Awesome Van't Kruijs !?!

Game Explorer gives bad stats for Van't Kruijs 1.e3 but Rybka evaluates it as high as +0.04. I'm trying to discover reasons of this discrepancy. Many good openings listed below have no names LoL.

2nd ply: 1.e3 Nf6 and 1.e3 g6

3rd ply: transitions to QP:Indian 1.e3 Nf6 2.d4 and 1.e3 g6 2.d4;  English transition 1.e3 Nf6 2.c4; 1.e3 Nf6 2.Nf3 and 1.e3 g6 2.Nf3

4th ply: transitions to QP:Indian 1.e3 Nf6 2.d4 g6 and 1.e3 g6 2.d4 Bg7; transition to Gruenfeld 1.e3 g6 2.d4 d5; transition to English: Great Snake 1.e3 Nf6 2.c4 g6; transition to English: reversed Sicilian 1.e3 Nf6 2.c4 e5; 1.e3 Nf6 2.Nf3 g6

... to be continued

I see that your rating is much higher than mine, so you certainly know that trusting a computer in such an early stage of the game is misleading.

And even if you want to take him seriously, what does +0,04 means? W has 1/20 of a pawn advantage? Nothing on  earth can calculate with such a precision the real value of a position.

And even if the program is right, then there is no difference betweeen +0,04, -0,07 and similar stuff. It always means equal, since such a little advantage is clearly irrelevant.

For example most programs evaluate 1 d4 d5 2 c4 dc4 as "-0,something". Do this means something apart that in such a positional context you can't really say what's going on after 2 moves? Most computer on this basis also prefer moves such as 2 Bf4. But if you feed the machine with more moves, it will start to prefer 2 c4 over 2 Bf4.

Bach to 1 e3, well it's hard to imagine that it's such a bad move that B is now better. Your program confirms that by saying +0,04, which means =. This is reasonable, since the move is not very flexible and is likely B can benefit from this to reach easy(er) equality.

You're certainly aware of all this, so my real question is: what are you trying to say? i can't figure it out. This is not a criticism, i only need an explanation because (probably because of my limited chess skills) i'm a bit cinfused,

 Not necessarily, on 1...d5 there is a good answer 2.c4.

Bresando: I work with given data provided by a well-known chess engine. If there are better evaluations, I'd be glad to examine them. With your approach of "almost everything is almost equal up to insignificant 1/n of a pawn" U can't go nowhere.

Conzipe: Hmm ... okay, it seems that U are essentially right.

1.e3 opens a diagonal for both the queen and the king bishop, so it does assist your development. The main flaw is that 1.e4 does just the same, and stakes much more of a claim to the center. The pawn on e3 also hinders development of the c1 bishop unless white plans to develop it to a3 or b2, both of which require b3 at some point.

Anyway, it's one of the openings that will really screw with your opponent's mind if he or she happens to be one of those who will only expect serious tries for an advantage from white and get all flustered if they are gifted equality. (Me, for instance, though I hope I am working on that.)

EDIT: You are trying to find reasons for the discrepancy between the Rybka score of an opening move compared to its scoring in a database? Easy. Humans. Starting from the hypothesis that 1.e3 does basically gift equality to black whereas black is usually accustomed to fighting for it, why does black score significantly more wins than white? I'd say it might mainly that under normal circumstances, it wasn't (databases record the history of an opening, not its objective worth) the kind of opening White would choose to win, not generally at least. Many players perhaps played it against opponents they believed superior in opening theory and perhaps hoped to evade that, some players might have played this in an attempt to snooze their way to a draw only to face someone who was not inclined to let them, and all other things being equal, including skill, my belief is that the player who genuinely wants to win is more likely to win than the player who is content with a draw. The White players who genuinely do want to win have a great many openings to choose from that do provide them with opening advantages of many kinds, and will tend to choose one of those; which is not to say that there aren't a great many White players who will play 1.e3 fully intent on trying to win - Kasparov, famously, did that against Deep Blue. It also looks like something Nimzowitsch might have played, often.

Silfir: I agree with many of your points, thanx for this detailed answer.

So I wonder why 1.e3 is named after Maarten van 't Kruijs (Kruys, pronounced vant cries)?  He played it by transposition, but after he played 1.a3, Anderssen's Opening, against Anderssen.  1.e3 was probably first played in a correspondence game between Amsterdam and Rotterdam in 1824.  Staunton was playing it in 1843 against St. Amant.  Zukertort played it several times against Anderssen and others in the 1860s.  It was also played by Bird, Chigorin, Blackburne, Schlechter, Maroczy, Tartakower, Nimzovich, Capablanca a few times, and Kasparov against computers.  Most books give van 't Kruijs name wrong.  It is van 't (with a space) and not van't (without a space).  the 't is short for het, so his name is Maarten van het Kruijs (of the cross).  He was the first person I can find who played the Larsen's opening, 1.b3.  He was the first to play, organize and win a random chess tournament similar to Fischer Random chess.  I cannot find any games where he opened up with 1.e3.  But he gets the credit somehow.  Against Anderssen in 1861, van 't Kruijs played 1.a3 e5 2.c4 Bc5 3.Nc3 a5 4.e3, and won, but not due to the opening.  And he did not invent 1.a3, as Anderssen played that opening against Paul Morphy three times in 1858 (and 1.a3 was played even earlier by Kieseritsky).

Ok i'm not saying that my argumentation proves something :)

There are alternative evaluations: the "verbal evaluations" given by masters for example.  Accordins to these, e3 is suboptimal because B can choose a setup in which e3 is not an usual W move. Looks reasonable.  Do this mean that we have to blindly trust masters word? Of course not, but if we want to argue we must provide a valid counterargumentation. It is widely recognised that computer's ability to evaluate a quiet position is not so hight. Given that practical evidence(databases) says that e3 is not such a great move, and that there is a plausible reason for this, simply posting "+0,04" says very little. 

This is not meant as an attack against your work! Only if you're trying to evaluate early openings moves via engine outputs, in my (maybe mistaken) opinion, you're wasting your time.

Billwall: thanx for this nice historical survey.

Bresando: No problemos. I just found it funny to look at this stuff today, I'm not planning to spend my whole life on these engine evaluaions.

Echecs06: always at your service, my general!

Your head is full of BS spaghetti. Don't troll in my awesome threads, go create yours. Hasta la vista, baby!

There isn't a sound in the English language that comes close to the pronounciation. The "j" in the word is just an old way of writing it and can be removed. So just the word kruis (cross).
Also not van't or vant. But "van het" (from the).

Catalyst

Kasparov did play 1.e3 against the computer (Fritz).  It wasn't silly as it got out of book.  It took me 15 seconds to find the games.  He played it three times, winning 2 and losing 1.  You did not do your homework.  Shame on you.  Here is one game.  You should be able to find the other two.

http://www.chessgames.com/perl/chessgame?gid=1070735

He also also played e3 by transpostion (1.b3, 2.Bb2, 3.c4, e3 - same as Fischer) twice.

QUOTE: Game Explorer gives bad stats for Van't Kruijs 1.e3 but Rybka evaluates it as high as +0.04.

A rating of 4/100th (1/25th)  of one pawn doesn't strke me as a very high rating for WHITE's first move!

Engines deal with centipawns and to trust their evaluations too much will lead one down the wrong path imo. There is a lot of dogma in chess and whether it's dogma of humans or engines it can only hinder one's progress. Engines also arent known for playing openings exceptionally well without their "opening book" so I don't personally put much faith in their evaluations of the first few moves in a game. I have seen strong engines give an advantage of as much as 2 pawns in dead drawn positions so when one tells me something like  = ( 0.21 )  I figure the position is equal. I prefer human evaluations such as : equal, unclear, white, or black is a little better, white or black is much better, and white, or black is winning...... I understand these evaluations, though I may not always agree with them.... I do NOT understand  - ( 0.45 )  WTF does this mean exactly ?? 

I was the dunderhead who wrote he used it against Deep Blue when he only used it against Fritz; I apologize. For the record, isn't 1.d3 even worse then 1.e3? :)

"too weak to challenge" - 1.e3 potentially gifts black equality, but it hardly gives black the advantage, I don't think. Equality doesn't mean that White cannot win the game anymore. Without a properly prepared opening book, it might well be that Fritz wasn't even capable of seizing equality back then, though.

Anyone know when was the last time a 2600+ player used 1.e3 against an equally strong opponent?

Catalyst,

Sorry to be so strong in my statement.  I just wanted to be accurate and did research 1.e3 as well as van 't Kruijs on what I could find. 

Silfer,

I couldn't find any recent 2600 players playing it against equally strong players, but did find at least six 2500+ players playing it against 2500 and 2600 players.   The highest rated was Bojan Kurajica at 2554 playing it several times against 2500+ players.  Next is Pavel Blatny at 2550, beating other 2500 players.  Aurellen Dunis is 2534 and he drew with Ivan Sokolv at 2689 and Dorfman at 2595.  Ronen Har Zri beat a 2562 player with 1.e3.  Other 2500 rated players that play 1.e3 and beat lowly 2400 players include Sinisa Drazic  (2533) and Tigran Nalbandian (2505).  And I have found dozens of 2400 players and a lot more 2300 players who play it and win.

There are some interesting comments in this four-year-old thread.  Using my own chess engine, I am surprised to see that 1. e3 gets the same +0.07 CAP assessment as 1. d4, implying that 1. e3 is just as good as 1. d4.  However, in my database of 3.6 million games, the winning percentage for 1. d4 is 55%, while the winning percentage for 1. e3 is a dismal 41%.  It is also interesting to note that the winning percentage for Black with 1...e6 in response to 1.e4/d4/c4 is 46%, giving the impression the move gives significantly better results when playing it a tempo down.  By the way, you can't attribute the low percentage for 1. e3 to a small sample size, as the database has close to four thousand games with 1. e3.

So what gives?  You can certainly argue that 1. e3 is unambitious, but how can such an innocuous move produce such a weak winning percentage?  I suspect the answer has to do with the type of player who will venture 1. e3.  Grandmasters would consider such a move to be much too undignified to play themselves, so I'd expect most practitioners of 1. e3 to be lesser players who want to experiment a little.  Of course, since these 1. e3 players aren't GMs, they bring down the winning percentage. 

There is also the issue of playing for advantage with White.  GMs would probably regard 1. e3 as immediately conceding any attempt to gain an opening advantage.  Many lesser players don't think that way, i.e. they believe that taking the game down a less traveled path gives a better chance to gain an advantage.  Unfortunately, in the case of 1. e3, it isn't a challenging move.  Upon seeing 1. e3, I'd imagine most opponents will simply transpose to an opening they are comfortabe with.  And, below the 2400 level, I think the player with the advantage out of the opening, winner isn't the one with the theoretically better position, it's the one who is more comfortable with the position.