides behind a building? He must be a dumb noob, camping here! a “purple” player hides behind a building? Ah, he’s waiting in ambush, surely because he expects the enemy, smart! – and so on.

    Yes, that´s because the red player gets buttfucked cause he has no clue what is going on behind him, and the purple guy can actually use the building as cover.

    Personally I am lazy to recunfigure XVM so it shows the stats again (config got broken like half a year ago), still I can identifiy noobs by their actions in battle, sometimes I look up their stats and voila – they are red.

    Let´s look, why the majority is red. “Average player” in this case is not the player with the most common performance. For XVM the “average player” is the player with average results. Since the average winrate is 49,5%, these players are yellow (neither red nor green = neither bad nor good).
    I really see no major problems with XVM. Players who drive off a clip right at start screaming “aaargh i was put in team with reds” will do that 2 minutes into the battle when the score is 3:0, so basically no difference, thos people screw up battles independant of XVM.

    The skill MM was demanded (noobs and whiners always demand) loooong before even the first versions of XVM appeared, so it´s really not an XVM issue.

  82. great POST

    i uninstalled VMX because i found i was playing to the Stats. When we should win i played differently than when we should lose. I was then not sure if we were wining or loosing based on my change in play style.

    i voted against skill MM

    I still thing MM is broken though and its a REAL problem with the number on the server drops. where the number of tiers on each side is un equal. i can’t see why once the teams ahve been assemsmed the MM could go back through them and swap people side to ensure equal teirs

  83. All of these points make a lot of sense, and I can agree that the color coding is skewed, not to mention that, indeed, winrate is relative.

    I personally have a WR that fluctuates between 52-54% (It was 54% before November, then as soon as November came around…well let’s just say people probably had a bit too much “Holiday Spirit”, and it dropped to 52% – it hasn’t been quite long enough for me to recover from that), which has me showing up in varying shades of green on XVM, stating that I’m above average (52-54% WR is in the top 10% of all players). The moment you hit 51%, however, it changes to yellow, then at 49%, orange, and red from 48% and lower, which of course doesn’t represent relative statistics accurately.

    HOWEVER, due to there being so many “averages”, the result is that “average” players are more or less unpredictable: are they genuinely players that know what they’re doing, but can’t pull it through when it really matters, or are they players who normally suck but would be sucking even more if they weren’t playing a particular tank or platooning with somebody? If I’m up against a Unicum they’re usually rather predictable, as the game has a meta and understanding that meta usually means one takes a certain course of actions, an “Average” player, on the other hand, is unpredictable, and thus, sometimes, even more dangerous, which is why I take into account more than just stats, such as what vehicle they’re driving, are they platooned or not, what map are we on, what side of the map are they spawning on, what gamemode is it, have they been spotted yet, WHERE have they been spotted, etc.

    This is the part that many players don’t get when they use XVM: there’s too many variables to take into account for the stats aspect of XVM to be truly reliable, which is why I tend to ignore them and just use the other parts of the mod.

    Of course that hasn’t stopped some unicums from using “Insane Troll Logic” to justify their egotistical behavior (one in particular on the NA server’s forums a while back – I think it was Kewei – used Ayn Rand’s “Atlas Shrugged” to justify their behavior and that of other equally egotistical players – not bashing anyone who agrees with that ideology, but seriously, as if there wasn’t enough controversy already).

  84. SS, let me start off by saying that I absolutely respect what you have done for WOT community and I have been reading most of your posts on this blog, even though I have stopped playing WOT for almost 8 months.

    Now, I’m very sorry to say this but I cannot believe how short sighted you are for making some comments about skill-based MM in this post. Here is what you have gotten wrong about skill-based MM: Almost everything.

    I believe like you, most players would agree that statpading is a major part of this game in high level matches. Quoting you, this is an issue that raises unhappiness. Issues like this including OP tank abuse, unicum platoon abuse, and gold shell abuse is making this game very “unenjoyable” for a couple group of players: 1. Starters, who may just want to try out this game and didn’t spend any money yet to get premium tanks, 2. Regular-skilled players, who may have been playing this game for a while and maybe trying to decide whether they need to spend more money in this game, 3. Lower-skilled players, like the one who loves complaining about losing yet never had an above-average win rate to begin with.

    SS, correct me if I’m wrong. I believe the 3 types of players I just mentioned makes up the most of WOT player base. You think that a skill-based MM cannot help them, that’s probably because you are too short-sighted to see how a skill-based MM can help new players, satisfy average players, and give low-skilled players a equal chance.

    You have made an example of creating a new account and pad one’s stats with it. Let me ask you, when you are so good, so pro with that newly purchased mini-maus, after 10 wins in a row, MM decides you should be matched with 29 other players just as good as you, are you still going to be able to own new players like that first 10 games? Hold the though, you may or may not depending on if you are really THAT good. Look, when you pad stats, there are as many of 15 players would be negatively impacted by your action because they will most likely to lose. This kept happening because there’s no mechanism in MM to prevent abuse of using OP tanks or abuse of skill against new player. A skill-based MM would correct this by simply making you try harder to win. Such MM match a new player with players at his/her skill level, just like a game with automatic difficulty adjustment. How would a MM like this not fix statpadding with new account?

    It is invalid to say that skill-based MM is making everyone average. How? Because if you are THAT good, you will win. Even in a match against 15 players who are equally skilled as you, the team works together has a better chance more to win. Such example exists in WOT, just look at the top clan’s CW win rate against other top clans. I can’t say for other servers, but at least in NA server, back when I was fighting for a top clan, we had more victories than any other clan on the server, it didn’t happen without a reason. Similarly, Top tourney team don’t have a 50% win rate either.

    If anyone still don’t get this, just reference a true E-sport, such as SCII, LOL, DOTA2, etc. Which one don’t have MM that considers skill? Why is LOL the most popular MMO out there? Even WOT has much better graphics, even WOT has much more playable “characters”? Even WOT has TANKS? Because a skill-based MM is essential for a game to absorb new players. A skill-based MM also forces a good player to play better if the player is TRULY good. This, is what makes E-sport a sport, not abusing OP tanks, not statpading, not taking advantage of rigged designs. Those are the reason for making money off players, we all knew that.

    And I know for sure a skill-based MM will lower some players’ stats, that’s because it is what a skill-based MM is designed for, to rule out those who are not worthy of their places. And before any “unicum” trying to bash me, I don’t hate any of you, but showing off that stat you never deserved, is wearing me off.

    SS, I assure you that there’s no good measurement of WOT skill due to many disadvantages of game design. But there’s nothing wrong about XVM, it attempted to provide a measurement to players who believes it. It was funny that you mentioned XVM should be banned in order to solve statpadding all together, because it was like saying guns should be banned because it can be used to kill people very efficiently. Here is my point, XVM is like guns, it was a tool, the goods and bads is all due to the person who uses it. A murderer will find another way to kill without a gun just like a statpadder will find another solution to check stats. In the end, it is available to everyone.

    Sorry for ranting so much. I assure you SS, I only made such reply wanting to show another train of thought that you might have overlooked. This is only a game that I have abandoned for over 8 months yet I do think it is worthy my time to type this out so maybe one or two people may read this.

  85. To me everyone below 52% winrate is mostly worthless in a random battle and thus coloured red in my XVM.
    That being said I have advocated a ban of XVM since more than a year now. It’s just as annoying to get focus fired every god damn battle than it is if abuse gets thrown at you every battle as well. Both will make the game significantly less fun.
    The really funny part is that as with everything else the baddies fail when focussing only on the unicums because those are harder to kill and you’ll more likely lose the battle everywhere else. So being good makes you win even if you don’t do a lot just because of the attention you draw. That’s good but not fun :/

    • Not to mention all the moron hightier-arties out there focus firing you all game long as the only T8 heavy in a T10 game, even though you are using every cover available. Most don’t even realize that they are actually aiding the enemy team by doing that.
      Envy is such a strong emotion and also one of the most disgusting.

  86. I had game which started with one player’s message: “7% win chance”. We won. So yes I dont want to use XVM :)

  87. Well SS, regarding your request for a coloring change – it’s user customizable. There are even on-line editors, so no need to wade through xml files.

    Furthermore, you are correct about global WR and EFF rating, there’s however a more interesting part, and that is the XVM enables you to display not just global ratings (that can be biased by stat padding etc), but also a stats for the respective vehicle the player is driving at the moment.

    That gives you a lot better information as you will filter out bias from other vehicles and can only focus on how the respective player with that actual tank. Try that out…

    What I however find more disturbing is the win probability prediction. That is the fun killing factor as when you see, that your team has 23% win chance ;). I can only recommend to disable that feature locally.

    • yep, i disabled it and only use xvm to identify the best enemies..and have a good laugh at some of those tomatoes tbh xD

  88. I don’t know if this was mentioned already but the color scale, at least on, is based on population percentage.

  89. half hour reading and still don’t get it
    why skillMM

    1. cos the MMker is working flawless and speedlight fast so new conditions won’t add to the waiting time – i don’t want that for sure

    2. cos WinRatio was not enaugh so we got efficiency rating but cos of cap mania it was not good – then we got
    WN1 to 6 and was not good either …then WN7 now WN8 , and in the mean time PR and others….so we make skill MM based on WHAT?????? – till we get a formula that will be good for all regardless of tier and tank type …big NO

    3. cos tomatoes are everywhere so if they implement it in the most reasonable way with similar skilled teams (2 unicum + 7 average + 6 tomatoes for each team with every tier having same weight in skill) – those tomatoes are still there…and they will “ruin” the game and fun anyway
    and if we have games of 15unicum vs 15 unicum then….some unicum will have to be the new tomatoes…..and for games wuth 15 tomatoes vs 15 tomatoes….they will kill each other and wont have an ideea of what the do wrong…and will complain that they are good and need to go to upper levels….but based on what ?…they are good at theyr level…how do you promote them?…and what you do once they promote and then are bashed…if you relegate them…will have new whinnings……wtf…you solve a problem generating others

    4. there always been a problem…type59 flood, kv1 he, bad moderators, rigged games, OP1S,nerf arty, nerf td, nerf anything, fv(183), he shells affair, russian bias, etc etc etc people love to complain and to find problems where there is no case…….let’s play the game…..for those that need to play against same level players…just go to clan wars and teambattles…..skill MM is IMPLEMENTED THERE ALREADY…so WTF is this DISCUSSION ABOUT?

    5. i had games with lots of greens and purples and blues in my team vs whites and yellow and reds….95% to win….result….3-15 in 3 minutes
    i had games with the opposite team colors…….15-0…….
    so what do people believe that skilled MM will solve?

    6. every system of calculating skill will have it’s padders….and they will be promoted by this……they are bad at all levels..i see nowhere and noone liking them …except themself

    7. an old player that played bad at the begininning..will have to stay at low levels even if he is much better now…..and if you take into account only last lets say 30 days….then if you take a break you are relegated …and if you take into account last 1000 games…..but you played last year…you still have to accomodate….so again NO NO NO….

    yes i know i am trolling with this post….but…get serious……it is not a real problem…..till now i see more problem arising from it than problems solved…so i say BIG NO…random battles have theyr name for a reason….THEY ARE RANDOM…..go to a dictionary and read the deffinition of RANDOM

    and last word to those unicums that get bored…..if you are bored it means you spent too much time in the game and neglected other activities……dudes…even SerB wrote at one point that this is WRONG


  90. I realize I’m a bit late to the party, but I still wanted to add something. I personally don’t use XVM, though I’m often in a situation where I wish I would (usually when I need to decide weather or not to rely on a teammate, I’d like to know if he has a vague idea of what he’s doing).

    The problem with the discussion of skill based matchmaking is one of definition. I don’t think that basically splitting the community by winrate, efficiency or wnX would help anyone in the long run, though it might make games a bit more even. Bad players, who are just relatively new but not ignorant won’t be able to watch better players to learn from them, for example (yes, I believe this happens).

    So, what about the definition of skill-based-matchmaking? For me this term just means that the matchmaker takes skill into account when building teams and that it’s a fundamental metric, not an afterthought, for it. I don’t mind playing with high-skill players in the other team or bad players in mine. What I do mind is playing against great players with only bad players on my side. The matchmaker just has to even out the teams, not just in terms of tank power and/or matchmaking weight, but also in terms of player skill. The way to do this can be as simple as adding up the [insert rating system here]-numbers for both teams and requiring them to be roughly equal, same as it is with matchmaking weight of vehicles. Many other variants are of course possible, just ANYTHING to make the teams skills be closer than they often are at the moment.

    The whole point is: I don’t (just) play WoT to win, I play to have a good fight. I don’t mind losing a game that is close when the other team just plainly did a better job (though I of course would prefer to win), but losing 3 close games is so much better than steam-rolling the enemy team 3 times for no other reason than them being screwed by the matchmaker (player-skill wise).

    I have basically stopped playing for now because of this. I kept track for a few days where I played multiple hours a day, and just over 2/3 of all matches were very one-sided – one way or the other. That is not fun for either team, and quite frankly not worth my time (or money). I’ll play a match every now and again, but that’s about it, and I doubt it’ll change unless there is something fundamentally changed in regards to the team balance.

  91. Hi SS,

    u czech, mate ?

    Btw…dont be hating on XVM stats, doesn´t matter what colour the player gets from the code, even different XVMs use different colors for same stats. Sometimes I´m orange, sometimes I´m red with GR of 25 (654) and WR 51. Btw none of those XVM´s seem to update on regular basis, because I believe my GR is now much higher, also my WR is now 53% and it shows me 0.1k battles in tanks I have 200+ battles with…
    What was my point however is, that if you can suppose the XVM is actual (see above) you can roughly say, what can u expect from that player…Though you ´re right that you can only differ total noobs from decent players by comparing their GR (15-20 vs 35+)