Using Fritz or Any Chess Analysis

Sort:
Pwnx

Well, my whole life I've sucked in chess to be honest. I still do. But I've decided I want to get better and be the best in my school so I can try to join some tournaments.

I browsed the internet and came across something called Chess Analysis Softwares. I eventually reached this forum and learned about some free softwares. I tried downloading Arena and Fritz 5.32. But I don't seem to get it.

In Arena, there are no suggestions whatsoever. At least in Fritz 5.32, there's this coach feature that gives suggestions every move. I tried pressing the Analyze Entire Game button in Fritz but nothing happened.

What really am I supposed to expect with these softwares? I assumed it would go through all of my moves and say like "Nf6 would have been better" or "You shouldn't have done that" but all I'm seeing in the section I suppose is the Analysis section are lines like:

"=(-0.19): 3.Qxd4 Nc6 4.Qa4 Nf6 5.Bb5 Bc5 6.Nc3 Qe7 7.Bd2 0-0"

How am I supposed to learn from this? :\

-waller-

The = means the position is roughly equal.

The -0.19 means that the engine evaluates Black to be 0.19 pawns ahead (although this is barely anything).

Seriously though, don't use engines, they are not helpful to start learning with. Try analysing your games yourself first if you want to learn from them.

A better way to continue is to find some articles on this site about tactics and study those for a while.

Shivsky

Good series of questions ... here are the answers:

1. Chess Engines have come a long way in terms of their ability to clearly pick out accurate moves to be played in any position, though they don't do a great job telling you why those moves are good nor will it "clearly" tell you why a move you think is good is not good.

2. What you can learn from them is the relative worth of moves (that you might have considered).

In other words, Computers rate  ANY CHESS position by a numeric value, often called an "evaluation score"..  Think of a number line as seen below:

 

Now the common convention most engines follow is such:

Really negative values => means Black has a WON position...reallly positive values =>  White has a WON position.

For everything in between, think of actual material count.

For example, an eval. score of +2.05 might mean "white is literally up 2 pawns" (1 pawn = 1.0)

Lower values like  +0.5 is half a pawn, meaning a small advantage for White.

0.0 would imply a dead even position.

Now how does this all help?  Well, in any engine, you should be able to set a position and then plug a move and get its corresponding eval-score...thereby determining the "Okay, if I play this move, who is winning (black or white) and by how much"

Therefore, when considering between moves A, B and C, you get to see what the engine thinks is the best with a given side to move.

Most modern engines allow you to look at N best lines where you can vary N from 1 to quite a high number, depending on how much time/CPU processing you are willing to spend.

For example in this position with White to move:

You can see that Nxe5 is the "best" line which (because of the tactic) wins material for White, where as something passive (Ne2, 3rd best) which just retreats the knight attacked by b4 keeps the position fairly even (0.0).

In summary, relative eval. scores is how one measures the value of one move vs. the other.  

For tactical positions (such as the one above), I don't think you need more convincing as to why one move is clearly best and if you played an inferior one, you SHOULD be able to learn from this.

For strategic/positional positions, you may not "know" why an engine preferred one move over the other or you may be a little unsure about the recommendation the computer makes  => in that case, get with a stronger player/coach or even post in this forum and ask. 

Engines don't do a great job of "human-like explanations of positions" so nothing beats good ol' fashioned "human explaining to human" learning :)

If you really want computers to speak to you (or are antisocial!) , I'd recommend a later version of Fritz which tries to explain (in plain english) what the merits/demerits of all legal moves are in a position.  Ubisoft's Chess Master series analysis does a feeble "human language" post-mortem of your game ... though these STILL won't really give you a solid explanation when you are stumped with a position.

Pwnx

Oh, ok then. I'll take your advice and not use engines. But by analyzing my own games, you mean writing down my games on paper then playing them out on the computer?

Lbrto369

Kapatid, 

 

I won't discourage you from using a chess engine....It helps... Particularly in searching hard to find "combinations" in a position...

However if you are aiming to improve your game, I suggest you first ground yourself with the basics... like the various theories of positional play.

 

Once you're sure you've understood it, then that's the time that you use the help of a chess engine to help you analyze the game... 

 

So in summary, it's much better to store books (for now) that discusses how to improve your assessment of the game instead of abruptly approaching a silicon friend to help you analyze games...

 Good luck! Smile

Pwnx

@Shivsky, wow long answer. Thanks for the info. I get it now.

 

@Alberto369, thanks for the advice. I have this chessbook. Not really a book per se but it has like 200+ situations and moves. I think it's a good start.


Random, why do people call it "silicon friend"?

oinquarki
Pwnx wrote:

Random, why do people call it "silicon friend"?


It's a friend made of silicon.

Shivsky

Silicon => Silicon associated with computer chips?

Friend => Because it's there for you ? :)

Lbrto369

I was thinking more of books like "chess fundamentals" though that might be too elementary for you since I assume you've already been exposed to the game for quite some time now...

(But it won't be bad to go back to the basics... Laughing)

For intermediate books, there are lots of books out that focuses on the various "laws" in chess but what I find very helpful for me was the classic "Think like a Grandmaster" by Alexander Kotov...

Most of his thoughts I think has already been incorporated in the much recent books...

You might also want to search for another classic like "Grandmaster preparation" by Lev Polegaevsky... Laughing

Also I think this one will help --- The Middle Game by Max Euwe....

And lastly Endgame Manual --- Mark Dvoretsky... (This one is very recent... Laughing)  

I don't actually own any of these books, but I was lucky to read through one of the old books in the past... And I swear it is a good help in grounding one's self in the theories of the game... 

You may also want to read through the various blogs written by masters here at chess.com.... Wink 

 

Some focus on a particular aspect of the game like endgames...

Pwnx

Friend was pretty obvious. XD

But I get Silicon now. :)

theturtlemoves
Alberto369 wrote:

I was thinking more of books like "chess fundamentals" though that might be too elementary for you since I assume you've already been exposed to the game for quite some time now...

(But it won't be bad to go back to the basics... )

For intermediate books, there are lots of books out that focuses on the various "laws" in chess but what I find very helpful for me was the classic "Think like a Grandmaster" by Alexander Kotov...

Most of his thoughts I think has already been incorporated in the much recent books...

You might also want to search for another classic like "Grandmaster preparation" by Lev Polegaevsky...

Also I think this one will help --- The Middle Game by Max Euwe....

And lastly Endgame Manual --- Mark Dvoretsky... (This one is very recent... )  

I don't actually own any of these books, but I was lucky to read through one of the old books in the past... And I swear it is a good help in grounding one's self in the theories of the game... 

You may also want to read through the various blogs written by masters here at chess.com....  

 

Some focus on a particular aspect of the game like endgames...


Don't buy dvoretsky's endgame manual unless you're already a fairly strong player - I have it and I'm finding it pretty tough going. Think like a grandmaster was good though.

Personally If I were you Pwnx I would study tactics e.g. chesstempo.com and buy Chess the Art of Logical Thinking for everything else, and if you really want something that covers massive amounts of stuff (including how to use an engine!) then the Mammoth book of Chess is immense.

alec39

Make these books you're personal friends for 10 months study them backwords and forwards with lots of practice you'll be ready for tournaments:

Learn Chess Tactics by John Nunn:

http://www.amazon.com/Learn-Chess-Tactics-John-Nunn/dp/1901983986

Chess-5334-Problems-Combinations-Games by Laszlo Polgar:

http://www.amazon.com/Chess-5334-Problems-Combinations-Games/dp/1579125549/ref=sr_1_1?s=books&ie=UTF8&qid=1314712650&sr=1-1

1001 Deadly Checkmates by John Nunn

http://www.amazon.com/1001-Deadly-Checkmates-John-Nunn/dp/1906454256/ref=sr_1_1?s=books&ie=UTF8&qid=1314712740&sr=1-1

The Game of Chess by Siegbert Tarrash:

http://www.amazon.com/Game-Chess-Algebraic-Siegbert-Tarrasch/dp/1880673940/ref=sr_1_1?s=books&ie=UTF8&qid=1314712819&sr=1-1

1000 Checkmate Combinations by Victor Khnekin:

http://www.amazon.com/1000-Checkmate-Combinations-Victor-Khenkin/dp/1906388709/ref=pd_sim_b_1

Super Nezh:

http://www.chesshouse.com/Super_Nezh_Hardcover_p/6247.htm

Attack with Mikhail Tal:

http://www.amazon.com/Attack-Tal-Cadogan-Chess-Books/dp/1857440439

Winning Chess Endgames Just the Facts 2nd Edition by Alburt and Krogius

http://www.amazon.com/Winning-Chess-Endgames-Facts-Second/dp/1889323152

Pwnx

Damn. That's a lot of books. I doubt I'll be able to buy them though. Considering the fact that I'm 15 and can't buy through amazon. I hope I find one of these in local stores though. Thanks.

Zelenkooo

try this web site chesstempo.com , it's free and they have a lot of puzzles ( interactive )

"Oh, ok then. I'll take your advice and not use engines. But by analyzing my own games, you mean writing down my games on paper then playing them out on the computer?"

Ye write them on paper if not playing on coputer :D and analyzie them ( write comments for every move ) post it here on forum it may help you alot 

AlCzervik

Pwnx, I think you might want to take Shivsky's advice regarding the human element. For what it's worth, I think that is the best advice. Make some friends here, play them and see where your game is. I believe there is no substitution for head to head play. Of course, I will admit that I've never used any computer program to play or move. I have found that practice, by playing, seems to be the best way to learn (initially). I would also recommend playing others that are slightly better than you. You may lose most games, but, you will learn.

This, as with the books, Frits, etc., takes time.