Forums

Best program for automatic game analysis / annotation

Sort:
hdjur_jcv
I have requested here:
 
 
an article about:
 
Comparison_of_Chess_GUIs with respect to automatic analysis features, reliable sources should be found in chess periodicals or portals, at least Fritz, Chessmaster, Arena, SCID vs PC, LucasChess, Tarrasch should appear in rows, and different important features as found by the author, in columns
 
I wanted to write it by my self, based on my own research, but realized that I don't count as a reliable source, and original research cannot be base for an article there. Although, that wasn't much of original research or something anybody cannot done, I just basically installed on my PC whatever I could for free: Arena, SCID vs PC, LucasChess, Tarrasch and started to compare them, started reading reviews on Fritz and Chessmaster I could find online (for free), came up with some ideas on my own about how this things should be implemented and so. But being self critical as I am, I decided not to try to offer my insights to any of such chess magazines, because I believe I would be rejected, and I really cannot find such a comparison in these reliable sources. If you do, point it out please. Also, if someone is able to write such an article for Wikipedia, I would appreciate it very much.
C-Crusher

Fritz 15?

hdjur_jcv
pfren wrote:
[COMMENT DELETED]

 What's that?

aman_makhija
hdjur_jcv wrote:
pfren wrote:
[COMMENT DELETED]

 What's that?

It means he deleted his comment.

And yeah, stockfish 6 is best

hdjur_jcv
aman_makhija wrote:
hdjur_jcv wrote:
pfren wrote:
[COMMENT DELETED]

 What's that?

It means he deleted his comment.

And yeah, stockfish 6 is best

 The question was:

http://www.urbandictionary.com/define.php?term=What%27s+that%3F

Like: "Come again?", "You said something?" etc...

And by the way, you compare apples and oranges. I may also think that Mercedes is the best, but why would I write such a thing here, in this thread?

hdjur_jcv

The main idea behind my posting here was hypothesis that not only computers are superior to humans in playing chess, but they can be also superior in chess tutoring. Although the former is actually more like a proven fact, the latter still sounds like a heresy to many. It basically means that learning chess would no longer require any interaction with people, some of it is already present on this site, at least for premium members.

It revolves very much around computer capability to analyze ones games, present the results in a humanly understandable and beneficial way, study her/his weaknesses and prepare adequate excersises, ... whatever is part of a normal training process. 

Analyzing games of master players in this context is for example a task similar, but not entirely identical to the one already mentioned. The difference is that the pupil already has some clue about ideas behind each move played in games he/she took part of, while that is not neccessarily the case when dealing with master games. So this requires a different approach with respect to "configuring the output" of analysis, but still, these both are basically the same "automatic game analysis / annotation" task.

 

I'm aware of the fact that what I mentioned there is just a tip of the iceberg, primarily built for Windows, and when it comes to chess GUI's for Android, there is a plethora of other programs, literally hundreds of them, that I didn't look at. Any sensible input here would be greatly appreciated...

 

For example, one formal point is with respect to the input and output of that feature. They both should be the most standard PGN's, not producing non standard characters/tags while annotating, so that afterwards you cannot import it anywhere without editing first to remove the excessive formatting, and such things...

 
Martin_Stahl
hdjur_jcv wrote:

...

It revolves very much around computer capability to analyze ones games, present the results in a humanly understandable and beneficial way, study her/his weaknesses and prepare adequate excersises, ... whatever is part of a normal training process. 

...

 

 

This is the problem. Most programs can present lines but in many cases, the reasons why that line is better than another is not clear and the program can't explain it to you.

hdjur_jcv

Martin_Stahl wrote:

hdjur_jcv wrote:

...

It revolves very much around computer capability to analyze ones games, present the results in a humanly understandable and beneficial way, study her/his weaknesses and prepare adequate excersises, ... whatever is part of a normal training process. 

...

 

 

This is the problem. Most programs can present lines but in many cases, the reasons why that line is better than another is not clear and the program can't explain it to you.

OK, so which programs did you test for this purpose?

Sommerswerd

No point analysing your games by yourself without an engine because you will be using the same thinking-process as you used in the game itself.

aman_makhija

sommerswerd @ post 57

  • You have unlimited time analyzing a game by your self, whereas in the actual game you had only 30 seconds or so on average per position. You ca go to the critical point and spend 30 minutes, and find out the truth.
  • There is no pressure on you while analysing a game, whereas there is during the actual game. A mistake in analysis costs nothing.
  • You can always give your games to a coach, or post it so people can explain to you in English where you and your opponent went wrong. Compuers can only give a confusing mixture of letters, co-ordinates and evaluations.
  • You can move the pieces around while analyzing, whereas whil playing you have to visualize possibilities

 

hdjur_jcv

OK, so I just realized Martin_Stahl already posted in this thread, and pretty much answered the question, but it was 3 years ago, so the question still makes sense.

gideonse

May I suggest that anyone who's considering using a chess analysis program to analyse their mistakes, would try doing it with decodechess.com?  The difference here is that moves and positions are being explained, in intuitive language and with a cool User Interface that is divided into tabs. They are providing explanations for Stockfish's suggested moves, so it's a double win here, because you basically get Stockfish with a cool user interface, and can understand what stands behind the suggested move, instead of just being fed an answer. Everything's online and you don't have to download anything to your computer which got me out of the tech mess...

 

student-t

Try SmallFish chess app (http://www.smallchess.com) to download.

There's a blunder check feature:

"Blunder Check" will give you analysis:

JogoReal

At my age (started playing amateur tournaments in 1978) I don't believe in and don't care about improving my strength of play. I use automatic analysis to see my mistakes in a fast way and to evaluate my level of play in the game and that of my opponent. I also go into the database of my own games searching for my best games. For this I use the evaluation of game strength for white and black, something many programs do nowadays and still not mentioned in this thread. The evaluation is usually made as a % or as an average of lost centipawns, or an Elo estimation.

And this even mentioning how useful the feature is for cheaters detection...

Installed software I know that can do this are: Fritz, Hiarcs and Lucas chess. There are more?