two points.

It is sad, that it's different from the European rule (that allows 2 stones).

But having Sandbaggers by design means, that people who do not like sandbagging, will not join the tournament.

When I promote Tournaments in France I always get complaints about the French rating and most players do not like sandbaggers.

As a tournament director I want players having good games and I accept ratingreset, if comprehensible (an online rating per se is not comprehensible).

Greetings from Germany

Wilhelm

To be more precise, the minimal promotion is

Dan - 4K : 2 levels
5k - 10k : 3 levels
11k - 30k : 4 levels.

You may think that 3 or 4 levels is too much, that's why I said that other countries could adopt a "similar" system. The main point is that you cannot self-promote, you can only apply and a committee decides.

What is the logic behind that restriction? On first reading your post, I thought that you must have meant maximum instead of minimum. I suppose the reason is mentioned somewhere https://ffg.jeudego.org/informations/officiel/cr.php but it might take a little while to find that.

_________________
North Lecale

I don't know, I didn't read all CRs.

Another thing is that there is an additional mechanism ("automatic readjustment"): if after a tournament your rating increases by a number R > 60 points then R-60 points are automatically added to your rating prior to the tournament and the algorithm is applied recursively. The process converges quickly. The consequence is that it is not uncommon for a DDK to gain 3-4 stones in a single tournament.

Because of that readjustment mechanism, applying for small promotions would give unnecessary work to the reevaluation committee.

On the other hand, the problem is the presence of sandbaggers in a tournament. For instance, player A registers at 12k but is 9k level. He wins all his 5 games, including against player B who is 10k. Player B feels sad after his game because he thinks he lost against someone who is 2 stones weaker, and needs to wait until the end of the tournament to know that player A's rating will be automatically reset at 9k before the tournament (hope this makes sense, this is not crystal-clear).

P.S. However that readjustment only applies to the FFG rating, not to the EGF rating, so if a European non-French 10k who participates in a French tournament loses against the 12k sandbagger, it will be counted as a loss against a 12k w.r.t. the EGF rating.

Thanks, that is a clear explanation, and that makes sense within the confines of the FFG.

_________________
North Lecale

Here's my two cents:

1.) The main purpose of a rating system isn't to motivate players but to provide evenly-matched opponents in tournaments. In fact, if a player grossly overestimates his strength, a proper rating system will (should?) have a demotivational effect. That's the fault of the player rather than the rating system. At this point, I think it's still too early to say whether this update is in line with this purpose, many more (in person) tournaments are needed.

2.) The fact that certain countries aren't implementing the rating system properly is also not a fault of the rating system but rather their national associations and the EGF for not providing/enforcing proper guidelines. Javaness has enumerated some of the key questions that need to be resolved, but I'm afraid that this will still leave us with a few issues:

a) National associations will require more people who are willing to take on the responsibility of managing other players' ranks (be it tournament organizers, rating committees, etc.). Finding enough dedicated and knowledgeable people, in particular in smaller associations, will be difficult.
b) From what I understand, a 5k, for example, can only double promote to 3k if he's never had a higher rank than 5k. I assume the reason for this rule is to prevent cheating but most players I know have taken a break from go at least once in their go career and were often several stones weaker when they returned. For this reason I have always felt that this restriction should be removed. Of course, this would make it ever more important that there's someone who's monitoring the players' ranks and strength.
c) I'm glad that double demotions will soon be possible but I wonder how many players will willingly take advantage of this option.

3.) We tend to conflate the terms 'rating' and 'rank'. To a certain extent they are the same if by 'rank' we mean the rank associated with a certain rating. However, there's also a player's declared rank at a particular tournament, which can differ significantly from their rating even if the player is not aiming for a double promotion. This is another matter where there's no clear guidelines. Fortunately, many tournaments these days use EGF ratings instead of declared ranks for pairings but it still rubs me the wrong way when I'm playing someone whose declared rank is 4 stones stronger than mine yet we're on par in terms of strength and rating.

In the Netherlands, there is a 3 person ratings commission, but it only regulates dan ranks. That commission has existed since long before the EGF rating system and they have promotion criteria independent of the rating system (using declared ranks instead of ratings). They keep records of tournament game results of Dutch players of 2k (declared) and stronger and they have criteria for dan promotions, which get tougher as the considered ranks gets higher.

For example: for promotion to 1d, you have to make a streak of about 20 consecutive tournament games where you score about 67% against 1k, about 50% against 1d and about 33% against 2d. Criteria for higher dan ranks are similar, but your streak needs to be longer (so you need to show more consistency for higher dan ranks). The commission also accounts for opponents who have clearly declined (so their rank has outstripped their level) by assigning less weight to wins against them. If dan players wish to demote themselves, they need to ask permission from the ratings commission (I did so in 2013, I requested a demotion from 4d to 3d and it was granted).

Kyu players are allowed to promote and demote themselves as they wish, but I think most of them just register with their club ranks, determined by their handicaps against stronger players with established ranks in their clubs. Most club players have quite an accurate notion of their level, even when they rarely (or never) play in tournaments. For players who only play online, it's more difficult to estimate their offline rank equivalent when they play their first IRL tournament. I suppose Dutch tournament organisations try to make a best guess from any established online rank that such players might have.

For kyu players who regularly play in tournaments, I think it's common practice to promote themselves when they win prizes in tournaments (for scoring 4 wins out of 5 games) in short succession. I think the reverse is also fairly common (demoting after scoring 1 win out of 5 a couple of times in short succession). I think it's also common that they get advice (or ask advice) from other players about decisions to promote or demote.

In recent years the Dutch ratings commission recognizes the EGF rating system more and more, allowing players without an official dan rank to register for a tournament with a dan rank, if their EGF rating clearly warrants that dan rank. An official promotion may follow soon after, if they keep doing well with that rank.

Another trend is that some Dutch tournaments started using the EGF ratings for the pairings, instead of declared ranks.

The EGD has a page where you can see how well declared ranks and EGF ratings are aligned in different countries. This is the alignment for Dutch ranks and ratings (after the system update):

Apart from rating resets for improving kyu players, there's the problem of what to do with mystery Chinese 5ds entering their first tournament. They can have a wide range in strength.

At the last London Open for example, Richard Chow entered as 5d and lost all but 1 (vs Toby 2137) of his 7 games https://www.europeangodatabase.eu/EGD/T ... n=19501075. He was retrospectively entered as 1d (2100). I think entering him as lower than 5d was correct, but 1d too low. This reset down irked Lukas Podpera 7d, whose strong 1d 2160 girlfriend Adriana was the weakest player he lost to: her good win against who she thought was a much stronger player ended up as a poor win rating wise. Based on watching his games, his shape knowledge was at least a British 3d I'd say, but he blundered a lot. Perhaps not used to the pressures of tournament play. I would have entered him at 2d or 3d.

Another new 5d at the London Open was Chi Chang. He won all 4 of his games (had to miss some days), and so his 5d initial entry stuck.

In his 2nd tournament was Zeyu Qiu 5d. He came 2nd, beating everyone (incl Lukas 7d and Lucas Neirynck strong 5d) except Daniel Hu the winner (who beat all except Lukas 7d). Lukas later expressed annoyance that he was only rated as 5d, and apparently is a famous 6d in China; Lukas could feel his strength in his and other games watched). Zeyu's first tournament was Wessex, and his first game was against me. I lost, but didn't feel overpowered and was leading until overtime mistakes. Zeyu also lost to another Chinese 5d in that tournament, who I beat (I had a good tournament, also beat Matthew Macfadyen), so until London I thought 5d was sensible for him, after London maybe 6d is better. Should Zeyu have reset up to 6d? Should Daniel (who is still 5d rating, but later beat Tanguy 1p and Pavol 2p at EGF grand prix)?

Another case is Kanno Hiroki. I know him from Biba, when in 2013 he was a Hong Kong 5d around my 4d EGF strength (I played and beat him in my first testing game there). He has got a lot stronger since, he played in the online US Go Congress and came 3rd. His first EGF tournament was the BGA online congress, where he was entered as 5d, and won all his games, strongest opponent being Zhang Xiang 6d from Singapore (Sadaharu one of top OGS players). So maybe a good 5d, or could change entry to 6d. Next he played in the online Paris tournament. He beat Kim Youngsam 8d (Youngsam blundered in yose, but was close before) and Artem 2p as well as 6d/5ds. So 5d (as Lukas said, annoyed with British entry grades from his first tournament, he thought 5d was official max though this isn't case as we had a mystery Chinese enter as 7d at London few years ago, and his joint 1st place finish with wins vs Haylee 4p justified it) is obviously wrong and 7d would be reasonable. Will he reset up? There wasn't really enough evidence from the BGA tournament, for 5d one win against 6d is possible, but wins vs Artem and Youngsam show much stronger than 5d.

I happened to be one of his opponents. I registered as 3d in that tournament, but the organisation changed it to 2d, because my rating had just dipped a bit below 2250 (old ratings) in the 2019 EGC. I wasn't doing well, even at this lower than usual rank, because in the first 3 games I played, I scored 1-1 against 2d and 0-1 against 3d.

Then I was paired against Richard Chow 5d in my 4th and last game (I skipped some rounds). At first I was a bit anxious that I was paired against a 5d after already having not so great results. But as the game went on, it became clear that my opponent was much weaker than a typical EGF 5d (and also weaker than EGF 3d, I would say). So after winning that game (quite easily), I was happy that at least I would not lose rating points from this tournament, because a win against a 5d would make up for my loss against a 2d. But to my disappointment, his rank was retroactively set to 1d after the tournament ended, so that win didn't compensate for my loss against 2d after all. I did feel that this action was a bit unfair, because it took away my chance of compensating for my loss against 2d and not lose rating points from this tournament.

Apart from having a conversation with the player about what their rank should translate to on the EGF scale, what can you do there?

It would be smart to mark on the wall lists that such players had a rank which was provisional, meaning that it was subject to an adjustment post tournament if it turned out to be OUT_OF_THIS_WORLD. That's a feature that no td software has today.

_________________
North Lecale

1) Ensure this conversation takes place. This is responsibility of tournament organiser.
2) Get help in above if player and TD don't understand each other. Geoff asks new players their KGS grade, Chinese players say what's KGS. At London I acted as go-between and asked Qiuchi Li if he played on Fox, he said yes, he's strong 9d, so I told Geoff 7d was reasonable. I knew to ask if just low 9d, as EGF 6d or maybe even 5d can be low 9d on Fox.
3) Where there is not enough information from the first tournament (e.g 5d? beats everyone but no 5d+ so infinite upper bound) then you need to pick some entry grade to submit results, but this should not be final and can be reset later, i.e. make Hiroki reset to 7d now we know after Paris he's not 5d.


Indeed, I mentally do this already myself being fairly familiar with the regular tournament players.

What about Japanese players? I think it's common knowledge that their (dan) ranks are weaker than European ranks but this usually isn't taken into account (correct me if I'm wrong). If the Japanese player is already in the database and it's a rating-based tournament this isn't much of a problem but if they're first-timers than it is an issue. I feel that organizers often aren't comfortable with downgrading declared ranks of foreign players, including players from another European country but even more so Asian players. There's always the fear that they'll be offended and wouldn't want to participate in future tournaments.

Do you think that it would be beneficial for the ranking system if there was a standardized way of promoting players in all countries? Either that or exclusively using ratings in official tournaments.

As far as dan ranks go, I think that a lot of countries require a formal promotion by a committee or the like but the requirements vary widely, not to mention that there’s countries with more precise requirements while others use more subjective methods. I believe that the fact that the Dutch association is starting to place more emphasis on the EGF ratings is part of a larger shift. Since a lot of players nowadays start playing go online it’s more natural for them to equate rating and rank. There’s also no special requirements for reaching dan level online, you simply rank up, the same way you did when you were a kyu. Most of these promotions of declared ranks by national associations are sticking around because they are part of a tradition. Before the ranking system was implemented they had a much larger role. Since dan ranks then were harder to achieve, there was more value placed on them and consequently dan ranks could even be seen as titles, i.e. something that was earned and can’t be taken away (as is the case with professional ranks).

My point is that while official promotions are a nice tradition, we should now start placing a greater emphasis on ratings. At least to the extent that declared ranks don’t interfere with the ranking system. (Special situations such as rating resets excluded.)

Schrody, my recollection from a few EGCs is that new Japanese players are entered at a EGF dan grade lower than theirs back home, as it is indeed well known their grades are inflated. Maybe new Japanese players at smaller less well organised tournaments aren't lowered like this.

Britain used to have a dan promotion committee like gennan describes in the Netherlands, but it's defunct now and we just go by ratings.

Declared ranks and ratings are different things in the EGF rating system. Once a player enters the system by participating in their 1st tournament, their declared rank and their calculated rating are completely decoupled. A single rank promotion (bumping the declared rank) doesn't have any effect on their EGF rating. Only in the case of a double promotion (or more), their rating is reset to their new declared rank. Using ratings instead of ranks for tournament pairings can help to match more evenly skilled players, but it doesn't fix the issue of underpromoted players, because they will not only be underranked, but also underrated.

Overall, I think that the EGF rating system is not that sensitive to differences in national restrictions on double promotions. Double promotions are probably quite rare for strong kyu and dan players all over Europe, so it mostly matters for lower ratings (EGF ratings below ~1500). If a country has very strict rules for double promotions of their kyu players, you can bet that their kyu players will be underranked/underrated compared to kyu players with similar ranks/ratings from countries with few restrictions. It can be annoying when such kyu players are paired in international events such as the EGC (it's not much fun to travel for hours to play mismatched tournament games, as @karlsgo said), but I don't think this has enough impact to compromise the rating system itself. The rating system held up pretty well with this for 25 years and kyu ranks are not that accurate anyway (a couple of 100 GoR spread in ratings of players of a specific declared kyu rank can be observed even within one country).

Considering how difficult it was to find agreement in a small commission on these relatively minor changes to the EGF ratings system, I think it will be extremely difficult (read: impossible) to unify the double promotion policies of all EGF member countries. I'm sure every EGF member country likes their own policy best, so they will feel that the other EGF member countries should adapt and use theirs, or they will simply refuse to participate in this policy unification process. EGF member countries don't even have to use the EGF ratings system and some even have their own rating system. In my view, the EGF rating system is just a service from the EGF to EGF members countries that don't have their own rating system.