|
< Earlier Kibitzing · PAGE 37 OF 63 ·
Later Kibitzing> |
| Dec-04-21 | | Messiah: The site is very slow today. |
|
Dec-04-21
 | | MissScarlett: E Cronhelm vs Harrwitz, 1849 A P+2 game that didn't make it either. |
|
Dec-09-21
| | MissScarlett: Steve, fell asleep at the wheel - replace current pun with <1551229|Dubois|Duke Karl / Count Casabianca|Dubois Stood on the Burning Duke (MissScarlett) > |
|
| Dec-10-21 | | diceman: The Laboratory Computer isn't really working at all: 1.c4 ♘otes by Stockfish 11 (minimum 30s/ply) Nf6 2.Nc3 e5 3.g3 Bb4 4.e4 better is 4.♗g2 O-O 5.♘f3 e4 6.♘d4 ♘c6 7.♘c2 ♗xc3 8.dxc3 h6 = +0.19 (37 ply) Bxc3 ? 4...♘c6 5.♗g2 d6 6.d3 ♗g4 7.♘f3 ♗c5 8.h3 ♗d7 9.O-O = -0.40 (19 ply) 5.dxc3 better is 5.bxc3 d6 6.♗g2 O-O 7.♘e2 b6 8.O-O ♗b7 9.d3 c5 10.f4 = +0.39 (21 ply) d6 = -0.25 (24 ply) after 5...O-O 6.♗g2 d6 7.♕e2 h6 8.♘f3 ♗g4 9.O-O ♘bd7 6.f3 a5 7.Nh3 O-O 8.Nf2 better is 8.♗e3 ♗e6 9.♘f2 ♘bd7 10.♕e2 ♖e8 11.♗g2 a4 12.O-O ♘c5 = +0.22 (36 ply) Be6 = -0.28 (32 ply) 9.Bd3 Nbd7 10.Qe2 c6 11.O-O Re8 12.Rd1 Qc7 13.Bc2 h6 14.f4 b5 = -0.23 (7 ply) after 14...exf4 15.gxf4 b5 16.♕d2 ♗xc4 17.♕xd6 ♕xd6 18.♖xd6 15.cxb5 cxb5 ? = 0.00 (7 ply) 16.f5 ⩲ +0.99 (11 ply) Bc4 17.Qd2 b4 = 0.00 (30 ply) 18.b3 Ba6 19.c4 Qa7 = +0.23 (17 ply) after 19...♗b7 20.♕xd6 ♕xd6 21.♖xd6 ♘c5 22.♗e3 ♘fxe4 23.♗xe4 20.Qe3 Qc7 21.♕d2 ♗b7 22.♕xd6 ♕xd6 23.♖xd6 ♘c5 24.♗e3 ♘cxe4 25.♗xe4 = +0.41 (11 ply) <25.Bxe4 = +0.41 (11 ply)> 11ply? <15.cxb5 cxb5 {? = 0.00 (7 ply)> 7ply? <14.f4 b5 {= -0.23 (7 ply)> 7ply? That's at max depth.
You usually get 18 to 20 ply with 4 seconds of fuel. |
|
Dec-10-21
| | MissScarlett: How can we accommodate illegal moves made in a game? With particular reference to odds games, where castling is permitted although the rook is missing. |
|
| Dec-13-21 | | diceman: <diceman: The Laboratory Computer isn't really working at all:> Any chance this will be working this year? :)
I think about Nov 2 2021 was the first time I brought this up. |
|
Dec-14-21
 | | stevemcd87: <diceman> I made an update to the engine/analysis to spot instances with a lower ply count and reevaluate, if necessary.
If you spot another occurrence, please leave it on your "RECENT ANALYSES" section and provide your "job #".
As of now, we would need to reevaluate the game for it to update.
We are going to update the Engine/ Analysis server very soon, to be able to assign more CPUs in each evaluation, to increase ply count.
Thanks for your help.
<How can we accommodate illegal moves made in a game? With particular reference to odds games, where castling is permitted although the rook is missing.> Feel free to upload the game and provide the gid.
|
|
| Dec-15-21 | | diceman: Thanks Steve.
It still looks too low to me.
<job# 9685069444400621> Some selected moves:
<19.Rf5 Nd7 = -0.06 (13 ply)> <24.Rxf2 Nd4 ⩲ +0.57 (14 ply)> With a quick look, it looks like move 10 (out of 20)
is the last deep move:
<10.Qxe2 {= +0.32 (38 ply)> Then till the end of the game, the ply drops by about half, or more. |
|
| Dec-16-21 | | Messiah: U should introduce Armageddon games @ the live game feature, with 5 min refresh interval. |
|
| Dec-17-21 | | Messiah: Terrible! |
|
Dec-20-21
| | stevemcd87: <diceman> <Messiah> <Game Analysis enthusiasts >
We've made the decision on which new server we will use for the Analysis server and got the approval. I am confident this will give us higher quality analysis than what we have been seeing. It will be done by the end of the week, for sure😬🤞.
<Question> Would you rather a more in-depth analysis(more than CG has ever been able to do previously) or the ability to analyze more games at once, as long as the ply count are in the 30s(as we were once able to do)?
It's a tough decision for me. Feel free to voice your opinion.
<Editors>
I've been working on the new gameloader feature.
First, I'm creating an evaluate button that would allow us to easily evaluate the PGN to be uploaded.
It will give us stats such as:
- Distinct players(letting us know if they're in the database(DB) or if it will create a new player)
- Distinct events (letting us know if we should add it to an existing event)
- Duplicate games (letting us confirm or deny if it is an actual dupe... with a force option)
- Perhaps some sort of validator to approve or deny games(Eg. maybe only allowing games with at least one player higher than 2000).
We can then use these evaluations to create a user interface(UI) that will allow us to review what others have been approving.
From there we could probably begin writing descriptive guidelines of what we'd all like our DB to consist of.
I'll provide a picture of a sample evaluation soon. As always, thanks for your help and patience. |
|
| Dec-20-21 | | diceman: <<stevemcd87: Question> Would you rather a more in-depth analysis(more than CG has ever been able to do previously) or the ability to analyze more games at once>I would <rather a more in-depth analysis(more than CG has ever been able to do previously)> option, as long as you can back off and still do mild analysis
when the position isn't that important to you. |
|
| Dec-21-21 | | Messiah: I have no strong preference. |
|
Dec-22-21
 | | jessicafischerqueen:
<Stephen>
< Perhaps some sort of validator to approve or deny games(Eg. maybe only allowing games with at least one player higher than 2000).> How are you going to calculate the ratings for older-era historical players? These are almost never included in the pgns I find. There also wasn't a standardized way of computing/awarding a numerical rating level until relatively recent times. Almost none of the historical games I have uploaded to cg.com have listed a numerical rating tag in the pgn. I fear that new "gatekeeping" software might prevent me uploading other historical games I might find in the future? ========
Thanks again for helping me out so quickly and successfully on the <"disappearing game" that "could not be put back"> episode... it felt like I was in a "Nancy Drew" mystery for a few minutes there... |
|
Dec-22-21
 | | OhioChessFan: My mobile throws out question marks instead of the apparent diacritical mark. See <Everett> Karjakin vs Rapport, 2021 |
|
Dec-22-21
 | | beatgiant: <OhioChessFan>
I already noticed the same issue with <keypusher>. Can you please go to my forum for a follow-up? |
|
Dec-26-21
 | | Stonehenge: Steve, could you please have a look at https://www.chessgames.com/playerli.... The order is a bit off:
10000
10001
10002
64624
10004
65645
10007
10008
57512
etc.
Also, can you get rid of player files with zero ( |0|||| ) games? A lot of them have been made lately.
I cannot really delete them all by hand because it's not allways clear who the player is with which the zero game file should be merged with. I merged some with NN but that was not such a great idea apparently. I have just uploaded this tournament Ermua (1990) and the Iruzubieta games appeared under NN. |
|
Dec-27-21
| | MissScarlett: <Also, can you get rid of player files with zero ( |0|||| ) games?> But not Woodrow Wilson |
|
| Dec-27-21 | | Messiah: On the page Player Directory there is an outdated warning that reads: <Chessgames.com will be unavailable Wednesday, December 22, 2021 from 4AM through 5AM(UTC/GMT) for maintenance. We apologize for this inconvenience.> Tango-Echo-Romeo-Romeo-India-Bravo-Lima-Echo |
|
Dec-29-21
| | Stonehenge: There's something wrong with searching for players with the same given name plus the same surname. A search for <Juan Garcia> only show one game: https://www.chessgames.com/perl/che.... A search for <Juan Reyes> shows zero games: https://www.chessgames.com/perl/che.... |
|
Dec-30-21
| | stevemcd87: <Stone> I believe I found one of the reasons why gameloader is acting up.
NN page has many aliases, Iruzubieta being one, that the gameloader takes into consideration.
Editors now have the ability to select an alias and remove/zap it.
https://www.chessgames.com/playerli... is "terrible".
What "columns" would you like a filter for?
I would think Pid, Name, and perhaps birth and death dates. Anything else?
I would think it may be useful finding dupe players |
|
Dec-30-21
| | Stonehenge: <Editors now have the ability to select an alias and remove/zap it>. That's brilliant work, thanks.
<I would think it may be useful finding dupe players> I don't know. I've always maintained the playerlist by hand. So entering the FIDE card number and entering the country and DOB. Thus I have found countless of duplicates. But with tens of thousands of games being uploaded lately, I cannot keep track anymore. |
|
Dec-30-21
| | Stonehenge: I don't mind the pink colour of the aliases but now the kibitizing area of player pages has turned pink as well :) |
|
Dec-30-21
| | Stonehenge: Fixed, thanks. |
|
Dec-30-21
| | MissScarlett: William Timbrell Pierce The knights in the biography look strangely depressed. |
|
 |
|
|
< Earlier Kibitzing · PAGE 37 OF 63 ·
Later Kibitzing> |
Kibitzing
