08 Jun 08
Originally posted by VarenkaI see what you mean, but at the same time a 560,000 game database won't even include many perfectly sound variations which a db 6.25 times the size does.
I suspect the quality of games used for the shredderchess database is much better than those at 365chess. I prefer if the game stats aren't skewed by low quality games.
It is up to the user (as always) to seperate the wheat from the chaff.
08 Jun 08
Originally posted by Squelchbelch365com includes games where both players are rated below 1000. Just how much separating the wheat from the chaff do you want to do? 🙂
I see what you mean, but at the same time a 560,000 game database won't even include many perfectly sound variations which a db 6.25 times the size does.
It is up to the user (as always) to seperate the wheat from the chaff.
I use a database where at least one of the players is above 2300. I may not have all "sound moves" but if a 2300+ player hasn't played a given move in the last 30 or so years then it's far more likely to be poor rather than sound.
08 Jun 08
1 edit
Originally posted by VarenkaReally?
365com includes games where both players are rated below 1000...
Where did you get that from?
********
Let's take one of my favourite openings; the Morra gambit
Shredderchess db has 136 games after 4.Nxc3
365chess.com has 4393 games after 4.Nxc3
For anyone who knows the opening at all, 136 games doesn't even cover a fraction of the main line defences for Black & tactical motifs for White.
Argument over as far as I'm concerned.
08 Jun 08
Originally posted by VarenkaYeah, but you also have to consider that your opponents are going to be playing "unsound" moves. Sometimes a larger database will show a refutation of a weak move, while the weak move would not even be included in the GM-only database.
I use a database where at least one of the players is above 2300. I may not have all "sound moves" but if a 2300+ player hasn't played a given move in the last 30 or so years then it's far more likely to be poor rather than sound.
08 Jun 08
Originally posted by Squelchbelch>> "Where did you get that from?"
Really?
Where did you get that from?
********
Let's take one of my favourite openings; the Morra gambit
Shredderchess db has 136 games after 4.Nxc3
365chess.com has 4393 games after 4.Nxc3
For anyone who knows the opening at all, 136 games doesn't even cover a fraction of the main line defences for Black & tactical motifs for White.
Argument over as far as I'm concerned.
While exploring the tree it mentions games where the players are unrated or both below 1000.
It's easy to construct large databases. At least on shredderchess, you know the moves are quality even if there are others missing. But on 365chess how do you tell if a move is sound or just popular at a low level? 365com allows members to upload their own games. If someone is rated below 1000 and plays hundreds of Morra Gambit games, you want them in your database? I'd rather not.
08 Jun 08
Originally posted by VarenkaI don't think it does, actually.
365com allows members to upload their own games. If someone is rated below 1000 and plays hundreds of Morra Gambit games, you want them in your database? I'd rather not.
I think you've misunderstood a function of the site, which is to allow you to create your own database of games, which you yourself can then explore, separate to those of the 365chess 3.5m database.
Another very useful feature, if you ask me!
08 Jun 08
Originally posted by incandenzaMy database is not GM-only games. It's games where at least one player is 2300+, not necessarily both. So there are games where weak opening play is being refuted.
Yeah, but you also have to consider that your opponents are going to be playing "unsound" moves. Sometimes a larger database will show a refutation of a weak move, while the weak move would not even be included in the GM-only database.
I agree that databases that are filtered may occassionally drop a useful game. But I think it's worth it to get rid of the poor games that skew the statistics. Look at the lines included in 365chess after 1.e4 e5 2.Qh5. There's even a Ke7.
08 Jun 08
Originally posted by VarenkaTo take the Morra Gambit example a stage further:
...At least on shredderchess, you know the moves are quality even if there are others missing.
1.e4 c5 2.d4 cxd4 3.c3 dxc3 4.Nxc3 Nc6 5.Nf3 d6 6.Bc4 a6!
Is a very tough defence in which it is difficult for White to make any inroads.
Critical is 7.Bg5! not the more intuitive 7.0-0?!
7.Bg5! appears 21 times on 365chess and err... zero times on shredderchess which just gives 10 examples of the incorrect 7.0-0?!
I'm sure advocates of other openings could find fault with a surface-scraping db like shredderchess, where critical lines are often ignored.
08 Jun 08
Originally posted by SquelchbelchOk, you're right, it is for personal databases. My mistake, sorry.
I don't think it does, actually.
I think you've misunderstood a function of the site, which is to allow you to create your own database of games, which you yourself can then explore, separate to those of the 365chess 3.5m database.
Another very useful feature, if you ask me!
But why not just use a database like Chessbase for such purposes?
I think you're still getting carried away with big numbers. I could construct a 10 million game database, but what's the point is a large percentage of the games are of a low standard? e.g. 1.e4 e5 2. Qh5 Ke7 on 365com. It's easy to get the numbers up by including such poor games.
08 Jun 08
1 edit
Originally posted by VarenkaI prefer the explorer-like features.
...But why not just use a database like Chessbase for such purposes?
chesslive.de has even more games, so I don't understand the numbers argument. I will probably use both anyway to be honest.
The way I use databases is I first look at candidate moves myself & try to narrow down to 2 or 3.
I then perhaps consult an openings book to see if the variation is any good or if I've missed something.
Lastly, I enter the top couple of candidate moves into chesslive.de (at the moment) & look at the results & also at the resulting positions several moves later to see if a blunder skewed results one way or another.
People may say that this isn't "real" chess, but to be honest I get much more enjoyment out of OTB & use this site mainly for refining my openings for club games.
Funny you should mention 1.e4 e5 2.Qh5?! Ke7?? because both chesslive.de & 365chess have the same single game; Gritchuk-Skrypin 2007 1-0
😛