Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / Freestyle Openings
- - By lkaufman (*****) Date 2007-06-23 16:52
     Two things I notice about the openings in the freestyle event: No one opens 1d4, and no one answers 1e4 with 1...e5, even though these two opens are extremely common at the highest levels of human chess. Can anyone knowledgeable about freestyle chess explain this? Is it just a preference of the particular individuals in the finals here, or is this generally the case in freestyle chess? Is there a feeling perhaps that Black can draw too easily against 1d4 (if so, how?), and regarding 1e4 e5 is there some reluctance to play the Marshall Gambit even though top GMs consider it completely sound?
Parent - - By Lukas Cimiotti (Bronze) Date 2007-06-23 16:59
maybe the intention is to throw the opponent out of his book as quick as possible ?
Parent - - By lkaufman (*****) Date 2007-06-23 17:16
Sorry, I don't follow you. How is 1e4 or playing the Najdorf "throwing the opponent out of book" any more than 1d4 or 1e4 e5?
Parent - By Lukas Cimiotti (Bronze) Date 2007-06-24 20:19 Edited 2007-06-24 20:22
Sorry, Larry,

i was doing too many tasks at a time - please forget about my crappy answer :(  , please ;)

in fact there are many recipes in the net how to make a good book - and most of them are limited to only a few openings

I donĀ“t find them all anymore - but here is one:

That might be the explanation
Parent - - By Harvey Williamson (*****) Date 2007-06-23 17:04
Hi Larry,

Not all - here is a nice win from the qualifying round with the Berlin which you suggest in your book. This is played by GM Tony Kosten. The notes are by him and its from Tony was assisted by Hiarcs but this Berlin idea is his!

[Event "Rated game, 60m + 15s"]
[Site "CSS Freestyle Tournament"]
[Date "2007.06.01"]
[Round "?"]
[White "Big MC"]
[Black "Tony Kosten"]
[Result "0-1"]
[ECO "C67"]
[Annotator "TonyK"]
[PlyCount "123"]
[EventDate "2007.06.03"]
[TimeControl "3600+15"]
[Source ""]
[SourceDate "2007.06.13"]

1. e4 e5 2. Nf3 Nc6 3. Bb5 Nf6 4. O-O Nxe4 5. d4 Be7 6. Qe2
Nd6 7. Bxc6 dxc6 $5 {I started looking at this idea of Trifunovic's after
seeing Nigel Davies' Feb '05 update, the great thing about this line is that
the 'theory' gives an inferior variation for White!} (7... bxc6 8. dxe5 Nb7 {
is the mainline.}) 8. dxe5 Nf5 9. Rd1 Bd7 10. e6 $6 {Tempting, but} (10. Nc3 {
is clearly better.}) 10... fxe6 11. Ne5 Bd6
12. Qh5+ g6 13. Nxg6 Ng7 14. Qh6 Nf5 15. Qh3 {
Playing for the loss! the major problem of this line (as a winning proposition)
is } (15. Qh5 Ng7 16. Qh6 Nf5 {
1/2-1/2 Simon, O (2388)-Kosten, A (2542)/Cap d'Agde 2006.}) 15... Rg8 16. Qxh7 Rg7 17. Qh5 Qf6 18. Qh8+ $2 {
A computer move! Still, it is important to know why this loses.} (18. Ne5+ {
is the best move, } Ke7 19. Ng4 {when} Qh4 $5 {
'may be worth examining' (Nigel Davies)} 20. Qxh4+ Nxh4 21. h3 e5 22. f3 Bc5+ 23. Kf1 Be6 24. b3 {(else ...Bc4+) } e4 $5 25. Nd2 exf3 26. Nxf3 Nxf3
27. gxf3 {is very acceptable for Black, with his bishop pair
and open files. }) 18... Kf7 19. Qxa8 Qxg6 {
Although White's 18th move is a novelty, I had actually wondered about this
possibility and had analysed the position at home some time ago.
} 20. g3 Qh5 {Intending the simple ...Rh7. } 21. Nc3 Rh7 22. Qxa7 {
The queen hastens to return. In my analysis I had concentrated on}
(22. h4 Nxh4 23. Qd8 { when Black wins by } Ng6 $1 (23... Nf3+ 24. Kf1
Be8 {is less clear }) 24. Qxd7+ Be7 25. Kf1 Ne5 {
(the point) and White can resign, as } 26. Qd4 Qh1+ 27. Ke2 Qf3+
28. Kd2 Bg5+ {leads to mate.}) (22. Rxd6 {is also possible, but after} Qxh2+
23. Kf1 cxd6 24. Qxb7 Qh1+ 25. Ke2 Nd4+ 26. Kd3 Kg6 {
White is hardly going to last very long, } 27. Kxd4 $2 c5+ {wins the queen.})
22... Nxg3 $1 {I had a long 'think' here, as} (22... c5 {is also tempting, } 23.
Qxb7 Nd4 24. Rxd4 cxd4 25. h4 dxc3 {and I suppose Black should win
comfortably as the 3 white pawns can hardly compensate the d7-bishop. Still, I
followed the old saying: "a queen is a queen"! }) 23. fxg3 Bc5+ 24. Qxc5 Qxc5+ 25. Kg2 Kg8 {I imagine this could be quite a
difficult position to win against a computer, if I didn't have the help of
another computer!} 26. Rd3 Qh5 {Other moves are also strong,
but I wanted to play ...c5 and ...Bc6+ when my pieces would all be working
well, and so I had to move the queen from c5. } 27. h4 c5 28. Bg5 Bc6+ 29. Kg1 Qg4 30. Rf1 Rd7 {
By swapping rooks Black no longer has to worry about checks. }
31. Kh2 Rxd3 32. cxd3 Qd4
33. a4 Qxd3 {
Having captured the d-pawn it remains to push the passed e-pawn.
} 34. Rf2 e5 35. Bd2 Qd4 (
35... e4 36. Nd1 {stops the pawn.}) 36. Kg1 b6 {
} 37. Bc1 Qg4 38. Kh2 Bd7 39. Nd5 (39. Kg1 Kh7) 39... Qh3+
40. Kg1 Kh8 (40... Qxg3+ $4 41. Rg2) 41. Rf8+
Kg7 42. Rf2 Bxa4 (42... Bc6
43. Rh2 Qg4 44. Ne3 Qxa4 {was also good. }) 43. Rh2 Qg4 44. Ne3 Qe4 45. Rf2 Bc6 46. Ng2 Qb1 47. Rf1 Qc2 48. Ne3 Qe2 {Might as well keep the white
pieces passive before pushing the extra pawns. } 49. Nf5+ (49. g4
Bf3) 49... Kh7 50. Ne3 Bb7 51. g4 Bc8 52. Ng2 Bxg4 53. Bg5 Qxb2
54. Rf7+ Kg6 55. Rxc7 Qb1+ 56. Kf2 (56. Kh2 Bf3 57. Rd7 Bxg2
58. Kxg2 c4) 56... Qc2+ 57. Kg3 Qe4 58. Ra7 Qf3+ 59. Kh2 Qh3+ 60. Kg1 Bf3 61. Ra2 c4 62. Rf2 0-1
Parent - - By lkaufman (*****) Date 2007-06-23 17:19
I've known for a while that 10e6? is not good, but after 10Nc3 I think White is better, so I guess this line is basically a gamble/bluff. My real question is why the freestylers don't play what the top pros play now, namely the Marshall. 
Parent - - By Harvey Williamson (*****) Date 2007-06-23 17:36 Edited 2007-06-23 17:38
Hi Larry,

It is certainly true that most, computer, games on Playchess and in tournaments start 1.e4 c5.

I think rightly or wrongly it is presumed that symetrical openings between strong engines give a high draw % Also any lines where Queen's come off in book early seem drawish between the top engines.  I do not have any data on this but would like to see it if anyone has some.

I think a lot of engine Chess at the top level is now down to opening books. I know that if you read the readme about Jeroen's latest release book that he has used a lot of games from Playchess. I know that a large proportion of these games are 1.e4 c5. There are certainly many traps to be set by the Playchess book cookers in these lines.

Of course there are novelties and traps to be found in many other openings as well!

Parent - By lkaufman (*****) Date 2007-06-23 17:46
It's true that 1e4 e5 gives a higher draw probability than the Sicilian. But in the finals of an event where all the players are using supercomputers, surely a draw with Black is a good result?
Parent - - By Banned for Life (Gold) Date 2007-06-23 17:24
I think Team Rybka opened 1. d4 in most of the previous free style events. Vas has indicated that the freestyle openings are Krasenkow's domain. Maybe Vas will let us know the rational for switching away from this.
Parent - By lkaufman (*****) Date 2007-06-23 17:44
Sure, but my question is general, not about the preferences of a particular individual.
Parent - - By Fulcrum2000 (****) Date 2007-06-23 17:54
I will quote Alkelele on this one :
"Not to mention 1.d4. It is a complete mystery to me why even GMs keep blundering with that move. Already Fischer told us that 1.e4 is best."
Parent - - By Harvey Williamson (*****) Date 2007-06-23 18:02 Edited 2007-06-23 18:19
I challenge Fischer to a match where Hiarcs will open 1.d4 in every game with White. I am sure Vas/Larry would be happy for Rybka to make the same challenge.
Parent - - By lkaufman (*****) Date 2007-06-23 18:08
Actually I would not want to waste time playing a match in which Rybka has White against Fischer or any other human, as with either 1 e4 or 1 d4 Rybka would win nearly every game. Only with the White pieces would the world's best players have even drawing chances. Any future Rybka-human matches must at least give them the White pieces and other advantages (as in our upcoming match with Ehlvest) to be at all interesting.
Parent - - By Harvey Williamson (*****) Date 2007-06-23 18:15
At the moment there is no proof that an engine which is best in engine v engine play will be best in engine v human. I doubt we will ever get enough games to prove which engine is best v humans.
Parent - - By lkaufman (*****) Date 2007-06-23 18:44
There is no "proof", since as you say there are not enough games. There is however strong evidence at least that Rybka is much stronger against humans than Fritz and other programs which have played recent matches with top programs, because Rybka successfully spotted Ehlvest a pawn handicap, whereas other programs until recently were just making even scores in matches with top players with no handicap. I would also point out that it's hard to even talk about relative strength against humans now, since I would guess that any of the top few programs running on a Quad would be virtually unbeatable by humans, it's just a question of draw percentage, which in turn is heavily dependent on opening choices. If you would like to try to make a comparison, I recommend that you have Hiarcs play handicap matches against GMs under the same conditions that we have Rybka play them. It isn't very expensive, as GMs under 2650 don't command large appearance fees.
Parent - - By Harvey Williamson (*****) Date 2007-06-23 18:55 Edited 2007-06-23 19:06
because Rybka successfully spotted Ehlvest a pawn handicap??

What does that prove - nothing.

The draw scores you quote were played several years ago on, relatively, inferior hardware until the recent Fritz v Kramnik match. The draws were also played in public against very strong GM's.

Yes I agree any of the top programs would probably win on something like a QX6700 quad core.

But some of them like Hiarcs play much more exciting and aggressive Chess to watch. ;-)

If you want to set up an event where we can try to prove which is best v strong GM's it sounds interesting but very unfair on the poor GM facing 2 silicon monsters!
Parent - - By lkaufman (*****) Date 2007-06-23 19:20
     The pawn odds result against Ehlvest was clearly the most impressive result for a computer against a grandmaster to date, except for Hydra-Adams, by any reasonable standard. If we call the handicap one class (which I believe is a fair estimate according to our tests), the performance rating for Rybka was about 3000. Now it's possible that Hiarcs or Zap could win a match against him or another 2650 player under the same conditions, but the burden of proof is on them to demonstrate a 3000 performance in an eight game match against a GM to claim equality, and a handicap match is the only affordable way to do this. I'm not suggesting we have to have the same GM play both (or all) programs for a test, just go by performance rating, with some fair estimate for the rating value of any handicap given.
     As for exciting and aggressive chess, if you have seen the Ehlvest handicap games, you would have to admit that some of Rybka's wins were among the most exciting, finest attacking wins on record vs. human GMs. I'm sure Morphy, Alekhine, or Tal would have been proud of a couple of those wins.
Parent - - By Harvey Williamson (*****) Date 2007-06-23 19:22
I don't think a pawn odds match is clearly anything. Rybka has to prove itself v a Super GM 1st at normal Chess. I look forward to seeing such a match.
Parent - - By lkaufman (*****) Date 2007-06-23 19:27
Well, we have the Ehlvest match next month. If you want a 2700+ GM, you will be disappointed, as they want too much money.
Parent - - By Harvey Williamson (*****) Date 2007-06-23 19:27
Indeed they do!
Parent - - By lkaufman (*****) Date 2007-06-23 19:42
That's why I'm suggesting handicap matches to you for Hiarcs, you don't need to play the super-expensive top guys to make your point. If you don't like pawn odds, go with our book-odds format of next month (Ehlvest all games White, double time, no book for Rybka past move three). I'm really trying to help you here, it's a great way to prove how good your program is against humans "on the cheap". I think Hiarcs deserves at least to be considered among the top three, not on a par with Fritz.
Parent - By Harvey Williamson (*****) Date 2007-06-23 19:49
That is a lot more appealing than pawn odds to me and I think a more interesting match. I will watch Ehlvest v Rybka 2 with great interest!

I just don't like claims that because of 1 pawn odds match Rybka is clearly best v humans this is spin pure and simple. But it's the Rybka forum so you are free to claim whatever you want. Pawn odds is something we may consider, in the future, it is not something that interests me.

Best Wishes,

Parent - - By Banned for Life (Gold) Date 2007-06-23 20:52
All of the writeups of the first Rybka-Ehlvest match at pawn odds suggest that Ehlvest's performance was decidedly better in the second half of the match, compared with the first half. Of course this reduces the number of games to an even less statistically significant number, but if one believes that Ehlvest's performance did improve during the match, it would be hard to conclusively state that this learning curve might not continue and allow Ehlvest to beat Rybka in this format, given enough practice games.

Parent - By lkaufman (*****) Date 2007-06-23 21:32
Certainly that is true, Ehlvest might indeed have a chance in a rematch. Of course, I also learned things that could improve the "minibook" used in the match, and as you know Rybka 2.3.2 is much stronger than 2.3.1 was. Anyway, unless something changes, the next pawn odds match will be with Joel Benjamin.
Parent - - By Uri Blass (*****) Date 2007-06-23 20:21
I think that we cannot compare results with different conditions.

It is possible that one program is relatively better against humans with pawn handicap and another program is better against humans in normal conditions.

Parent - By lkaufman (*****) Date 2007-06-23 20:35
Of course this is possible. It is just not very likely to be true to a substantial degree, unless someone has devoted massive effort to programming for pawn-odds chess, which no one would do under current conditions.
Parent - - By turbojuice1122 (Gold) Date 2007-06-23 23:38
"The pawn odds result against Ehlvest was clearly the most impressive result for a computer against a grandmaster to date, except for Hydra-Adams"

Let's not forget that Fritz basically spotted Kramnik perhaps a 200 elo point advantage with that opening book show and everything, since it allowed Kramnik basically to prepare much deeper than the Fritz book and come out of the openings with a decent advantage.
Parent - - By lkaufman (*****) Date 2007-06-24 01:25
Fritz definitely spotted Kramnik an advantage, but it was nowhere near the pawn handicap Rybka spotted Ehlvest, as I'm sure can be verified by comparing the scores of the first positions in each game where both players took time. If we're calling the pawn handicap 200 points, then I would call the Fritz/Kramnik handicap roughly 100.
Parent - - By turbojuice1122 (Gold) Date 2007-06-24 02:19
I guess I'm getting into a bit of a speculative realm here, but while I agree that the evaluation of, say, a quad Rybka/Zap/Hiarcs/Naum/Fritz left running for a few hours on the first positions in each set of games where both players took time would be greater for Ehlvest than for Kramnik, I would tend to think that the positions that arose in the Ehlvest games were much more dynamic.  Not only were there more chances for Ehlvest to make mistakes due to being in the types of positions that required very precise play due to their gambit/initiative nature, it would seem that Ehlvest knew far less about what to expect from Rybka than what Kramnik knew to expect from Fritz because he could often perform fairly accurate preparation 20-30 moves deep due to knowledge of exactly how Fritz would respond to most of his opening choices, and problems only arose in positions where several moves would have the save evaluations, in which Kramnik could still trace ahead and get a feel for the types of positions that Fritz would play into, even if he had to take time to figure out new moves during the actual game.  Being an international master, you would have a good vantage point on this--would you agree with the above logic?
Parent - - By lkaufman (*****) Date 2007-06-24 03:59
Yes I would. This was not primarily due to the terms of the Fritz-Kramnik match, but to the failure of the Fritz team to take them into account. If I were in charge of the Fritz team, I would have selected highly unusual openings every game, something like 1d4 d5 2a3 for example, to throw both players out of book as early as possible. This is not optimum strategy normally, but under the strange match rules it would have minimized Kramnik's advantages.
Parent - - By turbojuice1122 (Gold) Date 2007-06-24 13:11
That's true, but Kramnik would have been ready for that, because the Fritz team kind of doomed themselves in the contract (otherwise, Kramnik would not have played) by allowing also the rule that the operator and team cannot change Fritz's opening preferences after they give Kramnik the particular version of Fritz used along with the opening book that would be used.  Kramnik would have had 20 moves of preparation for 1.d4 d5 2.a3 if the Fritz team was going to use that in the match.
Parent - By lkaufman (*****) Date 2007-06-24 15:10
Oh, I didn't know the rule included "opening preferences"! Then I retract my retroactive advice!
Parent - By Vasik Rajlich (Silver) Date 2007-06-25 14:27
There is also no proof that Kasparov can beat me head-to-head :)

Parent - - By Uri Blass (*****) Date 2007-06-23 20:28
This is only an opinion.
You may be right but it is not proved.

GM's can learn and I am not sure that they cannot learn an efficient strategy to get draws against rybka even when they have the black pieces.
I am not sure if it is a question of rating and it is possible that some 2700 players may fail in drawing most of the games with black against rybka when a player with rating of 2500-2600 may succeed in the same task.

Parent - By lkaufman (*****) Date 2007-06-23 20:39
It is only "proved" with respect to Hydra; it's last five games with White vs. 2700 GMs were 5-0. Top GMs can't reliably draw with Black against each other, they often struggle, so I think it's hopeless against a 3000+ program. But of course some GM might find a "trick" to drawing with computers, then we would need to take countermeasures.
Parent - - By lkaufman (*****) Date 2007-06-23 18:10
Clearly 1e4 offers better winning chances, but also more losing chances than 1d4. With 1d4 you can play for the win "for free", without much risk of losing (if you are using Rybka of course), whereas with 1e4 you must have massive opening preparation. I guess these guys feel that the draw risk with 1d4 is just too high?
Parent - - By Nelson Hernandez (Gold) Date 2007-06-24 02:37
This is a very interesting question you've posed, Larry, and I think it goes beyond simple chess theory. 

The Playchess server is the biggest chess server in the world that encourages engine matches.  ICC is a worthy contender, but in terms of sheer volume there is no comparison.  I can't tell you exactly how many engine games are played there, but it has to be in the thousands every day, many times what you get on ICC.

Looking over games I collected in the early days of Playchess, back in 2002-2003, the opening fashion was quite different from what it is today.  Back then ECO B22 was very popular, matching the favorite line in most all-human books.  By early 2004 it had shifted and B66 was popular.  In late 2004 and 2005, B80 was most popular.  However by late 2005, B90 surpassed B80 and ever since it has increased to the point where it now must represent up to 15% of all games played in the engine room.  So fashion does change, but it requires some catalyst--usually several key players making a determined effort to play a particular opening.  Then the followers join in and you have a new trend.  So this is the larger context of what has gone on in recent years.

Now to some of the dynamics that produce this phenomenon.  A key factor is the prevalence of .ctg books among engine room players, which strongly skew towards the most popular (highest 'N') moves.  But beyond this developing a strong .ctg book is just damned hard work--ask Jeroen--and there are all kinds of practical issues when the book grows beyond a certain size.  These factors tend to discourage the development of sophisticated books and encourages the development of simpler, usable, time-effective books.  A byproduct of this is that the games that are loaded tend to be Playchess games, and these recursively play similar lines over and over unless somebody does something to stop it.  Most don't bother.

So against this backdrop you now must factor in the way people think in Freestyle.  It is very hazardous to generalize because there are many different skill levels; people have different resources, insights and amounts of time to allocate to the hobby.  However it is fair to say that most Freestylers are engine room players, and those that aren't have built opening repertoires that are at least in part based on engine room games.  So right now it stands to reason that most people have books that incline naturally toward B90 lines--hence the plethora of Sicilian Najdorfs in qualifiers.

However once in finals the game changes.  You know who your opponents will be and you know they are all tough bastards.  At that point people start thinking about cutting down their risks with black and trying something clever with white, either a strange variation or more often an improvement on a past line.  French Defense for instance is far more popular in Freestyle than in the engine room because it is perceived to be drawish.  At that level, most people are just trying to survive with the black pieces.

As for 1.d4, I think the general feeling is that it produces less chance of a clear win for white and is more labyrinthine than 1.e4, where the player at least has the comfort of a large body of theory to fall back upon. 
Parent - - By lkaufman (*****) Date 2007-06-24 04:14
     Thanks for your insights. I can certainly agree that 1e4 offers more chances of a clear win than 1d4, though also more risk of loss. But the choice of the French to reach a draw strikes me as a bit strange; generally speaking the Sicilian and French are played to win, while the Caro and 1...e5 are chosen when a draw is okay. The top GMs are in almost complete agreement it seems that the safest way to equalize against 1e4 is to aim for the Marshall, and so my #1 question would be why hardly any of the freestylers choose to copy them?
     Maybe you could shed some light on the level of play of the human players involved in the finals. I know the level of Vas and Krasenkow of course, but I don't know about the others. Are some of them really weak players who just rely on monster hardware, or are they all at least master or near-master level players? I ask because the opening choices of weak players might simply reflect ignorance, but if master level players choose to disregard the opinions of the superstars they probably have some reason. If the best openings for humans are really not the best openings for engines, that is something worth discussing and understanding.
Parent - - By Nelson Hernandez (Gold) Date 2007-06-24 12:26
Larry, I really can't speak for everyone.  Every player/team is unique.  All I can say is that most Freestylers are engine players.  Maybe 60-70% of the nicks playing are recognizable as such.  Within each entity there are different strengths, weaknesses and playing philosophies into which I have as little insight as you.  However, what skews opening play for many are the phenomena I alluded to earlier: shifting fashion, the limitations of .ctg books and the aversion of most people to get into time-intensive book preparation (or perhaps their lack of skill at it).  There are noteworthy exceptions, of course.

I think it is fair to assume that titled players are the exception and most Freestylers are mediocre OTB players.  What I think you may not fully appreciate, Larry, is that engine players have a drastically different way of approaching the game than OTB players.  Their decision-tree is structured along totally different principles.  Conventional theory may not enter into their considerations at all.
Parent - By Vasik Rajlich (Silver) Date 2007-06-25 14:35

you ask some interesting questions. I did some interviews with many of the top freestylers, have a look here:

Many of your questions are answered there.

There is definitely some sort of "fashion" in freestyle opening choices, just like there is "fashion" in top-level human chess openings. You see the same variations hammered for a while, blood is shed, conclusions are reached, and then everyone shifts to the new battlefield.

Parent - - By revengeska (**) Date 2007-06-24 05:16
I think you tend to generalize too much, Nelson.  Our team isn't represented in the playchess engine room, and we don't normally prepare openings either.  In general we're pretty lazy.  However, I'll agree with your last paragraph and last sentence.
Parent - By Nelson Hernandez (Gold) Date 2007-06-24 11:34
Always glad to interact with people who confess to laziness.  I find them the best thinkers, always trying to figure out how to accomplish something with the least effort!
Parent - - By revengeska (**) Date 2007-06-24 05:13

If you want to know why our team doesn't often play 1..e5 in Freestyle, take a look at Zorchamp-revengeska in the 2nd Freestyle, and either Nebula-revengeska or Poweronoff-revengeska(I forget which one we played e5).

As for d4.. d4 is too boring.  We'll have better scope to outplay weaker players with e4.
Parent - - By M ANSARI (*****) Date 2007-06-24 06:44
d4 usually requires much better understanding of chess fundementals and is better suited for long manouevers where minute positional advantages can be slowly built up.  Since most Freestyle players rely heavily on engines and strong hardware ... an evaluation that goes from say +.2 to +.25 can hardly be called exciting.  But I do believe that a strong GM with excellent positional understanding and even more importantly has the ability to properly utilize engines in aiding his analysis will have very very good success with d4 ... maybe even better than e4.  If you look at Flyingfatman vs. Cato ... this is a perfect example of exactly that.  Although white didn't win, it was a very close affair and I would say that with a little more time white could have had a nice win against an opponent that had not lost a game ever in freestyle.  Engines do not have long term strategy and look for concrete tactical variations ... therefore if you don't know what you are doing ... d4 could actually be a recipe for disaster, especially in the later rounds when you have some very strong chess players involved.
Parent - By nuff (**) Date 2007-06-24 11:42
Chess appears to be the loser if the freestylists are all going for tactical games. I would expect centaur skills to be evident. By the same token I'm surprised at the decision by the Fritz/Shredder/Mission Control in the poisoned pawn. Shredder played a funny opening in its win against Rybka last year and I thought they would go along the same route.
Parent - By revengeska (**) Date 2007-06-25 20:45
You're quite correct.  However,

1.  The strongest player on our team is a little over 2100 USCF(and a little over 2000 FIDE), which is good enough to recognize long term plan and ideas, but not enough to rely upon.
2.  Rybka is by far the strongest engine I've seen play in slow positional games such as d4, while other engines can be of better use in open tactical games such as e4.
3.  It's a waste of our time to play for such a small advantage where one tiny positional mistake could result in a draw.  The only Grandmasters that would play d4 in this type of a tournament would be more concerned about not losing than about winning.
Parent - - By lkaufman (*****) Date 2007-06-24 15:07
Thanks. Can you tell me where to find the games from the past freestyle events like this one?
Parent - - By nuff (**) Date 2007-06-25 15:43
What would happen if someone (about 2000 ELo) used an LK version, a Quad and a wierd opening book like the h3 seen in the engineroom? As long as their book contained sufficient entries for them not to make silly moves from the main players they could fare well. The problem with some of the engine only players is that they are 1500 and cannot recognise they they are being "sucked in".
Parent - By turbojuice1122 (Gold) Date 2007-06-25 23:09
The key words are "like the" in front of "h3".  I don't recommend using 1.h3 at all--there are a number of people who have a vast amount of theory in that line--I feel like I have quite a bit myself, and I'm definitely not willing to try it for fear of being blown away by someone with far more.  Try to look for something else--perhaps something based on 1.h4, or perhaps a decent diversion on the second or third move of a common opening--perhaps 1.e4 e5 2.Nf3 Nc6 3.h3 or something like that.
Up Topic Rybka Support & Discussion / Rybka Discussion / Freestyle Openings

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill