NeverAgain: "Moscow UM" was a minor 11-round international tournament held in January '68, according to RusBase: http://al20102007.narod.ru/it/1968/...
Most of the participants were from the Soviet bloc countries. Bagirov won, while the enterprising Simagin ended up a tail-ender.4.Qf3 - an unusual choice, aiming to unbalance the game. The standard 4.Bc4 leads to a more sedate play. In the 343 games with 4.Qf3 in Megabase 2012 White scored only 45% (+105=97-141), however one of the notable White wins featured Baburin (in his IM days) on the Black side - Z Basagic vs Baburin, 1991 - so the line shouldn't be dismissed out of hand.
4...e6 - 22 years later Bagirov essayed an immediate 4...Nb4 against lesser opposition - J Lapinski vs Bagirov, 1990
7...Na5 - quite non-standard opening play, wouldn't you agree? Believe it or not, this is by far the most common move in this line, and it has scored an impressive 60% for Black. Everything else (like the solid-looking 7...Be7) has fared worse and is much more rare.
11.Ne3 - not sure what the point of the whole transaction was. Black now got the bishop pair. Perhaps Simagin was aiming for an offbeat position and wanted to take his theoretically primed opponent into the woods? He may have been also counting on getting play on the queenside, with Black's two minor pieces poised precariously on the edge of the board; however, Bagirov expertly parried all his thrusts.
12.Qh5 - gaining time on the a5 Knight and edging closer to the black King. I would prefer 12.Qg3 instead, trying to hinder Black's kingside development; although after 12...b6 (or even 12...b5!?) and 13...Qd6 Black has no problems.
15...Qg6 - Black offers a Queen exchange to forestall any unpleasantness on the kingside. Note the unusual Knight deployment on both sides; Black's is even fianchettoed, although it's quite ready to jump into the center via d6. Here White might have accepted the offer and try to exploit the weakening of Black's kingside by positional means: <16.Qxg6 hxg6 17.h4 Δ Nf4>.
17.Nf4 - <17.0-0> looks more natural, but I guess Simagin didn't aim for natural with his opening play.
19.d3 - <19.Be5> seems slightly more preferable here, forcing Black to castle queenside into a pawn storm underway, or to abandon any idea of castling altogether.
19...0-0 - the immediate <19...f6>, preventing <Be5> may have been objectively better, but I think Bagirov was consciously baiting his opponent at this point.
20.h4 - Simagin goes for it and sends his h-pawn on a death march.
24.0-0 - content to create what he must have perceived as kingside weaknesses, White castles into it. True, Black's KS pawn formation looks like swiss cheese, however White lacks the light-squared Bishop to exploit it.
27.Ng4?! 
- still intent on an attack (and incidentally lending support to the forlorn h6 pawn). Despite his preceding leftfield play White could have kept things in the equality zone by starting to centralize his Rooks with 27.Rfe1 instead. Neither Komodo 9 (d=30) nor Stockfish 6 (d=35) seem to worry about the potential loss of the white h-pawn.
28.Ng3?! 
- a completely illogical retreat. Of course, something had to be done about the threat of ...Re4 and ...Nf5, but <28.f3> served that purpose much better: <28...Nf5 29.f3 Nf5 29.Qf2 Nxh6 30.Nxh6+ Qxh6 31.Rbe1 Qg7 32.Rxe8 Rxe8 33.Re1 Re6 34.Rxe6 Bxe6> with opposite-colored Bishops and a level position despite Black's extra pawn
Kibitzing
- the right pawn at the wrong time. White had to keep his back rank secure and dispel the looming threat to his King by exchanges: <30.Qf3 Qxh6 31.g3 Rh2 32. Rfe1 Bc8 33. Ng4 Bxg4 34.Qxg4 Rh1+ 35.Kg2 Qh2+ 36.Kf3 Rxe1 37.Rxe1 Kf7 38.Re6 Qh1+ 39.Ke2 Kg6 
