Id rather just fix the algorithm and let the ladder adjust over time.
I've seen this mentioned before (probably by you), and as far as I can tell, this is the problem with the rating in Team Games, but most of the time I've seen this come up, noone mentions the cause. I don't know if just using the average Elo instead of highest would solve it, but I imagine it would be much better.
I think most of the people whining about how "broken" the system is here on this thread are simply 1v1 nerds who have no idea how to play team games, lose the game for their team and then proceed to blame their teammates (since they are the highest rated 1v1 players, obviously it's not their fault right??) for the loss.
Just learn to play team games and if you are good enough your rate will climb. And if you don't want to queue with randoms (I do think there is a massive issue with the solo queue, since you only have one ban...), then try to make friends to play with.
So I believe that anyone's TG elo will always monotonically increase and never stagnate, the question is just how fast it will do so. I don't know whether it will create a specific problem at some point of time per se, but would still be nice if they fix it. I imagine it would be a pretty easy fix.
Like I suggested before, it does exist but only taks a very small % of the player base. How many players have you seen that hit 2k8+ tg, with less than 50% win rate? Looking at the tg ladder, I don't think it's very common, and you will only probably run into them every 1 in 10-20 games anyway (instead of "most of the games", as OP complained about).it is true, that not playing 1v1 in a long time, but playing a lot of TGs might improve your skill. but in that case the winrate of players should rise. and a 40% winrate for players with a 2k7+ rating should not be possible. it only shows, that you gain too many points from winning, but lose too little when losing. because once you hit your "real" elo, your win rate should be very close to 50%. that is just how elo works.
There should not be 40% win rates at any Elo over 1000. So that there are 40% players over 2000 should tell you how broken the system is.Like I suggested before, it does exist but only taks a very small % of the player base. How many players have you seen that hit 2k8+ tg, with less than 50% win rate? Looking at the tg ladder, I don't think it's very common, and you will only probably run into them every 1 in 10-20 games anyway
I'm personally happy with a system that's flawed, but works for 95% of the players, and wouldn't call it "broken" just because you spotted a few outliers.There should not be 40% win rates at any Elo over 1000. So that there are 40% players over 2000 should tell you how broken the system is.
Like other people already mentioned the only place where people should have win rates significantly over or under 50% is at the very top and bottom of the ladder but like I mentioned 100 times already the system is broken and why you think a broken system works most of the time is beyond me.
This just shows that you don't know how numbers work...I'm personally happy with a system that's flawed, but works for 95% of the players, and wouldn't call it "broken" just because you spotted a few outliers.
I literally just scrolled through #500~#1000 players (around 2k7~2k8), and only saw like 15 players with < 45% win rate.Like other people already mentioned the only place where people should have win rates significantly over or under 50% is at the very top and bottom of the ladder but like I mentioned 100 times already the system is broken and why you think a broken system works most of the time is beyond me.
Learn statistics, I beg you.This just shows that you don't know how numbers work...
If there is a 5% failure rate that means that on average there is one 40% player in 2.5 4v4 games and one in 3.33 3v3 games. That is way to high to be acceptable, but in my experience the failure rate is way over 5%.
And how many above 1k? my point was there should be none above 1k....I literally just scrolled through #500~#1000 players (around 2k7~2k8), and only saw like 15 players with < 45% win rate.
Okay please tell me what I did wrong grand master of maths.Learn statistics, I beg you.
Thats not how MM / probability / statistics work. Heck there were also <50% WR TG players on Voobly.And how many above 1k? my point was there should be none above 1k....
What kind of distribution do the players follow?Okay please tell me what I did wrong grand master of maths.
Thats not how MM / probability / statistics work. Heck there were also <50% WR TG players on Voobly.
So what you are saying that on average there is one 40% players in 2.5 4v4 games amusing a 5% failure rate just like I mentioned. Yes it possible that there are more or less in a given game but the average stays the same.Thats not how MM / probability / statistics work. Heck there were also <50% WR TG players on Voobly.
What kind of distribution do the players follow?
How do you calculate the probability of having one single 40% WR guy in a team?
What about the probability of having another 40% on the other team?
What about 2500 (40% + 55% *3) vs 2500 (50%*4)? Is that fair or not fair elo-probability wise?
etc etc
Its not like 5% * 4 *2 * 2.5 = 100%
Yes and no.Voobly hardly has a place in this discussion since you can choose your opponents there. Automated matchmaking should have different standards imo.