Question about PME position

Page 2 of 3 Previous  1, 2, 3  Next

View previous topic View next topic Go down

Re: Question about PME position

Post by Bilbo on July 1st 2009, 3:57 pm

I imagine this will be the case in many many complex positions. Just have a play through many mainline openings with Rybka and watch the evaluation score lurch around as it sometimes gives white or black an almost winning advantage in a position which theory attests to be completely equal or will give white no advantage in a position widely regarded as one of white's bests.

A computer doesn't understand chess strategy and positional factors, actually it doesn't think at all. All it can do is make calculations, abeit a lot of them.

It reminds me of a chessbase dvd by Rustam Kasimdzhanov on the Kings Indian Defense.

He recounts a story from about 10 years or so ago when in a certain well known line of the Kings Indian, the computer engine as white would play b4 and the human opponent would respond a5. As in the Kings Indian white is supposed to be opening lines on the queenside then he needs the exchange but the computer knowing nothing of positional subtleties would always advance with b5 thus closing the position.

Kasimdzhanov remembers how when this happened the Grandmaster would inevitibly laugh at he computers total lack of positional understanding, but would then stop smiling a few moves later when the computer beat him anyway!

The fact is that computers play incredibly strongly because they are tactically perfect and can therefore get away with moves that are positionally unsound, hence their preference in evaluating certain lines as better.

A human player is not a perfect calculating machine however and therefore has to play moves that have strategic and positional merit.

In a way neither is right or wrong, but both have different styles of play.

A human, even a great one does not play in the same style as a computer even though they may make many of the same moves.

Bilbo
International Master
International Master

Posts : 269
Join date : 2009-04-18

View user profile

Back to top Go down

Re: Question about PME position

Post by AoxomoxoA on July 2nd 2009, 5:57 am

Strad wrote:I think our teachers simply do not trust in engines opinion, here i quote a paragraph from month 7 ("e- and f- Pawn Phalanx"):
"In the game Botvinnik - Capablanca, 1938,... ."

Many excelent chessplayer think Capablanca was the best. Bobby Fisher and others. Gary Kasparov describes Capablanca with words: He contrived
to win the most important tournaments and matches, going undefeated for years (of all the champions he lost the fewest games)." and his style, one of the purest, most crystal-clear in the entire history of chess, astonishes one with his logic."

Capablanca was the WC who did play closest to a chessengine : http://www.chessbase.com/news/2006/world_champions2006.pdf
The average difference between moves played and best evaluated moves by computer analysis was with Capablanca 0.1008. ( Place 2 has Kramnk with 0.1058.). Interesting is that the "positional" players are closer to the computer analyses then the "tactical" players.

Maybe the text was written in a time where the engines where not that good. 10 jears ago the engines where not that clear superior, and they dont write their texts every jear new.

AoxomoxoA
Club Player
Club Player

Posts : 44
Join date : 2009-06-28
Location : Mannheim / Germany

View user profile

Back to top Go down

Re: Question about PME position

Post by Bilbo on July 2nd 2009, 1:19 pm

Those results are what you would expect. Capablanca avoided all complications where possible and so his games were less tactical than the other world champions. As a computer can be said to have virtual perfection in tactical calculation you would expect world champions of an attacking combinational character to score worse, hence Steinitz and Tal near the bottom.

You will see from your own games that in many strategic positions an engine will not score much difference at all between various strategic move choices. For example imagine a pawn exchange where the oppenent has the option of recapturing with cxd5, exd5 or maybe even Nxd5. Rybka or another engine may barely alter its evaluation score in any case, maybe 0.015 difference for example yet to a human player one move may clearly be more preferable due to the strategical nature of the position and the pawn structure, piece placement etc that may result.

Kasparov's position so high on the list clearly attests to his brilliance as he was a ferocious combinational attacker, so to have such a high computer agreement percentage even with such a high level of tactics in many of games probably marks him out as the best player of all time in real terms.

Capablanca would obviously struggle today being nearly 80 years out of date on opening theory and some key changes in the ideas of dynamics in chess.

If you took Capablance from 1920 now and put him straight into a big tournament in the 21st century he would struggle greatly as the play of his opponents would have improved tremendously and he'd be well behind in the advancement of chess knowledge.

This is important to note as it can be assumed that Capablanca's opposition would likely make much more mistakes than Kasparov's given 80 years of chess theory and so Capablana's games would have been easier to play in a sense.

This is not minimising Capablanca's briliiance, I'm sure if he had been born in the 1970's or 80's he'd be one of the absolute world's best today possible even world champion (I doubt though because of his lack of preparation and home analysis) but clearly the Kasparov of the 80's and 90's would have played better chess overall than Capablanca in the 1920's despite Capablanca having a higher computer agreement percentage.

Bilbo
International Master
International Master

Posts : 269
Join date : 2009-04-18

View user profile

Back to top Go down

Re: Question about PME position

Post by Strad on July 2nd 2009, 4:07 pm

AoxomoxoA wrote: Maybe the text was written in a time where the engines where not that good. 10 jears ago the engines where not that clear superior, and they dont write their texts every jear new.
Not at all... take a look what he says on month 7 when annotating Karpov - Stein, 1971 :

" 22.Qc2 a5?!
Black's text move and from now on, all Stein's moves are extremely precise if we consider what the best chess engine of our times thinks. As a short parenthesis, Kramnik was accused of cheating when he played the same moves that the chess engines indicate as best?!. The bad thing for Stein was he eventually lost the game and we can conclude that opening the position where the opponent is stronger cannot be good"

....and the final comment of this game :

"47.Nxf7+ 1- 0
This game will remain for a long time a masterpiece of a very wise plan in the middlegame and perfect transformation of the advantage.
The computer could not believe his eyes".

Conclusion : he talk about present chess engines (kramnik's incident reference) but anyway he does not trust it's opinion

Strad
Scholastic Player
Scholastic Player

Posts : 22
Join date : 2009-04-19

View user profile

Back to top Go down

Re: Question about PME position

Post by Bilbo on July 2nd 2009, 5:06 pm

Indeed it should still be noted that computers havn't yet had the truly decisive match victory over a world champion opponent yet.

Most such events Kasparov vs Deep Blue and Kramnik versus Fritz were inconclusive.

There is a VERY important thing to consider here. A human world champ may lose in a match against a computer NOT because the computer is better but rather because a computer always plays to its best, it is not fatigued, it doesn't make blunders based on complete oversights etc.

Against Deep Blue for example Kasparov missed chances he would ordinarily have found, and importantly he lost his nerve and offered a draw in a winning position.

Wladimir Kramnik lost a famous game to Fritz because he lost to a mate in one that a 1400 player would have been expected to see!

Why?

Because playing a computer that is tireless, never makes a mistake, never miscalculates etc over a series of games in a man vs machine match is incredibly draining to a human player.

We a fallible, and we can miss easy tactical shots that cost us games especially when playing a machine that is tactically perfect.

However none of this has anything to do with how well a human understands chess in terms of positional play and strategy versus a computer.

This means that in positional matters even Rybka's first choice move isn't necessarily always the best. A world class grandmasters' intuitive understanding of chess strategy and positional play IS STILL better than even the world's top computers.

The reason computers now win most of the man vs machine matches however isn't to do with positional understanding as much as it is to do with the perfect calculation ability of a computer coupled with its inability to fatigue or make glaring tactical blunders.

When world class humans lose to computers it nearly always a result of these latter factors rather than being outplayed positionally hence it is simply wrong to always accept that Rybka's best move according to how it runs on your pc is the definitive or objectively best answer to that position.

I would always choose the advice of an experienced grandmaster, like the ones on our course over Rybka running on my dual core pc.

Bilbo
International Master
International Master

Posts : 269
Join date : 2009-04-18

View user profile

Back to top Go down

Re: Question about PME position

Post by AoxomoxoA on July 2nd 2009, 6:05 pm

Strad wrote:
Black's text move and from now on, all Stein's moves are extremely precise if we consider what the best chess engine of our times thinks.
Conclusion : he talk about present chess engines (kramnik's incident reference) but anyway he does not trust it's opinion

Hehehe ( LOL ), i wrote my first chessprogram in 1982, on a Honywell Bull 66/80 in Algol 60, a stack of almost 200 cards. Every chess player said: "never will a computer beat a master , no creativity , no intelligence. They did want me to stop. It was simply blasphemic and everyone writing a program had to be burned or taken to a house for mad people. But its easy: computer double their speed all 2 jears and what doubleling in chess means we know from the "creator" of chess and the rice grains on the squares. Sooner or later even the stupiest program would beat every human. But no Chessplayer i met was able to see that. Chess is seen as an example of intelligence and intelligence is only for humans. Animals have no intelligence, only instincs, and machines only smell bad. ( Nice to read the essay: Maelzel's Chess Player from Poe). Many fights of religion against sicience in the old times ( and today ) you can see again in the fights against Artificial Intelligence. These things scratches on some peoples self-identity.

I wonder what Computer and Program they are talking of. Rybka 3 tells of bad bad bad blunders. In move 43 Black did throw away the value of 3 pawns. Move 39 1 Pawn, Move 37 0.75 and so on. About 30 + % of blacks moves differ more than 0.4 pawns from Rybkas best. Maybe they are right: we can not trust this Computer and this program ;-)

Thanks for the Info :-)

AoxomoxoA
Club Player
Club Player

Posts : 44
Join date : 2009-06-28
Location : Mannheim / Germany

View user profile

Back to top Go down

Re: Question about PME position

Post by AoxomoxoA on July 2nd 2009, 6:35 pm

Bilbo wrote:Indeed it should still be noted that computers havn't yet had the truly decisive match victory over a world champion opponent yet.

This means that in positional matters even Rybka's first choice move isn't necessarily always the best. A world class grandmasters' intuitive understanding of chess strategy and positional play IS STILL better than even the world's top computers.

The reason computers now win most of the man vs machine matches however isn't to do with positional understanding as much as it is to do with the perfect calculation ability of a computer coupled with its inability to fatigue or make glaring tactical blunders.

Yes a move from Rybka is not allways the best. It would need to run millions of jears and have a gigantic memory to prove the best move on the computers of our days. But the probability that rybkas move is better compare to a human is extreme high.

Rybka is a better chessengine than cyclone or toga. But toga is calculating 30 times more knots than rybka in the same time. Why is rybka better? because it uses much more time in "positional understanding". all good engines uses positional algorithms some more, some less. rybkas positional algorithms are written from/with 2 IMs and 1 GM. Some known positional faktors from Toga are Piece Activity, Piece Square Activity ,King Safety,Pawn Structure, Passed Pawns.

AoxomoxoA
Club Player
Club Player

Posts : 44
Join date : 2009-06-28
Location : Mannheim / Germany

View user profile

Back to top Go down

Re: Question about PME position

Post by Bilbo on July 2nd 2009, 6:37 pm

AoxomoxoA you are so missing the point here. Yes computers are getting faster and faster and able to calculate deeper and more complex variations but that doesn't mean they know chess strategy better than a grandmaster.

Chess computers DON'T beat top level grandmaster by outplaying them positionally and knowing more about strategic play. They beat them because they are tactically perfect and never make mistakes or get tired.

It's like a boxing match between Manny Pacquaio the current world's best p4p fighter against Sam Peter, a big slow plodding lumbering heavyweight with very little basic boxing skill.

Who would win in a boxing match? Sam Peter would! Why because he is almost twice the size of Manny Pacquaio ( who started out at flyweight) and his sheer size and strength would be too much for a little man like Manny Pacquauio to handle.

But even though Peter would win the fight, he is no way even close to being as good a boxer as Manny Pacquaio. He wins the fight due to the non technical aspects, the main factor being sheer size.

This is how it is with computer chess. A computer not only has an entire opening book programmed into it, millions of moves already stored, it also has perfect tactical calculation and endgame tablebases so it knows how to play any position with less than 5 or 6 pieces on the board perfectly.

It's doesn't know more strategy or have more positional skill than a top grandmaster!!!

It's winning because of the OTHER factors.

Put it this way, imagine a man vs machine match where the human was given fairer odds, and so allowed access to his own opening reportoire database during the game, he had endgame table base books to play those final positions perfectly AND he was allowed to check all of his moves prior to making them to eliminate tactical errors.

This would then make the battle purely a matter of positional and strategic understanding.

In such an event I imagine a player of the calibre of Kramnik, Kasparov, Anand etc would trounce any current computer chess engine.

That would be like turning the tiny 5 ft 6 140 lbs Manny Pacquaio into a 230 lbs 6ft 5in monster and then sticking him in with Sam Peter. In such a case Peter is getting annihilated!

You have to drop this ridiculous idea that because a computer can beat a human over a series of matches its therefore better at absolutely everything in chess than a human because that is NOT true.

Bilbo
International Master
International Master

Posts : 269
Join date : 2009-04-18

View user profile

Back to top Go down

Re: Question about PME position

Post by AoxomoxoA on July 3rd 2009, 8:28 am

Bilbo wrote:
It's doesn't know more strategy or have more positional skill than a top grandmaster!!!
It's winning because of the OTHER factors.

Your assumption is difficult to prove. Do you have and evidence for this? The engines simply try to find the best possible move. the moves get better as deeper they calculate or as better the positional and other ( heuristic ) algorithms get. If they ever will be able to calculate till the end of the game (maybe quatumcomputer will do?) then they are perfect.
Why are we trying to get a better position? A better positon gives us better chances for a tactical stroke. better activity of pices, open lines and files for towers and so on increase the chances for a successful tactical stroke. Even if the engine only would see the tactical stroke in 15 moves he will improve (most times) the position. and if you can mate in x-moves, your position is optimum anyway.
That people like Kramnik play like Fritz shows that Fritz "positional" understanding at least not much worse than the one of Kramnik what is not thaaat bad.
To make a good positional move and loose because of a tactic is no good idea to.

I check ( some ) positions not only with Rybka but with other engines to. the algorithms are different so the chances to get a "right" opinion increases.
and these algorithms have proved to be good (sucessful). If several Grandmasters tell me ( with 600 Elopoints less ) after minutes of analysing : a move is bad ( for what reason ever ) then i better stop dsicussing. If severals engines tell a grandmaster (with 600 Elopoints less ) ...

AoxomoxoA
Club Player
Club Player

Posts : 44
Join date : 2009-06-28
Location : Mannheim / Germany

View user profile

Back to top Go down

Re: Question about PME position

Post by Bilbo on July 3rd 2009, 9:16 am

Ok so let's take your stance. The chess computers know more opening theory, they play perfect tactical calculation, they are better in endgames thanks to their tablebases AND they play better positionally, to say nothing of the facts they never get tired, never make mistakes etc.

So effectively they are better than humans in EVERY SINGLE aspect of chess.

Yet Kasparov was able to beat Deep Blue in 96, lost in 97 only due to a terrible mistake in the final game and of the other 5 games was tied with 1 win apiece and 3 draws.

Anand lost to Rebel in Blitz in 98 but beat it over regular controls. Kramnik draw with Deep Fritz 4-4 in 2002 whilst Kasparov drew with Deep Junior in 03 and had a winning advantage in the final game.

He also drew with X3d Fritz in 03 and would have won possibly won the match but for a terrible blunder in game two.

Kramnik lost 4-2 to Fritz in the last man vs machine match of significance but missed a win in game 1 and lost one game thanks to the worst blunder probably ever by a GM to a mate in one.

Take away the blunders in all these matches and the humans might not have lost any of these matches.

Yet they are inferior, according to you in EVERY aspect, openings, positional play, endgames, calculation, tactics, errors and mistakes, tiredness, emotional distress etc.

So how did they win any games at all and get great positions and draw in most games?

Bilbo
International Master
International Master

Posts : 269
Join date : 2009-04-18

View user profile

Back to top Go down

Re: Question about PME position

Post by AoxomoxoA on July 3rd 2009, 12:04 pm

Bilbo wrote:
So how did they win any games at all and get great positions and draw in most games?

The engines did improve the last 10 jears and the computers are bigger (in memory) and faster now.
The speed of the computer did increase 32 times since 1999 and the programs improved with each new version 50-100 elopoints at the same hardware by new (positional) algorithms.

* 1984: Novag Super Constellation 6502 4 MHz (1631)
* 1985: Mephisto Amsterdam 68000 12 MHz (1827)
* 1986: Mephisto Amsterdam 68000 12 MHz (1827)
* 1987: Mephisto Dallas 68020 14 MHz (1923)
* 1988: Mephisto MM 4 Turbo Kit 6502 16 MHz (1993)
* 1989: Mephisto Portorose 68020 12 MHz (2027)
* 1990: Mephisto Portorose 68030 36 MHz (2138)
* 1991: Mephisto Vancouver 68030 36 MHz (2127)
* 1992: Chess Machine 30 MHz Schröder 3.0 (2174)
* 1993: Mephisto Genius 2.0 486/50-66 MHz (2235)
* 1995: MChess Pro 5.0 Pentium 90 MHz (2306)
* 1996: Rebel 8.0 Pentium 90 MHz (2337)
* 1997: HIARCS 6.0 49MB P200 MMX (2418)
* 1998: Fritz 5.0 PB29% 67MB P200 MMX (2460)
* 1999: Chess Tiger 12.0 DOS 128MB K6-2 450 MHz (2594)
* 2000: Fritz 6.0 128MB K6-2 450 MHz (2607)
* 2001: Chess Tiger 14.0 CB 256MB Athlon 1200 (2709)
* 2002: Deep Fritz 7.0 256MB Athlon 1200 MHz (2759)
* 2003: Shredder 7.04 UCI 256MB Athlon 1200 MHz (2791)
* 2004: Shredder 8.0 CB 256MB Athlon 1200 MHz (2800)
* 2005: Shredder 9.0 UCI 256MB Athlon 1200 MHz (2808)
* 2006: Rybka 1.2 256MB Athlon 1200 MHz (2902)
* 2007: Rybka 2.3.1 Arena 256MB Athlon 1200 MHz (2935)
* 2008: Deep Rybka 3 2GB Q6600 2.4 GHz (3238)

So Kramnik lost against Fritz in 2006. fritz was in 2006 a program with elo 2830, lower than the leader rybka 1.2 with 2902. That Kramnik did not lost 6-0 is because Kramnik plays like a computer - not that bad ;-) Now Rybka is an other 330 Elopoints better ( and I am using a slightly better machine as mentioned in the list above ). Kramnik said after analysing: "Deep Fritz (8.0!) is stronger than Deep Blue" , so Deep Blue was not thaat good either.
That a good GM had a better positional judgement in 2000 seems to be clear ( same rating but bad tactic = better posiitional understanding : sound reasonable ) but now 2009 we compare a car against a horse on an highway.

AoxomoxoA
Club Player
Club Player

Posts : 44
Join date : 2009-06-28
Location : Mannheim / Germany

View user profile

Back to top Go down

Re: Question about PME position

Post by Bilbo on July 3rd 2009, 12:22 pm

I agree completely with the overal superiority of today's engines over human players. But in the area of positional play and strategic understanding humans are clearly slightly superior due to the results I showed above.

If they wern't it would impossible to ever even draw a single game as they would be inferior in every single aspect.

That clearly hasn't happened yet so considering humans can still get results in individual games against computers they obviously have to have an advantage somewhere in those games or else they would lose every match 6-0 every time.

Especially in closed positions human's can still hope to outplay or at least survive even against the best computers today.

It may not remain that way for ever but it is the case now.

Anyhow, this debate is going way off topic. The original post by you was that 29.b4 as suggested by ICS was an error. I have explained countless times that it was not, and showed how human's play to differnt plans and strategies than computers due to having different strengths and weakenesses and how 29.b4 is a sound practical idea for White in the position, and one that is in accordance with ICS's criteria of evaluations for PME.

All ths further talk is neither nor there regarding the original point which is that 29. b4 is a perfectly ok move. Nobody (not myself or ICS) said Rc1 was an incorrect move or even that 29.b4 was better, but rather that the suggested move represents an excellent practical play by White and is a valid alternative to the text move.

I don't think even you are any longer arguing this point, but instead preferring to take the discussion way off topic.

So let's just end this discussion now and move on with the rest of our lives. Smile

Bilbo
International Master
International Master

Posts : 269
Join date : 2009-04-18

View user profile

Back to top Go down

Re: Question about PME position

Post by AoxomoxoA on July 3rd 2009, 2:07 pm

Bilbo wrote: .. due to the results I showed above.
If they wern't it would impossible to ever even draw a single game as they would be inferior in every single aspect.
The chessegines had been better and they did gain several hundrets elopoints since then. Show me a draw in 2009. Your arguments are outdated.
A draw from a 2600+ against a 3200+ is very seldom. Our diskussion did make me look at the position after b4 and it looks bad to me. and it looks bad or even extreme bad to several chessengines elo 3000+.

Bilbo wrote:
Especially in closed positions human's can still hope to outplay or at least survive even against the best computers today.
Example ? Please 2008 or later. ( one of the rara example you find here : http://susanpolgar.blogspot.com/2008/01/benjamin-rybka-draw-odd-match.html )

AoxomoxoA
Club Player
Club Player

Posts : 44
Join date : 2009-06-28
Location : Mannheim / Germany

View user profile

Back to top Go down

Re: Question about PME position

Post by Bilbo on July 3rd 2009, 2:54 pm

Well in that case it is clear that you can learn nothing from this course being taught by foolish humans so why are you even here?

Just fire up Rybka and away you go, you'll be a Grandmaster in no time as long as you stick to its variations.

Bilbo
International Master
International Master

Posts : 269
Join date : 2009-04-18

View user profile

Back to top Go down

Re: Question about PME position

Post by Strad on July 3rd 2009, 3:05 pm

Bilbo wrote:Well in that case it is clear that you can learn nothing from this course being taught by foolish humans so why are you even here?

Just fire up Rybka and away you go, you'll be a Grandmaster in no time as long as you stick to its variations.
imo if we are working hard on improve our play we must turn off our Hiarcs, Fritz, Rybka, etc...and start to think by ourselves, we will be alone in our next OTB tournament... Very Happy

Strad
Scholastic Player
Scholastic Player

Posts : 22
Join date : 2009-04-19

View user profile

Back to top Go down

Re: Question about PME position

Post by Bilbo on July 3rd 2009, 3:28 pm

It is true.

Today I've been working through two heavily annotated games in Artur Yusupovs Build Up Your Chess relating to openings. The games were Meek vs Morphy from 1855 and Bueker against Yusupov himself from 1992.

In both games there are moves that Rybka marks as virtual blunders, greater than 30/100ths or even 50/100ths of a pawn move difference yet Yusupov says these are the best moves with analysis of why other moves are bad, often Rybka's choices are deemed as mistakes, blunders even.

One example is when the king is stuck in the center on the d file and the d pawn is under a pawn attack Yusupov cites as a blunder the opponents decision to play f3 instead of castling to safety.

Rybka regards f3 as the 'best' move however. The point is that to Rybka is might well be because it see all the tactical permutations that will follow and can keep the king safe from attack because it can defend perfectly, but such an approach would be suicidal for a human player.

In the other game he criticises as a bad mistake white's choice of pawn grabbing with his queen whilst his king is already shaky and he is behind in development, yet once again Rybka regards that plan of action as the best move.

Again if you can play tactical positions with absolute perfection, never miscalculate and never make an oversight then yes maybe you can afford to take the pawn, but for a human player such a path of action is incredibly risky especially against a stronger opponent than yourself.

We don't need instruction as to what are the best moves in an absolute sense providing we know how to play any resulting positions perfectly as there is not the slightest margin for error.

We need to know the best moves that can actually help us play safely.

The best way I can describe it as follows.

Imagine a general is fighting an army across a minefield and by marching across the field by night hecould take them by suprise and defeat them battle whilst they were completely unprepared it providing nobody stepped on any mines and blew themselves up and then alerted the defenders who mowed everyone down in a hail of bullets.

Now in an absolute sense, if the soldiers knew exactly where to tread to avoid all mines and they could be certain of crossing without incidence the march across the minefield at night strategy IS the best one.

BUT if you don't and of course there would be no way you could, then to send your troops to march across a minefield at night would be SUICIDE!

That's why we need to learn positional chess from human grandmasters and not chess engines

Bilbo
International Master
International Master

Posts : 269
Join date : 2009-04-18

View user profile

Back to top Go down

Re: Question about PME position

Post by AoxomoxoA on July 3rd 2009, 4:09 pm

Bilbo wrote:
That's why we need to learn positional chess from human grandmasters and not chess engines
Did i ever said Chessengines are good teacher? Did i say engines should play for us? Do you think i learn at ICS because i dont know where to put my money?
No!
I say: chessengines are now jear 2009 extremly better in chess than every human, thats it.
It might even be, that in a heavy tactical situation the best moves of rybka are not playable from any human. thats the problem with all far better players. Even the suggestions of our teachers might be to much for me. but the position after b4 is not thaat tactical.

AoxomoxoA
Club Player
Club Player

Posts : 44
Join date : 2009-06-28
Location : Mannheim / Germany

View user profile

Back to top Go down

Re: Question about PME position

Post by Bilbo on July 3rd 2009, 4:14 pm

Yes I know the position after b4 isn't tactical, that's the point!

Just look at the board, white has a passed pawn and the black knights are pinned to the defense of each other and king to the defense of a knight too.

White doesn't have any winning chances, unless Black is too aggressive but he can hold by keep his rook on the second rank pinning the knight and the other rook on the d file pinning the knight.

Run it through Rybka, it holds.

That's all the ICS team said, that it's a 'practical' move to play.

RC1 might have been better but it allowed black to play a5 so stopping white from having a passed pawn and it allowed the knights to move so now Black could play for a win.

It's really not that hard to understand.

At least we agree on the last part of your post, Rybka's tactical suggestions are bascially unplayable for maybe even the top grandmaster to go into over the board.

Bilbo
International Master
International Master

Posts : 269
Join date : 2009-04-18

View user profile

Back to top Go down

Re: Question about PME position

Post by AoxomoxoA on July 3rd 2009, 4:39 pm

Bilbo wrote:
At least we agree on the last part of your post, Rybka's tactical suggestions are bascially unplayable for maybe even the top grandmaster to go into over the board.
You know already my opinion to the other points. Here you seem to missunderstand me. almoust all suggestions of rybka are playable. Kramnik usually does and even i had 1 blitzgame against a master where i played 100% rybka ( short game, the master was playing bad in this game). if you are a GM and you analyse hundred games with an engine you will find a game where this problem might happen. then you write it down in a book because its an interesting exception.

AoxomoxoA
Club Player
Club Player

Posts : 44
Join date : 2009-06-28
Location : Mannheim / Germany

View user profile

Back to top Go down

Re: Question about PME position

Post by Bilbo on July 3rd 2009, 5:55 pm

Of course all suggestions of Rybka are playable, but not for every chess player.

When Rybka recommends a dangerous complicated sacrifce for example or leaving your king in the center in an open game because you can exploit a tactical advantage with perfect play those suggestions arn't wise for all but the most talented players.

It's better for us amatuers at least to concentrate on getting a good grasp of positional chess and playing logical sound moves in accordance with the position.

You just seem to be deliberately antagonistic on this issue.

Bilbo
International Master
International Master

Posts : 269
Join date : 2009-04-18

View user profile

Back to top Go down

Re: Question about PME position

Post by AoxomoxoA on July 3rd 2009, 6:58 pm

Bilbo wrote:
It's better for us amatuers at least to concentrate on getting a good grasp of positional chess and playing logical sound moves in accordance with the position.
You just seem to be deliberately antagonistic on this issue.
No, there i agree for me, not for Kramnik though.

You can put tactics ( combinations would be the better word ) in different categories. You might start with a fork then a pin and the result might be a won pice. these categories are digital. "there" or "not there" ( they may apere in an other line but its almost "digital" ) and tactical motivs are sequential: first, second ...
with positional play and strategies its far more complicated. you have dozends of pattern :Majority attack, minority attack depending if you have a majority or a minority at the queenside. isolated pawns might be advanced. weak fields might be unimportand and so on. in every position you find patterns as much as you want and they are often controvers. how relevant is each? are the 2 rooks able to intrude the position. are they able to protect the b-pawn against the attack of 3-4 pices. will the attack of the pawn open the position for the rooks. Questions wich could be only answerd with lots of experince and/or Elopower. every player answers the questions ( if been asked ) in a different way. Steinitz - Nimzowitsch thought queensidemajority is a clear good thing. and then did capablanca invent the minorityattack ( see the secrets of John Watson ) and swoop a minority of queensidepawns is good. The problem of working with pattern like we humans do is, to weight the patterns correctly: the pawn is isolated is not that important therfore ... is important.
we need an eye for these weights like we need eyes for forks.
we will develope and improve our weights if we constantly think about them and ask our self about these faktors as told at our school.
but these weights ( like: open file is more important then an bishoppair, at least in position xxy) still differ from Master to master. and if they are perfect then you are WC like Kramnik ;-) I am almost shure b4 was in the candidate - list of Kramnik, Kramnik knows about queenside majoritys and he had a reason to decide to move different. the position is not that dynamic, the engines decide at once what they want and what not in this position, so i guess kramnik saw whats wrong with b4 as the engines do at once.
I think ( and would like to see it done ) it should be possible to do now a reverse enineering. you know the move b4 is bad: what is the positional reason for it. where are the wrong weights. at least a GM should be able to analyse a position with the aid of an engine and find the important faktors and their correct weights. This would mean an improvment of knowhow. when i think about it: the people who resist learning new things the most are teacher (well not really ;-)
Thats all i have to say about b4, more will not come!

AoxomoxoA
Club Player
Club Player

Posts : 44
Join date : 2009-06-28
Location : Mannheim / Germany

View user profile

Back to top Go down

Did I miss something?

Post by hoopy on July 4th 2009, 1:18 pm

I started this thread by raising my concerns that we had not been given a complete list of equivalence material for PME. Innocent enough I thought. When I come back from vacation it seems that I set off world war 3!! Some great stuff some overreaction, some emotion. Unfortunately my chess skills are not yet strong enough to completely follow the discussion but I reckon there are some gems I will come back to when my game improves. Thanks guys. Very Happy
avatar
hoopy
National Master
National Master

Posts : 194
Join date : 2009-05-04

View user profile

Back to top Go down

Re: Question about PME position

Post by Bilbo on July 4th 2009, 9:01 pm

I sound more aggressive online than I actually intend to be!

AoxomoxoA exasperates me more than annoys me Smile

Bilbo
International Master
International Master

Posts : 269
Join date : 2009-04-18

View user profile

Back to top Go down

Re: Question about PME position

Post by Blue Devil Knight on July 4th 2009, 11:44 pm

This subject turns out the emotion. See the crazy comments here at J'adoube's blog:
http://j1mb3d3nbaugh.blogspot.com/2007/02/play-like-grandmaster-or-chess-engine.html

I think everyone reasonable would agree with some truths:
1. An unhandicapped computer can beat any human in an N-game match.
2. A good human is much better than a good computer for help with postmortem analysis, but both are useful.
3. There are lines that a computer ranks highly that would be silly to play in practice, even though the computer playing that same line would be fine.

__________________________________________
http://chessconfessions.blogspot.com
avatar
Blue Devil Knight
Grandmaster
Grandmaster

Posts : 616
Join date : 2009-04-12

View user profile http://chessconfessions.blogspot.com

Back to top Go down

Re: Question about PME position

Post by AoxomoxoA on July 5th 2009, 7:04 am

Blue Devil Knight wrote:
I think everyone reasonable would agree with some truths:
1. An unhandicapped computer can beat any human in an N-game match.
2. A good human is much better than a good computer for help with postmortem analysis, but both are useful.
3. There are lines that a computer ranks highly that would be silly to play in practice, even though the computer playing that same line would be fine.

Yes true
At Point 2 there might be a change in the future. There are no programs for analysis at "this" time. the engines are programed to play the best move not to explain a position to a human. I am thinking of implementing some analysing technics though, ... but this will not be better than an good playing human very soon :-(
Point 3 is true, i guess ( would be interesting to see one ), but my projected 1911 ELO ( about 2011 UCSF ?) tells me: this is not the case here and my experience witch computeranalyses of positions tell me the same, espiacily for player like Kramnik. I would expect these rare cases in extrem sharp situations. Off couse i have no proofs for this statement though :-(

AoxomoxoA
Club Player
Club Player

Posts : 44
Join date : 2009-06-28
Location : Mannheim / Germany

View user profile

Back to top Go down

Re: Question about PME position

Post by Sponsored content


Sponsored content


Back to top Go down

Page 2 of 3 Previous  1, 2, 3  Next

View previous topic View next topic Back to top


 
Permissions in this forum:
You cannot reply to topics in this forum