Missed Opportunities

TODAY was the double-round day at Jersey - necessary to fit nine rounds into a Saturday-to-Saturday schedule - and it turned out for me to be a day of missed opportunities.
First up this morning in round three I had Black.

Black to make his 31st move in Alek Safarian (2005) - Spanton (1914)

*****
*****
*****
*****
*****
*****
*****
31...Rh2??
A ridiculous move that removes any danger from White's king while exposing Black's weak kingside pawns.
I wanted to play 31...g5 but rejected it because of 32.Ne4. What I missed, but Stockfish10 and Komodo9 see instantly, is that Black has 32...Nxe3! as the obvious 33.Bxe3 runs into 33...gxf4. Black is then temporarily down a bishop for two pawns, but threatens to win the e3 bishop or fork White's king and g2 rook, in either case giving Black a winning advantage.
After the text, the game continued ...
32.Rxh2 Nxh2 33.Ne4
… when White's bishop-pair, combined with Black's weak pawns and uncoordinated pieces, gave White a winning advantage (1-0, 53 moves)
***********************************
MY reward for being the only player in the Open on nul points was to be upfloated against the 12th seed in round four.
Spanton (1914) - Paul Curtis (2186)
French Rubinstein (Fort Knox Variation)
1.d4 e6 2.e4 d5 3.Nc3 dxe4 4.Nxe4 Bd7!?
The Fort Knox has a super-solid reputation but seems a strange choice against someone rated 272 elo lower. Stockfish10 at first rates Black's fourth move as giving White a winning position, but soon calms down.
5.Nf3 Bc6 6.Bd3 Nd7 7.0-0 Ngf6 8.Qe2
It is more popular to avoid exchanges with 8.Ng3.
8...Be7 9.Re1 Nxe4 10.Bxe4 Bxe4 11.Qxe4 c6 12.Ne5!?
The engines do not like this, preferring 12.Bf4.
12...Nf6 13.Qf3!?

Position after 13.Qf3!?

This seems to be a novelty. The idea is to meet 13...Qxd4 with 14.Nxc6, when the engines agree with what I thought was the main line, namely 14...Qd5 (14...bxc6?? loses the a8 rook after 15.Qxc6+) 15.Qxd5 Nxd5 16.Nxe7 Kxe7, when White has bishop-v-knight and a queenside pawn-majority, but Black has no weaknesses.
13...Qd5 14.c3
The engines reckon 14.Qxd5 cxd5 is fine for White, but I suspect PC would have fancied his chances of turning a minority attack into a full point.
14...Qxf3 15.Nxf3 0-0 16.Bf4 Rac8 17.Rad1 Rfd8 18.h3 Nd5 19.Be5 Bf6 20.Bxf6 Nxf6 21.Ne5 Rd6 22.Nc4 Rd5 23.a4 Rcd8 24.b4 b6?!
This may look aggressive because of the line-up of Black's rooks on the d file, but in reality it is weakening.
25.Ne5
Even stronger is 25.a5, which I looked at but was not sure about after 25...c5? 26.axb6, although with the help of the engines it is fairly easy to see that Black is in big trouble, eg 26...cxd4? 27.bxa7 wins for White, not least thanks to the threat of Nb6. The engines reckon Black's best after 25.a5 is the horrible 25...b5, when White is clearly better after 26.Ne5.
Maybe if I had not had such a disappointing start to the tournament, or had been playing someone nearer my rating, I would have looked harder at these lines instead of concentrating on avoiding defeat.
25...R5d6 26.Nc4 Rd5 27.Ne5 R5d6
While I was thinking about my next move, which would almost certainly have been 28.Nc4, PC offered a draw, which I accepted.