Considered a dan in OGS

[Maharani]

@yahel.or.3

I understand that Snow White is at a low level

All the dwarves are dans, though.

[Maharani]

@yahel.or.3

have a question, what level is considered an OGS dan? Like records or before 1920?

To answer your question, yes. On OGS, dan level is considered to be a Glicko2 rating of 1919 or greater.

[Counting_Zenist]

@yahel.or.3

Alpha Zero plays Smoltani against Maktshanim

No so sure what you are talking about, Alpha Zero never plays against a human in Go in an official setting. Unless this is some lost in translation problem.

[yahel.or.3]

That’s fine. I just said if I were to represent AlphaGo Zero, I would invite the 10 best players in the world, plus 5 Europeans and 5 Americans, a first game without an advantage and a second game according to what he values, how much he should give the professional in front of him.

[yahel.or.3]

Thanks for this information, didn’t know they flew there.

[Counting_Zenist]

Most open source AIs nowadays are based on the AlphaGo Zero paper (not Alpha Zero which is a framework which extend beyond Go games), so anyone can challenge them as they like in any format. Even close source Go AIs like FineArts and Golaxy has public interface that anyone can play against online.

BTW, Go AIs that challenged pros like Lee Sedol and Ke Jie back in 2016 and 2017, they were using different structure than current Go AIs (they are supervised learning model, the model played Ke Jie has a bit self-learning and modification compared to AlphaGo Lee). And most pros today agree that AlphaGo Lee from today’s perspective wasn’t that much of a leap compared to pros todays (hence why Lee Sedol was able to won one game against it)

[yahel.or.3]

I understand. I’m happy that in a year it will be 10 years since AlphaGo, which was a special event with professionals all over the world.

[jlt]

@pwsiegel

The reason is probably that stronger players tend to prefer other servers

Yes exactly, many EGF dan players I know don’t play on OGS, or play unranked, so it’s hard to compare EGF ranks with OGS ranks at dan level. Among people I know who play a lot on OGS:

• Player 1 is EGF 4d, and OGS 3d
• Player 2 is EGF 3d, and OGS 3d
• Player 3 is EGF 3d, and OGS 2d
• Player 4 is borderline EGF 1d-2d, and OGS 2d
• Player 5 is borderline EGF 1k-1d, and OGS 1k
• Player 6 is borderline EGF 1k-1d, and OGS 2k.

We would need more recent data, but I think that OGS dan is about the same as EGF dan.

[yahel.or.3]

Real give me an idea that can last a week send people’s games and mine too ask them guess the level of players in chess make crabs about it what do you say want to send 2 games of players tell them what they think then reveal to them their real level

[yahel.or.3]

I agree with you. But there are those who are not willing to reveal the true power there in a computer game. You must remember this. I can say about myself that I do not play aggressively in every game because of this my level is unstable and in my past my level would not stabilize at all only in 2018-2024 something released me. I saw this only in official tournaments where I played. It is true that I said I did not play in tournaments since the end of 2020 and in 2023 I decided without thinking about it to change my approach until 2025

[yahel.or.3]

@GreenAsJade

It’s great to have other people respond of course, but still I feel compelled to question this.

What do you mean by “Dan in the real world”? There is no single worldwide Go rating.

It is true that there is no one true dan level, but if we consider that a person gradually improves his understanding over a period of 2-4 tournaments in a whole year, we can say that he is a dan based on how many tournaments he played in that year.

[gennan]

@Counting_Zenist

If the rating formula is different between each association and with EGF, then wouldn’t the rating of each of them drift apart over time?

@Groin

Could be. I m no expert in rating system, I think @gennan has more expertise as me.

The thing is that some national federation like to have their own and some don’t care and use the EGF one.

@Counting_Zenist

It’s fascinating to me to see a Go system that is still in the process of evolving, like watching history unfold in real time.

In the decades before the EGF rating system existed, ranks were determined in different ways in different countries in this region of the world. But there was some rank consistency because stronger players from those countries played each other fairly regularly in European tournaments such as the European Go Congress (held since 1957, which is also the founding year of the EGF), and handicap was widely used to determine ranks of club players, and to track their progress. As far as I know, European clubs still use handicap to determine IRL club ranks relative to stronger club players with an established rank.

The top (amateur) rank was 7d, which was also considered to coincide with the lower bound of pro level. A player of that level would be expected to hold their own with a 3 stone handicap against a top pro (note that 3 stones was the proper handicap between a 1p and a 9p in the Edo period and I think also in the Oteai system). So by extension, a European 1d would be expected to hold their own against a top pro with 9 stones handicap.

I think Ronald Schlemper was the first to receive a European 7d rank (in 1985, awarded by the Dutch go association’s dan rank commission). Among his achievements are 3 times European Champion (1982, 1985, 1986), reaching top of his Insei class in the Tokyo Nihon-Kiin, reaching 3rd place in the 1988 WAGC and tying for 1st place in the 1991 WAGC.

The EGF rating system started collecting data in 1996, so at this point it’s almost 30 years old. In the beginning it was perhaps more seen as just a register of European players’ declared ranks and tournament game results, but over time the calculated EGF ratings are considered a fairly authoritative reflection of players’ strength. Tournaments increasingly use EGF ratings rather than declared ranks to generate pairings (at least for established tournament players).

Some EGF member countries have their own rating systems. I don’t know how closely they align with the EGF rating system, but I would assume they don’t deviate too much. I also don’t know for what they are used exactly. Perhaps to award national ranks and/or to generate pairings for domestic-only tournaments?

[jlt]

Yes, the FFG rank for instance is used for pairings in domestic tournaments (if a player from another EGF country joins the tournament, they have to register with their EGF rank). The FFG rating system has several advantages:

• automatic adjustment system in case a player performs in a tournament much better than his rank.
• 13x13 and 9x9 games can be rated.

Otherwise the FFG and EGF scales are pretty similar.

[GreenAsJade]

@yahel.or.3

It is true that there is no one true dan level, but if we consider that a person gradually improves his understanding over a period of 2-4 tournaments in a whole year, we can say that he is a dan based on how many tournaments he played in that year.

Earlier you concluded we have a language gap, and this may well be true. Maybe also a philosophical gap in how we view this question.

“we can say that he is a dan based on how many tournaments he played in that year.”

I don’t think you can have this discussion without saying who is “we”.

Sure you can say that a person is a Dan based on how many tournaments.

But you asked “when is someone considered an OGS DAN”.

Did you mean “What do each of us participants in the OGS Forum each think about this, based on our experience out there?” ?

Some people will only have OGS experience, so a Dan is a [d].

Others will play in Japan or China or Korea, and have very high impressions of what a Dan is.

I have come to think this is your question - seeking input about how OGS ranks are viewed from different perspectives,

That’s very different to “What does OGS consider to be a Dan”, which is how the original question reads (to me).

[Counting_Zenist]

@gennan

3 stones was the proper handicap between a 1p and a 9p in the Edo period and I think also in the Oteai system).

In the Edo period, it was 4 stones between 1d and 9d. After Nihonkiin was established in 1924, there were several proposals, such as 3 ranks per 1 stone and 4 ranks per 1 stone, and they eventually settled for 3 ranks per 1 stone, (hence 3 stones difference)

Although in practice, here in CJKT, the strength difference within 7d amateur is pretty wide (the strongest is pro level, even stronger than weaker pro, but weaker 7d, would be like maybe 2 stones weaker, and it differs from country to country, region to region, or if they were ex-insei)

It is interesting that the EGF associations use a more top-down approach reminiscent of the traditional ranking by handicaps, while CJKT systems were sort of both bottom-up and have a “quality check” at the top when the strongest of them became pros. Although getting certified by pros (usually by playing handicap games) still existed, it was rarer in terms of amateurs (mostly only happened for special occasions, or for those who already knew pros, like families or friends with them). While the majority would have to mostly first join some Go classes and get some types of initial rankings from them and then join tournaments to “certify” these Go school ranks, and most tournaments don’t use handicaps at all, but using top % of winner in tournaments in each bracket to advance to the next rank (bracket). If you don’t go through the Go school systems, you have to play from the very bottom rank bracket tournaments and work your way up. This sort of guarantees players experience in tournaments as well as their blitz game ability (tournament games are super fast, like usually finishing in 1 hour and 4 to 5, or more rounds per day, if you ever wonder why foxwq players always like to play so fast, this is one reason). However, it was also fluid, depending on the “competitors”, and the rank-up percentage and regional strength differences. The gap within ranks didn’t get closer, but wider entering dan levels here.

[gennan]

@Counting_Zenist

here in CJKT, the strength difference within 7d amateur is pretty wide (the strongest is pro level, even stronger than weaker pro, but weaker 7d, would be like maybe 2 stones weaker, and it differs from country to country, region to region, or if they were ex-insei)

I don’t think 7d EGF has such a wide spread that a stronger one could consistently give 2 stones to a weaker one. There are a few CJK former insei (mostly from South Korea) active in Europe who are clearly stronger than 7d EGF, but they are recognised as 8d EGF.

But as you go down to kyu ranks, regional differences become larger. Like around 3k EGF you can find stronger ones from one country who can give 2 stones (or perhaps even 3) to the weaker ones from another country. France in particular is known for having strong 3k players. And around 1990 the Netherlands was known for having strong 1k players (until many of them were eventually promoted to dan).

I’d say that we observe the widest spread in Chinese 5d visitors in recent years. Some of them are as strong as 6d EGF while others are only as strong as 1d EGF.

[Counting_Zenist]

The relative strength of proportional promotion contributes a lot to the spread. 7d is sort of a gatekeeper rank, since it usually requires a national-level tournament, and likely the champion or the 2nd place, like 1/64 to 1/100, which is usually the upper limit of a day’s match for the 6d bracket. Think of 7d like the distribution of the strongest players in a province/major cities in CJKT, and they reflect the overall local player pool cap. However, for ranks below that, it becomes a sampling of the top players. And as the base grew larger, some of the regional tournaments would be unlucky (6d players are like 10 times more than 7d) to include “weaker demographic”, while some with stronger demographic (usually related to Go schools in each region). But they usually promote the same percentage of players in each tournament. And this chain goes all the way down.

[yahel.or.3]

@gennan

In the decades before the EGF rating system existed, ranks were determined in different ways in different countries in this region of the world. But there was some rank consistency because stronger players from those countries played each other fairly regularly in European tournaments such as the European Go Congress (held since 1957, which is also the founding year of the EGF), and handicap was widely used to determine ranks of club players, and to track their progress. As far as I know, European clubs still use handicap to determine IRL club ranks relative to stronger club players with an established rank.

The top (amateur) rank was 7d, which was also considered to coincide with the lower bound of pro level. A player of that level would be expected to hold their own with a 3 stone handicap against a top pro (note that 3 stones was the proper handicap between a 1p and a 9p in the Edo period and I think also in the Oteai system). So by extension, a European 1d would be expected to hold their own against a top pro with 9 stones handicap.

I think Ronald Schlemper was the first to receive a European 7d rank (in 1985, awarded by the Dutch go association’s dan rank commission). Among his achievements are 3 times European Champion (1982, 1985, 1986), reaching top of his Insei class in the Tokyo Nihon-Kiin, reaching 3rd place in the 1988 WAGC and tying for 1st place in the 1991 WAGC.

The EGF rating system started collecting data in 1996, so at this point it’s almost 30 years old. In the beginning it was perhaps more seen as just a register of European players’ declared ranks and tournament game results, but over time the calculated EGF ratings are considered a fairly authoritative reflection of players’ strength. Tournaments increasingly use EGF ratings rather than declared ranks to generate pairings (at least for established tournament players).

Some EGF member countries have their own rating systems. I don’t know how closely they align with the EGF rating system, but I would assume they don’t deviate too much. I also don’t know for what they are used exactly. Perhaps to award national ranks and/or to generate pairings for domestic-only tournaments?

I liked your answer. Readers, you remember quite well. Check out the history of this rating system.

[yahel.or.3]

@GreenAsJade

Earlier you concluded we have a language gap, and this may well be true. Maybe also a philosophical gap in how we view this question.

“we can say that he is a dan based on how many tournaments he played in that year.”

I don’t think you can have this discussion without saying who is “we”.

Sure you can say that a person is a Dan based on how many tournaments.

But you asked “when is someone considered an OGS DAN”.

Did you mean “What do each of us participants in the OGS Forum each think about this, based on our experience out there?” ?

Some people will only have OGS experience, so a Dan is a [d].

Others will play in Japan or China or Korea, and have very high impressions of what a Dan is.

I have come to think this is your question - seeking input about how OGS ranks are viewed from different perspectives,

That’s very different to “What does OGS consider to be a Dan”, which is how the original question reads (to me).

I also asked when the OGS rating was raised and I also put Dunn outside. Note that this discussion still evokes memories in people in their 70s-80s…Didn’t think this was a discussion on this topic.