KoKko wrote on 02/21/10 at 10:10:49:


With 1.a3 you may lure the machines in some open variations of the sicilian, where the move a2-a3 is plausible. But always remember that the sicilian is a typical counter opening.
I was quite impressed when I once read about the “Reflex-Theorem”, a term invented by Rainer Schlenker, a German off-beat-guru (once publisher of the mag-/fan-zine “Randspinger”). I found an explanation in Keilhack’s “Die Tarrasch-Verteidigung”, only available in German - Uuh oh (you know what I mean if you are familiar with the discussion right now taking place in the 1.d4 d5-forum on Tarrasch as black and white).

That “Reflex-Theorem” resembles me to the martial art aikido where you use the power your opponent exerts to defeat him.
In chess that means: Every move you make has pros and cons (e.g. g2-g4 may attack but at the same time it weakens king’s side). So with every move you give informations to your opponent, you show your active ideas and you concede slight weaknesses (you may not feel at that time!).

But strong players can base their whole game on the cons. I think Petrosian played like this. With prophylaxis he tried to nip any attacking ideas of the opponent in the bud, and then he shifted his attentions to the weaknesses left behind...

Coming back to 1.a3 e5 2.c4 Nf6 3.Nc3 (which in fact is a position known to theory by an other move order) and now say 3...d5 4.cxd5 4.Nxd5 we have reached a sicilian reversed with extra a2-a3. Wow! Wow? An aware player on black’s side knows what this means: Things to come like b2-b4. A ...0-0-0 would be very dangerous up to selfmurder. You gave him these informations right in the first move. So he will play patiently castling short, being aware of attacks on e5 with b4-b5 thus perhaps playing f7-f6 and so on. In the sicilian you need an opponent who overextends. With a tempo down simply the lack of free time makes this less likely. So you lack possibilities to use the "Reflex-Theorem" regarding his overextentions. 

But of course you can still play on for a win. I was totally stunned by the game Fischer - Andersson, Siegen 1970 (not at the olympiad but at an exibition), where the future WC scored a wonderful win.  No, Fischer didn’t play 1.a3, he was a bit more to the right, it was 1.b3... and only 4.a3

My experience is that humans behave quite differently when playing white or black. As black they play much more careful as with the white pieces. It is interesting how rarely especially in the lower to middle ranks a player with black takes the whole center, when he is allowed to do.  TN exactly highlighted this when he asked what you would do after 1.a3 g6.

Yes once and a while there are players who take the center with e5 and d5. More often they play some sort of slav or Tarrasch set-up, that is pawns on d5 and e6 combined with c6 or c5. As said, playing with black is much of psychology. With black you draw (hey, I was a tempo down...), with white you win (hey, I am a tempo up). So they are (too) careful, play modestly and your extra tempo may evaporate (which is not a total breakdown - just wait for the zeitnot and see...).


KoKko wrote on 02/21/10 at 10:10:49:


Despite the above I continue to be somehow “infected” with 1.h3. But frankly said I wouldn’t play the exact continuation your engine gives as critical, cause I think it isn't. 1.h3 e5 2.e3 d5 3.d4 is a reversed french. And I feel white has yet given quite useful informations to black (Reflex-Theorem!). With 3...e4 it is an advance variation. You can proceed in normal manner (c2-c4 etc.) but you will have problems to nibble e4 with f2-f3 (which in the normal advance variation is on the schedule in some lines) due to the weakened h4-e1-diagonal - you gave the first part of this information right in your first move! 1.h3 weakened your control over g3 a bit. 
Besides black could also play 3...exd4 4.exd4 thus transposing into a french exchange variation with the unusual move 4.h3. 

But I think you can choose many other approaches - nearly all open or semi open openings at hand in a reversed form. 1.h3 e5 and now:

2.c3 (Caro-Kann, there the move h7-h6 is often useful)
2.d4 (scandinavian, there black’s queen’s bishop often needs a refuge on h7)  
2.e4 (all sorts of open games, and h3 stops the possible pin Bc8-g4)
2.c4 (all sorts of sicilian, h3 might be useful in quite some of the variations)  
2.Nc3 (Nimzovich, in variations comparable to ...d5, e5 Bf5 you have a refuge for the bishop; in variations comparable to ...e5 the pin Bg4 is prevented)
2.Nf3 (Alekhine, h3 stops Bg4)
2.d3 (Pirc, hm, at least h3 stops reversed Bg5 variations, but on the other hand it gives rise to a reversed 150-attack...)
2.b3 (well... perhaps hippopotamus)
2.g3 (well... perhaps hippopotamus)
2.a3 (Aaaah... yes, Basman live)

An equivalent list can be made for 1.h3 d5 or 1.h3 Nf6. But the more we move towards a noncomitting way of play by black (regarding the center and the resulting levers etc.) the more it is just muddy waters. Again 1...g6 may be a very good try against 1.h3 - you just spoiled a tempo if you want to lever the future fortress of black’s king with the h-pawn. On the other hand you’re right in the tracks for g2-g4... Oh, my dear...

One way or another, your opponent knows that you have played h2-h3. He will not give you all the cake... Reflexem-Theorem!

cheese

So what are the critical lines after 1. e4 and 1. d4 according to Rybka (after a week, or perhaps after months as in the earlier 1. c4-versus-1. Nf3 analysis), and what fraction of a pawn is White up in each case?  Enquiring minds need to know!

It might be mentioned that "regular" opening books sometimes mention 1. a3 (sometimes called "Anderssen's Opening"; he played it a few times against Morphy).  Incidentally ECO did give 1...g6 in reply to it.

Against KID players I just pretend to play black and patiently wait for thier plans to show...maybe trying to transpose into some english lines.
1.a3 Nf6 2. Nf3!? g6 3. g3 (or c4 followed by Nc3)

I analyzed some positions with engines, and it seems a3 and h3 have the same "strenght"  maybe a3 is more balanced against all lines, h3 is likely to be more effective in positional games (d4/Nf3 lines); its only drawback... it can't avoid reversed Ruy Lopez.

For the sake of statistic Rybka shows about +0.04 for a3 (critical lines...a3 e5 c4 Nf6 Nc6 and a3 d5 d4 Nf6 Nf3 e6 Bf4), and 0.00 for h3 (critical lines h3 e5 e3 d5 d4 Nc3 and h3 d5 d4 c5 e3).
One week analysis.

Anyway yesterday at the club I easily beaten some unaware opponents, in one game I even won early his queen thanks to b4 push in a Queen's gambit 

Maybe you guys are right, I got to just paint black my white pieces

I had a 2300ish team-mate called Xavier Delebarre who always opened 1.h3 with White. Perhaps you can check his games in a database? he is very experienced as a player, knows classical theory pretty well and worked hard to send opponents into foreign territory (with good success except against semi-slav iirc).
As for 1.a3,  you may end up a full tempo down in most KID lines, so perhaps you can show a little flexibility and play 1.h3 against KIDsters?
These little tricks are more likely to bamboozle an e4 player relying on "book theory", while a d4 player doesn't care which side of the position is equal. 

Realistically if one just wants to play chess then 1. a3 followed by treating the game like you're black isn't a bad thing.

That said, I'd rather try to put pressure on black. If you're scoring that much higher with black it's probably a good sign that your ability to play with the initiative needs work.