I could use some advice on this one...

Here's my analysis of a game I just completed.  My main concern is how my position was so poor up until move 15.  I really thought I was about even.  The evaluations are houdini's but that's the only engine analysis done, I didn't look at the moves, just had it add the evaluations.

Any thoughts would be greatly appreciated. 

I would suggest you make your own analysis of the game and forget about Houdini at the moment

You'll probably be wrong tons of time, but you'll learn and improve much more...

So to get started : at which point do you think you made a decisive mistake, and what else could you have done ?

Wait, I have to actually do some work to learn this game? 

I actually go over it on a board before letting houdini have it.  I should probably do that before asking for advice though, I just really don't understand the positional considerations in this one.

Houdini is very helpful for identifying certain tactics themes, etc. However, identify the mistakes yourself is invaluable. However, how do you procede when you can't figure them out? :)

4. ...Nc6 is a line, however, the postion you reached is not a favorable one. The reason the computer rates White's position as much better around move 13-15 is because White is very active and Black has no active prospects. Generally in this structure, it is favorable for Black to play ...c5 if possible early (even on move 4!). With this knight stuck on c6 not doing too much, black simply has a really bad version of some Queen's Gambit Declined.

TLDR: The resulting position from 4. ...Nc6 is playable but not with allowing cxd5. Look at 4. ...c5 or 4. ...0-0 for black!

I think it's best to round off the human analysis first (including asking in forums), and go over the game with Houdini at the end.

Makes sense, thanks very much.