Here's a game I played recently, I'm interested in hearing your thoughts on it. I'll readily admit I'm lost in Sicilian play, and it seems to me like this game is lost by black, rather than won by white.... but what do I know?

Anyways, I'd like to hear your thoughts, with improvements for both sides welcome; I've already run it through the compute analysis, but humans are better at explaining their thoughts.

Don't like Bb7 or Bxf3 by black.  Black needs to develop more effectively, one move threats that cause white to gain even more of a developed vs undeveloped piece ratio are terrible.  Maybe as early as move 9, something like h6 or Bc5... unless this is all book... Qc7 looks odd to me though.

I thought white played fine.  Black was outclassed.

I'm not familiar with the Najdorf, but at a glance I like Bh6 better than Bb7.

Stops white from castling long and the Queen looks less dangerous on the light square d3.

I don't think the attack on the a8 rook along that diagonal is anything to worry about as it can be developed quicker from a7 in any case.

But what do I know with my 1500..

FWIW I was thinking along the same lines, about it not being a big threat, and the rook being able to go to a7 to develop.  Maybe even instead of taking on f6 (where white is just developed like crazy).  Moves like Bh6, Bc5, or Ra7 and black is fighting back.  The position after Bb7 and I'm not surprised if white isn't just winning already.

Your position is already quite difficult after 12...Qxf6.  Have you checked the chess theory to see where things deviated?  I know there is a line with plenty of games up until a certain point, but I don't know how deep; my suspicion is you deviated somewhere between moves 7 and 12.

I appreciate the feedback everybody, so thanks! ...but is there any way in which white's play could have been improved?

Well sure, but it might take some serious analysis to crunch out the details.  Off the top of my head it all looks pretty good.

The move Ne5 though stands out as iffy, I don't think you should let black off the hook with a queen trade.  With reduced material, and no targets this is looking like a dead even endgame... the slight advantage of the bishop on an open board is perhaps compensated by the slightly better white pawns.  But basically this is just equal after a queen trade (17...Qxd6 and all you get is a check and probably a draw).

Hmm, which actually suggests white probably isn't winning when I thought he was.  Probably as early as move 13 white needs to be looking for improvement, it seems 13.Qd2 (which looked good at first, sure) is too slow and gives black a chance to fight back.

This is how you do analysis by the way.  You backtrack and erase your old evaluations (one side winning or just better or equal) and find specifically where you need to look for improvements.  Like I said if move 17 is too late, we have to go back further.

So lets say 13.Bf3 or 13.Rf1 or maybe even Qd3 or a knight move.  If it turns out black is fine here too, we go back further.  You might even decide at some point that black is equal or better from the opening, and this is how you get better... so anyway... back to move 17, I like Qe5 better but geez, Rg8 and black is developed with tempo and everything.  So at least a move earlier on 16.  It seems 16.Qd6 was bad.  How about 16.Qd3, you're eyeing h7 possibly making black think twice about Rg8 and e4 with the light square diagonal.

But giving up those dark squares is rough... black isn't so bad off really.  When I first saw the game I was thinking of a developing move instead of getting rid of your dark square bishop on move 10 (that f6 knight is not so important) but looking again it's hard to decide how to defend e5... but you may have to go back even to this point where you dark squares become weak with an open game for black's bishops.  Which also suggests Qc7 (the move I didn't like) is actually good.

I definitely like 8.e5 because 7...b5 just has to lose time... sure it's dynamic and ballsy, but e5 is the principled reply for sure... so I'd start there at move 10.  Look at positions with 10. Bf4, 10.exf6 or even 10.Qe2, see if you like them.  Are they better or worse than what you got in the game?

So in a nutshell this is what I'm saying.  It's not just about good moves it's about good evaluations.  What I notice is after move 15 "something's gone wrong" for white... by which I mean my evaluation goes from better for white inexplicably to kind of equal really.  So we have definitely:
Move 16 (try 16.Qd3 16.Qe2 or something else you like) and
move 17 (can't allow a queen trade.  Look at 17.Qe5 or other)

Only if you want to go deeper
Move 13 is bad by means of deciding the position at move 16 is equal for black.  Look at 13.Bf3 13.Rf1 13.Qd3 or other.

And deeper
Move 10 is bad by means of deciding the position at move 13 is equal for black due to his bishop in an open position and lack of targets for white.
Look at 10.Bf4 10.exf6 10.Qe2 or other.

But only this far
Because move 8.e5 is definitely the principled reply to a tempo waster like b5 by means of reasoning pawn moves are not development... ahead in development means open lines.  (But in true analysis actually you'd challenge this move too... challenge every move regardless of how good it looks is thorough analysis).

Analysing is tough work, I've given you an outline to start with :)  I should note if you disagree with any of the moves I've marked... then that excellent!  You should always work to find the truth yourself.  Don't believe me, a GM, or a computer, work hard to decide what moves are good and why.

For your level you might feel this is too rigorous (or impossible even!) to do on your own.  That's fine, mark any major mistakes you feel you made and move on to new games and new lessons.  If you don't see any major mistakes, move on to your losses (there must be a mistake in there :).  These wins though you have to dig deeper sometimes.  If I had to give you a main take away for this one it'd be your willingness to trade queens was bad.  You seemingly played well by conserving a tempo with Ne5 but failed to evaluate the position after a trade, 17...Qx Qx Nxd6+ and Ke7 for example... the resulting position is just even. 

For reference, the analysis I've outlined is probably mostly practiced by experts and up (although some A players too maybe) and should take at least a few weeks to do a full blown game.  Even so it's useful to understand how evaluations play a role.  You can discover seemingly innocent moves (like 13.Qd2) are not as good as they look.