Hello everyone, what follows is a part of discussion between the developers about the skill MM. It has been redacted by the Insider and by me of course, names and such. It’s here only to illustrate the way the developers are thinking. It’s also very long, so… make sure you have patience. Developer 1: All right, new preliminary results… once again, how do we…. Measure…… skill??? Developer 2: What players’ don’t get is that skill is a variable as well, let’s assume for a moment Mr. 63% wins gets into the ultra-unicorn league, well, now what? He goes and does his thing in the game, therefore, as all skills differ with different players, his particular skill is that so some of the players go around and push a certain area of the map and he takes his shots from a distance and scurries into a flank whenever a breach is open, but while then, “oh no, why is Mr. 63% camping? He must be a noob who purchased his account!” Skill is also part of RNG, because it’s impossible to measure how any player will behave in different scenarios, there’s no set rules of “oh you are a certain tank so you must behave exactly as you’re told to”, and the tank who breaks the rule and grabs their side by surprise is now a “hated noob because he killed me” If Players want to live by the knife’s end on harsh matches every day, all the time, most of them will be pleading for a rollback of the system, if they want difficult matches where coordination & cooperation is everything, there’s Clan Wars or Platoons. Developer 3: The current matchmaker encourages the use of strategy and learning abilities, we can’t help those, who lack either (and is way less than the average player thinks there are, about 7% of the estimated server population on the RU cluster has a below 47% while a staggering 77% has a ratio of acceptable terms between 48%-52%), but we can help them understand why driving on an open map with a heavy tank is a bad idea. There are, in reality many variables to determine skill, and none of them can absolutely measure a players capability (thus that’s why we (you *redacted*) laugh at the XVMmod predictions and models, its faulty for light tanks damage but not assists, faulty for personal abilities for those who exclusively platoon, or shoot premium ordnance, not to mention if its implemented, those with higher skill according to XVMmod will have the harshest times because they will always be priority targets for 15 enemies). Starting for the type of tank they drive, the equipment they currently have, elited or stock, or in-between, the ordnance, the gun characteristics and the variables on how other 14 players behave for him and how 15 enemies will behave against him, each with its own set of independent skills, different tanks and ordnance, some will shoot premium, some wont, some will even shoot HE, so: Are we to punish those who have an apparent “augmentation” of skill (SS: stat padding) because they decide to spend on premium shells (“oh, he is skilled because he kills tanks”) oblivious to the fact he used premium shells, or likewise because he thought the situation needed to shoot a different ordnance than a regular shell. And top of it, the +/-25% shell penetration RNG Developer 4: Ohh but we know what some want, they want “to face enemies without despairing matches, like KV-1 Vs KV-1, metal storm! well of course!, but they forget that there are 14 other players under them, and for those on the bottom -IT WILL ALWAYS BE UNFAIR-, sure no complains when the said player suddenly gets at the TOP of its list and just tells the bottom pit “deal with it!”. And repeat over and over I a circle of never-ending whine? This “equal pairing” is the equivalent of a +1/-1 Matchmaking and that is not going to happen, the primary tests on live servers several patches ago showed a staggering amounts of “fight your clone” matchmaking, and flushing the “unwanted artillery kids” into terrible matches with 7v7 artilleries and 3 or so unlucky tanks to fight for them. We are not going to let that happen again. Some of the tanks would blatantly destroy its own peers if a 1-1 MM was introduced, leading to more nerfs, more whine, having a 2+- MM ensures players will always get a challenge that’s designed for the tanks they drive. It’s not our fault if they expect a Comet bring down an IS-8 by shooting it at the front and then call it unfair because they can’t kill it, where a side shot (or 10) is enough to do the job, but yeah, “our fault that they can’t pull a miracle” Developer 5: Forums are unreliable to some degree, when do we have to wait for the poll turnabouts? Developer 6: Until the 14th of November, and when the poll out for the 200,000 interviews is over we have to wait for the results, the highest answer who matches a criteria for all polls will be the one who’s put into investigation Developer 3: Sigh… Why are forums SO unreliable? :) Developer 7: Trolls, goblins, hobgoblins and even orcs, however if we ask in a more personal way via a poll they tend to get more polite and express their true nature without all the “internet troll” drama, not to mention that if the poll shows THAT IF ONLY a minority actually wanting this change and our models show that the bigger larger player pop is fine with this, we can bury this line of thoughts for a while. Of course, some smart pants will think that unicorns will be at the top of their matching’s, and all the zombies will be rock bottom, not going to work like that :) Players with overly inflated wins, as example, there are several IS-6 players with 92% win ratio on these tanks, mostly because they spend big on the game, shooting APCR 99% of the time, will suffer if we pair them on their “equal peers”, and we’re going to take the blame for listening to people that “doesn’t know what they want but want it anyways” Developer 8: Oh there are some nice solutions that could work, the match threshold doesn’t have to be harsh at all, it’s still going to be a +2/-2 MM, and we can “equalize” player matching’s by having a personal rating threshold (+/-2200), of course, we would need to test it first, and there’s a lot of points to think of: Continue reading →

More...