How fair Chessbase is in including poorly annotated games in Megadatabase

[bmp1974]

I am against removing annotations because it may be interesting to see also the mistakes that people did in the past in the analysis.

I think that it is better to add new annotations without removing old annotations so people can see 2 versions of the game(with old wrong annotations and with new annotation that is verified by stockfish).

I agree your point for the chess books. But for the digital DBs one is better off seeing correct evaluations.
One thing that stood out in the recent WCC is most of the commentators including Vishy (who has annotated so many games for Informants) was itching to see what the engine thought about a particular position. I feel the Chess Engines with NNUE technology have earned that respect which was not the case few years back.

[Cornfed]

I am against removing annotations because it may be interesting to see also the mistakes that people did in the past in the analysis.

I think that it is better to add new annotations without removing old annotations so people can see 2 versions of the game(with old wrong annotations and with new annotation that is verified by stockfish).

I agree your point for the chess books. But for the digital DBs one is better off seeing correct evaluations.
One thing that stood out in the recent WCC is most of the commentators including Vishy (who has annotated so many games for Informants) was itching to see what the engine thought about a particular position. I feel the Chess Engines with NNUE technology have earned that respect which was not the case few years back.

If I may butt in - your comment about Vishy runs counter to what someone else 'complained' about on this very forum - that Vishy did not seem to want to refer to engines!

I did not watch Vishy - preferred Fabi on chess.com's broadcast, so I do not know whom to believe.

In any case, if you or anyone else wants a little non-verbal 'engine truth', you are welcome to it. It's simply not what you are paying for though...and even engines can 'disagree' a little. Perhaps you would do better with Bigbase or downloading a bunch of TWIC's and letting an engine of your choice annotate everything.

History needs to be kept. Some complained about Nunn's 'engine approved' footnotes to his Alekhine books (I think it was Alekhines...maybe Keres) and Fischer was dead against revisions to his annotations. So there is two sides to the argument. I think most authors of annotations would err on the side of respecting the history.

[dkappe]

If I may butt in - your comment about Vishy runs counter to what someone else 'complained' about on this very forum - that Vishy did not seem to want to refer to engines!

I did not watch Vishy - preferred Fabi on chess.com's broadcast, so I do not know whom to believe.

In any case, if you or anyone else wants a little non-verbal 'engine truth', you are welcome to it. It's simply not what you are paying for though...and even engines can 'disagree' a little. Perhaps you would do better with Bigbase or downloading a bunch of TWIC's and letting an engine of your choice annotate everything.

History needs to be kept. Some complained about Nunn's 'engine approved' footnotes to his Alekhine books (I think it was Alekhines...maybe Keres) and Fischer was dead against revisions to his annotations. So there is two sides to the argument. I think most authors of annotations would err on the side of respecting the history.

I shopped around and settled on the broadcast with Vishy and Anna. I liked it better without the engines. Anand said it best when he pointed out that engines cut off debate and thinking. When an engine says -2.5, you just stop looking. When Anand wanted to check a deep tactic with “the engine,” you could just as well have used SF11 or even Crafty.

[Cornfed]

UI do however create my own 'Reference base' (despite the fact that 'big data' generally points you in the right direction) with players just over (depends) 2200 or over 2400 where I augment with annotated data from books/magazines and older good players 'pre-elo'.

Could you give some details how to do this? I’m having a hard time figuring out how to extract the really “good/quality” games. After 1970 one could take Elo. But I’m more interested in the “classical era”. There are many games in Mega from the 19th and 20th Century by club player etc which I would somehow like to filter out.

I plan on taking off a few days after the first of the year and this is one of the things I was going to work on. I had a hard drive crash earlier in the year and had no back up for several files....otherwise I would just be 'adding' high quality games to what I had already established.

Basically though, I start with an empty database into which I am going to bring only high quality games. I call it 'Quality Base'. Then I take multiple sources - Megabase, Corr Base, ebook pgn/cbv file, Chess Informants, Chess Publishing files, etc....and spend time "data washing" of the files into 'test bases'. I also like to filter out draws before move 20.I filter out games by rating - I think last time I only kept where 1 player was >2400 and his opponent at least 2200. For "Correspondence" games, I set the low bar at....think it was 2500 for both opponents, I then combine those into Quality Base, where I do one last bit of filtering so as to keep the 'better' game in case of duplicates.

All I am trying to do is create a really good Reference Database that I can use within Chessbase instead of just having something like 'Mega' as the default. I really could care less if two 1500's played a different move than I did when checking a game I played or working on opening files.

[KLc]

Basically though, I start with an empty database into which I am going to bring only high quality games. I call it 'Quality Base'. Then I take multiple sources - Megabase, Corr Base, ebook pgn/cbv file, Chess Informants, Chess Publishing files, etc....and spend time "data washing" of the files into 'test bases'. I also like to filter out draws before move 20.I filter out games by rating - I think last time I only kept where 1 player was >2400 and his opponent at least 2200. For "Correspondence" games, I set the low bar at....think it was 2500 for both opponents, I then combine those into Quality Base, where I do one last bit of filtering so as to keep the 'better' game in case of duplicates.

Ok, great, this is what I'm doing right now. I'm using PGN files (I really dislike Windows) and use pgn-extract to remove duplicates after adding new games.

All I am trying to do is create a really good Reference Database that I can use within Chessbase instead of just having something like 'Mega' as the default. I really could care less if two 1500's played a different move than I did when checking a game I played or working on opening files.

That's exactly my point as well.

[Cornfed]

All I am trying to do is create a really good Reference Database that I can use within Chessbase instead of just having something like 'Mega' as the default. I really could care less if two 1500's played a different move than I did when checking a game I played or working on opening files.

That's exactly my point as well.

I know truly big data analytics might imply the end results might be fairly reliable even with that stuff thrown in...but so many lines I look to research and play have comparatively small representation that it just makes sense to build my own reference database.