You're utterly unqualified in this matter, despite previous attempts to explain it to you.

You cannot do this within the constraints of the PGN standard.

If Alekhine was the book, I am the standard.

Putting the site of a game in the field marked 'Site' is not the radical idea you pretend it to be.

<A match of 14 games between Max Euwe and Paul Keres staged >

was staged

<in various cities in Holland (Amsterdam, Hilversum, The Hague and Rotterdam) >

I'd prefer the list after "cities", not "Holland".

<Prior to this match, Euwe and Keres had played five times since their first meeting in 1936. >

This implies they'd played 6 times. Maybe "Prior to this match, Euwe and Keres had played five times, with their first meeting in 1936."

< Euwe establishing an early two point lead>

established

<+5-1=0 in games 5 and 10.>

What's wrong with putting the actual location as a comment in the PGN?

Or having an auxiliary Location tag in the PGN?

Please, be explicit in explaining the pros and cons you might anticipate.

Oh, and where is your authority derived from?

I tend to rely on actual de facto practices of <ChessBase> and <SCID>.

(What about <NIC>?)

http://en.chessbase.com/post/tata-s...

Here's a link to a game from the round:

blob:http%3A//en.chessbase.com/aaec189b-1329-41e-
7-9828-20f3a78dec99

This round was actually not played in <Wijk aan Zee>:

<The 10th round of the Tata Steel Masters was played in the Dutch Railway Museum (Spoorwegmuseum) in Utrecht. The round was opened by Utrecht mayor Jan van Zanen. >

http://en.chessbase.com/Portals/4/f...

I'll save you a click, here's the solution <ChessBase> came up with:

<

[Event "78th Tata Steel GpA"]
[Site "Wijk aan Zee NED"]
[Date "2016.01.27"]
[Round "10"]
[White "Karjakin, Sergey"]
[Black "Adams, Mi"]
[Result "0-1"]
[WhiteElo "2769"]
[BlackElo "2744"]
[ECO "A45"]
[EventDate "2016.01.15"]
[PlyCount "62"]

>

I'm just listing the PGN header tags as an example.

So, one can list all the tournament games as being played at the main site.

Which is more acceptable to me than <MissS>'s approach (since it keeps the tournament normalized), but <CB>'s PGN is both inaccurate, and slightly misleading.

But maybe <MissS> knows better how to normalize a tournament than <ChessBase>?

chessgames.com chessforum (kibitz #24166) chessgames.com chessforum (kibitz #24170)

and <MissS>'s misconceptions...

chessgames.com chessforum (kibitz #24172)

and previous pronouncements:

chessgames.com chessforum (kibitz #24177)

and my follow-up immediately thereafter:

Although is seems that all the tags are related to the game itself, not the event, i.e. the date is the date the game was played, not the starting date of the event (this is why there is a EventDate tag). But one could argue either side of this...

My personal preference is that the SITE tag should be game related, not EVENT related. You can use the EVENT tag to add multiple cities, i.e.:

[Event "Holland Match (Amsterdam, Hilversum, The Hague and Rotterdam)"]

[Site "Amsterdam NED"]

[Date "1940.01.03"]

[EventDate "?"]

[Round "8"]
etc.

Try your proposed approach on a multi-city tournament, like <Tata Steel (2016)> and watch what happens.

I'll even prepare the PGN for you in the interest of doing "the experiment".

OK, suppose you have such a PGN.

What happens when you load it into <ChessBase>, or <SCID>, and attempt to generate a crosstable?

,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,

Again, I ask the same question from flip-side of the experiment.

Suppose you have a PGN prepared as I propose, using the inclusive Site tag, and a comment.

What is wrong with having the Site tag as "(various) NED" with a comment in the PGN indicating where the actual game was played?

Really, when you want to best communicate the locations of the games, you want this view:

<
R1-4, R6-9, R11-13 Wijk aan Zee

R5 Amsterdam

R10 Utrecht
>

Of course, one could imagine the tournament splitting into two groupings, each playing in different cities on the same day.

Just speaking hypothetically.

My approach is messy in this case, but not too much more than we've already discussed.

The other approach, putting the actual location in the Site tag, gets very messy in such a hypothetical case. I think the GameList window would be quite confusing in such a case.

As I've said many times, the two tags are connected if one wants to work with properly behaved tournaments, i.e. properly "normalized" tournaments.

In fact, the EventDate tag should also be included, but being so misunderstood, <SCID> and <ChessBase> aren't fully compliant.

There is a parallel between the dichotomy of the Date and EventDate tags that should have been mirrored in a Site and EventSite tag pair, but given a lack of foresight in 1994, wasn't.

If you don't understand the Date and EventDate tags, and want to gain more insight, I recommend reading Tim Harding's treatment here:

http://portablegamenotation.com/Lit...

I suggest scrolling down to <Can We Extend/Improve/Replace PGN?> about 3/4's of the way down.

I'm trying to educate people in a similar fashion about the proper use of the Site tag.

It's a more difficult task, because the need is rarer, and the standard lacks the EventSite tag.

http://www.chessgames.com/perl/nph-...

Wow, what a piece of dung the PGN tags are... WTH is Philadelfia? No mention of Montreal where the game was actually played??

This is a very unfortunate feature where the original standard is clearly lacking.

In principle, we could add extra tags that <CG> knows how to process for displaying individual games on a webpage.

Of course, one should try to avoid making design mistakes all over again.

* * * * *

One could even add some syntax in the comments for such purposes. I do this with stub games in Z-base.

There, I've added two extensions.

First, I've created a <Stub> tag for the PGN.

Second, I inject a <@stub - ...details..> comment into the pgn.

E.g.

<
...
[Stub "missing game"]

{@stub - missing game}
1.e4 e5 ...
>

or

<
...
[Stub "forfeit"]

Actually, I like the look of it all the more just typing it out.

The trouble, at the moment, is outside of <CG> such a tag basically is invisible inside of <SCID> and <ChessBase>.

Which is OK, since they don't strip it out. They just carry it along.

So, you found a good example of a tournament needing retro-fitting.

But some patience is needed, as people are slowly learning the need for respecting the PGN properly.

* * * * *

If I may ask, what software are you using for your PGN?

Are you collecting your games into a great big database - like Megabase or MillBase?

Do you use your database software to generate crosstables (xtabs)?

http://www.mark-weeks.com/chess/pgn...

R1-8 New York
R9-11 Philadelphia
R12-19 Montreal

Now, it's a match, and the xtab is I get from <SCID> when viewing the first game looks like this:

<
wcc
New York, 1894

Age Nat Score 12345678
,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,

1: Lasker Em 25 GER 5.0 / 8 1010==11 (+4 -2 =2)

2: Steinitz W 57 AUT 3.0 / 8 0101==00 (+2 -4 =2)

,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,

8 games: +5 =2 -1
>

I use ChessBase 6.01 still for game entry, then I modify the PGN manually before I add the games to my master database. For me, event driven sites are "worthless" so all of my Site tags are treated as GameSite tags and the Event tags contain all of the non-game data (per the example I gave before). I have never needed to generate historic crosstables, but I always need to track player locations.

Well, such a system clearly works well for you.

But here on <CG> we have a greater challenge, given the openness and scope of the project.

http://theweekinchess.com/chessnews...

http://theweekinchess.com/assets/fi... (pgn download - grpa)

He follows the tournament source, and only uses Wijk aan Zee as the site for all the games.

Generally, <TWIC> is very good about keeping the tournaments normalized.