1. e3

As in chess, the moves 1. e3 and 1. e4 in atomic open up diagonals for the queen and kingside bishop. In chess, however, 1. e4 is to be highly preferred over 1. e3 because of principles involving occupation and control of central squares. This may or may not be the case in atomic chess, for which the verdict is still out as to which is the better move (if either).

With 1.e3, White is making an immediate threat to win the game. The threat is 2. Qh5 g6 3. Qd5. On d5, the White queen will be attacking d7 and f7 simultaneously. White can also get to d5 via the f3 square in some lines. This is one advantage of 1. e3 over 1. e4 (a pawn on e4 would block a queen from moving to d5 from f3). Another advantage of 1. e3 over 1. e4 is that 1. e3 is also a defensive move in that it blocks the a7-g1 and h6-c1 diagonals leading to the vulnerable attack squares d2 and f2. Yet one more advantage is that an attack on a pawn on e4 can stop a piece (like a queen) from advancing to a central square (like d5), lest it get caught in an explosion. All that being said, 1. e4 has its advantages over 1. e3 as well.

Of Black's twenty legal responses to 1. e3, the following ten do not address the threat of 2. Qh5 g6 3. Qd5: 1. ... a6, 1. ... a5, 1. ... b6, 1. ... b5, 1. ... c5, 1. ... f6, 1. ... h6, 1. ... Na6, 1. ... Nc6, and 1. ... Nh6. All ten lose to 2. Qh5 g6 3. Qd5 1-0. Also, losing outright is 1. ... g5 after which 2. Qh5 wins immediately.

Two more moves that grant White a quick mate are 1. ... g6 and 1. ... h5. Let's look at

1. ... g6

Here White wins with a quick attack on f7: 2. Qf3. Black can now protect f7 by either advancing the f-pawn or by playing Nf6. But if Black plays 2. ... f6 or 2. ... f5, then White wins with 3. Qd5 (attacking d7 and f7 as mentioned earlier). So: 2. ... Nf6 3. Nh3. Now, White is threatening to move his knight to g5, and the boxed-in Black king won't be able to run from the ensuing explosion centered at f7. Futile is 3. ... h6 (to "protect" g5) because White would play 4. Ng5 anyway, because if 4. ... hxg5, then the Black knight on f6 is gone and White captures f7 with his queen. Also futile is trying to make some breating room for his king with 3. ... d5, since White could continue with 4. Ng5 Kd7 5. Bb5+ (or 5. Ne6) 1-0. The best Black can do is prolong with
3. ... g5 4. Nxg5 f6 (or f5) 5. Qh5#. Note that 1. ... h5 leads to the same kind of winning attack: 2. Qf3 Nf6 3. Nh3 (actually 3. Bd3 wins here too) g5 4. Nxg5 f6 (or f5) 5. Qd5 1-0.

So far, we've seen that 13 of Black's 20 legal responses to 1. e3 end in a quick mate. The other 7 don't lead to quick forced mates, but some lead to losing positions, nevertheless. For example, consider:

1. ... c6

This move does protect the d5 square, from which we've seen that the White queen can wreak havoc, but it gives White another nice tactic: 2. Qh5 g6 3. Qe5 e6 4. Qc7 Qxc7 +- Here, all of Black's moves were forced and he finds himself down a bishop and knight without any compensation. Note that 1. ... d5 leads to a similar fate: 2. Qh5 g6 3. Qe5 Be6 (this saves Black a piece over the inferior 3. ... e6) 4. Qxc7 +-. This time Black is down a pawn and knight for no compensation (or almost none).

The next move we'll consider is

1. ... e5

This move is a bit tricker to dispense with, but we shall dispense with it on the grounds that it also leads to noncompensated material loss. First, do not be tempted to play 2. Qg4 f5 3. Qc4, attacking both c7 and f7 because now Black doesn't have to respond with d5 to guard f7, instead he can respond with 3. ... Qh4, attacking f2 so that White has no time to move his queen to f7 for the mate. So, after 3. ... Qh4, Black has forced the trade of queens for a roughly equal game. Instead, White should proceed with 2. Qf3 Qf6 (2. ... Nf6?? 3. Qf5 1-0 or 2. ... f5?? 3. Qd5 1-0). Now, despite the superficial appearance of equality, White has an enormous edge because of the privilege of the first move. He continues: 3. Nh3 White is threatening to move his knight to g5 from where it will do much damage. Black can do nothing to stop it. For example:

(A) 3. ... Nh6 4. Ng5 Kd8 5. Nxh7 (or 5. Nxf7 Ng4 6. Qxg4 +-) 5. ... Qh4 6. g3 Qxg3 +-.

Or:

(B) 3. ... Bb4 4. c3 Qf5 5. Ng5 f6 (or Kd8) 6. Bb5! (not 6. Nxh7?? Qc2 0-1) 6. ... c6 7. Nxh7 +- .

Now, if 7. ... Qc2 then White has 8. O-O. Or, similarly:

(C) 3. ... Qf5 4. Ng5 f6 5. Bb5 c6 6. Nxh7 Bb4 c6

and we have transposed to (B).

If you're counting, we're down to four responses to 1. e3. All 16 considered so far have been shown to be bad at best. Another response that is probably not ok (but just by a little) is

1. ... f5

This move, played by mrundersun for a while gives White a couple of good options to proceed. He can play into a Vlasov position (which normally arises after 1. Nf3 e5 2. Ng5 f5) with 2. Nf3 e5 3. Ng5, from which White can probably gain a comfortable advantage if he isn't careless, or he can win a quick pawn like this: 2. Qh5 g6 3. Qh4 Nf6 4. Bd3. Now, Black must play e6 or e5 in anticipation of Bxf5: 4. ... e6 5. Bxf5. Now White can probably hold on to his slight advantage. Also worth exploring is 4. Nf3 instead of 4. Bd3, but don't think it is as good.

Next up:

1. ... d6

This move has an interesting history. For a long while it was considered playable until a very interesting tactic arose that destroyed the move's credibility. Today it is still considered bad and no one plays it, but not many realize that it's probably not as bad as they think it is. In August of 2006, mrundersun (at FICS) played 1. ... d6 against me several times in quick games (3 0 or 2 2) and won 40% of them (mostly due to my blunders). He was altruistically testing the opening out to see if it could be saved, and his conclusion at the time was that at best White could gain a pawn for no or little compensation. This opening is a little involved and we will explore it in a chapter of its own.

The final two moves to consider are 1. ... e6 and 1. ... Nf6, which are virtually the only two moves anyone plays anymore with 1. ... e6 being highly preferred over the other. 1. ... Nf6 needs to be analyzed more, but right now, it seems that it confers a slight edge to White. In most lines, White gains a pawn and hangs on to it comfortably. We'll explore it in its own chapter. We will also explore 1. ... e6 (the symmetric variation of the Rekursiv-Hooby opening) in its own chapter, which will undoubtedly give rise to many other chapters since there is a lot of known theory after 1. e3 e6. For now, if your opponent plays 1. e3, I recommend responding with 1. ... e6. The reasons why 1. ... e6 is the most solid response is because of its defensive powers. It protects d5 and it closes off the dangerous diagonals leading to d7 and f7. It also opens up the Black queen and king's bishop for development and counterstrikes.