Insider Returns to Talk About Skill MM

Hello everyone,

the Insider returns to tell us what the developers are thinking and talking about. And I think it’s very interesting.

The developer topic of the month… you guessed it: skill MM.

Lots of developers don’t want skill MM, because they are afraid it would break the game for a lot of people. Here’s one example of such a post to illustrate:

“This will make players with high battle count but not so great stats the boot on many ways, because theres a point when they realize they get thrown with the “special kids” class and he cant climb the ladder to be with his friends on the next class, whose clan only recruits on certain ladder class, this will also deprive the game of the ability of players improving, and would end up profiling most of players, not to mention, what would happen if an unicorn found himself not in the same “unicorn” league as his friends? this ought to create something bad, we’ll be excluding a large amount of players like this, and some might call our numbers biased, are numbers really biased? what happens when players that only want skill ladder matchmaking find themselves out of it supposed skill level because they happened to platoon to inflate their own stats and turns out his achievements are but a “fleeting dream?”

The list of such posts goes on, some developers are truly dug in against it. It’s not just anti-skill-MM rage, they are also brainstorming the ideas of how some of the disadvantages of skill MMM could be fixed. It is confirmed that the developers are actually really experimenting with skill MM on internal server. This is however NOT something that is sure to come and even if it does come, it will not be anytime soon, definitely after Havok. Nowhere near the foreseeable future anyway. Developers will wait for the poll results (the poll I posted yesterday) and they will run many closed tests to.

Here, the insider emphasizes: it is just an experiment, the chance of this being implemented is very low. Here are a few points from the testing phase of skill MM:

- skill MM won’t work in platoons (for obvious reasons)
- performance will be based on the personal score of your account (no XVM predictions and such)
- however it may also count the number of times you have platooned
- hard to balance players who are superb at one tank (“light tank god”), but mediocre at others (“heavy tank piston-burner”):
this would mean they have to create a dynamic or at least a flexible match maker for these situations, which very common actually, and could open an exploit, so they will make the requirements stricter
- artillery will suffer again, because they have a lot of RNG in matches, and likewise, TD players may get unexpectedly harsh battles
- if there’s not enough skill players for certain skill ladders, they would be added from the next ladder in line and so on
- skill MM could be actually used as an activate option, but it won’t be very popular after a few weeks, or so they think
- this may create a rift in clans, if a player intents to recruit players based on personal score, it may not turn for the best, so they will rework the player rating formulas (if it’s given a good to go)
- balancing skill MM means also balancing tanks to a degree (49%-50.1% global is mentioned), however its not currently viable for certain tanks, not without causing a whinestorm

(that point was mentioned by a developer with a quote, that apparently is very significant)

“One part of the game is RNG itself, whoever thinks that can win based only on skill will not triumph nt the game, RNG is what gives the ability to a regular player to outscore a rampaging player, likewise, it’s the same RNG that allows a player to become a Monster itself, but they don’t realize it until it’s taken from them”

- bad players are part of RNG and so it’s part of a match variable, take that out and the match may become dull, likewise, fill a match with underperforming players and it’s gonna be very hard for a player to get out of the hole, or someone may actually exploit it
- even if it’s implemented, some players will call “conspiracy”, because whether they like it or not, there’s ALWAYS gonna be bad matches where someone is steamrolled, or some good matches where they will steamroll, thus giving the apparent impression that it was never “implemented” to begin with
- it might encourage players with regular skills to leave the game because they can’t “get better”
- the game was designed for a regular family member, not hardcore players, who can only think of min/max, having skill MM goes against the principle of what the game was designed for, however it could be used in league games effectively
- currently, there’s an active ingame “limited MM”, but only for new players and players who bought new tanks: for the first 3-10 matches upon purchasing said tank and playing said tank, they will get preferential MM to “learn the newly acquired tank mechanics” and to make stock grinds easier. However, this special rule is negated if the player platoons or if there’s not enough players to fill the MM criteria during said matches, afterwards it’s a regular matchmaker and this rule only applies to the purchased tank to prevent exploits, “so someone cant purchase a Luchs and get special MM on its tier 8 tank” – devs dont see any problem with that current rule.

The discussion goes on, it’s beeing going for past two weeks.

As a last point, remember how it was written that the developers are looking for good players? They found a candidate on US forums: flakker2.

130 thoughts on “Insider Returns to Talk About Skill MM

      • If unica guys don’t realise they’re unica because of stompping pubbies and not because they beat other unica then they’ll regret the day skill MM goes live

        • That’s the exact reason why I think you’ll find most unica (and generally most above average players) don’t want skill-based MM. You’re only so good as the opposition allows you to be.

        • playing with the clone of yourself = bad

          divide for unicums 2 vs 2 instead of 4 vs 0 = better

          What I would like to see implemented would be:
          after mm chose their team do last 1 check – in each group of “same” vehicles (“same vehicles are lights of the same tier, arty of the same tier, other tanks of the same tier, so there are up to 9 different groups) make swap of good and bad player between teams so overall tank balance will remain the same and skill balance will get closer to being the same.

          some advantage of 1 team is not a probem but roflstomps are no fun

      • This kind of “skill MM” is not what the people were asking for, SS. People asked for a MM that balances teams, not individual players. No one wants those 15-2 or 2-15 steamrolls that happen most of the time. Developers did not understand their playerbase once again…

        • Steamrolls will happen with identical balanced teams because of the snowball effect.

          You lose one tank then you are weaker, so you lose another, and another.

          The only way out of this trap is not to have equal skilled teams to start with.

          • It is more complex than you put it. The snowball effect only happens at one flank, not at the whole map simultaneously. Also the effect only matters, if both teams lemming over the same flank and fight each other in a 10vs10+.

            The blue and green players would suffer the most. But everyone would gain a better gaming experience, because the matches would be more fun.

        • Its the same thing, having the same amount of skilled players and bad players in both teams will make the win chance 50% anyways, which in return will level out the WR of all players more evenly.

      • There is very good points in this post, and I agree many of them.

        I think workable compromise would be to put under 50% WR players to their own league, and over 50% WR to their own. If player gets better, he would graduate and be promoted to the adults league, and if it would be too challenging for him, he would be demoted back to the kindergarten league. Also to get promoted, player should have played maybe 4000 battles, so the beginners would not be promoted accidentally, if they happen to have a good streak. What do you think, would this make the game more fun for the beginners and also better players? This would also prevent the seal clubbing.

        • As has been pointed out before, this won’t really work because the players on the edge will keep flipping back and forth between the two levels. As soon as they get good enough to top 50% they get put in with better players, start losing again, and drop below 50%.

        • No offense, but the queue times are already long, the majority of the player base is sub 50% and that would me you would be unable to play wot say at 2 am when all your “Unica” go to bed, World of tanks should be world of tanks not Warthunder, which the queue times for the highly skilled last usually 20 minutes or so. Just putting that out there.

      • And how is that? Think about it like boxing. Heavywight boxer can’t fight with featherweight. Is supposed to be fun for a tomato to fight with a purple? That’s bullshit. LEAGUES, or teams consisting of players of the SAME SKILL, are the only choices.

      • The problem is not to have NO Skill MM or Full skill MM (like Starcraft, Lol etc) the problem is to have the SAME number of skilled players ate the SAME tier in each team. For example having 2 blue tier 10 in each team, 3 gean tier 9, 4 yellow tier 8 and the rest orange/redish in EACH team, it would change the games SO MUCH.

        Look what i had in my last game:

        https://www.dropbox.com/s/tugq2x2k37v4d5s/worldoftanks%202014-09-27%2015-03-29-31.jpg?dl=0

        The result ? 15-2. plain and simple, got steamrolled because of our 42/42/43/42% top tier, yes, you can understand that 42% top tier vs 56+% top tier = total rape.

      • I’m not a unicum, also not a bad player (WN8: ~2025). I don’t want a skilled MM they think about, I just want to make both teams equal. Same amount of Tiers per battle. Like 5x Tier 10, 7x Tier 9 and rest Tier 8 for both teams (numbers are just an example). And the “skill” of all players per team should be at least almost equal. The amount of battles which end with 15 to 3 or 4,5 something below 8 are to high

      • Now here comes an idea:

        Dont “copy-paste” teams onto both sides, but make the following:

        - implement 3-4 Types of players:and 3 basic types: Below-Average – > Average – > Above Average

        and maybe Newbies (unter 4k battles) with a special group….

        U can descend/ascend if u meet the requirements on atleast 2 of the following 3 things:
        - Winrate
        - Kill/Death-ratio overall
        -Damage ratio overall

        (numbers are JUST made up! thing to balance)
        47-54% Winrate is Average >54 is above 1.3 is above 1.3 is above <0.8 is below

        so a player with 56% winrate, 0.9 Kills per death and 1.0 damage dealt/recieved is average.
        while a player with 51% winrate 1.6 Kills per death and 1.4 Damage dealt/recieved is above average

        NOW, you have 3 types of players, let the Matchmaker make EQUAL teams on the above mentioned terms, which means 5 below average 5 average and 5 above average players per team for example. or 10 below average and 5 above average.

        how good or bad they are is something which can be RNG, but yet you atleast dont have total fuckups in terms of teams because u always have some player against you who can play and if ur team is bad the enemy is not extremely superior.

        What do you think?

        • the numbers got fucked up! somethings not right with using the “bigger than”, “smaller than” and “horizontal line” symbols.

          47%-54% winrate is average
          over 54 is above average
          below 47 is below average

          0.8-1.3 damage ratio is average
          over 1.3 is above average
          below 0.8 is below average

          0.8-1.3 kill death ratio is average
          over 54 is above average
          below 47 is below average

      • Most people have the wrong concept of skillbased matchmaking, I agree, but you are one of those people, too, Silent. It really isn’t necessary to create a ladder-environment where you only fight a certain bracket of the players. What is necessary, however, is creating an environment were both teams have a fair composition. Distribute tomatoes and unicums evenly between two teams and you’ve solved or side-stepped a lot of the issues that are mentioned up there, while improving matchmaking a thousandfold for everyone. Bad players playing with good players and maybe even learning something from them instead of wading in their own sea of red after their own early demise for eternity and good players not having to fight their own team on top of the enemy team, because the enemy team faces the same issues as your own team faces.

      • Here’s the thing, SS. Many other games have skill-based matchmaking, and no one complains about it, unless the Dunning-Kruger effect is coming into play ( http://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect) . Almost all MOBA’s, StarCraft I and II, some MMO’s in their PvP, etc etc, incorporate this sort of matchmaking.

        Now, it can be argued that the MMO pvp has RNG to it, but skill is also a huge factor (far more important is when or how one uses abilities, rather than if it hits for 100 damage more or less).

        There’s literally no reason skill-based MM can’t be implemented into WoT; indeed, if the devs want WoT to be taken seriously as an E-sport, they nearly HAVE to implement it. At least 50% of the whining will be because it’s a new thing and people hate change. And the other 50% will be the devs whining because it means they will have to drastically reduce or remove the RNG in one or both of penetration and damage rolls, which they inexplicably love so much.

        To all the naysayers saying unica won’t be unica. This is true from a stats standpoint – superunica like Garbad or Sela (US server… sorry I do not know the names of any EU or RU unica by heart) will no longer have win rates or WN*’s literally dozens of percentiles higher than the average. However, their skill will place them in a much higher bracket. The new unica will be Platinum or Diamond or higher (to use SC2/League of Legends terminology), and all the bads will be pushed down into whatever the lowest bracket is.

        TL;DR Skill-based matchmaking is a typical case of “naysayers don’t actually understand skill-based matchmaking.” However, we all know that it is very hard for Wargaming to “properly” implement something, so unfortunately the naysayers will get to say “we told you so.”

    • Fix MM first so no un-balanced teams (ie 5 v 3 top tier tanks….).

      Skilled MM…… maybe it is as simple as each player has a score associated to their ability and each team must have a minimum total score…

      eg Unicum = 100 but team of 15 players must have a total team score between 750 and 800? A platoon of unicums will have a team of tomatoes to play with….

      I am not interested in a laddered system where I would only play against players of my skill level… on SEA server we only have ~20,000 online at peak….

    • We need this skill based mm so bad. Biggest complaint everyone has in official forums is mm. Make it optional so we don’t get saddled with a team of special needs players. When I’m #1 in damage in a tier 8 scout wtf.
      The best part of this game is those few matches where my opponent is competent and through teamwork and tactics we win.

    • Im just curious if this so called “skill MM” is the same as “ladder” in the heads of all the nay-sayers. Because I am not really sure what this “skill MM” is, I just know that “ladder” does not have any of the disadvantages the nay-sayers are bitching about.

      • As I see it a main problem with “fair skilled teams” is platooning. How much time will it take to balance platoons between teams?

        And from personal perspective: There’s hardly ever a unicum playing alone. A Unicum platoon is 20% of a team and 10% of the overall tanks. That’s were a lot of the wipes I experience come from (a load of the others from “bot level” platoons)

  1. Why don’t they think about calling it Skill Balanced MM?

    Just trow the same people you have in the same matches as always, but with ”skilled” Players to balance the other one out. This way it can’t ever be a green team against one tomaytoe + 1 green team.

    • Agreed. I only want “skill MM” is for the MM to reduce the performance gap in players between two team. Not for the separation of the playerbase based on their performance.

    • That is the worst idea of all, your tomato never has to try because there is a tomato on the other team to balance him out, your unicum has to carry every battle but always has a unicum on the other team. Bads stats get better, goods stats get worse. Eventually you can’t stat match at all.

  2. Interesting discussion. I think it would be sufficient if they just made sure it’s not a complete green/blue/purple against red/orange game.

      • you make obvious mistake.
        If everyone has 50% win ratio, then there is no skill balancing so some will start to win more then others

        win ratios will get more flat, but if WR will be important part of your skill assesment, structure will remain as it is, only denser

  3. About the survey. I got this email and did the survey. But, the first three questions were removed from the survey. i did not get to see them. That is strange.

    Btw, skill MM will be boring as hell. just watch eny CW or tournament ware players ratings are equal. they tent to stay and camp in cover till the last minute. With skilled MM, every game will be like that!

    • I usually see bads camping after the good players who tried to get momentum going died….

  4. Yes but how can u win when in one team 6 in a row u have 5 -6 red players or white players and in other all pink, blue ,green and 2 red players how can u win????? and get more exp ????

    • Well, you try your best and probably lose anyway (but sometimes the enemy team gets overconfident and make stupid mistakes). There are some matchups you realistically can’t win, but they happen to the enemy just as often as they happen to you, so focus on your performance and in the end you’ll win more often than not.

  5. i am for another type of MM , battles participation MM including overall battles ,for example u are stock and u have 0 battle in tank then u play with other stock tanks tier 5-7 for example

  6. I would like to see another SKILL MM. Same amount of each level player, each team. Same number of red/yellow/green/purple player in each level both team.
    That would be nice, maybe balanced, and not hard to level climb the ladder.

    • This is what most people would want I expect.

      But MM couldn’t cope with it I think,
      and developers would ague that taken over 1,00′s of games that is what you have already. not through planning but by the law of averages.
      So it will be the game with clones line they go down.

      I wish people would stop asking WG to do some thing as what people ask for and what WG give us are never the same and never for the better. take game modes all failed. either had to be removed or reworked heavily.

      People you have a game that works and that you play if WG break it you will only have your self’s to blame.
      Be carful what you ask for you might just get it.

      • No this is not what most people want, in fact I’m sick of hearing about it.

        All you do is make it so good players have to carry every… single… game…

      • There is no way matched teams could make the game any worse, except for slightly longer queuing (depending on how strict the MM is programmed). Matched teams means more games are challenging and reduces inevitable and impossible wins, causing no net WR change but a more reliable experience.

        … unless you derive enjoyment from being ineffectual in battle and waiting for chance to throw you wins.

        • I just don’t get you. You make sure the enemy team has a good player to balance me in EVERY match I play. how can you argue this won’t cause my WR to decrease?

          • If you have to rely on the enemy team having worses players than your team and yourselv, you are not a good player and should shut the fuck up about such issues.

            • Rofl, a player is only good because he’s better than other players. By definition he’s better than bad players.

              I’m guessing you’re bad.

  7. preferential MM to “learn the newly acquired tank mechanics”

    I’m calling bullshit, my very first game in the T-43 was a tier 9 where I was the only tier 7 tank on our team, completely stock tank, I did fuck all in that one.

    • “preferential MM to “learn the newly acquired tank mechanics”

      I’m calling bullshit, my very first game in the T-43 was a tier 9 where I was the only tier 7 tank on our team, completely stock tank, I did fuck all in that one.”
      ———————————
      “…However, this special rule is negated if the player platoons or if there’s not enough players to fill the MM criteria during said matches…”

      • I’m pretty certain whatever devs are talking about that don;t know it was removed from the game.

        It was fun while it lasted, get a new tank, pimp it out, get a win rate above your ability.

  8. As many have already said, most of us do not want a ladder-based MM. If that’s what they’re chasing, then they’re even more out of touch than I thought they were. What the majority want is a Skill-*BALANCED* MM. Even using the crummy built-in skill scores to balance the two teams to within 10-15% would be a huge difference. It’s been frequently pointed out that most people don’t even enjoy winning the 15-1 stompfests.

      • No it doesn’t. It has no net effect, because the average match (statistically, not literally) is balanced. It only reduces impossible wins and losses in equal proportion.

        • What? You balance matches, skill-wise, so there is an even chance of success for both teams.

          How the hell can you argue it has no net effect? What sort of statistics are you smoking?

  9. Couldn’t they make it like it is in CS:GO: you have casual modes with no skill rating (=regular random battles) and you have a separate competitive mode with its own set of stats and ratings (=skill matchmaking). Specificly, CS:GO uses Elo rating, which is something WG also uses for the clan ratings so I don’t see the problem implementing the same for individual players. That way you can play for fun and grind in the regular battles and when you want to compete and test your skills you switch to “competitive” mode.

    • its not the same a 5vs5 game where everyone has the same guns, and a 15vs15 with 10 different tiers and a lot of variables like stock or crews.

      even an unicum cant carry a t9 game with a chi-ri or the opposite, an average player could carry as top tier under certain circunstances.

      also the core of the randoms is grinding tanks, i wouldnt like to be matched agaist good players only in a stock/bad tank

      • its not the same a 5vs5 game where everyone has the same guns, and a 15vs15 with 10 different tiers and a lot of variables like stock or crews.

        1. It’s not the same guns: playing with an SMG is not the same as playing with a sniper rifle. In WoT everybody could play with IS-3 if they wanted to.
        2. 3v3, 5v5, 32v32… it doesn’t matter, it’s team play

        even an unicum cant carry a t9 game with a chi-ri or the opposite, an average player could carry as top tier under certain circunstances.

        Never say never. A tier 7 can carry a T9 game. I’ve been in several matches where a tier 7 tank came on top of the team with kills and damage.

        also the core of the randoms is grinding tanks, i wouldnt like to be matched agaist good players only in a stock/bad tank

        Read my comment again: I never said skill MM should replace what we already have. It should be separate thing with separate stats that people could opt in and out.

  10. And how the hell does meeting purple player, who ownes like he wants, helps me to develop? xD Bring the skill MM!

    • “Lets say a new player has started his career in WoT. With skill based MM he will end up with 600-700 WN8 players with battlecount of more than 10000 battles, aswell as with players with 1000 battles , WITH SAME SKILL LEVEL. And from who will he learn, if hes not pitted against blue/purple players , whether they are on enemy team, so that he can observe them , or in his own team, so when he dies, he could switch to their tank to watch their moves, their tactics and their decidions?”

    • “And how the hell does meeting purple player, who ownes like he wants, helps me to develop? xD Bring the skill MM!”
      - Well because you might learn a thing or two, you know experiance is everything in wot. and still havent improved. If you inte future would become a decent to good players, would you want that advantage to be taken away from you becasuse you suddenly have to meet equal players? Good luck fighting them because you wont have any skill advantage any And the more you play and learn and the better crew skills you have, you will soon be able to do similar stuff as the better players. That is killing players that are worse than you longer.

      Many people dont realise that unicums and good players also has been new “noobs”, but they figured out the game, thats all. They reason they are good is because they learn the hard way. Now, wot is really too noob fridnly and easier than ever for beginners.

      Its all simple, if you are better than average you will have better WR and since 80% of the playerbase are 43-49% WR players, that means that if you perform better than them you will have better stats. In randoms, the chances are very big that 80% of the players you face are those having below or exactly 50% WR. Current stats are only because of random battle mechanics, just like battlefield stats.

  11. This SHIT is soooo big that i don’t even have words… Much better would be what i proposed earlier: skill-based REBALLANCER. In short:

    1. Leave current MM as it is.

    2. Once two teams are matched against each other reshuflle members slot after slot to rebalance for equal total skill. Works for platoons too, only they are reshuffled together.

    3. Skill indicator value is PER TANK, with LATEST PERFORMANCE in mind (average from last 20-30 battles). Proposed base xp as easiest, but it’s up to testing (and rebalancing xp earnings maybe, too).

    Whole this will make battles more tense and more demanding instead of smash&grab fests. While it may seem to “punish” unicums, it will not – it certainly levels WR even more, but WR stopped being skill indicator long ago – and if you are good, you will still be good, doing dmg, spotting, tracking and whatever is needed to win the game – which counts to wn8 ;) .

    • I don’t think this would be a good idea. The problem is that with such a system good players would always be balanced against themselves. Now my stats are nothing exceptional, but still in most matches they make me the best player on either team. This would mean that, instead of the occasional skill mismatch as happens currently, in most matches I would play I would have a team of handpicked tomatoes just to compensate for the fact that I am in it. Which means that I will need to do even more carrying than I already do. And obviously for better players the problem will be much worse.

      What I would like to see instead is a separate MM for the bottom 15-20% of players, who just fail to grasp the basics. They can play among themselves, so that it would not be possible that my team has to depend on a tier IX heavy with a 41% win rate. This should help to make your teammates a bit more reliable, making lopsided outcomes more rare. The rest can stay as it is now.

      • No, your idea doens’t cure the problem – you’ll get as much unbalanced battles as now.
        Mine will only remove rolfstomping which you advocate for, and yes – it will make battles more challenging for good players: you will still get your portion of unicums to fight against, portion of enemy tomatoes to punish and portion of allies to carry.
        Maybe it will be time to reduce loss punisment/win reward then, too.

        • Why should good players get more challenging battles than bad ones?

          It’s like saying, “Hey, you’ve managed to get good at this game, be it through hard practise or innate ability. As a reward, we’re going to make it ten times harder for you to earn your EXP and Credits.”

            • Who says that good players dont have it challanging alreaddy? Why should good players have it harder than they have now? Watch unicum streams of solo players its not like they mangicly get good stats “easy”. In wot everything can damage you even low noobs. And how fun would it be to get equally challangin battles all time, thats not the point with randoms. What makes people play are both challanging and easy games. A balanced between both from time to time. Leveling out the experiance is mediocracy.

              • Sure, rolfstomping yourself is much better than being rolfstomped :) …
                If unicums have hard battles with losses, than maybe some of them were imbalanced too and they would benefit in those cases from a balancer.

  12. WG as usual… if there are 2 ways to do something they select the middle way, AKA soviet style lol.

    Is so hard made a team has 4 unicums, 6-7 green players and 5 other type players??? the problem is in how MM put in a team all good players and bad players in the other AND how when teams are balanced on this MM put the best player in a team in the bottom and best player in the other as top tier (clasic is be best player in a team as scout tier 6 and other team has the best player as tier 10).

    Really WG with the lot of feedbak they have WTF do allways the worst shit!!! well, i know reply, when a Serb says that 90% of players ideas are shit and the ideas are ok are not new for them… well at least looks like AW is forcing WG move problem is that after years they forgive how do it out of add new umbalance lines of tanks.

      • Plus, what does the matchmaker do when there’s a platoon of three of them? That platoon would almost always end up playing with a team full of the deepest of reds.

  13. what a bunch of morons!

    for a lot of games, skill based MM actually works, (LoL, DOTA, MW:O and whatnot) and WG is complaining?!?!? they even actually tried?
    since when skill MM is just based on win-rate?

    instead of finxing and polishing this shitty engine, they gather data to prove skill MM is wrong!? WHAT THE ACTUAL FUCK!?

    • Quit moaning, dude. Too many XVM statwh0res have been complaining about “ZOMG WE WANT SKILL-BASED MM!”. Now that WG is actually looking into that, they’re morons? You have to appreciate that they are a business and they have to be VERY sure about their findings on changing one of the KEY mechanisms in their flagship product, or else it might run off the customers and that would be a financial disaster.

      I picked up a new clan member who had gotten himself in an E-75 and didn’t know how to sidescrape if I hadn’t showed him after seeing him get his ass kicked severely by a lower tier tank he should have definitely been able to kill. If it wasn’t for the current MM, he might not have figured that out. That is why I think it is bad to just separate numpties from good players. Balancing good and bad players equally over teams could be a solution, but is likely going to make the queue times much longer.

      • how are they looking into it? by making a case against it!?
        the fuck?!

        have you ever played a game with skill based MM, how the hell did we even get to things like noobs will be completely separated to stuff like queues will be longer
        HOW!?! because WG told you so? WG has no fucking clue how to do a proper software piece and they already know how’s it gonna turn out?!

        bunch of fucking morons

        if you can’t adapt by plain observation and learning from surroundings and enemy, you don’t deserve attention
        if your E75 buddy couldn’t see enemy doing sidescraping and learning from it!? I can’t even ….

  14. I am really curious about what all of you guys proposing the balanced numbers of players in each team would have to say, while (if this ever be implemented) you would have to face unicums in every goddamn match. If a unicum gets top tier and isnt in a terrible tank, he can easily carry the game single handedly. Basically what you mean is, that green players would be never “on the top of the foodchain” and would have to battle better players all the time. On the one hand, it may help them to improve, but on the other one, it may them lose interesnt in the game, because they would get stomped right away. I am not arguing about stompfests like 15-2 games are not fun (even I am usually on the winning side) but one of the fun factor of the game, as it is now is, that everybody can sometimes have a really good games by meeting players equal/worse than they are. Even a tomato can carry sometimes, if he has enough tomatos to face him. I mean, would it be fair to them? Everybody was a tomato once, and the thing is, everybody has a different learning curve. Somebody got the game mechanics after 3k games, somebody after 8k. So should we automatically say that 8k battles tomato is hopeless and put him into match where most players are better than him? Where is the fun in that?
    I think that while having more statistical sample of battles (1k lets say) the teams finally even out. And if you do your part in the battle and dont suck out, it should be no concern of yours if you lost or not, just focus on your own performance and hop into another battle.

    I openly admit, as a unicum player, that having equally good players against me would be much lesser fun. If there is a unicum platoon against ours for example, you have to take care of them and prioritize them (they do the same) because your half brained team wont do so. So it usually turns out into a battle of two platoons/players. After one wins, they stomp the rest of the enemy team. I dont really get all the advantages in that.

    Imho, the roflstomps with results like 15-1 or so are more caused by the terrible map design of new/reworked maps than such an imbalanced teams. Cannot remember last time when such a thing happened on Prokhorovka for example.

    • For starters, you’d only have to face unicums in every match (if you aren’t a unicum yourself) if unicums made up 6.7% of active players at every tier. Is that reality?

      But more fundamentally, matched teams would, assuming a reasonably meaningful skill metric being employed, force a unicum to have a similarly skilled opponent on the enemy team. Of course, they very well mightn’t meet each other in battle, but that is no different from now. At least both teams will have someone who might be able to carry them, instead of one team being stacked with a unicum platoon.

  15. Terrible idea and I hope they stand firm and don’t cave in to the vocal minority like they did with the arty nerfs and then TD nerfs.

    Can you imagine a “unicum” game… 15 campers on each side with no-one willing to go spot and possibly sacrifice their tank for the team. Remember those league matches with “pro” teams… LOL. Bunch of camping bush-clickers every single time.

  16. So much for anonymity – maybe you just ruined flakkers chance of actually having an impact on the game… I would be pretty sad if I had a chance to give my opinion to devs and it got cancelled because someone leaked my name :-(

  17. RNG is what gives the ability toa regular player to outscore a rampaging player,

    exactly, he just confirmed that battles are preworked, one team gets better players, more toptier, rng = 15-1

    • This has the one blatant flaw… Why does the current 40-70% winrate is considered a good grouping? It would be forced to closer to 49% from both ends, but it would only mean that smaller numbers mean much more. Say 60% is a 55% at the new system. That means that the 59% to 60% is not much these days, but that 54% is really horrid when compared to the 55%. The system would change, but the difference would be there. Random will average out, if you can beat more than 49% with that random draws here and there. You WILL get higher WR. It is law to being better, even when matched with the same better gamer one of you will be better and thus granted the win.

      All stats would get closer, much much narrower than now. And there would be less tomatoes all around… nope, there would be no less tomatoes, but the numbers indicating a tomatoe would just change.

      Think about that.

      • If all teams would be balanced with the equal amount of skilled players and every win chance would be close to 50%…stats in the long run would level out to the point there all would be the same more or less. Good players would meet good players all time, average would meet average all time and bad would meet bad all time…50% WR more or less for all players. What motivated them from progressing?

        Also the current stats players have are only because of the current MM. How can we balance players with a skill MM if the current skills are not the product of skill MM? Nobody becomes above average by facing equal players most of the time.

        • How do you decide who is better? Who wins more when you pit two of the best players against each other… ? One is better than the other, even if by that 0.01%. The imposed 55% could drop down to 51% or so.

          Balance? What is there to balance when WG introduces Recent WR among stats… Also the average is not 50% … not even by long shot. 49.3% something as IIRC.

          There will always be differences between players and others are simply better than others even when they have same WR it could be they are worlds apart on said match they are into.

          The idea of Skill Based MM is to bring matches closer in skill instead of large luck of draw… Who will pick the enemy team first is still much more than those Unicums could alone achieve. Every red dot on the screen is potential to them, every green.. is merely a tool to be used for those reds to be seen. Tactical usage is ever more important when you know your opposition is closer to your skill. There is someone to know how to answer your moves and you must be ever vigilant… Still there are routes and the 25% pen/dam RNG to ensure no match is the same. And tanks are still worlds apart how they handle. So unless we mirror teams… we will not have that 49% WR to everyone, ever. The amount to which single player can currently change the play is somewhere between 10% ( more for super unicums ) be it for better or worse.

          To use your team is every unicums way to deal with the enemy. It is in their nature to utilize them… So better packleader will always handle the herd better. To increase your chances are not that much to improve yourself, but to improve your teammates to your cooperation and vision of the perfect play. How your enemy team better players try the very same. Tactics counter-tactics. There will always be someone who can topple a plan into pile of rubble, but will your enemy see it in time, or did their plan ultimately fall into laid ambush of likely minded.

          You have to draw a card… it will either save your hand or destroy it.

          Even if the differences would go down from current over 20%, the difference in players will always be there to influence a game. No same WR% is the same even when taken from exact 0.001%. So if the differences could be ironed out to say less than 10%, it would be big difference in how battles are fought and how much more would each win mean. Up to the challenge to be in the 50% or drop to the sea of 48%.

  18. - performance will be based on the personal score of your account (no XVM predictions and such)

    Does WG even realise that their rating does NOT actually show performance? You can have a high rating just for having played a lot of games.
    Ofc people with a high WG rating but bad XVM stats will get smashed. Those are these kind of people who write you after the battle what a noob you are and tell you their WG rating is better than yours when in reality it doesnt mean shit.
    Please implement this “skill based MM” i will laugh all day, beacue a 35000 games, wn8 400 idiot will be my enemy.
    You can cleary see that WG implementet their Rating in a way that people actually believe that they are improving, they dont wont to have their customers realise that they actually suck at the game.

  19. FFS for not even understanding that MM should take player skill in consideration when it is building teams!

    Skill based MM SHOULD ACTUALLY BALANCE OVERALL TEAM SKILL NOT INDIVIDUALS.

    So in that sense bads would still fight unicum, only team overall skill would be more equal!!!

    Leagues based on skill like in other games just would not work, unless “league” is actually directly related to each Tier of tanks.

    In that sense WoT would have 10 leagues. 1 league for every tier, where unicum/sealclubber would be on top and tomatoes at the bottom.

  20. What is this, game for touchy feely pussies?

    “get thrown with the “special kids” class”
    I thought that’s exactly what you do with the disabled kids?? If you can’t compete in “real” world, you get sent to place with less challenge!

    “what would happen if an unicorn found himself not in the same “unicorn” league as his friends?”
    OMFFFFFG! Do you change their diapers as well??? Get a grip of yourselves, idiots!

    The biggest problem of this game IS and always has been the lack of skill based MM! Lets be real, how many battles have any of us been in a team in which a lead tank (or the nightmare 3 tank toon who get slaughtered in 3 min) is a fucking plant?! I’ve been in a lot and I’m fucking sick of it! WG is more concerned about some whiny bitches who will, god forbid, lose their unicorn (or whatever the fuck) status because they wont be able to play in their OP toon together or the idiots who probably shouldn’t play this game in the first place than about people who actually play this game the best of their skills not depending on anyone! Fucking babies!! Sometimes I understand SerB even though, he as well, is a fucking asshole!

    Furthermore, there is one more thing – with skill based MM, they would have work on another way of rigging the battles so they could contain overall 50% win ratio for everyone. Now it’s quite easy, they just put you in a team of 10 retards and you lose 100%!

  21. I fear the main reason players want skill MM is because when they lose a round, they blame their teammembers and skill MM would “solve” that problem. They’ld be surprised how wrong they are.

  22. A terible idea…It will destroy unicums, and all good players… imagine one unicum with 14 tomattoes vs ballanced lemons..= GG
    It will personaly destroy my game, i dont wana play vs clones… playing like that 50-50 will bring you 50% win rating…no one likes that…
    only 2 thing need to be balance are tomattoes… that one team cant have twice as much tomattoes than other team… coz tomattoes are so ussless…more like touchpad players… u cant left Mause 50 battle effi…it will lose from IS.
    And this MM balance is for Tomattoes ath first… Like they care to win a game…they are even to retarded to watch map…and to decap base…i lost xxx games due cap…coz all those noobs rush and base gets caped, and im to slow to get back.

    Only good idea is to balance stock tanks in game…that one team cant have 5 stock toptiers and other one fully upgraded.

  23. I love to see that anti-skill-MM-rage which obviously comes from the deserved members of Red army.

    Play what you want and let us play what we want, because yes, we dont want to play WITH YOU.

  24. Pingback: Of Basic World of Tanks Principles | For the Record

  25. I don’t understand what the devs are discussing here… They lost the plot it seems. People do not want 15 tomatoes vs 15 tomatoes and 15 unicums vs 15 unicums. They want BALANCE. They want the 15-5 roflstomp games minimized. So if you platoon with your unicum buddies then the other team gets a high skill platoon as well, and so on and so on. That is what skill-based matchmaking should be, and I would welcome it a great deal.

    Give us some form of team balance… anything really… stop these 15 reds vs 15 greens matches and you stop 50% of the frustration WoT often generates.

  26. At least I’m not the only one that thinks this way.

    And if they cannot find a way how to do it so the kiddies can play their “skilled” matchmaking, just add it as a separate mode togheter with separated stats.Problem solved

  27. I stopped playing WoT about half a year ago because there is no matchmaking that is aware of skill. I intentionally don’t call it ‘skill based’ because it implies something that I don’t think the game needs. It’s not necessary to limit enemies to a certain skill range (what games like Counter Strike, League of Legends and DotA 2 have), but we need to have teams that have comparable skill. The difference is quite massive.

    I am/was a slightly above average player with 54% win rate. I don’t mind playing with or against 60%ers and don’t mind playing with or against 48% players. The point is that I’d like to be in games with two comparable teams. Don’t have the 55+ percenters in one team and all the <50% players in the other. This approach also works with any platoons, if great players team up with bad players, just have comparable players in same-tier tanks in the other team. Make sure both teams have comparable strengths in the tier spread, i.e. don't have comparable teams, but put the good players in the top tier tanks in one team and in bottom tier tanks in the other.

    What stopped me playing the game was that the majority of games were incredibly one-sided (as in the winning side had 8-10 tanks left at the end). THAT has to be fixed. The matchmaker just has to create comparable teams, it's fine if one team stomps the other every once in a while, but it's not OK if that starts being the norm. I don't mind losing often as long as it's a close fight (that's why I play), but I do mind constantly losing or winning by a huge margin, that's not fun for anyone…

  28. Skill should have no effect on the M.M, One, all of your quote on quote “baddies”will just become Unica by stomping over the super baddies, which henceforth puts us in the same position before skill mm. Two, there is no room for these bad players to grow and gain actual combat experience where they have to use tactics, good players become good by having to adapt to other good players, not by just rolling over others. Overall this is a very, very, bad system and would be severely flawed upon release.

  29. skilled based as ur trying to understand it, isn’t the correct way to do it. all they need to do, is balances the 30 players selected for that match, by PR.. and u will have more balanced team IE:skill mm

  30. Yea, look how great the MoE formula worked out. I don’t want the guys that thought out the MoE math to mess with the teams and skill MM.

  31. after reading all these comments one thing is for sure, current MM is broken and it’s broken badly! It can be manipulated quite easily if you have no life and simply want to dig for stats but at the same time it creates the idea that you constantly flip a coin and hope for the best and no matter what you will do in that game it still won’t matter enough to change the outcome because the outcome was already decided at the very moment you pressed the random button.
    if not skill MM at least same number of games MM would be decent enough! Leagues are left aside of this discussion and only WG themselves know why they are doing this. it’s as if they want to have some sort of control over how and what kind of games you can have which makes it feel like you have no freedom and 0 importance in the team at all unless you’re top tier and platooned.

    • another interesting thing would be (for hardcore gamers ofc) to have way more training options and to be able to play against bots for training! so battles that test your skill against bots just like LOL has. that way even noobs will be able to finally learn something and at least after 1000 battles learn weak spots and how to shoot/kill . current training mode is also a very big fail.

  32. Just base MM on something like average base xp and introduce high and low skill caps, so highly skilled and unskilled players have something to match them to.

    I don’t think it would cause problems such as clans splitting etc. It’s each player’s choice where he will play.

    There is also some players who can’t be bothered to play randoms much because they are too easy and are searching for other things, If they like randomness and bad team mates they can always switch back to non skill based random battles. There would be randomness in positioning, timing, line-ups, player performance in skill based games, you might just have more elements of teamplay.

  33. One thing that many topics of skill based matchmaking skip over is getting rid of the landslide/steamroll battles and the battle with predetermined outcomes due to very large skill differences.

    In an analysis of over 2,000 solo pub battles, this is what i found:
    Players lose less than 3% of the battles with a greater than 60% XVM chance to win.
    Players win less than 3% of the battles with a less than 40% XVM chance to win.
    the battles outside of the 40-60% chance to win range account for about 30% of the average player’s battles.
    So basically 30% of the average player’s battles have a predetermined result due to the difference in skill between the teams.

    If we could get rid of the battles outside of the XVM 40-60% chance to win range the game play would be more dependent on layer skill and teamwork. Yet it wouldn’t force everyone to have the same win rate.

  34. One way to give the “give us skill based MM” people sortof what they want would be to create a Tier X league that you can turn on and off. Turn it on and when you play a tier X tank you are sent to games with either +50% winrate players (50% overall or on the specific Tier X tank can be debated) if you are above it yourself and if you’re not you get to play with the 50% winrate or less players. It’s like an extra dimension of the end game play.

    That way randoms and Tier 1-9 are still free for all and are just that “Random” and there’s a higher area of gameplay for those that qualify for it. If you enable it by tank % then it could also give the benefit of not excluding every player that has below 50% winrate overall.

    I’m sure this idea is full of holes but feel free to point them out to me,

  35. There are 2 kind of MM problems …
    1. Two min battles with results 15:2 or 2:15. This problem is caused by WoT action for example the Medium and Light tank this month.
    2. This problem came up from the type of tanks in team. Last month I play very often the games where my team has 5 medium tanks in the top and the oponent team has 3 Heavies and 2 TD. This is the worsed scenario on the huge amonut of maps. You have maybe the speed but they heve fire power and armor and if oponents heavies are not fully tomatoes and go in one group on right place it is the lost game in 90% of cases.

    Conclusion > If my team heve 5 mediums on the top enemy team shoud have 5 mediums on the top too. Its fine that the weight of tanks on both side are same but the setup is totaly wrong …

    • I’ve not even played for a year… yet for those 2½ I did… Playing a medium even when it was T9 tops didn’t mean bad… mostly mediums stomp heavies and TDs, but it was back then more about having that arty to take care of the said heavy opposition, but still a smart heavy was smart just as bad medium was bad.

      Perception of mirroring teams is just bad, but having same amount of OP tanks would be nice… Oh, but according to WG there is none…