SCID vs PC

Is there a way to maintain Statistics of Training (mates n puzzles) - in scid vs pc. I like the way the program can mark tactical exercises when annotating games & puts it under tactics. 

But i would like the program to keep a track of these exercises...

Is there any way to do that ?

Are you refering to the find-best-move feature ? If so, maybe you just want to search the database for tactics flag ? In the general search, click on "Find games with Flags" and then click on tactics.

Under Play there is Mate N Puzzles  > that takes us to tactics which can be played, but the program does not keep any stats of those we have played .......

Thanks for pointing out "games with flags" 

Iam having this error, so I am unable to use scidvspc, any advise please

Try uninstalling and reinstalling the program.....

I did that and dit not work

I'm still having trouble with the jagged sets and the prog refusing the switch to the colour scheme I instruct. I've tried resizing the board and reinstalling but to no avail.

Which version are you using ? i am on 4.12 

Uninstall should be clean. Lot of times files are left behind. You will need to delete those manually and then reinstall. 

 

Thanks, that works, now i can use scidvspc. I prefer this one over chessbase reader. thanks alot.

 

I'm having the color issue on linux so Idunno what's going on there. You should be able to just drag those lines to make the board larger.

Hi, folks! We are working on using Machine Learning to suggest next best moves. We have planned to use the a part of SCID 1.2 million games database, annotations by the Stockfish engine and the result of computer vs computer as the training for our algorithm. I was trying to automate annotating all the games one by one and collecting them in a folder. The analysis and annotation, I believe, gives us the relative score of the position of white. The analysis engine searches to a certain depth and then return us the calculated score. Since we've to process thousands of games, I want each game to be completed under a certain time. Actually game completion is not that important, annotation of each moves played by top players are. I read your help guide and it really helped me getting started to all. Actually what I want to do is run annotation only for certain depth. I want the engine to provide me a score for every move played. What now the engine is doing is that it is starts analyzing the first move and goes to a certain depth, then calculates the score and plays the next move by black itself and then does the same work till one side (computer vs computer wins the game) and then does this for the next move (played by the actual players) I dont want the engine to play on its own. Can't the engine just provide me the score for each move played in the current game only?

You want to use the auto-annotate feature, once the engine is loaded and you are on a database of games. The icon looks like a notepad. Make sure to chose all the appropriate options and verify you have set the process to do all the games you want. It works very well; that is what I do after loading a number of games into my database. Or at least, it has worked in the past when I used SCID; I now used SCID vs PC and it works there too.

@Martin_Stahl Thanks for answering, buddy! I have already read the manual and know about that annotation symbol. The problem is that the analysis for each move gives the next best moves till the game is finished by the computer and then moves to the actual next move in the loaded game. I just want a score for every move, nothing else. Hope you got me now.

I don't have the newest version of SCID, so it may not do what you want. Here are the options in SCID vs PC.

 

Under Add Score, All Moves

Under Add Variations, No

Under Which Side, Moves by both sides (if that is what you want). 

I think you can get something similar with SCID but it has been a while since I've tried.

That is how that annotator box looks like. Click on no variations and you will get annotations for only the game that you are analyzing.  

One piece of advice though - "Knowing "Why" behind particular annotation is more important than annotation itself"

 Stockfish or Komodo will tell you if your move is good or poor or blunder, but it won't tell you why it is so.....

You will need chess book - esp Positional play which is tough to understand; tactics are easier to understand - you will know why missed fork or pin is marked as ?? by komodo or stockfish but why simple rook move onto a open file is marked  "!" is tough to understand - unless you follow principles of positional chess 

Books that you need along with computer analysis - 

My System - by Aaron Nimzowitzch 

How to Reassess you chess - by jeremy Silman

Improve Your Positional Play by C.Hansen 

(There are zillions of books out there - and i am only suggesting these because i am using them, surely there must be many more good books out there)

 

Going through these books will help you understand why particular move is made. 

This is slightly off topic comment but just running chess engine through your game will not improve your chess. Of Course it is great way to spot missed tactics and build your own tactical database.

 

Thank you everyone for helping. You're right. I've just figured it out.

Just a last thing. If someone could help me with this, I would be able to complete my work. 

I want to annotate all of the games in my database. I've figured out how to run the analysis for all of the games in SCIDvPC but it doesn't edit the database. How do I make the edits reflect in the database? I can see only the options like Replace Game and SAVE PGN that would only save the modifications done to the displayed game but I want to re-save the whole database (with all the modifications done by the annotations to all games in a database). Am I missing something?

You can't do most work on PGNs. You need to import them into a SCID databse first. When you use the auto annotate feature you can batch annotate and it will auto save the changes. When you are done you can export all games to a PGN if you need them in that format.

@Martin_Stahl Thank you, buddy. "Export All Filter Games" was the option I was finding.

I want to annotate maximum possible games in coming 20 days. The database consists of 127,811 games. I can configure the engine according to depth per move or seconds per move. Can anyone tell me what should be a preferable depth for an engine per move? I'm thinking of giving 2 seconds per move. On my gaming laptop, after 2 seconds, I can see the engine going through 15-20 depths. Can someone guide me over this? Actually we are thinking to apply linear regression (a machine learning algorithm) over this dataset. Can you suggest me any reasonable amount of time or depth I should consider? I mean I want a trade-off between time per move, precision of the annotation score per move and number of games analyzed. Will 100 games would do considering 20-25 depths/move or would 10,000 games with 2 seconds (possibly depth 15) per move is better? I hope you understand my query. Just need a suggestion by the people here.