Handicaps

[Javaness2]

The EGD is not allowed to accept games played with more than 9 handicap stones for rating purposes.
On the other hand, tournaments can run with whatever komi they like, even systems such as auction komi are valid.

Is it allowed to do something like
10 difference units = 9 handicap stones + 6 points to black
11 difference units = 9 handicap stones + 12 points to black (etc)
and then include these games for rating purposes?

I just though I would hide this information here: HA[ ] : the standard handicap reduction adopted, prefixed by "h"; so "h2" means "handicap dropped by 2 stones", "h0" means "full handicap", "h9" always means "no handicap".
The system will implicitly evaluate colors and handicap depending on the rank difference, for all the games for which neither color nor handicap have been explicitly specified.
The evaluated handicap will never exceed (9 - the specified reduction), so for instance in a "h2" tournament the implicitly calculated handicap can't exceed 7.
N.B.: when the specification of handicap is at level of game (see below), it overrides the standard specification

[Matti]

The EGD is not allowed to accept games played with more than 9 handicap stones for rating purposes.

Has there been a decision about this and by which instance? We had many years ago a tournaent by Ing rules, where the handicas were unlimited and there were some of games with 20 stones or so. Ales Cieply refused to include the tournament to ratings even there was no rule about it.

[Javaness2]

Actually, I couldn't find any written decision about this point. I think it is an unwritten rule.

[Matti]

Actually, I couldn't find any written decision about this point. I think it is an unwritten rule.

I was a bit annoyed, when I find this out. I could have organised the tournament differently, if I would have anticipated it.

[BlueWhale]

Javaness2 said:

The EGD is not allowed to accept games played with more than 9 handicap stones for rating purposes.
On the other hand, tournaments can run with whatever komi they like, even systems such as auction komi are valid.

I do not think this is true. When you submit a tournament to EGD you need to specify a tag ;HA[hN]. N can range from 0 to 9. The system will then deduce a handicap for the game based on the grade-difference between the players. However if N=9 and if you specify explicit handicaps in the games then your handicaps are taken as is. So you could enter a handicap of as many stones as you like. See for example http://www.europeangodatabase.eu/EGD/To ... y=T191117E. In the UK we regularly use a combination of 9 stones and komi with 15 points per handicap stone above 9. The system does not record the komi.

I think a handicap rule based on grade-difference is wrong. In all McMahon rules I have ever seen, the handicap is based on the current difference in McMahon score (with some possible reduction). This handicap could be a lot different to the one based on grade-difference, and is more in keeping with the McMahon spirit of trying to give all players fair games.

[gennan]

The EGD accepts handicaps over 9 stones (see for example this tournament, but such games are ignored in the rating updating process.

I don't think the EGD has mechanisms to processes komi handicap (AFAICT komi is assumed the same for all games in a tournament except for komi = 0.5 in games with stones handicap).

[Javaness2]

Yes, so I suppose I could rephrase my initial question. Is it correct to place some limit on what komi can be for rated games?

[gennan]

I'm not sure what you mean. Are you proposing that the EGD should support arbitrary komi per game?

[Javaness2]

I think that I am proposing the opposite to that, some sort of valid range for komi.

[gennan]

The EGD supports one value for komi per tournament, which is usually 6.5 in recent years

Komi is one of the fields in the tournament info header (KM), see for example:

Code: Select all

; TC[T200223C]
; CL[C]
; EV[Steenen Kruis toernooi 2020]
; PC[NL, Deventer]
; DT[2020-02-23,2020-02-23]
; HA[h2]
; KM[6.5]
; TM[30]
; CM[ (submitted by gbouwman)]
 1 Yan Zihan                5d NL  xxxx  40   5    40   186  937  11+/b   18+/w   8+/w    4+/w    6+/b     |19498325
 2 Pomstra Koen             5d NL  Arnh  39   4    39   190  943  7+/w    3+/b    4-/b    5+/w    10+/w    |10249602
 3 Verhagen Rudi            4d NL  Ensc  39   4    39   189  944  5+/b    2-/w    10+/w   12+/b   4+/w     |10213610
...

But this komi info field is not actually used in rating calculations, so I assume you can put whatever you want in there without any consequence.

[Javaness2]

The komi for each individual game is not recorded anywhere, all you have is the komi for even games in the tournament.
For Auction Komi tournaments you will find ";KOMI: Black named bids for komi - jigo possible"
the system is not set up to detect something like -15 komi and 9 stones