We look for all the obvious things (cr, corr, corres) as well as organizations (ICCF, IECC, Golden Knights) and a few other miscellaneous keywords like "postal" or "email". Of course that doesn't mean we found them all. Some abbreviations demand extreme care, e.g. "CC" can mean "correspondence chess" but it can also mean "computer chess".

For the time being, if you can find a correspondence game not labeled as such (or one labeled as such, which isn't) we'd very much like to learn about it. Feel free to come back here and post your findings.

Other notes:

• For purposes of "EZ Search", the statistics will lump correspondence and classical games together, rather than putting correspondence games into the "rapid, exhibition, etc." group.

• We now label the Chessgames Challenge games as "EXHIBITION"

ID#s 1322135--1324004

So looking at the events for these games, I find cryptic codes that I don't understand, like these:

EM/MN/008
EM/M/GT/A011
EM/M/GT/A003
EM/MN/008

Even without understanding the, it's pretty obvious that "EM/" at the start of the event name means email, so we reprogrammed our software to look out for that. It fixed some 1,885 games.

<gids 1508861--1508934>

They're from a tournament of the CCCA (Canadian Correspondence Chess Association).

I've noticed a few scattered ones with another problem, for example T Wibe vs G Timmerman, 1991. The event field reads <NED jub25 corr9193>, so apparently your searches aren't picking up "corr" as part of a "word". Might have to be careful though, since there could be an event like the "Corregidor Open".

EM/MN/008
EM/M/GT/A011
EM/M/GT/A003
EM/MN/008 >

Those are ICCF codes. EM=email, MN=master norm, M=master class, etc.

<NED jub25 corr9193>

• CCCA is a keyword for correspondence

• BPCF is a keyword for correspondence

• "corr" followed by any number of digits is a correspondence game

That is picking up a ton of new correspondence games, from tournaments like these:

BPCF jub
CCCA-60 (-1985)

as well as scores of tourneys in the "corr" + # format:

URS-ch07 corr6566
olm5 fin B2 corr6568
WK2 Latvian gam corr9799
W-ch4 corr6265
Shakmatnoe Obozrenie 7th corr0304

(Who knew that there would be so many in that format?)

"JUB" as a keyword by itself is very reckless, just look at a random sampling of that keyword:

50 Years BdF Jubilee Invitational
Jubilee Open Tourney
25 years jubilee
JUBS
Ljubljana YUG ch
Novi Sad olm JUB-USA
Stockholm jub 6667
Jubilaeumsturnier Wiener Schachklub

If I can't tell by looking at it whether or not its correspondence I don't expect computer software to do any better. If some of the tournaments above are truly correspondence, some renaming may be in order. At the end of this process we'll have some stragglers that we'll just rename to fit our conventions.

Here's one: Willi Dürig

Great work by all in house and cg members.

My question is...

How does "chessbase" sepeerate and define all these subsets?

There may clues in their way to help here.

When we merge two player records, all the games are combined but the kibitzing on one is lost. I can move kibitzes "by hand" if necessary before merging, though.

When we merge two *games*, all kibitzing is combined, entries in game collections are combined, and the alternate score is preserved (there's a small link under the game saying "View alternate score" which links to a page highlighting the differences). So it's not necessary to look at kibitzing or game collections to see which one of a pair of duplicates to keep. However, as the database rejects duplicates when new games are submitted, there should rarely be exact duplicates (it goes by moves only, regardless of players). Exact duplicates can happen if one score was submitted with incorrect moves which were then corrected, or in situations like those discussed recently at User: chessgames.com (there are sometimes two ways to write a move when a piece is pinned, and Chess Viewer Deluxe even displays them identically, though the PGN is actually different). But usually there is a difference, often an inconsequential transposition of moves. So if you want to help me out with deciding which copy to keep, look carefully for differences in the score.

Santasiere vs Kashdan, 1938

The event is listed as "US ch fin-A." This is the wrong event for this game. The crosstable from the US championship (Chess Review, 1938, p 112) shows Kashdan <won> against Santasiere, not <lost>.

Santasiere vs Kashdan, 1938

This is the correct game. The game score was published in the New York Times, 3 Apr 1938. While Santasiere <did> have to qualify in a preliminary, Kashdan was directly seeded into the final, so they only played this one game in the championship.

The original event misclassification looks like it came from 365chess. http://www.chesslab.com/PositionSea... lists the game as from the American Chess Federation Congress in Boston, 1938. Chess Review's report on the ACF Congress wrote Kashdan "only lost one game [against Santasiere], and that one took four sittings and 127 moves!" (1938, p 185)

Anyhow, today's trivia problem: some games - eg, O Bernstein vs Capablanca, 1914 - are Exhibition games played (usually) under tournament conditions (see the kibitzing on the Capa game).

"Exhibition" is one of a group of terms describing the general conditions under which a game was played. Analogous terms would be "Tournament", "Match", "Team Tournament", "Blindfold", "Simultaneous". An exhibition is a chance for a master to demonstrate skill without the distractions of a competitive environment.

"Classical" relates to the time control used. Analogous terms are "Rapid", "Blitz", "Correspondence".

pgn generally has its own input/output standard protocol, and is there anything analogous yet for creating a ranking/round-by-round cross-table output as text reports (Canadian fed uses 'swiss-sys', and there are many other popular pairing programs)? Sometimes one wants to include the results of games without pgn in certain x-tables, and could do that with a 0-move game and result report. Also, when a result is reported one way in the header but different in the textual end, where does the game go to, and what takes precedence for the viewer(s)?

World Junior Championship (2010)/Raja Panjwani has 11 of 13 games linked to the page, and another was uploaded later when it was found (same event/locale info were submitted there, but it didn't batch in with the World Junior Championship (2010) or World Junior Championship (2010) TIDs).

Can an uploader submit a pid # in place of trying to repeatedly spell out his name (which format? it used to be last-name, 1st-name; now is 1st-name mid-initials/name, lastname) when submitting the black/white fields, and are they also tracking extra tags like [Rounds "# of total rounds"] or team names for FIDE team-type pairing events (or does it slow the upload process to allow extra tags - it might be good to update some instructions on the pgn-upload FAQ pages for examples of good & bad pgn issues again)?

For the bio-writer admins out there, is it good or bad to duplicate the same bio on duplicated pid pages with separate pid numbers (or keep a birthday or country origin/rep of one blank and leave the other as a primary bio-summary)?

Most definitely "EXHIBITION" is the correct designation.

Each player in events such as Chess Olympiad (Women) (2010) (keywords: women, team, etc) would indicate that: (i) they can be catalogued by country teams, (ii) their profile's gender can be tagged. Perhaps a script for (ii)'s purpose can switch a bunch of biographies at a time.

In addition to batching games by tournament in their own viewer, FIDE's site usually suggests country info, and when one also finds the tournaments played in the past 2-year cycles by juniors, the birthyear of a pid can also be evident.

When a correction/elaborate form is filled out, is it mandatory to click the 'other (specify)' box when one wants to add extra info, or does the form still expect an elaboration in the box (without checking other) if the only correction request is a "duplicate pid" or "spelling" adjustment?

For the topic of when to mention an annotator of a game with notes, it mainly seems useful to record sentences or punctuation (?, !, etc) associated to lines & sidelines, rather than mentioning processes like "draw claim by repetition, but unsuccessful", or offers of draw as part of the annotator's notes (gid=1581553 was an early attempt of adding a transpositional sideline that quickly went back into the main graph-line of the game; not much notes needed, can't change the move # the 2nd time through when viewing the main line (in contrast to gid=1132687 or gid=1285148, this way makes for ease of user navigation when viewing the game moves, along with doing a chessbase "merge games" call), and the person viewing probably doesn't care who the annotator was in that case of notes).

Meanwhile, a trivial merge issue which I've also posted in the Biog Bistro:

Bjorn Nielson (1 game only) was the same person as, and should be merged with, Bjorn Nielsen. The ‘aka’ link on the latter’s Biog can then be removed.