OGS Ranking

I prefer to play against stronger opponents since I find that I learn more from such games but I have realized that that has a bad impact on my OGS ranking. What percentage should one try to have with playing stronger and weaker opponents to get a realistic ranking on OGS?



So I should start looking for weaker opponents… :thinking:

I think playing too strong opponents actually less helpful than playing equal opponents
And you can’t actually test your understanding without actually winning at least sometimes.


Just create open challenges or auto matches and play whoever accepts. You should get a reasonable mix of players that are higher or lower rank than you.


Maybe play even games ranked and stronger opponents unranked?


There are many reasons why it’s good to play weaker opponents along the way to getting your rank as quickly as possible to as accurate as possible.

Not the least of which is, if everyone only played stronger opponents, then no-one could get a game, right?


But is it about perspective? I find I generally lose to stronger opponents and my ranking seems unaffected.
What if it’s that ones ranking was somewhat of an overestimate and having played stronger opponents that ranking is now more “realistic”? It seems it’s just a matter of how you look at it.

What I mean is that when you say

I assume you mean your ranking went down but if the impact was to make your ranking more accurate then that is a good impact.


… but if you continually play only stronger players, your ranking will asymptotically approach the “lowest estimate”.

Similarly, if you only play weaker players, it will asymptotically approach the “highest estimate”.

That’s why playing half up half down is a good idea.


… and it’s why I put the pie graph next to the ranking graph so you can see at a glance where you stand in playing up vs down :slight_smile:


So really the answer is to play more games!

I note that if you are on mobile you never normally see the pie chart!


Turn it sideways :wink:

1 Like

Yeah, same. Still no pie chart. History graph just gets wider
Edit: chrome on Android if that helps!


I mean you can go into “desktop mode” but I don’t usually think to do that.


Weird. Must be a display ratio thing? Works for me…

Is that really your phone @Bhydden?

TBH I remember when I was doing this I couldn’t find a decent way to make it lay out properly for mobile :smiley:


The goal of getting a higher rank can conflict with the goal to be a better player.To progress at times you need to change ways of playing and that is going out of confort a while and lose more.
Playing weaker player is something which can be fruitful for strength. The fact that you have to explain the reason of your moves and the ones of your opponent is still an interesting process.

I think it’s good to play everyone in fact. It’s good to meet a lot of different ways to play go and be ready to deal with.

Of course playing stronger can bring some clue on what is wrong in your go if you care enough to review the game.
But it’s a bit egoistic to play only them and playing weaker can help progress too.
Last playing fair games is surely an important part in the study and fun of the game.

Myself i like to be an interesting opponent for whatever level i am going to play with. If not, i feel to have a fresh rank not really well secured.


You can get some insights more easily when your play actually works, rather than is countered, as was said, and when you’re not having your head bashed against a wall :smiley:

Yeah of course. Pixel 6. Look at the top it’s got time and battery and reception and all that jazz… and AR is way off to be a tablet…

(Full disclosure, I set my display and font and all that stuff to “smallest” on all my devices)

Well there ya go, you barbarians playing Go on a phone should all get Pixel 6s :smiley:


I thought there was no difference between playing stronger players, weaker ones or even ones due to the server using glicko.

How is it a disadvantage to play stronger players? It is harder to win but I’d be risking fewer glicko points and earn a lot if I win. Shouldn’t it even out in the end? Isn’t that the point of elo-glicko, to be able to match players of different strengths and still get legit conclusions from the match-up?

I would even say that playing stronger opponents should be better to get your rank up, because it is the same risk/gain ratio for all opponents but there is the extra factor of games not being finished properly which works better for you when you play stronger opponents. I mean, against stronger players, if your opponent abandons the game or times out, you gain substantial rating and if you time out, you lose little. Also, it is much likelier to find a sandbagger in a lower ranked opponent and risk a lot in a game where youre actually the underdog. So, when playing stronger players, at least these details work to your advantage.

Could you elaborate on this? What are these estimates, actual technical terms or are you using them to just notate high and low ends?

Am I wrong in thinking, that in theory, with glicko, my rank should be the same whether I play dramatically stronger opponents or dramatically weaker ones?