ARCHIVED POSTS
< Earlier Kibitzing · PAGE 726 OF 1118 ·
Later Kibitzing> |
Jul-20-14
 | | chessgames.com: <The error message should flag the missing player if possible.> You're right, it behaves oddly in that regard. <This one's a little more serious - doing an <Adv Search> on <Yu> + <Yudin> yields no hits, same for <Yudin> + <Yu>.> If you try just doing a search on "Yu" alone, you'll see the problem: <Sorry, we have no names that exactly match '' in our database.> Problematically, it doesn't allow for two-letter searching, for technical reasons. Fortunately in the case of players of GM Yu's calibre you can use the drop-down menu, but it doesn't let you just type it in. You are right, that one is more serious, and will take more effort to remedy. |
|
| Jul-20-14 | | zanzibar: <Chessgames> OK, thanks for the explanation. BTW - FIDE is embedding player ID's into the PGN.
It would be wonderful, but probably impossible, if <CG> could use FIDE pid's directly (so <CG> pid = <FIDE> pid when applicable). Short of that, I can't help but ask if we could we do a search on FIDE pid. Maybe something like <#14907526> + <#5007003>. (The '#' being an escape for FIDE id.)
I know, I know, if wishes were horses...
(As my mom always said - it never hurts to ask). |
|
Jul-20-14
| | chessgames.com: Searching on FIDE ID numbers is a very good idea. The introduction of FIDE IDs to our database is relatively recent (in the grand scheme of things) so the software doesn't take full advantage of them yet. You don't happen to know the numerical range of the FIDE ID numbers do you? If they are all orders of magnitude larger than our own PIDs (as in your example) it might be possible to allow the end user to simply type a number, and the software will figure out if it's intended to be a FIDE ID or a Chessgames PID simply by how many digits it has. That would eliminate the need for the # prefix. |
|
| Jul-20-14 | | zanzibar: I believe most of the "modern" FIDE id's follow a format I wrote about on my blog, back in May: http://zanchess.wordpress.com/2014/... and
http://zanchess.wordpress.com/2014/... Quoting from the first post:
<[Suppose his] FIDE ID is 61400220. A simple explanation is that the ID is a combination of Federation identifier + player identifier:CCCnnnnn
Where nnnnn is a 5-digit number, and CCC is a 1-3 digit Federation identifier. The CCC is often called a country code, but I believe it really just identifies a Federation within a country. Or, perhaps it’s redundant, since some countries can have several CCC numbers, e.g. Turkey, Germany, USA, Russian, France, Spain, etc. It appears that the nnnnn part of the identifier is deliberately kept less than 10000, and so always leads with at least one zero.> However, there are older FIDE id's which don't follow this scheme. And a few exceptional modern ones as well (a couple of which I make note of in my blog). Putting them aside for the moment - the general FIDE id will involve 6 or more digits. |
|
| Jul-20-14 | | Alien Math: Most FIDE ID searches appear between 6 and 8 numbers, have not find numbers outside those limits, are information from self searches and time and not official numbers |
|
| Jul-21-14 | | zanzibar: Yes, <Alien Math>, you may be right for "active" codes. I didn't reread my articles too closely, but it looks like the exceptions I cite are cases of "weirdness" with the country codes. And I think the other cases, from my memory, were on Olimpbase, and involved id's for players who are no longer active. |
|
Jul-21-14
| | chessgames.com: <It would be wonderful, but probably impossible> No it wouldn't be impossible at all. That's why I am asking some questions to try to come up with a plan. However, our PIDs are up to 5-digits now so it's not impossible to foresee 6-digit PIDs one day, especially with our propensity to waste PIDs when we merge players. So perhaps the #-prefix is a good idea after all. |
|
| Jul-21-14 | | heuristic: CG:
for the Opening Explorer; how do we :
- ask for the #games of a position to be re-calculated?
the total #games for a position (under the board) does not equal the sum of the games of each alternative line (right column) - notify the authorities of errors?
for instance;
Opening Explorer the 15...gxh6 line references a game that has a different move 13 for BLK. |
|
Jul-21-14
| | chessgames.com: <heuristic:>
<- ask for the #games of a position to be re-calculated?> That happens automatically, usually a few times a week. <the total #games for a position (under the board) does not equal the sum of the games of each alternative line (right column)> We actually just went through this the other day. It seems like a bug, but actually there is a logical reason for it. For more information see the post by John Barleycorn (here: chessgames.com chessforum) and the ensuing discussion. Also see Opening Explorer Help Page under the section "When I add up the number of games in the moves column, it's bigger than the total number of games that contain the position shown. How can that be?" <notify the authorities of errors?> Assuming the problem is with a specific game, the correction slip process. (if you don't know what that means, see Chessgames Help under "I found a mistake in your database that requires correction.".) If the problem is more systemic, please explain the problem here or via email (chess@chessgames.com). <for instance; Opening Explorer
the 15...gxh6 line references a game that has a different move 13 for BLK.> Yes it does, but that's not an error. I think you have some misunderstandings about the way we handle transpositions. The game linked to in the 15...gxh6 line contains the exact same position as the page you were on. That's what's really important: what to play in a given position, regardless of the moves that led up to it. |
|
| Jul-21-14 | | heuristic: <#games>
Doh! my fault...
<game linked to in the 15...gxh6 line contains the exact same position as the page you were on> Ok. if I understand your last sentence, then the move number is superfluous. so why show it in the right column? you need the move number in the top banner. but the move number in the right column is misleading (and redundant!) I admit it might seem disconcerting at first; but no move number reinforces that this is a position; not a sequence. IMHO! |
|
| Jul-21-14 | | zanzibar: <RE: Dubai 2014 Blitz (and Rapid)> I'm just curious about cases like this:
http://www.chessgames.com/perl/ches... Does <CG> intend to delete the erroneous game at some point? |
|
| Jul-21-14 | | notyetagm: <CG.COM>
Can we please have a forum for the just started <2014 POLITIKEN CUP>? I am sure that many will find the next game to be of interest. 
click for larger view4 ... ♘b8-d7??
click for larger view5 ♘e4-d6#! 1-0
click for larger view[White "Rapport, Richard"]
[Black "Rogers, Tim L"]
[Result "1-0"]
[Round "01"]
[Date "Mon Jul 21 2014"]
[WhiteClock "1:31:00"]
[BlackClock "1:25:00"]
1. e4 c6 2. Nc3 d5 3. Qe2 dxe4 4. Nxe4 Nd7 5. Nd6# 1-0 |
|
Jul-21-14
| | chessgames.com: Ha, that's reminds me of F La Rota vs J Sarwer, 1988. Sure, we'll get those games in soon. |
|
Jul-21-14
| | chessgames.com: <Does <CG> intend to delete the erroneous game at some point?> Of course. Somebody should make a correction slip. And one must be careful, because especially in blitz, there's no guarantee that the side who is winning in the final position didn't get flagged. |
|
Jul-21-14
| | chessgames.com: Politiken Cup (2014) We'd love to open discussion and put this on our homepage but I'm having trouble getting the "five Ws" for the event. Well in this business it's really only three Ws: Who, where, and when. And I already have the "who"—but what city, what start date, what end date? Official site: http://www.ksu.dk/politiken_cup/ |
|
Jul-21-14
 | | SwitchingQuylthulg: As stated in the PGN, the location is Helsingør in Denmark. The dates can be found here: http://www.ksu.dk/politiken_cup/tur... |
|
Jul-21-14
| | chessgames.com: Thanks SQ that's exactly the page I was looking for. |
|
| Jul-21-14 | | zanzibar: RE: Politiken Cup (2014) To save a hop and translate of
http://www.ksu.dk/politiken_cup/tur... <Because of the Olympics in Tromsø and the Norwegian and Swedish championships move the tournament in 2014 for new dates.
In 2015 and 2016 we are back in week 31
Monday, July 21 to Tuesday, 29. July 2014.
The tournament is open to all.
Timetable:
Opening: Monday 21 July at. 13
Round 1: Monday 21 July at. 14
Round 2-3: Tuesday 22th July at. 10 and 17
Round 4-9: Wednesday, July 23 to Monday July 28 at. 13
Round 10: Tuesday, July 29 at. 10 Prize giving at. 16.30. Tournament:
10 rounds FIDE Swiss.
Reflection:
90 min for 40 moves + 30 min for the rest of the game + 30 seconds increment for every move starting at row 1 Other rules:
In the Politiken Cup has been allowed to come 1 hour late before losing the game.> |
|
| Jul-21-14 | | zanzibar: <chessgames> <And one must be careful, because especially in blitz, there's no guarantee that the side who is winning in the final position didn't get flagged.> One thing I was wondering - for these top-level blitz tournaments... what is the percentage of games decided on time? And how does this vary as a function of average player rating? Just curious... in case anybody knows. |
|
| Jul-22-14 | | zanzibar: Actually, DGT games should also record the remaining time on the clocks in the PGN stream. That would be additional confirmation for flagged games, and would be otherwise be interesting anyways. |
|
Jul-22-14
 | | Stonehenge: CG has been slow for a while and now I see there are more than 5700 users online. |
|
Jul-22-14
 | | Chessical: <chessgames.com> Sorry to be a real pest but driven by an overwhelming compulsion to complete: Game Collection: Hastings (1980/81); I see that Andersson-Sunye - R7
Liberzon-Brito - R7
don't appear to have uploaded from the 19th (probably my fault with the rogue ½-½). Should I resubmit them? |
|
Jul-22-14
 | | OhioChessFan: User: den
User: nile
User: Ove
User: ser
User: vice
User: Attack Wesley So
User: Stupid |
|
Jul-22-14
| | chessgames.com: <Chessical> No need, I reprocessed them. Ulf Andersson vs J Sunye Neto, 1981
V Liberzon vs L Brito, 1981 |
|
Jul-22-14
| | chessgames.com: That's odd, thousands of people are visiting the site today mostly via http://sportgeza.hu/sakk/2014/07/22... all to see the
R Rapport vs T Rogers, 2014 game. If we get some new Hungarian members let's welcome them with open arms. |
|
 |
|
ARCHIVED POSTS
< Earlier Kibitzing · PAGE 726 OF 1118 ·
Later Kibitzing> |
Kibitzing
