50% winnrate bullcrap

Looking at streamers like Chu8 or Fan who intentionally rank up from the very bottom of the ladder to master, this slowdown for the very top players happens at around high platinum to low diamond.

They still do get a massive winrate at this point, but it drops from an outrageous number like 97% to 65%. Which means after this rank, luck will play a considerable role in the outcome of the match as even a single AFK might be too hard to overcome.

(a low diamond is not equal to Chu8 or Fan in skill, but they’re good enough at the game to deny them a win with decent soaking, objectives and not feeding if they have a 1 player advantage).

For an average player the cutoff point where they will start winning a lot less will obviously be lower. Most peoples’ skill cap is around silver to gold (there are MMR charts on Reddit). They are not ex. HGC players who got paid hundreds of thousands in tournaments.

Statistical cherry picking.

The algorithm for match making is built towards 50% winrate.

Not because of it being forced but because once you are at a 50% win rate you are playing at exactly your skill level with equally skilled peers

So if you are winning 100% of your games you climb
Then people start getting better as you climb
Now it’s only 75% WR
Still climbing
People still getting better
Until eventually you don’t climb any more.

It’s fairly straight forward but also not 100% perfect in Landing everyone at 50% WR because there’s too many unpredictable variants that it can’t take into account.

These systems have existed for a while in many games
And
People tend to blame losing streaks on things like this because it’s easier than understanding how it actually works

I agree that a top level player would not face the same problems. My post was rather long but I did say that within a given matchmaking bracket if you are better than the best opponents within that same matchmaking bracket you would not face the same issues. However, if you at the level with the best opponents (and yet better than say a median person within that same bracket) you would not rise, although it will feel as though you should.

I would also add that the problems really come in with the pyramid nature of the ranked system when combined with an mmr/elo system. So effectively you have more bad players than good players within almost any given matchmaking range. This is because the top end of every range shrinks in number of players.It also means as say someone falling within the wost 50% of your bracket but not the worst 10% you won’t fall at all. This does presupose that you are not often picked as the best player within a lower bracket which is harder to ascertain.

Here is an actual real life example (I can provide screenshots although I am wary about naming the person). This person has two alts they played regularly, though one has about double the number of games:

Alt1:
Platinum 5
Started: Preseason (when game started)
Games Played: 1.585
Win Rate: 49.7%
KDA Ratio: 3.2

In 2017 Season 3 he was was Bronze 2

Alt 2:
Silver 2
Started: 2017 Season 1
Games Played: 710
Win Rate: 49.2%
KDA Ratio: 3.3

In 2017 Season 2 he was Gold 5

So oddly, the same person playing with the same KDA and the same winrate is climbing on one account and falling on the other. Does this mean that climbing is difficult or that within say Silver to Platinum there is no difference? Perhaps it means the difference between the best and worst players in silver is greater than that between a median silver and a median plat player. Not really enough information to form a solid conclusion but enough to show that the ranked system as it stands either cannot recognise the skill levels or players or that a simple win/loss elo system is flawed, or both. Either way from the perspective of the player - he is stuck at the 49-50 mark, and there are flaws within the ranking system.

From my personal perspective I have no axe to grind - I have climbed from Bronze to Gold twice so far and have been pushed down during placements back to Bronze twice (another element which contributes to the elastic nature of ranks within hots) - I don’t play rank seriously though and it’s not a concern to me.

Why then is this an issue? It’s an issue because it’s the main reason that people that the ranked mode is, and has been for a long time, dying. When I started playing I joined a clan which was advertising for people to group together to play hots. That clan, along with many others, moved to other games.

What about the tens or hundreds of thousands of players whose maximum lifetime skillcap is silver or gold and they will never advance no matter how many games they play?

Or those who have a maximum skill cap of Bronze 2 due to their knowledge of the game mechanics and micro skill with individual heroes?

Your entire post implies people can keep climbing infinitely - something which isn’t possible for anyone but the very top players (grand masters).

The vast majority (+70%) of the player base have a skill cap somewhere between bronze and platinum.
If they perpetually stay in that league with 50,0% win rate, nothing is flawed. They are where they belong.

1 Like

That is a solid point - people aren’t going to rise indefinitely, although should they be able to rise within a decent time to play with and against people of their level? Also, how do you know someones level when people who are platinum and people who are silver are often the same person? The argument used to be you will rise or fall if you play enough games. Is 700 not enough games? Those figures were just for storm and heros league by the way.

Also, players should get incrementally better with every game until they reach a level beyond which they are limited by mental or physical attributes. This is true of every game and also every sport. At the levels you have suggested as a skill-cap players often don’t time camps, don’t lane properly, don’t rotate, don’t pick talents based on the map or the opponents, spam abilities, and so on. Why is it that players appear to improve less at this game than my experience of any sport I have played or any other online game I have played?

Is it the lack of feedback loop? If you play well and lose, or you play badly and win, you lose the connection between reward and correct play. If I log on and play TF2, even in pub games most players can use skills that take practice and timing. If I log on here I have to keep pinging the tank to stop chasing someone solo for 10 minutes while a camp is taking out our keep?

I have seen accounts with 10,000+ games who are still stuck in Bronze 4 or 5. One of them is harassing the EU forum (largely inactive now) almost daily but I will not name him as there is no point in doing so.

Not everyone is going to improve just by playing a lot, as in sports which you mentioned (most people will forever stay in the casual divisions, not play for money or even against semi-pros).

I know that player as i have seen him report people for abuse chat if they dont agree on his ideas so people like him can have a easy time climbing out of B5. When i checked his profile i saw he was lifetime Silver 5/Bronze 4-5 player.

Each time he logs in he start calling the devs names and degenerate stuff towards them. Sometimes he would even /w you and call you degenerate stuff.

1 Like

Yeah, that’s the one, br_____ot.

I rarely if ever block people but had to make an exception for that one when logging on EU.

Healsonheels on US general chat is nowhere as toxic as that person though they also spam nonsense daily. He/she mostly just posts their tumblr link and political ramblings.

Also saw the other player that defend his ideas on eu forum are B5 too. No idea if its his alt account but they both promote his ideas on eu forum.

:roll_eyes:
So typical.

What is typical here?

You’re attempting to debunk the OP for no reason.

i found the blizzard dev here

WELL-- I have a theory about that too, and it revolves around grouping vs solo play…

A friend of mine, after having been stuck in platinum for years, became a one-trick Cho main last season and is now on the verge of hitting Master. The Gall they play with had also never been past gold (or so I am lead to believe). What really blows my mind, when it comes to their progress, is that they have a 72% win rate (with several hundred games played ofc) on Cho, but their next highest rated hero is only 50%-- and that 50% was playing at a platinum level, vs where they are in diamond now.

Several seasons back, I had something similar happen to myself. I encountered a Malfurion who was gold at the time, and ended up playing with them for 6 hours straight, and in that time we won almost every single game we played… At the end of it, he thanked me for carrying him-- Which I thought was silly, but he came across as entirely sincere.

I would tie this into my previous post in that, the more people you can confidently group with, the more your win rate will increase due to the match making system being forced into a more limited role… For every slot you fill on your team, the match maker will have a harder and harder time “averaging” you against an opposing team.

1 Like

Being wrong is a reason to debunk misconceptions and conspiracies.

1 Like

I just want to point out that Chogall is op, so that helps at climbing.

So you entered the thread with implicit bias and immediately began attacking OP to find the chink in his armor because you’ve already concluded that he is wrong.

I don’t disagree there. LOL

But even if Cho’gall was at pre-nerf Zed levels of OP, it wouldn’t explain that kind of sudden progression on its own.

I guess if I enter a thread about “The Flat Earth” I also guilty of the same.
Being “biased” and considering the @Op wrong.

But… what is your scenario? Did you come here with the bias that he’s right? Or that I’m wrong? Or why did you addressed my comment from the many?
Do you want to find the chink on my armor?

The only place where 50% is forced is guessing the outcome of the matches, since that is indeed 50% thanks to the dualist/binary nature of how matches can turn out.

1 Like