Mysteries of the Jerome Pawns

I wish that I had won the following game because I solved the "mysteries" of the "Jerome pawns" discussed after move 20. Alas, I took advantage of my opponent's mistake and slid into a drawn endgame, feeling lucky... 

perrypawnpusher - chingching
blitz, FICS, 2011

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+

4...Kxf7 5.Nxe5+

According to The Database, which I consulted after the game, up until now, chingching was 9-1 against the Jerome Gambit. However, every one of those games saw 5.d4 instead of 5.Nxe5+, so my move was going to at least give my opponent something new to look at. 

I was interested to see that chingching had also played 16 games (scoring 75%) defending an Italian Gambit, 1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.d4, which transposed to the Jerome with 4...exd4 5.Bxf7+ (an idea as old as Wright - Hunn, Arkansas, 1874). Among the Jerome Gambit Gemeinde, DragonTail is the top practitioner of this line, and we have previously seen DragonTail - chingching, blitz, FICS, 2009.

5...Nxe5 6.Qh5+ Ng6

7.Qd5+ Ke8 8.Qxc5 Qe7 9.Qe3 d6

10.0-0 Nf6 11.Nc3 Rf8 12.f4 Kf7

Continuing to castle-by-hand. Instead, 12...Ng4, as in perrypawnpusher - MRBarupal, blitz, FICS, 2010 (0-1, 31) is not as strong.

13.f5 Ne5 14.d4 Neg4

As often happens, it is tempting to harass the White Queen. I suppose this is because White's opening seems so artificial that certain allowances can be made in how Black responds. Still, 14...Nc6 was for choice, as White can now develop some initiative.

15.Qe2 Kg8 16.h3 Nh6 17.Bxh6 gxh6 18.e5 dxe5 19.dxe5

19...Qc5+ 20.Kh1 Nd5

This is a critical position for the game, as Rybka 3, in the post mortem, saw Black's last move as a blunder (recommending 20...Ne8 instead).

The point is that the "Jerome pawns", now connected and passed, work very well with White's heavy pieces after  21.Nxd5 Qxd5 22.Rad1

analysis diagram






See?

I admit that I still did not, and it took some work to see why Rybka saw White as more than a piece better.

For starters, White is threatening e5-e6 and f5-f6 which combine with Qg4+ in a mating attack. That suggests that Black's best retreat for his Queen will be 22...Qf7, even though that still allows White to play 23.e6.

If Black uses his Queen to blockade the forward pawn with 23...Qe7, then 24.Qe5 will force 24...Bxe6 25.fxe6.

If Black uses his Queen to blockade the trailing pawn with 23...Qf6, then White has a tricky combination: 24.e7 Re8 25.Rd8!? when 25...Rxd8 is met with 26.Qc4+ and Black has to either give up his Bishop with 26...Be6 or face the other advancing Jerome pawn with 26...Kf7 27.exd8/Q Qxd8 28.f6+ Kf8 29.Qc5+ Qd8 30.Qe3 Be6 31.Qxh6+ Ke8 32.Qxh7 Qf8 33.Re1 Qxf6 34.Qg8+ Qxa8 when White is up the exchange and two pawns.

Wow. That is so far over my head...

21.Qh5 Nxc3 22.bxc3 Qxe5 23.Rae1 Qf6

White's position is only a shadow of what it was 3 moves ago. His Kingside "attack" is also doomed to failure.

24.Re3 Bxf5 25.Ref3 Qe6

An odd slip by opponent, dropping a piece, when 25...Qg6, instead, would have sealed his advantage.

26.Rxf5 Rxf5 27.Qxf5 Qxf5 28.Rxf5 Rd8

We have reached a drawish looking endgame.

29.Rb5 Rd2 30.Rxb7 Rxc2 31.Rxa7 Rxc3 32.a4 Rc1+ 33.Kh2 Ra1 34.Rxc7 Rxa4 35.Rd7 Ra2 36.Re7 Game drawn by mutual agreement