New players and the rating system.

[RealityCheck]

1500 is not even an average rating. Most players have less than 1000 rating probably. That number 1500 was chosen years ago for some reasons that were relevant then.

[Katharsas]

1500 is not even an average rating. Most players have less than 1000 rating probably. That number 1500 was chosen years ago for some reasons that were relevant then.

As shown in the link in my first post the average rating was 851 in September 2017 if the stats were correctly made.

[Wesmania]

IIRC we had a talk about doing some weighted average for newbies that would only affect their matchmaking, just so instead of losing 20 games in a row, they can at least win 10 out of their next 50 games. I checked just now, and the logic for choosing a 'similarly rated player' is done client-side. So, I did a little patch that you can talk about:

https://github.com/FAForever/client/pull/1055

I want to hear devs' and players' opinions on this change. Java client devs can also chip in regarding equivalent change in Downlord's client. I suppose eventually we should move the logic for this to the server, but that's what we have right now.

If merged, this change will both go into the develop client and a stable 0.18.1 release.

[Katharsas]

It's a good change, but, you know, why not fix the obvious things first? If starting rating would be fixed, there would be less need for tricks like this.

[Wesmania]

We actually had a whole discussion during dev conference regarding that (I believe it's posted somewhere on youtube). Changing starting rating for new players won't change anything - eventually this starting rating will become the new "average" as they play games against players with established rating. That's one of counter-arguments brought up on the conference.

[PhilipJFry]

mazernoob is right
by lowering the rating (and deviation) that users start with on faf we'd just end up lowering the overall rating and the relative difference between newbs and established users would not change

[FtXCommando]

We actually had a whole discussion during dev conference regarding that (I believe it's posted somewhere on youtube). Changing starting rating for new players won't change anything - eventually this starting rating will become the new "average" as they play games against players with established rating. That's one of counter-arguments brought up on the conference.

I don't think the point is to change anything in regards to the current system. Why would having 0/0s in ladder facing 600 rated ladder players rather than 1400-1600 cause any sort of shift in average rating? Chances are these new players will still lose 50-60% of the time but that's still better than losing 95% of the time.

[Katharsas]

We actually had a whole discussion during dev conference regarding that (I believe it's posted somewhere on youtube). Changing starting rating for new players won't change anything - eventually this starting rating will become the new "average" as they play games against players with established rating. That's one of counter-arguments brought up on the conference.

That would ONLY be relevant if the current starting rating was actually the average rating!
WHICH IT IS CURRENTLY NOT! (in the dev discussion, you guys assume that it is. i have watched all of them)

Which is the whole problem, thats the whole thing that i am discussing since the second post in this thread! Thats how our True Skill system is SUPPOSED to be setup! There really should be nothing to discuss. The current starting rating needs to be the average rating, it is not, so it needs to be adjusted!

And yes, it is true, if we wait like 10 years (what you describe as "eventually"), until we have enough inflation accumulated to bring the average rating up to 1500, the problem has solved itself. Until then, all new players suffer. Why 10 years? Well, because in the last two years, our average rating has not yet become 1500. Far from it. So this "eventually" does not happen fast enough to correct this!

To make this clear:
I am not trying to adjust the starting rating AWAY from the average rating, im trying to adjust the starting rating TOWARDS the average rating faster than it would happen by itself!

Edit2:
Okay, i looked at the code. Looks like axle (the guy who implemented TrueSkill for FAF) simply set the average to 1500 (he could have set it to anything, because i guess at that time there was no rating system, right?) and then calculated the optimal TrueSkill parameters for that. So we need to change mu to whatever the current average is, then re-run his machine learning code to get the new optimal parameters, check if they make sense and if they are similar enough to the current parameters to not break stuff, and then make the changes to the parameters.
Then we need to do CTRL+F and search for "1500" in any server/api code and replace it with the new average, because it is, at least in the python server, hardcoded in a bazillion places (though most of them are tests). Doesn't sound too hard.

By the way, weren't people complaining that the rating fluctuates too much at some point? I imagine that things like that could happen if the actual rating average was different than the average that TrueSkill assumes, and that is currently the case.
Well, the current settings are:
(mu=1500, sigma=500, beta=240, tau=10, draw_probability=0.10)
We should recalculate axles stuff and see what we get with an updated mu. Or just follow the recommendations about how to set them. Beta and tau are mostly dependent on sigma anyways, and draw-propability we can simply calculate from games played. Or we can just leave them as they currently are, since they have been adjusted already by other people.

For anybody interested in what these parameters mean: http://trueskill.org/
beta – the distance which guarantees about 76% chance of winning. The recommended value is a half of sigma.
tau – the dynamic factor which restrains a fixation of rating. The recommended value is sigma per cent.

[Katharsas]

Double post.

[Apofenas]

I have win rate of 7% in ladder. I don't waste neither my time nor time of those new players who happened to go for a competitive game mode rather than playing gap. All you get from such games is a frustration for both players. But if 1500 player would just search for another opponent, the 0 may never return to ladder thinking it is hardly unbalanced game mode. If you know russian you can scroll through this https://vk.com/scfaf . That group is full of complains about ladder and "advices" to play crater instead.

After that we wonder why there's so many gaps and craters and so few ladder games?

One of main problem that makes it hard for new players is separated rating systems. When i joined FAF i got 1k from 1st 3 ladder games(got lucky to win 1500 Steinklotz with Jester rush). New players have hard times finding any game because they got no rating. And they got no rating because they can't find any game. Problem is they get kicked even from all welcome games unless it is some heavily modded unrated cancer. That is absurd how our community doesn't like new players.

800-900 rating is where most players and most games are. Making new players to count as those will make it much easier for them with opponent and let them get games faster due to amount of potential opponents.

Another problem that needs to be solved is separated rating for ladder. Complete bs. If a player farms single 1v1 map he gets overrated AF; if player plays 30+ different maps he doesn't get any global rating to "prevent him from being overrated in team games". A lot of sence here!

System needs to go back to single rating or at least make ladder affect global again. Although knowing how things are done in this community we'll more likely to sit with broken things for years and ignore them.