The opening is worth examining, being a good example of when the opponent (Black in this case) varies from standard opening play with a solid and reasonable, yet theoretically inferior move (4...d6). There is of course no immediate punishment for the move; in fact, it transposes into an opening formation for Black - both knights developed plus the d6-e5 pawn chain - which can result from the Black Knight's Tango, Panther, or Old Indian Defense (the root of the previous two openings). The difference between this game and the main line of the other defenses is that White has played e3 instead of d5; for more on the Panther setup, there's an interesting series of articles at the Kenilworthian.
Returning to the game, the first major decision occurs on moves 7-8, with Black exchanging central pawns on d4 and White deciding to recapture with pieces rather than the e3 pawn. It's these sorts of decisions that will greatly affect the structure and flow of a game, although they may appear innocuous at the time. It was interesting to see that the choice I made was supported by 100% of the master games in the database. At the time, I went that route largely through a greater familiarity with the central structure in question, which is in fact not a bad reason. From an objective standpoint, the alternative central pawn structure (pawns on c4+d4 with an open e-file) is a little loose for White and I think Black finds it easier to play against, with the idea of undermining White's pawns and eventually putting a rook on the e-file.
The early middlegame features some silly maneuvering on both sides, although Black's turns out to be a bit sillier, since he maneuvers his knight into a pin against g7 on the long diagonal from White's Q+B battery. The threat of mate on g7 dominates the rest of the game, although analysis points out where both sides could have profitably broken out of this situation. Things are finally brought to a head when White goes caveman starting on move 20. After a rather lengthy sequence he manages to exchange off Black's defending knight, although analysis shows that by this point it had no real impact on Black's defenses.
One of Black's problems throughout this game was that he was evidently thinking only one ply ahead on a number of his previous moves - the antics of both the knight and the light-squared bishop are witness to that - and missing obvious replies from White. Unfortunately for Black he does the same thing again on move 27, attacking White's "caveman" rook but failing to count the number of attackers against g7. The end is brutal, with White throwing stones and then clubbing Black to death.
Moral of the story: brains can beat brawn, but only if they are used for thinking ahead more than one move.