Re: New players at OGS and how strong they are

Hey, me here. Remember this one.

I decided to do the same but with new ranks. I also think it’s better to plot ranks upon reaching certain deviation level, namely 100 and 80 (because I don’t know which is better still). This gives us a vision of ranks once the system is certain in them.

Cons: only survivors are plotted.


This is super interesting! Thank you @PetAccount :grin:

Two things I notice:

  1. Relatively few beginners (20-30 kyu). Is this because we genuinely don’t have many beginners? In which case discussions of how to be more welcoming to beginners (e.g. Glicko-2 does not require accounts to start at 12 kyu. So why do we do that?) are maybe not as important as we thought. Or is this because:

and OGS is so unwelcoming to beginners that few survive? In which case, those discussions are even more important.

  1. There appears to be a steep drop from 2 kyu to 1 kyu and from 1 kyu to 1 dan. Why is this? :thinking: Maybe there is some great leap in how players think about their game that is necessary to break through these levels that makes it inherently more difficult than progressing at lower levels. Or maybe, so many SDKs striving for shodan get annoyed being stuck at 2-5 kyu that they get demotivated to put the effort in to keep improving. Or a bit of both. Or something else… :man_shrugging:

Anyway, cool plot. Thanks again :smiley:


Hi PetAccount, welcome on the forum!

Not what i see on the graph. Still quite a big group.
If you enlarge a bit to ddk players, then we even find the dominant part of users.
What worry me is a not so neglectible number of players under 30k. Are they losing systematically all their games?

As for the small and big drops if i am surprised it would be by how well synchronize they are with the ranking systems OGS was aiming to fit so that’s quite a success.

You here?


It’s harder to statistically justify using accounts that are not “stable” (have low deviation). I’m not sure if deviation is dependent on account age. I can think of some situations where they would be independent, so it makes more sense to not include them.

You’d have to look at a bunch of other thing din order to find a correlation between account age and rank deviation.

1 Like

Interesting indeed. I had a theory for a long time that around the dan boundary is where many people leave the game (except uninterested beginners or early quitters OC). After the “kyu illusions” get dispelled, not everybody like what remains.

I don’t think so. In my opinion from 3k up, the steps are much harder until becoming dan, and that’s where we lose some players (not all, some accomodate to be that level), not by the discovery of a new world but by being tired of trying to reach it. Solid 1 dan will quick go to 3d.

1 Like

Remember it’s a histogram of new accounts, not of all accounts. Most dan players either got an account on OGS long ago, or keep their habits elsewhere and will never open an account on OGS.


Coming back here as this sounds interesting but i have no idea what are these “kyu illusions” ? Do you have anything in mind?

I am not sure what it means, but I think people may quit for a variety of reasons:

  • Improving further requires too much work. Or no matter how much you study, you eventually reach a ceiling and don’t improve further.
  • Circumstances of life: job, marriage, children,…
  • Interest in a hobby naturally wanes after a few years anyway.
  • Found other interesting activities (chess…)