PGN for tournaments

I. Tournament PGN tags:

Let’s start by my acknowledging that I haven’t really probed how <ChessBase>‘s software determines what games belong in a tournament. As <chessgames> noted, <SCID> is apparently much more strict. I like to err on the strict side myself, after all, all the games in a tournament should exactly share certain PGN headers; the games are from the exact same tournament after all.

Obviously some computer heuristic must be applied to group games in a PGN file into tournament sets. The natural minimal set of  PGN headers which can be considered for selecting common games in a tournament are Event / Site / EventDate.  Here is an example from <TWIC>’s PGN for the recent <New Zealand Open (2015)>:

TWIC example:

[Event "122nd ch-NZL Open 2015"]                                [Event "122nd ch-NZL Open 2015"]
[Site "Auckland NZL"]                                           [Site "Auckland NZL"]
[Date "2015.01.01"]                                             [EventDate "2015.01.01"]
[Round "1.56"]                                                  [EventType "swiss"]
[White "Aldridge, A"]
[Black "Chandraveer Singh Chauhan"]
[Result "1-0"]
[WhiteElo "1890"]
[ECO "E61"]
[EventDate "2015.01.01"]
[Opening "King's Indian defence, 3.Nc3"]
[WhiteFideId "4301528"]
[BlackFideId "5097401"]
[EventType "swiss"]

The PGN headers are show on the left, those in blue are the tournament-specific headers, the rest I’ve grayed out. I selected the tournament headers only to show more clearly on the right. You can see the content of information provided in the PGN headers by TWIC is excellent.

As a biographer, I’ll make the remark in passing that I would like the EventDate tag to denote not just the starting date of the tournament, but also the finishing date. Or, as I commonly refer to it, the bracket date. E.g. [EventDate “xxxx.xx.xx-yyyy.yy.yy”]. This additional information can be very useful, and it naturally fits in this tag.

I haven’t mentioned several supplemental tournament tags that that appear in the PGN standard, but might not be so familiar: EventSponsor, Section, Stage. My view, and de facto usage I believe,  are slightly different from the strictest reading of the standard. So I’ll quote from section 9.2:

9.2: Event related information

The following tags are used for providing additional information about the

9.2.1: Tag: EventDate
This uses a date value, similar to the Date tag field, that gives the starting
date of the Event.

9.2.2: Tag: EventSponsor
This uses a string value giving the name of the sponsor of the event.

9.2.3: Tag: Section
This uses a string; this is used for the playing section of a tournament (e.g.,
"Open" or "Reserve").

9.2.4: Tag: Stage
This uses a string; this is used for the stage of a multistage event (e.g.,
"Preliminary" or "Semifinal").

9.2.5: Tag: Board
This uses an integer; this identifies the board number in a team event and also
in a simultaneous exhibition.

As a biographer, I’ve felt the EventDate important enough to promote to mainline tournament tags. This makes sense if you consider an annual Event at the same given Site, if the PGN maintainers don’t wish to enumerate each and every version of the tournament in the Event tag (which is quite common). Generally I like to denote both the version and year in the Event tag for such tournaments, but it’s a matter of taste.

For example <CG> uses a terse naming convention (e.g. [Event “Tata Steel”] for years 2011->2015 and beyond).  This means that a date, presumably EventDate (or game dates within a window) must be used to group the tournaments into respective years.

Now, <TWIC> used a tag which I’ve also seen used by <FIDE> – EventType. Examples of this tag might be “swiss”, “round robin”, “double round robin”, “knockout”, “match”, “tourn”, or “team”. (I’m sure I missed some.) I consider this a good tag, informative and unintrusive.

On the other hand, Board doesn’t belong to the meta-Event tags. Instead, I view it as  more closely associated with game data. Generally, I find that this information is quite often stored in a compound round number by sites like <OlimpBase> (well, I know it uses compound round numbers in a form like this: [Round “”]).

I’ve never seen the two other tags mentioned here used, i.e. Section and Stage. In fact, I view them with suspicion, as they might even break some of the database software. Not having seen it, I really should conduct some “experiments” with SCID and ChessBase 2009 Light to actually observe the behavior. Maybe later I will, and then I”ll come back here and update the page.

Instead, let me note that the usual “de facto” method for handling both the different sections of a tournament, and different stages – is to embed this information in the Event tag. Then the database software has an easy time of it for separating the games. And people too, as the Event tag is always prominently displayed for all database software, but not these other tags.

Let’s look at another example of some actual PGN, this time from a game exported from <FIDE>:

FIDE Example:

[Event "BUL Championship boys12"]
[Site "Sunny Beach"]
[Date "2009.03.28"]
[Round "1.1"]
[White "Manev, Aleksandar"]
[Black "Daskalov, Dimitar"]
[Result "0-1"]
[ECO "B30"]
[BlackElo "1933"]
[PlyCount "64"]
[EventDate "2009.03.??"]
[EventType "swiss"]
[EventRounds "9"]
[EventCountry "BUL"]
[SourceDate "2008.07.28"]

There are two new meta-Event tags used here, EventRounds and EventCountry. The EventRounds tag is nice, even if slightly redundant. I believe some redundancy can be beneficial for error checking, and it means every game carries a little bit more unintrusive information.

But I think the EventCountry tag can be problematic. This information should be placed in the Site tag, its natural home. Why not [Site “Sunny Beach, BUL”] or somesuch? Normal database software prominently displays the Site tag, but not the EventCountry tag. I believe by removing information from the Site tag this is allowing unnecessary problems to arise in case of conflicts. Unlikely, yes, but this increases the risk with very little, if any, benefit.

The main adopter of this convention is FIDE. I can’t say for sure, but I have the general feeling that <ChessBase> and <TWIC> rarely use this tag. It would be OK if it were redundantly pulling the country from the Site tag, but that’s not how FIDE uses it. Oh well, they also delete and recycle FIDE ID’s for deceased players. Both of these conventions should that FIDE doesn’t fully understand stable database management, when designed to last decades and decades with as few problems as possible.

Let me end by noting that at least FIDE doesn’t use the Section tag, and instead denotes the fact that this game belongs in the boys12 section (12 and younger(?)) by embedding the information in the Event tag: [Event “BUL Championship boys12”]. A convention widely adopted, as already said.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s