Thursday, November 29, 2007

Funny Fritz

White to move (a position I reached in a game I just played as white):
Pretty much any move that doesn't give away the queen for nothing wins for white. I made a reasonable human move, Qf2+, forcing him to trade Queens, giving me an easily winnable endgame. Fritz, on the other hand, doesn't even rank that move in the top 20! Indeed, it says of that move 'Gets refuted' (and the "refuting" move is Qxf2+).

The lesson: Fritz doesn't play like a person, and Fritz doesn't like to simplify to an easily won endgame.

Note I'm not saying Qf2+ is theoretically best, but I like it because it involves the least amount of thought, follows the old saw "When ahead, make even exchanges, especially of Queens."

P.S. My play is better now that I've stopped worrying about ratings again. The cycle continues.

16 Comments:

Blogger Temposchlucker said...

but I like it because it involves the least amount of thought

I bet that is no criterium for a chess engine:)

11/29/2007 03:25:00 AM  
Blogger likesforests said...

What a refutation?! One of my favorite Heisman articles is called "Going to Sleep In The Endgame", which explains how to win endgames without thinking. (Don't get your hopes up... almost every real endgame I've played required some alertness!)

11/29/2007 03:40:00 AM  
Blogger Glenn Wilson said...

Note I'm not saying Qf2+ is theoretically best

If the "theory" involves practical winning chances against a strong player (especially with limited time per move).

Of course, most people usually mean "theory" in the sense of leading to mate in the fewest moves with perfect play.

So, what I might be tempted to say is that Qf2+ is the best practical move for many/most humans in this position.

11/29/2007 06:12:00 AM  
Anonymous Anonymous said...

"Least amount of thought" is a motto I use myself all the time, and is, I think, a perfectly reasonable strategy (unless, perhaps, mechanical play decreases your general alertness). Not the least because in pressure situations (i.e. tournaments), when the result is more important than the beauty of the game, creative brilliancy may escape you, but you can more or less always rely on well-honed endgame/mating skills.

11/29/2007 08:48:00 AM  
Anonymous Anonymous said...

Glenn stole my thunder somewhat. Qf2 is absolutely the best move for a human.

The only thing that might be better is if there were some sort of mating net (but Fritz may have caught that, yes?) But humans aren't perfect calculators, so keep it simple. Save the energy of tough calculation for when you need it.

11/29/2007 09:37:00 AM  
Blogger Unknown said...

why not Qc5 and the rook to f1 and close in on him?

11/29/2007 11:08:00 AM  
Anonymous Anonymous said...

I was thinking of Maintaining a page with Chess Blogger information, that way if we play on FICS etc. or visit a Blogger's city we can meet up
[if we want to :) ]


The first step is to decide what to include:


Blog Name:
Real name: (optional)
Blogger Country:
Blogger City:


Handles :

FICS
ICC
CTS
what else?


Rating : (optional)

National:
FIDE:
FICS:
ICC:

Please email response to : iwijetunge[at]yahoo[dot]com


Getting to 2000

11/29/2007 12:20:00 PM  
Blogger Blue Devil Knight said...

Chessloser: that would work too. As I said, my move just makes things easier. Why give him a chance to come up with some swindle with his queen when I can trade queens and be up a rook? In practice this is (for me) the best way to play. When ahead, simplify, when behind, complicate and give your opponent opportunities to blunder. When ahead, I prefer the easy to the clever. I have lost too many won games trying to be clever. I don't do it anymore :)

Anonymous: interesting idea, though I prefer semi-anonymity.

11/29/2007 12:39:00 PM  
Blogger drunknknite said...

So, when I initially looked at this position I liked your move (Qf2 should prompt an immediate resignation from Black so it should be fastest), but then something occurred to me. In this particular position you can easily win Black's queen for the Rook. If you start with Qc5+, where does Black move? The lines are actually very simple.

1 Qc5 Kg7

(1...Qe7 2 Qc8 Qe8 [2...Kf7 Rd7] 3 Rd8; 1...Kg8 2 Qc8 transposes; 1...Kf7 or Ke7 2 Rd7 Kf6 3 Qe7)

2 Rd7 Kh6 3 Qf8 Kg5 (3...Kh5 4 Rh7) 4 Qd8! Kh5 5 Rh7

All of these lines net the poor Black Queen and should be fairly simple to see in a longer game. So maybe Fritz is right....

11/29/2007 03:09:00 PM  
Blogger Blue Devil Knight said...

Kevin et al: perhaps it's useful to recognize this pattern so if it were to come up in a different context it would be helpful. In this context, even if I saw that set of moves I'd still play the one I did. Less thought, less need to trust in my crappy visualization skills.

11/29/2007 05:17:00 PM  
Blogger likesforests said...

1.Qc5+ Kg8 2.Qc8+ Kg7 3.Rd7+ Kh6 4.Qf8+ Kg5 5.Qd8+ Kh5 6.Rxh7#

Qc5+ is a great line, but I'm not going near it with a 10-foot pole if I see a simple one-mover like 1.Qf2+ that guarantees the point.

"Do not calculate complicated lines before you are absolutely sure it is necessary" -- Dvoretsky

11/29/2007 05:36:00 PM  
Blogger drunknknite said...

What I love about the blogosphere is that attention is drawn to really abstract theoretical considerations. In a recent post from liquideggproduct I commented that one time I sacrificed two queens for simplicity. This is an even more extreme example of what you did here. I took a knight with my queen, promoted, took the other knight with my new queen and then had two pawns against a king and converted that with no thought (I was a little annoyed that my opponent hadn't resigned yet and wanted to show him how badly he was losing, also I was very bored and figured he would move faster if he only had a king). And I may well have snapped off Qf2 without any thought. But then I saw that the king could only escape via f6 or h6 and that f6 allows the deadly skewer. So I was intrigued.... It is because I saw that the f6 square was inaccesible that I started looking at variations at all. And the only variation I had to look at was with the king on h6 and the rook already at d7 and then Qf8 (forcing move (also from the LEP comment), interestingly Qc1 forces Qg5 and then Rh7 also wins the queen) and Qd8 follows very naturally. To tie up loose ends we must of course consider that the queen can block the initial check, but if this is the worst case we can simply take the queen on e7 and accomplish the same thing as qf2. So really before I even started visualizing I had it narrowed down to a very specific position. Just a little insight into my thought process and why I would even take the 30-45 seconds to trap the queen when I could just trade it.

11/29/2007 06:43:00 PM  
Blogger likesforests said...

It's neat that you can work out all three variations (Qe7, Kh6, Kf6) and prove that Ke8 and Kg8 transpose in 30-45s. It took me 3-5 minutes.

11/29/2007 07:53:00 PM  
Blogger Glenn Wilson said...

There is an interesting dynamic between playing the "best" move and a simple move that also wins. If you see a mate in four should you look for a mate in three? If you are annotating a game do you give a mate in four a "?" because there was a mate in three?

BDK's position is not a white to move and win type of position. White is already won. At this point it is a matter of technique. From a technique perspective I prefer Qf2+. I see that technique (in general and not just this position) as the surest way of picking up the full point.

11/30/2007 07:54:00 AM  
Blogger Robert Pearson said...

Funny, after reading this post last night I went to the club and won a game where afterword the spectators were showing me a move that would have left me a full queen up (queen against nothing), but instead I had played a line that left me a knight and three pawns up against nothing and a pawn sure to queen...I didn't feel too bad about the missed tactic, but I think it's good to go back and find these in the postmortem--sometimes there is only one win available, after all.

I remember reading about an adjournment the great Keres had where he found four different wins, then decided to play a fifth, more "artisitic" win...which was a draw. Nowadays, I'm from the "practcial" school of chess!

11/30/2007 12:48:00 PM  
Blogger Polly said...

As somebody told me last weekend, Fritz doesn't go deep into end game motifs. Fritz will look for tactics and brute force. For us mere humans, simplifying by trading queens eliminates possible counterplay or other complications. Fritz goes by KITS, intead of the human method of KISS.

11/30/2007 10:58:00 PM  

Post a Comment

<< Home