Valuation of W's Opening Move 1.e4?

[Kevin Pacey]

Re: Valuation of W's Opening Move 1.e4?

Here is what IM John Watson writes concerning 'arithmeticm' in chess, in 'Secrets of Modern Chess Strategy':

"Especially with the appearance of chess-playing computers which update a numerical assessment of the position on every half-move, there are players who tend to think in terms of arithmetic advantages, e.g., "White is better by 0.33 pawns'. This has its uses, but can lead to a rather artificial view of the game. What happens when both sides make a few moves which are the best ones, and suddenly the 0.33 pawns is down to 0.00, or full equality? The defender of this point of view will say: 'Well, I didn't see far enough ahead. If I had, I would have accurately assessed the original position as 0.00.' The only problem with this point of view is that chess is a draw, and all kinds of clear advantages (in the sense of having a good probability of winning a position in a practical game) are insufficient to force a win against perfect defence. So most positions would be assessed as 0.00, which is not very helpful. In the extreme, we have the same problem when we claim, for example, that 1.Nf3 is 'better' than 1.e4, or 1.d4 is better than 1.c4. These are rather meaningless statements, unless we put them in the context of 'better against opponent X' or 'better from the standpoint of achieving good results with the least study' or some such. As for the objective claim of superiority, what would be our criterion? I would suggest that only if a given first move consistently performs better than others against all levels of competition might we designate it as 'better' in a practical sense. Since all reasonable first moves lead to a draw with perfect play, a claim of ultimate theoretical superiority for one of them cannot be justified."

[Bob Armstrong]

The " Perfect Ideal Chess Game" ?

Hi Kevin:

Excellent find!

As with the thread on " Is chess a sport ", this " perfect chess game " concept is a perennial discussion, just that it only surfaces from time to time, as the new generation chews it over, and realizes it is an issue, but has never had access to the ancient writings.

I respect my elders, such as IM John Watson, and so will tank for a few months on this paragraph, to see if I can at all still defend my position that there is a " perfect idealized chess game ", apart from all the down in the muck variables that occur in " real " chess.

Bob A

[Kevin Pacey]

Re: Valuation of W's Opening Move 1.e4?

Hi Bob

Here is something else IM Watson writes in his book, regarding what is the proper result of a perfectly played chess game:

"As you probably already know, it is a draw. There is no need to wait for computers to solve chess in order to understand this; it is crystal clear from a study of theory. Of course, I can't prove this, but I doubt that you can find a single strong player who would disagree. For those who like authoritative sources, I remember Kasparov, after a last-round draw, explaining to waiting reporters: 'Well, chess is a draw'. If one of the most dynamic players in history (with the highest winning percentage of his time) believes this, we can probably assume that it's true. And yet, how many sources simply neglect to address this subject! One feels that they don't want to present chess as a boring game which is only made competitive by human error. But for one thing, even 'perfect' play can result in wild shifts in momentum; see the next chapter for a discussion of swings of initiative which mark dynamically-balanced chess games. Moreover, any number of positions assessed as 'equal' in books are so unbalanced and complex as to make that assessment only a guide to one's practical chances. Tal once remarked that most of the '=' assessments in opening books were actually '=/+'! As an author of opening books, I know what he means. Because of the presumption of White being better, the juncture of the game at which Black frees his game or neutralizes White's plans has often automatically been assumed to give him equality, even though in dynamic openings, the exhaustion of White's initiative very often means that Black has seized it with advantage."


Note that like for my previous quotation of Watson, I don't agree 100% with what he writes.

Regarding the above quotation in this post, I seem to recall an interview Larry Evans gave on the internet in which he stated Fischer believed White ought to win. It wasn't clear at what point in Fischer's career he believed this, or whether he believed it at the end of his life. However Steinitz, who was a hero figure for Fischer, based his theories about chess on the axiom that if no one made a mistake then a game of chess should be drawn. Hans Berliner, a correspondence world champion, on the other hand, wrote a (suspect) book claiming 1.d4 wins.

Much farther back, American Weaver Adams tried by various means to prove White should be winning straight from the opening phase. Nowadays, even, GM Rowson and GM Adorjan have written books sympathetic to Black's chances of ultimately winning. So there is not complete agreement among strong players that a draw should result. Note, however, that some matches between the very best modern computers have featured a high ratio of draws, just like matches between top humans, even though top programs beat top humans consistently now.

Another point is that Kasparov has been proved incorrect in some of his past pronouncements concerning chess openings, alone (besides the [remote?] possibility that he might have told reporters that chess is a draw purely to avoid showing his disappointment at failing to win). It is heartening that Kasparov has scored well even with Black; like Gligoric he was one of the few top players who could make even the QG Tarrasch look like a terrific weapon against top competition, for example. Regarding Watson/Tal's assertion that many '=' positions are '=/+', I think they exaggerate. For one thing there are more dull, drawish positions that can arise than I care to guess at.


In the previous post I made, concerning Watson's opinion on arithmeticism in chess, I disagree with using practical results as the standard to determine what the best first move might be for practical purposes, as Watson puts it. One or more defences, if not single critical variations, can score better against a given first move than the average score that Black has against it, not to mention that a given player can tend to score worse than the average with a given first move for White, or with a given defence against a given first move. In short, even mega-database statistics for a given move will almost invariably never be completely reliable.

[Bob Armstrong]

Is Chess - Draw? or Win for One of the Sides?

Hi Kevin:

I need to consider more these esteemed opinions, based on elite chess experience.

But it does not deter this class player from holding an opinion:

" A chess game is a draw. The reason is that Bl can never improve his position unilaterally. He moves from the opening position valuation, to an improved valuation, only to the extent W makes a weak move. Similarly, W can never unilaterally increase his advantage, unless Bl makes a weak move.

So the perfect game result all depends on 2 things:

a) what is the " perfect " first move in chess, assuming there is one, and what is the valuation of that position after W's first move?
b) assuming W has some type of benefit from the first move [ maybe initiative, which I value between 0.00 and 0.25, or maybe a " slight " advantage at least ( 0.26 or over )], is it sufficient to cause a win for either side , on perfect play by both sides?

Bob A

[Kevin Pacey]

Re: Valuation of W's Opening Move 1.e4?

Not sure how 0.25 got to be the upper limit for '='. Why not, for example, 0.3333333...? The designer of Fritz13 has no monopoly on setting the limit.

Even if we assume that chess is a draw, there still could be a White first move that's objectively best. I'll personally base this on the length of the game before a draw must be agreed, rather than anything else (someone once calculated that the maximum length of a legal chess game was just over 5000 moves [that was before certain endgames were allowed to be played out to 100 moves, so I don't know if this has changed]). That and the objectives of each side.

If both sides are satisfied to draw, the perfect game is a handshake before White even makes a first move, assuming the organizer of an event allows this. :)

If White wants to win and Black wants to draw, White must play the first move that forces the longest possible game against Black's attempt to draw the game as quickly as possible (nevertheless, a draw should still result). What first move is that? A computer won't provide even a clue, at the moment.

If White wants to win and Black wants to win, both sides 'must' drag out the game as long as possible before a draw is grudgingly forced by one side or the other (e.g. 50 move rule). What first move serves White best for this purpose? Again, a computer won't help with a clue, at the moment.

If White wants to draw and Black wants to win, White must play the first move that forces the shortest possible game against Black's wishes. What first move should White play? Again, a computer is no help.


Rather than game length, you might prefer to make finding the most difficult, forced moves the criteria for the perfectly played game, but then I or someone else might disagree with that as the criteria for perfection.

[Bob Armstrong]

The Perfect Chess Game(s )?

Hi Kevin:

I hope you and I are not the only ones interested in this issue!! But it is fun to examine.

Why 0.25 ? Simple - I have Fritz 13 and nothing else.

But I would be loath to lift the equal range limit beyond that. 1/4 P advantage seems manageable for Bl, and he can draw. 1/3P difference starts, in my opinion, to get Bl into heavier waters. We need some demarcation line so that beyond it, W is considered to have a " substantial " enough advantage ( 0.26 & higher satisfies me ).

Whether a human is playing for something ( win, loss, draw ) does not affect the theoretically perfect chess game ( or as you opined, the theoretically perfect chess gameS ). And if we can get the world's strongest computer to play itself from all 20 opening moves ( it is 20 isn't it - my mind is totally fried with this thread ! LOL ) to the end of the game, we can then see the results and compare first moves. We will also have to agree to some convention as to how to judge among draw results, which opening move of them was " best " ( eg. your offering of no. of moves it took to get the draw - but I'm not sure I agree this would be a good criterion - I have problems with it which I must consider further ).

Bob

[Kevin Pacey]

Re: Valuation of W's Opening Move 1.e4?

Another point may be worth mentioning: even computer playing programs can have a setting which determines how much they like playing for a win/draw. The extreme is: always be happy with just a win/draw, depending on the number of the setting.

Also, humans can be less extreme in their result preference as well. Quite often we are happy with a draw should it happen, but we don't rule out giving ourselves a practical chance of winning. Hence against 1.e4 a human might play a Petroff if a draw is 100% satisfactory in a tournament/match, but in cases where a win would be nice, but a draw is otherwise okay, a human may prefer an opening with medium 'risk' of losing such as the Caro-Kann.

However, the Caro-Kann might not lead to a draw with the minimum number of moves if White is both stubborn and plays otherwise error-free (perfect?) chess. For that, the Petroff might fit the bill after all, but who knows?

[Zachary Dukic]

Re: Valuation of W's Opening Move 1.e4?

Another point may be worth mentioning: even computer playing programs can have a setting which determines how much they like playing for a win/draw. The extreme is: always be happy with just a win/draw, depending on the number of the setting.

Also, humans can be less extreme in their result preference as well. Quite often we are happy with a draw should it happen, but we don't rule out giving ourselves a practical chance of winning. Hence against 1.e4 a human might play a Petroff if a draw is 100% satisfactory in a tournament/match, but in cases where a win would be nice, but a draw is otherwise okay, a human may prefer an opening with medium 'risk' of losing such as the Caro-Kann.

However, the Caro-Kann might not lead to a draw with the minimum number of moves if White is both stubborn and plays otherwise error-free (perfect?) chess. For that, the Petroff might fit the bill after all, but who knows?

Not to get off topic, but in my opinion I find the Caro-Kann hard to play against, because it's very easy for both sides to mess up!

[Kevin Pacey]

Re: Valuation of W's Opening Move 1.e4?

The Caro-Kann would completely qualify as a solid choice, in top level play at least, were it not for the Advance Variation. In that line the percentage of decisive results tends to be relatively high in most databases.

The Petroff could be avoided by, say, the King's Gambit, but in top level play that would be too risky for White. Decisive results, though high in percentages, tend to favour Black at least slightly, so White possibly has to be pretty desperate to win to play with it. More importantly, the whole opening is suspect, and quite possibly losing with error-free play by both sides.

I still value White's (temporary/long-term?) edge in the position after 1.e4 as quite possibly 0.33333333..., based on my interpretation of Classical principles regarding the value of an extra tempo open positions (e2-e4 is virtually undeniably a great use of a tempo in trying to steer for an open position). Unlike Bob, I don't believe this is scary enough for Black to value the position as undeniably +/=. Even a study of theory (as in books) leads one to believe that Black should at least reach an unclear, if not equal, position eventually, though possibly after quite a number of moves.

The French Defence is a way to try to avoid open positions after 1.e4, as White's best lines tend to involve a closing of the centre. However in the case of the Tarrasch Variation at least, the position soon re-opens, after what is regarded by theory as best play for Black (at least) , after the sequence 1.e4 e6 2.d4 d5 3.Nd2 Nf6 (3...Nc6 is another way to close the game, but theory regards it as more suspect) 4.e5 Nfd7 5.Bd3 c5 6.c3. That is, best play (for Black at least) is considered 6...Nc6 7.Ne2 cxd4 8.cxd4 f6, when the game opens up again. Here I would guess White's (nominal?) edge at the moment, at least, is worth at least 0.33333333...

[Gary Ruben]

Re: Valuation of W's Opening Move 1.e4?

The French Defence is a way to try to avoid open positions after 1.e4, as White's best lines tend to involve a closing of the centre.

The Winawer should be banned. The only time to play that is if your opponent is a fish or you think he's a fish.

[Kevin Pacey]

Re: Valuation of W's Opening Move 1.e4?

The Winawer should be banned. The only time to play that is if your opponent is a fish or you think he's a fish.

In my largest database, overall White scores a bit better than is average for all 1.e4 lines/openings after 4.e5 against the Winawer. I'm sure Winawer/French Guru IM John Watson would find a way to spin this, even after what he suggested in 'Secrets of Modern Chess Strategy' about finding a (first) move superior in practice if it scores consistently well at all levels.

[Kevin Pacey]

Re: Valuation of W's Opening Move 1.e4?

It isn't right. And that's not surprising. Engine evaluations are ordinal numbers -- all they do is help it rank the variations so it can choose the highest-ranked move.

I think what you are looking for is a cardinal number evaluation: an absolute measure of the advantage/disadvantage for W/B. The only time engines give you that is when they find a forced draw, in which case (by the usual programmers' convention) the ordinal number matches what its cardinal numbers evaluation should be (0.00).

FWIW, at 20 ply, Houdini ranks White's best moves as:

1st : d4
2nd : e4
=3-4: Nc3 and Nf3

In cases of positional (or even forced?) draws, a purist might give the superior side (e.g. White) a nominal numerical edge. For example, in a pawn ending that's drawn because of the defender having the opposition, the 'superior' side might have its king on the fourth rank, vs. the defender's being on the third. Computers assess such positions as 0.00, though perhaps they shouldn't if the programmer is a purist.

In such a case, perhaps the program should really display both the cardinal/(absolute result) evaluation with a word "draw" (rather than with the number 0.00) and, I would suggest, an ordinal/(symbolic) evaluation of, say, 0.3333333..., which we can still debate whether is +/= or =. Such an endgame could even be the result of the/a 'perfect' game, whatever that is. In that case, the evaluation of White's initial 'edge' could actually stay the same from the start of such a game to its finish.

Note that there are some endgames with material advantages for the 'superior' side that cannot be won, against good defence, which even today's programs fail to assess as 0.00, at least until a repetition has occured, if not 50 moves without a pawn move or capture.

[Kevin Pacey]

Re: Valuation of W's Opening Move 1.e4?

Yes, the opening book is off. I am just using the usual analysis feature in the opening.

For the " perfect " game after 1.e4, Fritz gives ( using depth 20 or more, unless otherwise indicated ):

1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.0-0 Bc5 5.Nxe5 Nxe5 6.d4 a6 7.Be2 Bxd4 8.Qxd4 d6+/= ( 0.64 - depth 20 )

For what it's worth, the latest ECO gives 4...Nxe4 (leading to the Berlin Lopez main lines) as the only move that avoids letting White get any sort of clear edge with best play. In the Berlin main lines, White can avoid '=' in ECO, but has to settle for positions evaluated as unclear, with good play by Black. 'Unclear' is one evaluation that computers don't clearly have a numerical evaluation for. [edit: nor do they ever show an unclear symbol in place of, say, '=', as far as I know]

[Dan Scoones]

Re: Valuation of W's Opening Move 1.e4?

[QUOTE=Gary Ruben;47154]The Winawer should be banned. The only time to play that is if your opponent is a fish or you think he's a fish.[/

If the Winawer is so bad for Black, shouldn't White be trying to promote it instead of having it banned? Just wondering.

[Gary Ruben]

Re: Valuation of W's Opening Move 1.e4?

[QUOTE=Dan Scoones;47197]

If the Winawer is so bad for Black, shouldn't White be trying to promote it instead of having it banned? Just wondering.

It's often OK for higher rated players against lower rated players. It's also OK against players who have some problems playing against that defense for whatever reason. Mostly when playing the Winawer against a higher rated competent opponent who knows what he's doing it turns into a routine loss.

Would you play it against an IM or GM or would you find something else?

Maybe you mean why would I not try to promote the Winawer. The reason is when I watch a Canadian playing a foreign player I like to see the Canadian win. So why not pick something which gives better chances?