Well, for starters, have you ever played with someone in Bronze 5? The common theme I see among them is that they have chat turned off, and all whispers disabled. They go about the match doing w/e they want with no regard for anyone, anything, or even their own well-being. While some of that might be chalked up to naivety or inexperience, many of them choose to simply not deal with the other players in game… As a result, they end up playing essentially a solo game (this is why Butcher is considered the hero of Bronze).
Then there are also some players who actively CHOOSE to lose. I had a game last season (plat 1) where another player picked Mediv, and spent the ENTIRE match in bird form taunting us, his own team… When the system threatened to kick him for being afk, he would attack a wall, or a well, and then immediately go back into bird form. Checking his profile a few days later, he had dropped to a solid 15% win rate. Sadly, this is not an isolated example ether.
Barring VERY new players who have few games played-- The fact is, if someone has a win rate as low as “32%” it is likely because at some point they made the active choice to LOSE.
I don’t think its cheating. More like abusing game mechanics. If someone is good on QM it is impossible he would suck in any comp. Also I think it is better than everyone starting from bronze.
I agree with the majority of your post, except for this part here… Of late, I have been looking at other player’s profiles after QM games, and the majority of them do not hold a rank-- Matter of fact, I can check back several years in some cases, and still never find one that even ATTEMPTED ranked.
When I do find a player that has played ranked, it is VERY rare that this player is, or has been, diamond, or even platinum.
How do you know it doesn’t work like that? They know the comps and outcomes of every match. They can work out a hero’s chance to win vs another hero the same way that hotslogs could.
talks with QA about matchmaker and reading posts that development team has made about matchmaker.
The matchmaker overall:
tries to find a tank and healer for both teams.
if a tank/healer cannot both be found, it will settle for one.
it has a range of MMR values that gradually expands that need to be close enough so that each team has as close as possible to a 50% chance of winning.
If a player has a hero with lower level than 5 (or maybe it is 10?) the system downgrades the players MMR slightly, as they’re (presumably) playing on a hero they’re not familiar with.
I’m sure that there are other rules also, but this is a simple overview.
But to get into counters… you would need to put a ton of work into the matchmaker, and in order to fit those rules, you would drive queue times up a ton. Which would drive players away.
The insane amount of work, combined with the public comments from developers, and private comments from QA lead me to stand firm that they don’t have counters coded into matchmaker for low level heroes. My own personal experience also!
You’re wrong! I have nearly 3k games played and my win rate is much lower than 50%. I’ve started to win games recently so I’ve been moving up but for the most part every game I played was a predetermined loss. For a while my win rate was at 46% and there was nothing I could do about it. If i played one of my mains the game would make sure to give me a comp that would hard counter that character. When I was playing with a premade of friends to have them learn the game… they got paired against a team as if their levels and accounts were disregarded and to be put at my level of play resulting in pure stomps. When I started an alt account so I could play with them so that wouldn’t be accounted for it still happened, even accounting for I did not have my full roster.
I don’t mean that they specifically added in constraints for counters. But, they could do something like take the average of your hero’s win rate vs all the enemy heroes. Then set a constraint that both team’s averages of those values are pretty close. I’m sure there’s a better way to compose those values than a simple average, but just as an example. It shouldn’t be too burdensome on MM times as there are many ways to satisfy such a constraint.
If they would like to ensure that a new player on a hero has a pleasant time so that they are more likely to buy heroes in the future, they might add another constraint to minimize that player’s expected contribution to his team’s chances of winning (i.e. get him carried). But how the matchmaker satisfies that constraint isn’t transparent. It might find it easiest to match the player vs counters and then satisfy the fairness constraint by giving teammates with average win chances vs the enemy comp.
You already get matched with players whose experience on that hero is on a roughly similar level as yours - so if you just bought a new hero, you will generally go against enemies who are very lowly experienced on the hero they queued with, or they just bought them as well.
If you queue on a level 50+ hero you can expect more difficult matches than on a hero you just bought and have 23% win rate with.
“Players should be matched with and against players of slightly lower skill when playing a Hero for the first time, and slightly higher skill when playing Heroes with which they have extensive experience.
This results in a higher percentage of evenly-matched games, in which each team has a 45% - 55% probability of victory, and should encourage more players to try out a greater variety of Heroes in Quick Match games.”
couldnt imagine I’ll fine the fine tuning argument here hahaha
in a perfectly balanced matchmaking, the win/lose chance is 50%, so the more you play the more you go toward 50% winrate
since its not perfectly perfect, you may just hover over it going up and down between 49% and 51%, sometimes you go on losing/winning strikes to get to your extreme threshold
it does not apply to people on the extremes because they cant climb/fall more in the ladder to find their peers
See, I dunno that this is true-- at all. Yes, I skimmed the link, but I am very doubtful of its accuracy, because my experience has been to the contrary.
I have a few heroes above level 100, but if you were to ask me: if the games I play on them are harder vs the ones I play as lower level heroes, I would say that the game seems to expect me to preform on par with how I would while at the top of my game, at ALL times-- which can be tough on a level 2 hero. lol
I suppose, as a result, this does make said link’s conclusion wrong (if nothing else); I don’t try many new heroes in QM.
You know, I admittedly do not actively check the level of enemy heroes 100% of the time when I play against them, but of late, I have been checking rank. As far as rank is concerned, there is no deviation from playing Chromie at level 122 vs playing Artanis at level 4 (now 10) where my opponent’s “skill” is concerned.
Looking at my stats atm, I have 5278 games played, and tbh, if you hadn’t linked that page, I would have never even guessed that matchmaking was supposed to take hero level into account.
I hope you against the “forced winrate” argument because if ppl can choose to lose, there is nothing forcing an outcome.
If you can choose to lose, you can choose to win. (Ofc, throwing is easier than carrying.)
Anyway, to the @Op:
What some ppl think:
I won a lot so the MM will force me to lose.
What really happens:
The MM gives you points after every match, because if you win and win and win but wouldn’t gain mmr, you’d just stomp ppl worse than you, which is not fair for everyone.
So if you win a lot, you get higher and higher because the MM wants to match you with ppl who have similar success as you.
And after a point, you reach a height, that you can’t keep and you lose, because you’re at a lvl higher than where you belong.
But the MM doesn’t want you to lose, it doesn’t force you.
It just tought you should be at a higher lvl cuz you won a lot, but you couldn’t live up to the challenge.
The same true for losing too.
It’s not the MM that holds you back, but your own skills.
See, the thing is, the Mediv in my example likely played the same way in every game he got into, but he STILL had a 15% WR… This means he got into a few games where despite him trying to lose, his team was simply SO MUCH better than the other team that it didn’t matter that the game was basically 4v6.
A game that is THAT stacked in one team’s favor is the perfect example of a “forced win.”
And i’m going to guess you also have a few heroes with very high winrate.
Personally, I’ve got a few heroes where i’m like 1-7, 0-9, etc.
But the thing is… your “bad/unlucky” games can be a combo of being on a hero.
if you have more than 20ish games and it stays there, and you play the hero good… at that point i start to worry/place yourself outside of the 3rd standard deviation.
For example, i’ve got a 35% winrate on tracer over 20 games as Ana. I think i’ve just been unlucky. I’m by no means good at her, but if bad luck with players due to matchmaking meaning i lost 2 games? If you flipped those, I would be at a 45% winrate, acceptable.
Meanwhile, I’ve got a 22.2% on Malth… I’m just bad at him. Although mostly this was prerework.
I’ve got a whopping 4-16 record on Joh!
Flip it around, i was 6-1 on Xul before his imba rework. I’m not that good at Xul.
I’ve got a 5-17 on Raynor also lol! That’s just me playing him poorly. Somehow.
