Separate ratings between board sizes

[JohnnieDarko]

@shinuito

It could just be that you do a calculation to estimate the bots rank and freeze it, or take a middle value of sorts if it’s past rankings before making future games unranked. Though you still need a way to rank new bots.

I think my suggestion for bot ranks in general (in the other thread) could also apply to only new bots in your example.

TLDR: games vs bots exchange only a fraction of Glicko points compared to games vs people. Impact per person is small while bots play enough games to get a rank.

Static Ranks for Bots

How about having that low rate of rating change all the time for bot games? Let’s say 10% of a human vs human game. The results would be: Normal users around the bot rank cannot gain much from repeat games. The most enthousiastic people play 50 games in a row, just for reference. For the people in the 40-69k range (quite a lot), a first win against a 25k bot doesn’t insanely change their rank, but enough to stay motivated to play more and get into 25k territory. Bots stay useful, but …

[shinuito]

I guess this would also be a relevant comment for this thread

I am foxGo 5D but I cant win 2 OGS 1k player in a row

I do think a calibration is in order. I want to better understand the causes of the deflation before making any adjustments though. I have a couple of strong suspicions, but much analysis and simulation work is necessary before taking action to re-calibrate. This is in my plan for this year though fwiw.

If there’s recalibration, maybe part of the considerations can be the effectiveness of overall rank and individual ranks, and for the inflation/deflation considerations, the role of bots rankings etc

[anoek]

Yes I’m definitely going to be looking into splitting ranks as well during that process. I suspect we might be at the point where it makes sense now (we have a lot more games going on these days).

[garlicsoup]

I don’t have a strong opinion on how the server should handle the ratings issue as I can’t say it’s created any noticeable problems for me, but I will confirm that I can steadily improve my rating for 6 months playing 19x19 correspondence, then tank it in one night playing blitz 9x9 on tilt.

You could say I should try not to lose a 10 out of 12 consecutive 9x9 games but it’s not like I lost them on purpose, maybe I forgot how to play fast because I got used to correspondence speed.

[shamisen]

There is someone who purposely loses 9x9 games and then uses that to sandbag on 19x19 though:

I apologize if this is meaningless information. I don’t want anyone to waste time fighting sandbaggers needlessly – there will always be malicious actors in any system.

[shinuito]

You can report them on the main site and they’ll likely be banned (if they are sandbagging)

[Uberdude]

Did you check they actually purposefully lose?

I didn’t, I just only played a few 9x9s and aren’t as good at them:

[shamisen]

I’ve never reported anyone (maybe 1 time max when someone made a racist comment in my game), even those who cheated in my games (happened at least 20 times). While I do think mitigation helps, I mean that on a systemic level, not always worth individual effort. but next time it affects me, I’ll consider it, if it helps OGS.

[shamisen]

Can we report this guy too? joking

[jlt]

It’s not specific to 9x9. People may timeout or resign 19x19 games while being ahead on purpose too.

[qnpnpmqppnp]

@shamisen

I’ve never reported anyone (maybe 1 time max when someone made a racist comment in my game), even those who cheated in my games (happened at least 20 times).

Please do. Problematic players cannot be identified if they aren’t reported.