- About Us
- Chess Improvement
- Chess Puzzles
- Chess Research
- College Chess
- General News
- Home
- Major Tournaments
- News
- Polgar Events
- Privacy Policy
- Scholastic Chess
- SPICE / Webster
- Susan’s Personal Blog
- Track your order
- USA Chess
- Videos
- Women’s Chess
- Contact Us
- Daily News
- My Account
- Terms & Conditions
- Privacy Policy
I have bought Fritz 8 & 9 & 11. So I like Fritz.
I routinely use menu Tools > Analysis > CompareAnalysis.
For every ply, it gives me
(A) an evaluation number, and
(B) the move Fritz would have recommended.
But in Fritz_11 CompareAnalysis is badly broken, at least on my computer. Here is a .Jpg image showing that lots of plies are being *skipped* (no eval, no recommended move):
Fritz 11 Compare Analysis bug?
I have been repeatedly asking what happens when other people try/test CompareAnalysis on Fritz_11, but so far I have not heard back.
Does CompareAnalysis skip plies in your Fritz_11?
Thanks.
GeneM
CastleLong.com , FRC-chess960
Without a doubt, and after buying it, Fritz 11 is not nearly as strong as Rybka. Don’t waste your money. Really….if you want the best analysis, buy the best program: Rybka.
but, the graphics are nice.
I have Rybka multi-processor running from the Fritz 11 platform.
And its fast. Now I get to watch some of the strongest players ever, what a blast!!!
I have Rybka running in a tube of toothpaste. I will be brushing my teeth between moves at the next tournament.
If you can get fritz into an endgame with 10-12 pawns and 2-4 pieces, then Fritz plays like a chump. Fritz will be out of book and It has a limited horizon. In these positions the computer has no common sense. The trick is getting to the 12 pawn endgames.
I got a few months ago in Vienna.
It’s hard to beat, I think it’s obvious. But much depends on the construction of the computer Fritz11 runs on.
I have two different computers and I installed the program on both ones.
I could achieve only a draw against the one whose RAM memory is lower (144MB) than the ideal according to the descrpition (around 256MB).
Against the computer with (more than 800 RAM) I have no chances, I always loose.
But it is worth buying.
>>
Fritz 11 is not nearly as strong as Rybka. Don’t waste your money.
>>
Funny. Can you explain why a club player needs, say, a 2900 program more than a 2850 program? If your object is to learn the game better, both are about equally good. The only reason you might need the 2900 program is if you’re cheating and playing against someone who is also cheating, but only using the 2850 program.
To the club player, the real value in a program is not in how well it plays against other programs, but rather how well it teaches you to play the game.
Actually you need the strongest engine available for analysis — to get the best results in the shortest possible amount of time. Every little bit helps. Having said that, it would seem that Rybka is stronger in straight computer vs computer games against Fritz; but Fritz is better and faster at tactical analysis. So it is the choice of serious chess players looking for analytical help.
Everyone with any chess sense knows that you want and need the best chess software for analysis….when I made my post I clearly stated that I bought Fritz 11 and it is great. I love it.
Yet, it isn’t as strong as Rybka…and that’s just the plain truth.
Cheating?? I don’t even play at the club level…I don’t play correspondence chess…so how could I “cheat?”
I am merely stating a fact that everyone knows from the GM level down….Rybka is better than Fritz 11 and when you want to analyze a GM game….then what I do is let both Fritz and Rybka analyze them and just compare the differences.
Cheat? You’re speaking way out of bounds because you have no clue what I do nor what I use Fritz or Rybka for.
Ever thought I mith just be some guy with a trust fund that can afford to buy all this stuff and just sit and enjoy the programs?
I don’t play competitive chess. I don’t belong to a club. I’m not a member of the USCF any longer.
So, if I’m just analyzing games of the super GM’s…how could I “cheat?”
You really need to stop assuming too much. My reply that Rybka is stronger than Fritz 11 is just a fact. I didn’t make it up. Everyone knows it.
So, just go away and leave your childish assumptions in the strange “brain” that sits between your ears.
Not everyone plays competitive chess nor belongs to a club nor plays Correspondence chess where everyone cheats.
Some of us just really do have the money and time to have fun with all the new programs.
If “cheating” is the first thing that comes to your little mind…I truly feel sorry for you.
But, fact is: Rybka will eat fritz in a match every time. Everyone knows it and every GM I bet has a version of Rybka to analyze.
I’m not creating anything…it’s just the facts.
Live with it.
All chess software is a waste of money. McDonalds is more better. Thanks for asking me, I am the expert here.
Who cares about a Fritz 11, when she lost to Zappa?
Three technical Fritz-Rybka differences that matter for analysis:
(1) Whereas Fritz shows a nodes-per-sec. rate over 10x higher, Rybka often gets to high search depths (> 20) much sooner than Fritz, especially in endgames.
(2) Going from search depth d to d+1, Rybka seems to change its mind less often than Fritz. Wanting to quantify this held up my posting my Ejsmont cheating case data, before my Fall term got insanely busy—I hope to complete this by month’s end.
(3) Fritz (9 & 10 so I presume 11 too) understands immediately about Rook’s pawns and wrong-color Bishops even when its tablebases are switched off, whereas Rybka often has no clue even with them on. The following test case is extreme, but many others have frustrated using Rybka for my (still-ongoing) analysis of the Kramnik-Grischuk endgame, so I’ve used mostly DF10 plus some HIARCS 11 MP.
White: Kd4, Be2, Pe4,f5,g3,h4
Black: ke7, nh2, pf6,g4,h5
White to move, FEN = 8/4k3/5p2/5P1p/3KP1pP/6P1/4B2n/8 w – – 0 68
After 1.Ke3 Kd6 2.Kf2 Ke5, Fritz 9 w/o EGTs says “=” right away, since after 3.Kg2 Kxe4 4.Kxh2 Kxf5 Black cannot be prevented from …f5.f4 exchanging White’s g-pawn, and 5.Bd3+ Ke5 6.Bg6 f5 7.Bxh5 f4 is too late. But even with all 3-4-5 Nalimov EGTs switched on, my Rybka 2.3.2a gives “5.09” and higher. Eventually at some very high depth after 4…Kxf5, the eval falls, but is still over +2.00 (I got 2.11 just now). When I post this at the Rybka forums, this bit of “chess knowledge” will be #1 on my wishlist.
As a bonus, see if you can find White’s win with 1.Kd3!? Kd6 2.Ke3 Ke5 3.Bb5 after the initial skirmish 3…Nf3 4.Be8 Nh2 5.Bxh5 Nf1+ 6.Kf2 Nd2 7.Bxg4 Nxe4+. Then find how Black can defend against it prior to move 7. Unlike Susan’s standard stipulation for her puzzles, engine use is not only permitted, here it is practically mandatory…!
>>
Everyone with any chess sense knows that you want and need the best chess software for analysis….when I made my post I clearly stated that I bought Fritz 11 and it is great. I love it.
>>
If “Don’t waste your money” means “It’s great. I love it.” then you need to express yourself better.
And I repeat, that there are other things besides getting the best and quickest analysis that might recommend a program to the club player. Especially those under 2000 might well be advised to sacrifice a little analysis speed if they gain something else.
So why no by Chessmaster if you’re a “club player?” Club players benefit more from obtaining lessons from a human coach….Expert level or above…and not a computer program.
Computer programs are primarily (not this very important word for those that can really understand English: PRIMARILY) for analysis and Correspondence chess cheating. Everyone with a brain knows CC players cheat with computers though it is “officially” against USCF policy.
That is a joke in itself and not a single member of the current USCF regime will even approach this topic. When it is brought up the answer is: How can you prove someone used a computer in a CC game?
Well, either they’re are thousands of CC players that play at Rybka’s rating level or they cheat.
Pretty simple to prove.
Anyway, Fritz 11 is great. Rybka is better and that is that. People shouldn’t purchase Fritz 11 for anything other than analysis (which isn’t the best compared to Rybka) or for any other reason.
Spend your money wisely and either pay for a human coach to improve your game or, if you’re set on a program, then get Rybka.
But, if you have lots of money buy Fritz 11 and Rybka and all the others.
It’s a matter of choice. Fritz 11 is fine….but for the money….it’s not worth it. Get Rybka.
I don’t even understand why Fritz 11 is such a big deal???? it’s not stronger than Rybka, Shredder, Hiarchs, or Zappa….in a match.
So, who cares? No 2700+ human GM will ever play Rybka on even terms. Why? They couldn’t even get a draw.
For those that want to waste their money on Fritz 11 go ahead. it helps the economy. Please, by sever copies and spend lots of money on it.
12:02 PM wrote:
{when you want to analyze a GM game….then what I do is let both Fritz and Rybka analyze them and just compare the differences.
}
Yes. Telling the Fritz UI to run a CompareAnalysis with two engines is the best, if you can afford to wait twice as long.
When you see that they often disagree on the best move, it sure eliminates your impression that these engines all play “perfect” chess.
G
Grandmasters CAN get draws against Rybka. GM Benjamin [2574] just got 2 draws in 8 games. Conditions were even with the exception that a draw counted 1 point for him (‘draw odds’). Time control was 90m+30s and Rybka was running on a quad core pc.
See http://www.rybkachess.com/
Ah! The fault seems to lie only with Rybka 2.3.2a, which has been current since early July. Rybka 2.3.1 (both the regular and L(arry)K(aufman) versions) and Rybka 2.2n2 get the point here, even without tablebases, on my single-core laptop computer. I’ll have to try the multi-core versions on my 4-core machine when I catch the next break in my analysis.
Without tablebases, some engines also think the protected passed h-pawn after 4…Kxf5 5.Kg2 Ke4 6.Bxg4 hxg4 spells trouble for Black, but that K+P position is drawn without too much bother.
I have found that the better the engine, the less predictable it is concerning live GM games.
I am using a rather old engine during relays, and the reason is that I found it most accurate to actual GM play. That way I was able to beat most other Kibitzers fo the correct game-outcome-prediction, because they were using stronger engines that found better moves than the GMs OTB and therefore their engines evaluation did not count much during the game.
I do not know about analysis though. This was just once small thing I noticed.
I have to agree with someone else said before me: make up your mind about what you want and need and then buy the program.
Indeed, the “wrong color Bishop” bug against Rybka 2.3.2a has been noted on the Rybka forums already, Nov. 11–16, here.
On my old slow pc P4 3.2 duel, Fritz 11 takes Rybka 2.3.2a. Most recently a 10 game, 30 minute per game match. Fritz 5.5 Rybka 4.5.
I haven’t bought Fritz 11 yet, though I plan to. I’m interested in the Calculation Training feature – how do people accomplish the same thing now, without Fritz 11.
[For those not in the know, the feature allows you to record a variation from a given position, but keeps the pieces in place. This encourages you to develop board vision. When you submit the line for analysis, Fritz evaluates your chosen moves, offers alternatives, and then awards/subtracts points for each move in your line.]
Anyway – on the point about “missing analysis” – I use Fritz 9, and regularly analyze my games with a threshold set to 1/3 of a pawn. That means, when the player found the best move, there are no analysis lines. There are likewise no lines offered by Fritz when the move played was within 1/3 of a pawn’s evaluation of the best line Fritz found.
Ergo, plenty of moves with no analysis at all. I find that an enormous convenience, not a bug.
Could the default thresholds for Fritz 11 be causing this behavior?