Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / Anthony Cozzie speak !
1 2 3 4 5 6 Previous Next  
Parent - - By Harvey Williamson (*****) Date 2008-02-12 16:40
Don't you think Fruit did more than anyone to push 'our' competitors forward? Stefan and Mark manage to work full time on Computer chess - I don't see why being number 2 should stop Vas. He may not be able to afford to pay his team anymore of course.
Parent - - By lkaufman (*****) Date 2008-02-12 17:17
I guess you are right about Fruit. As for Stefan and Mark, I don't know what percentage of their time is spent on engine improvement, and what percentage is spent on other issues like GUI or opening book or website or clerical and sales tasks. Currently Vas and I both spend nearly all of our computer chess time on engine improvement.
Parent - By Harvey Williamson (*****) Date 2008-02-12 17:28
You could always help voluntarily like I do ;-) But it does seem to be the way with most commercial engines at the moment that they are only able to carry on with the goodwill of people willing to help them. i work full time for BBC News but I probably spend more time working on hiarcs each week - I hope my boss is not reading this. :)
Parent - - By Nelson Hernandez (Gold) Date 2008-02-12 17:00
What a frank yet dismal admission.  Only the #1 ranked program generates enough cash-flow to fund full-time improvement.  It's almost all-or-nothing.

May you always have several rivals nipping at your heels, as that is what is best for the hobby, and may you guys manage to keep the #1 rank for a very long time.  I do worry what would happen if your team had to disperse and work at full-time jobs.  After you've spent years optimizing the world's top chess engine almost anything else would seem incredibly drab and routine. 

Here's a question, what would Vas do for a living if he didn't have Rybka?  Bartender?  Nightclub bouncer?  :)
Parent - By Harvey Williamson (*****) Date 2008-02-12 17:06
Here's a question, what would Vas do for a living if he didn't have Rybka?  Bartender?  Nightclub bouncer?  :-)

I think we should open a Fish Restaurant together!?

But seriously all these guys at the top of the Computer Chess world could earn a lot more money out in the real world of programming.
Parent - - By lkaufman (*****) Date 2008-02-12 17:21
If Rybka lost top spot more or less permanently, I imagine that Vas would still work on Rybka, but would have to work on GUI and features and sales etc. quite a bit instead of just on engine strength, and I might have to retire.
Parent - - By George Tsavdaris (****) Date 2008-02-13 23:33

>If Rybka lost top spot more or less permanently, I imagine that Vas would still work on Rybka, but would have to work on GUI and features and >sales etc. quite a bit instead of just on engine strength, and I might have to retire.


If moon stops moving around the earth, then earth will lose its stability and i might have to move in mars. :)

Seriously, in the old days when someone on CCC had made a claim that there is an engine that can resist on a head to head match against the monster of those days: Shredder 9, i didn't believe him. Yet there was, Fruit 2.1. It was so much the prevalence and the beating Shredder 9.0 gave to its opponents that i couldn't believe it can lose.
The same effect happens with Rybka but this time magnified many many times. I know about the great law of life that things rise but inevitably there comes one moment that they set, but i still can't believe Rybka will set..... :)
Parent - - By NATIONAL12 (Gold) Date 2008-02-13 23:44
on this one i agree with you,i cannot believe any prog can touch rybka in next 10 years. my reasing is this vas talks to his customers every day and some of them are extremely brilliant brains which give feedback to vas all the time.
Parent - By oudheusa (*****) Date 2010-02-09 14:37
i am reading this post almost 2 years later and boy did things change!

this may well be in the top 10 of all time forum posts!
Parent - By lkaufman (*****) Date 2008-02-13 23:59
Rybka will get better and better, but there is always some chance that someone will make a great discovery that allows his program to leapfrog to the front.
Parent - - By Mark (****) Date 2008-02-12 17:06
Although this info is probably not available, it would be interesting to see how many copies of each program are sold each year.  Does the number 1 ranking really generate that many more sales than number 2 or 3? 
Parent - - By Harvey Williamson (*****) Date 2008-02-12 17:09
I suspect sales at the moment are:

1.Chessmaster
2.Fritz
3.Rybka

So being number 1 helps but the marketing of Ubisoft and CB also helps a lot.
Parent - By tomski1981 (*****) Date 2008-02-14 06:05
chessmaster wouldn't be #1 if it had competition. If you make a easy to use GUI, and add a few working/easy to follow training videos.. then you can say bye-bye to chessmaster. UBI isn't taking care of it's product... it's not even marketing it at all. it's just going off of repeat customers and word of mouth. chessmaster XI is just a redecorated CHX. i think that teaming up with a grandmaster would be also a plus (but he can't be boring). or.. even better (just a thought) someone from Chess.fm! like pete tamborou (sp?).. "openings for amateurs" style.

the possibilities are endless. (CHXI is helpful to amateurs, and 'serious' progs like rybka and other top dogs are best for very skilled players, but the intermediate level player [like me] is left out!!)
Parent - - By Highendman (****) Date 2010-02-09 21:18
As this thread resurfaced, I'm curious if your estimate of sales has changed since your guesstimate from two years ago?
Parent - By Harvey Williamson (*****) Date 2010-02-09 21:23
Recent sales of Rybka will almost certainly be > Fritz. Total sales by any one engine Chessmaster will still be way out in front.
Parent - - By lkaufman (*****) Date 2008-02-12 17:24
You have to have something special to sell a lot of copies. For Rybka, it is top strength. For other programs, it may be special features, special style, or tie-ins with other products.
Parent - - By NATIONAL12 (Gold) Date 2008-02-14 00:13
you appear to be very despondent in the last few days is your personal life affecting your normal judgements.
Parent - - By lkaufman (*****) Date 2008-02-14 00:30
No, I'm not despondent. Sorry if I gave the wrong impression. Work on Rybka is going very well, I'm in good health, no major problems. I'm just answering questions as truthfully as I can. I have no expectation of Rybka losing top place for a long time, but some people wondered what would happen if she did.
Parent - By turbojuice1122 (Gold) Date 2008-02-14 01:03
Haha, perhaps someone could liven things up and ask you what you think would happen if she became last in the rating lists among commercial programs :-D
Parent - - By George Tsavdaris (****) Date 2008-02-11 11:33

>Some are worth very little if anything, but some are really big (like one added today!).


When funs of Rybka hear something like this, they are dying to know more details about it :-)

Can you elaborate more on this "thing" you added today? :)
Parent - By lkaufman (*****) Date 2008-02-11 16:26
See response to debaser.
Parent - - By Banned for Life (Gold) Date 2008-02-09 00:06
See? I told you that Vas Rajlich (if that is his real name) is the Antichrist!
Parent - - By Anthony C (*) Date 2008-02-09 00:19
Yes, it seems I had a little bit too much fun with that paragraph, and it is distracting from the very concrete arguments in the rest of the post.
Parent - - By Carl Bicknell (*****) Date 2008-02-09 00:31
so....you're basically saying, I guess, that Rybka is called "Rybka" because it means little fish and it has the appearance of being a slow, high knowledge program when actually it's a fast, low knowledge program and, therefore, Rybka is an amusing name for it.

Rybka = little fish / little knowledge.

ah just some fun. I hope rybka is high knowledge though.
Parent - By NATIONAL12 (Gold) Date 2008-02-09 00:40
so why is rybka ahead of zappa in all rating lists if cozzie thinks so little of vas,s work.i lost the plot somewhere.
Parent - - By Anthony C (*) Date 2008-02-09 00:40 Edited 2008-02-09 00:46
OK, it is obvious I am not making myself clear, so let me state things in the most in the most boring language possible.

1. Strelka (Rybka 1.0) is an extremely heavily optimized but otherwise relatively standard (although obviously very effective) chess engine.

2. Later versions of Rybka almost certainly include more code (usually this is the way things work, it's why Tord is always rewriting).  Since I have not seen the code, I am not qualified to comment on this.

3. Later versions of Rybka are still very heavily optimized, based on the fact that Rybka's NPS hasn't slowed down *that* much since 1.0.

4. Who cares if Rybka is a high knowledge program or a low knowledge program? It plays chess.  The distinction is arbitrary.

5. Mr. Rajlich's made a long statement about how he does not optimize the code of Rybka.  Based on 1,2, and 3, I find this impossible to believe and insulting to my intelligence.  Which irritation prompted my previous paragraphs.

6. I am not criticizing the Rybka engine here. I am criticizing Mr. Rajlich.

anthony
Parent - By NATIONAL12 (Gold) Date 2008-02-09 00:50
we are obviously going to have to wait for vas reply to this but as he lives in europe like myself it could be another 12 hours.
Parent - - By SR (****) Date 2008-02-09 00:52
You claim that Rybka's code already is heavily optimized, but you do not believe that Vas does not currently work on optimizing the code. 

Maybe I have blind spot, but if Rybka's code already is near optimal (your claim), would it not be waste to try to optimize it even further?

I am not sure I understand your rage...
Parent - By NATIONAL12 (Gold) Date 2008-02-09 01:03
i think we should give everyone 18-24 hours to calm down , at the moment i feel like you.talk about a hornets nest.
Parent - By SillyFunction (**) Date 2008-02-16 08:37
Well, see this. Anthony is one of my idols.
But it is good to see that an idol can also be emotional :-)
It means that an idol is a human.
We're not talking to God.
And it's funny to have good time chatting with great humans.

PS: I think Anthony is truly scientific.
I really respect him for this.
But to be scientific may disturb someone sometimes.
my 2 cents :-)
Parent - - By RFK (Gold) Date 2008-02-09 01:31
Anthony,

Why do you portray having such disdain toward Vas? For all the members of this forum know, you may very well have a cordial relationship. But, your choice of words signal otherwise; " Twinkies " ? Is it that he is a graduate of M.I.T-the Harvard school of Technology!

sidserious-
Parent - By cma6 (****) Date 2008-02-12 02:22
Anthony is abusing Vas' courtesy: allowing a competitor to use the Rybka forum. If Anthony wishes to engage in ad hominen attacks, he should at least have the class to use his own forum for that purpose.

  It is a shame that Anthony is wasting his and our time with personal attacks, when he has created a fine program, Zappa. His time would be better spent working on Zappa than on posting on the Rybka forum.

At a minimum, Anthony needs to attend an ethics class at the University of Illinois. Rule 17a, he would learn, is "Do not personally attack your superior on his own forum."
Parent - - By billyraybar (***) Date 2008-02-09 22:04 Edited 2008-02-09 22:37
Didn't this guy teach an ethics class at University of Illiniois? I must be mistaken.
Parent - - By turbojuice1122 (Gold) Date 2008-02-09 22:40
I don't see that Anthony has really given 'ad hominem' attacks to anyone other than in a very minor way, so no reason to do it in reverse.  I will be most interested in Vas' answers when he returns from the weekend.
Parent - By NATIONAL12 (Gold) Date 2008-02-09 23:00
so will i and i imagine a lot of other people will be as well. i notice that the scandal magazine talk chess is waiting as usual to see vas reply, before commenting.
Parent - - By billyraybar (***) Date 2008-02-09 23:21
Coming to Rybka forum to call the owner of the Rybka forum a liar --pitiful efforts ... the jig is up ....we lower our opinion of him, etc etc?  If these were anonymous posts I would think they had come from a 10yr old brat.  I wouldn't call it ad hominem.  It's something born of jealousy.  "P.S. I have never understood why it's important that Rybka is a high knowledge program anyway.  If it works, who cares?" -- Obviously Anthony cares or else why would he come in this forum to argue (to himself? Us? Vas? ... Unclear) that Rybka is not knowledge based and ugly to boot? 
Parent - - By turbojuice1122 (Gold) Date 2008-02-10 00:35
He is arguing to try to disseminate truth.  It's obvious that the true node counts were hidden.  As far as we can tell, it sounds like everything he has said so far is correct (except the part about it being four times faster than anything out there, which was an obvious exaggeration).  Thus, we need to wait for the discussion to continue upon Vas' reply.
Parent - By BB (****) Date 2008-02-10 00:55

> it being four times faster than anything out there


I think he said "super-optimised to be 3 or 4 times faster" originally (and I jumped on him about the 32/64-bit nuance with bitboards). Later he said 3x faster when mentioning that this (alone) could explain the 80 ELO difference.
Parent - By Henrik Dinesen (***) Date 2008-02-10 08:24
True - not much to say, except that it sound to me like Anthony has done a longer study of the Strelka code than indicated int the readme we got from latest Zappa release, given it more thoughts than I initially was let to believe - anyway, apart from the at times hard wordings, the discussions shouldn't be harmfull :)
Parent - - By billyraybar (***) Date 2008-02-10 21:18
Is there 1 way that all chess engine programmers count nodes?  If the answer is no, then I would say the true node counts were not hidden.  Vas just counted them a different way.  Anhhow, I don't see why an end user would care about node counts. 
Parent - By Uri Blass (*****) Date 2008-02-10 21:38
There is no one way that programmers count nodes but most of the ways are close to each other.

number of nodes is basically the number of legal moves that you make in the tree that you search or at least should be close to it.

Strelka does not count nodes in that way and it increase the number of nodes everytime that it call searches but the difference is small
because in most cases you make a move after you call search or qsearch.
Parent - By Patricio (***) Date 2008-02-12 02:53
Completely agree.

For sure Anthony is a great member of this Forum, and is good to have his opinion here. Of course, we can or can not agree with him.

Furthermore, Vas replied very well Anthony's opinion.  So, no need to go deeper on this subject: "attacks". The technical discussion is the real important thing.

Best Regards.

Patricio.
Parent - - By Banned for Life (Gold) Date 2008-02-09 00:38
This is the most brilliant defense of trade secrets I've come across since the episode in Get Smart where they hide the top secret information in a place where nobody would ever look for it, the public library!
Parent - - By BB (****) Date 2008-02-09 03:29

> since the episode in Get Smart


I lived and breathed this show back in high school. Strangely, when I was in France once, some technology magazine wanted to have a nifty cover for its gadget issue, so it had Agent 86 speaking into his shoe phone [I couldn't parse the headline, nor am I sure if French technophiles would be GS-aware].
Parent - By Mark (****) Date 2008-02-09 03:50
Get Smart was one of my favorites also.  Many times I've wished out loud for the "cone of silence." The family never did figure out what I was talking about!
Parent - By NATIONAL12 (Gold) Date 2008-02-09 21:59 Edited 2008-02-09 22:48
i have never heard of the tv prog you are referring to, ie living in uk. in 1970 i had some very confidential information and went to the local library, obtained  a novel that was reasonably obscure that only an intellectual person would borrow, steamed open the inner cover and inserted a 35 mm negative of the info and destroyed the original. i was terrified that i would be burgled at the time but as it was of a technical nature there was no way i could remember the details.can i sue. there is a follow up to this,i got my money at the time and went to the libary  to retrieve the book. all deleted books were put on display so you could puchase them for next to nothing,well it was no longer in the main library and it was not on display in the deleted section, i enquired to the desk girl wehat had happend to the book, you could not make this up. she said that books that nobody read they sent to the local prison. as i said truth is stranger than fiction.
Parent - - By Sesse (****) Date 2008-02-09 01:25
I'm a bit intrigued about your comment about "[fooling] people who don't read C++". (Ignore for a moment that Strelka is C, not C++.) I can read C++ -- I've been using it since cirka 1997 (C before that, and other programming languages before that again), and I'm programming in it every day at work. Yet, the Strelka source is pretty much a mystery to me. There's lots of comments, but all of them are in Russian (a language I do not speak at all). The variable names are in English, but terse and full of jargon. (What is "entry->white_sheet_king[2]"? Or "QuadWhiteKing[square][Board->turn]"?) There's also a ton of magic constants which I recognize as "pretty" bit patterns (ie., I might guess that 0x3FBFBFBFBFBFBF80, being 11111110111111101111111011111110111111101111111011111110000000, in binary, somehow represents a rank and a file being masked out), but they make no intrinsic meaning to me at all.

Given all of this, I am not sure how you expect me nor any other casual observer to judge between you and Vas in who tells the truth about how Strelka works and to what degree this is normal for a chess engine (given that my only attempt at one ended up at about 1500 Elo on a fast machine :-) ). I'm sure that if you live and breathe chess engines all day long this just comes naturally (just as I can look at a piece of DSP code and pretty much instantly tell you what it does), but this just isn't that clear-cut to me unless I were to dig much deeper into the material than I've got the interest for.

/* Steinar */
Parent - - By BB (****) Date 2008-02-09 03:26
As someone who has written a bitboard engine, it took me about 2-3 hours (well, 2 or 3 boring seminars during which I looked at the code) to figure out almost completely what Strelka was doing. Indeed, Osipov missed a trick/simplification with the pawn shield/storm (these are the king-sheet stuff - there are three, for abc/def/fgh files) that would generate it from about 4 lines rather than 125 lookup entries.

If we are going to dispute VR's claim of no optimisation [and I would not have initiated this], I must agree with AC that the Strelka code has many tricks for speed (I noticed the sorting idea, but I think I recall correctly that boolean SEE is already in Fruit - I happened to like the PV/non-PV split from the coding standpoint, and maybe even a threefold PV/CUT/ALL split would be good, as extensions/reductions for these can be distinctive). However, as UB said, some of the speed is from data structures (for instance, save the attacked squares that are computed in eval, so as to avoid computing IN_CHECK and IS_LEGAL). Without further investigation, I would not be willing to say how much of this could be "normal" C code being optimised by the compiler. Incidentally, I get that Crafty has a higher NPS number (it has no POPCNT in mobility, while a Strelka profile will show eval time to be the majority, with most of eval being POPCNT [and possibly missed branch prediction]).
Parent - - By turbojuice1122 (Gold) Date 2008-02-09 04:55
I think that this "speed" question must go a bit beyond nps counts.  For example, Anthony has said twice now that Rybka is way faster than anything out there, but looking at Strelka, with the "unlocked" nps count, it's only about 50% that of Fritz 5.32 and about 30% that of Goliah Light 1.5.  Or perhaps Strelka still has stuff "locked"?
Parent - - By Uri Blass (*****) Date 2008-02-09 05:40
I think that you look at the wrong version of strelka and the wrong positions.
You should look at the 64 bit version of strelka and in this case strelka may be only slightly slower than Fritz5.32 in nodes per seconds in the opening position.

Probably the opening position is wrong and there are positions when strelka is faster than Fritz.
Here is an example for position that even the 32 bit version of strelka searches more nodes than Fritz5.32 on my hardware

37411 knodes in 14 seconds for strelka relative to 37398 knodes in 16 seconds

New game - Strelka 2.0 B
4k3/p6p/8/8/8/8/P6P/4K3 w - - 0 1


Analysis by Strelka 2.0 B:

1.Ke1-e2
  =  (0.22)   Depth: 1   00:00:00
1.Ke1-e2 Ke8-d7
  =  (0.03)   Depth: 2   00:00:00
1.Ke1-e2 Ke8-d7 2.Ke2-d3
  =  (0.09)   Depth: 3   00:00:00
1.Ke1-e2 Ke8-d7 2.Ke2-d3 Kd7-d6
  =  (0.03)   Depth: 4   00:00:00
1.Ke1-e2 Ke8-d7 2.Ke2-d3 Kd7-d6 3.Kd3-d4
  =  (0.09)   Depth: 5   00:00:00
1.Ke1-e2 Ke8-d7 2.Ke2-d3 Kd7-d6 3.Kd3-d4 a7-a5
  =  (0.15)   Depth: 6   00:00:00  2kN
1.Ke1-e2 Ke8-d7 2.Ke2-d3 Kd7-d6 3.Kd3-d4 a7-a5 4.a2-a4
  =  (0.09)   Depth: 7   00:00:00  4kN
1.Ke1-e2 Ke8-d7 2.Ke2-d3 Kd7-d6 3.Kd3-d4 a7-a5 4.a2-a4 h7-h5
  =  (0.15)   Depth: 8   00:00:00  7kN
1.Ke1-e2 Ke8-d7 2.Ke2-d3 Kd7-d6 3.Kd3-d4 a7-a5 4.a2-a4 h7-h5 5.Kd4-e4
  =  (0.09)   Depth: 9   00:00:00  12kN
1.Ke1-e2 Ke8-d7 2.Ke2-d3 Kd7-d6 3.Kd3-d4 a7-a5 4.a2-a4 h7-h5 5.Kd4-e4 Kd6-e6
  =  (0.15)   Depth: 10   00:00:00  22kN
1.Ke1-e2 Ke8-d7 2.Ke2-d3 Kd7-d6 3.Kd3-d4 a7-a5 4.a2-a4 h7-h5 5.Kd4-e4 Kd6-e6 6.Ke4-d4
  =  (0.09)   Depth: 11   00:00:01  41kN
1.Ke1-e2 Ke8-d7 2.Ke2-d3 Kd7-d6 3.Kd3-d4 a7-a5 4.a2-a4 h7-h5 5.Kd4-e4 Kd6-e6 6.h2-h4 Ke6-d6
  =  (0.15)   Depth: 12   00:00:01  70kN
1.Ke1-e2 Ke8-d7 2.Ke2-d3 Kd7-d6 3.Kd3-d4 a7-a5 4.a2-a4 h7-h5 5.h2-h3 Kd6-e6 6.Kd4-e4 Ke6-d6 7.h3-h4
  =  (0.09)   Depth: 13   00:00:02  229kN
1.Ke1-e2 Ke8-d7 2.Ke2-d3 Kd7-d6 3.Kd3-d4 a7-a5 4.a2-a4 h7-h6 5.h2-h3 h6-h5 6.Kd4-e4 Kd6-c5 7.Ke4-e5 h5-h4
  =  (0.15)   Depth: 14   00:00:02  407kN
1.Ke1-e2 Ke8-e7 2.Ke2-e3 Ke7-f6 3.Ke3-f4 h7-h6 4.a2-a4 Kf6-g6 5.Kf4-g4 h6-h5+ 6.Kg4-f4 Kg6-f6 7.h2-h3 a7-a6 8.Kf4-e4 Kf6-e6
  =  (0.15)   Depth: 15   00:00:02  1193kN
1.Ke1-e2 Ke8-e7 2.Ke2-e3 Ke7-f6 3.Ke3-f4 Kf6-g6 4.Kf4-g4 Kg6-h6 5.a2-a4 Kh6-g6 6.a4-a5 Kg6-f6 7.Kg4-h5 Kf6-g7 8.Kh5-g5 Kg7-f7
  =  (0.15)   Depth: 16   00:00:03  1399kN
1.Ke1-e2 Ke8-e7 2.Ke2-e3 Ke7-f6 3.Ke3-f4 Kf6-g6 4.Kf4-g4 a7-a5 5.a2-a4 Kg6-h6 6.h2-h4 Kh6-g6 7.h4-h5+ Kg6-f6 8.Kg4-f4 h7-h6 9.Kf4-e4 Kf6-e6
  =  (0.15)   Depth: 17   00:00:03  1819kN
1.Ke1-e2 Ke8-e7 2.Ke2-e3 Ke7-f6 3.Ke3-f4 Kf6-g6 4.Kf4-g4 a7-a5 5.a2-a4 Kg6-h6 6.h2-h3 Kh6-g6 7.h3-h4 h7-h5+ 8.Kg4-f4 Kg6-f6 9.Kf4-e4 Kf6-e6 10.Ke4-d4
  =  (0.09)   Depth: 18   00:00:04  3698kN
1.Ke1-e2 Ke8-e7 2.Ke2-e3 Ke7-f6 3.Ke3-f4 Kf6-g6 4.Kf4-g4 a7-a5 5.a2-a4 Kg6-h6 6.h2-h3 Kh6-g6 7.h3-h4 Kg6-f6 8.h4-h5 h7-h6 9.Kg4-f4 Kf6-e6 10.Kf4-e4
  =  (0.09)   Depth: 19   00:00:04  5547kN
1.Ke1-e2 Ke8-e7 2.Ke2-e3 Ke7-f6 3.Ke3-f4 Kf6-g6 4.Kf4-g4 a7-a5 5.a2-a4 Kg6-h6 6.h2-h3 Kh6-g6 7.h3-h4 h7-h5+ 8.Kg4-f4 Kg6-f6 9.Kf4-e4 Kf6-e6 10.Ke4-d4 Ke6-d6 11.Kd4-e4
  =  (0.09)   Depth: 20   00:00:04  9476kN
1.Ke1-e2 Ke8-e7 2.Ke2-d3 Ke7-d7 3.Kd3-e4 Kd7-e6 4.a2-a3 h7-h6 5.Ke4-f4 Ke6-f6 6.Kf4-g4 Kf6-g6 7.a3-a4 h6-h5+ 8.Kg4-f4 a7-a6 9.h2-h4 a6-a5 10.Kf4-e5 Kg6-f7 11.Ke5-f5 Kf7-e7
  =  (0.15)   Depth: 21   00:00:10  25893kN
1.Ke1-e2 Ke8-e7 2.Ke2-d3 Ke7-d7 3.Kd3-e4 Kd7-e6 4.a2-a3 a7-a6 5.a3-a4 Ke6-d6 6.Ke4-d4 Kd6-c6 7.Kd4-c4 Kc6-b6 8.Kc4-b4 h7-h6 9.h2-h4 h6-h5 10.a4-a5+ Kb6-c6 11.Kb4-c4 Kc6-d6 12.Kc4-d4
  =  (0.09)   Depth: 22   00:00:13  33871kN
1.Ke1-e2 Ke8-e7 2.Ke2-d3 Ke7-d7 3.Kd3-e4 Kd7-e6 4.a2-a3 a7-a6 5.a3-a4 Ke6-d6 6.Ke4-d4 Kd6-c6 7.Kd4-c4 Kc6-b6 8.Kc4-b4 h7-h5 9.Kb4-c4 Kb6-c6 10.Kc4-d4 Kc6-d6 11.h2-h3 a6-a5 12.h3-h4
  =  (0.09)   Depth: 23   00:00:14  37411kN

New game - Strelka 2.0 B
4k3/p6p/8/8/8/8/P6P/4K3 w - - 0 1


Analysis by Fritz 5.32:

1.a2-a4
  =  (0.03)   Depth: 1/1   00:00:00
1.a2-a4 a7-a5
  =  (0.00)   Depth: 2/2   00:00:00
1.a2-a4 a7-a5 2.h2-h4
  =  (0.03)   Depth: 3/3   00:00:00
1.a2-a4 a7-a5 2.h2-h4 h7-h5
  =  (0.00)   Depth: 4/4   00:00:00
1.a2-a4 a7-a5 2.Ke1-d2 Ke8-d7 3.h2-h4
  =  (0.03)   Depth: 5/6   00:00:00  1kN
1.a2-a4 a7-a5 2.Ke1-d2 Ke8-d7 3.h2-h4 h7-h5
  =  (0.00)   Depth: 6/6   00:00:00  2kN
1.a2-a4 a7-a5 2.Ke1-d2 Ke8-d7 3.h2-h4 h7-h5 4.Kd2-d3
  =  (0.00)   Depth: 7/8   00:00:00  5kN
1.a2-a4 a7-a5 2.Ke1-d2 Ke8-d7 3.h2-h4 h7-h5 4.Kd2-d3 Kd7-d6
  =  (0.00)   Depth: 8/9   00:00:00  13kN
1.a2-a4 a7-a5 2.Ke1-d2 Ke8-d7 3.h2-h4 h7-h5 4.Kd2-d3 Kd7-d6 5.Kd3-d4
  =  (0.00)   Depth: 9/11   00:00:00  31kN
1.a2-a4 Ke8-f7 2.a4-a5 a7-a6 3.Ke1-e2 Kf7-e6 4.Ke2-d3 Ke6-d5 5.Kd3-e3 Kd5-e5
  =  (0.03)   Depth: 10/12   00:00:00  69kN
1.Ke1-d2!
  =  (0.06)   Depth: 10/12   00:00:00  105kN
1.Ke1-d2 a7-a6 2.a2-a4 h7-h5 3.h2-h4 Ke8-d7 4.Kd2-e3 Kd7-e6 5.Ke3-d4 a6-a5
  =  (0.00)   Depth: 11/13   00:00:00  238kN
1.Ke1-d2 Ke8-f8 2.a2-a4 Kf8-e7 3.Kd2-c3 h7-h5
  =  (0.06)   Depth: 12/14   00:00:00  491kN
1.Ke1-d2 Ke8-f7 2.Kd2-d3 Kf7-e7 3.Kd3-e3
  =  (0.03)   Depth: 13/15   00:00:00  960kN
1.Ke1-d2 Ke8-d8 2.Kd2-e3 Kd8-e7
  =  (0.06)   Depth: 14/17   00:00:01  1619kN
1.Ke1-e2!
  =  (0.09)   Depth: 14/17   00:00:01  1798kN
1.Ke1-f2!
  =  (0.13)   Depth: 14/17   00:00:01  1982kN
1.Ke1-f2 Ke8-e7 2.Kf2-g3 a7-a5 3.Kg3-g4 a5-a4 4.Kg4-g5 Ke7-f7
  =  (0.16)   Depth: 15/18   00:00:01  2430kN
1.Ke1-f2 Ke8-f7 2.Kf2-g3 Kf7-g7 3.Kg3-h3 Kg7-f7 4.Kh3-h4
  =  (0.16)   Depth: 16/20   00:00:01  3434kN
1.Ke1-f2 Ke8-f7 2.Kf2-g3 Kf7-g6 3.Kg3-g4 a7-a5 4.a2-a4 Kg6-h6 5.h2-h3
  =  (0.06)   Depth: 17/20   00:00:02  5082kN
1.Ke1-e2!
  =  (0.09)   Depth: 17/20   00:00:02  5314kN
1.Ke1-e2 Ke8-e7 2.Ke2-d3 Ke7-d7 3.Kd3-e3 Kd7-e7 4.Ke3-f3
  =  (0.06)   Depth: 18/21   00:00:03  8290kN
1.Ke1-d2!
  =  (0.09)   Depth: 18/21   00:00:03  8581kN
1.Ke1-d2 Ke8-d8 2.Kd2-e3 Kd8-e7 3.Ke3-f4 Ke7-f6 4.a2-a3 a7-a6 5.Kf4-g4 a6-a5
  =  (0.06)   Depth: 19/22   00:00:05  11259kN
1.Ke1-d2 Ke8-d8 2.Kd2-e3 Kd8-e7 3.Ke3-f3 Ke7-f6 4.Kf3-f4 Kf6-g6 5.Kf4-g4
  =  (0.03)   Depth: 20/23   00:00:06  13490kN
1.Ke1-d2 Ke8-d8 2.Kd2-e3 Kd8-e7 3.Ke3-f3
  =  (0.03)   Depth: 21/24   00:00:07  15928kN
1.Ke1-d2 Ke8-d8 2.Kd2-e3 Kd8-e7 3.Ke3-d4 Ke7-d6 4.h2-h3 h7-h6 5.Kd4-e4 Kd6-e6
  =  (0.03)   Depth: 22/25   00:00:09  20873kN
1.Ke1-d2 Ke8-f7 2.Kd2-d3 Kf7-g6 3.a2-a4 Kg6-f5 4.Kd3-c4 h7-h5 5.a4-a5 Kf5-g4
  =  (0.00)   Depth: 23/27   00:00:12  28100kN
1.Ke1-d2 Ke8-d8 2.Kd2-e3 Kd8-e7 3.Ke3-d4 Ke7-d6 4.h2-h3 h7-h6 5.Kd4-e4 Kd6-e6
  =  (0.00)   Depth: 24/29   00:00:14  32972kN
1.Ke1-d2 Ke8-d8 2.Kd2-e3 Kd8-e7 3.Ke3-d4 Ke7-d6 4.h2-h3 h7-h6 5.Kd4-e4 Kd6-e6
  =  (0.00)   Depth: 25/31   00:00:16  37398kN

(,  09.02.2008)

Uri
Up Topic Rybka Support & Discussion / Rybka Discussion / Anthony Cozzie speak !
1 2 3 4 5 6 Previous Next  

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill