If any of us were asked "What makes some chess players better than others?", we might come up with a number of different responses. Some replies might be: They are better at seeing tactics, They have a better memory for openings, They can calculate better, They can see further ahead, They play more, They practice more, They read more chess books, They have a coach, etc.  All of these reasons have some validity, but I believe they overlook one of the primary reasons stronger players are stronger: They evaluate positions better.  That is, they may see ahead no further than their opponent, but judge the resulting positions better.  They avoid positions that they judge to be inferior for them and try to move towards positions that are superior.  The weaker players just play up to any position they have evaluated as equal using only crude techniques such as "Is the material equal?" or "Do I have doubled pawns?" and the like.  Stronger players use these also, but supplement them with evaluation parameters based on more dynamic criteria.  That is, they may say "I am a pawn down, but I have a huge kingside attack" or "I have doubled pawns, but the open file is worth it"  or "I am up the exchange, but his knight is so powerful on d5 I should probably sacrifice the exchange back if I want to have any winning chances" or "Material is even, but I have to trade queens somehow or his attack will be crushing."  As you can see, the stronger player stays aware of the material balance and the state of the pawn structure, but evaluates them not as always good or bad (as the weaker player does), but within the context of all the other factors of the position.  The weaker the player, the more rigidly they evaluate static factors.  For example, some very weak players are almost ready to resign after a queen trade because they "lost their queen" (the best piece).  Or their opponent plays BxN and they don't recapture the bishop because "I didn't want to get doubled pawns." 
I would say that, at lower ratings, tactical skill (not losing pieces) is the predominant factor in the outcome.  At the much higher ratings, positional evaluation predominates, since at that level, tactical skill is almost a given for both players.  That is why professional chess players don't usually play sharp, tactical opening gambits.  They assume the other player will see all of the tactics and when they are over, the resulting position will be somewhat lifeless and easy to evaluate, negating the stronger player's main advantage.  Rather, they tend to play slower developing, more positionally complicated openings where the emphasis is on the evaluation of each changing position (should I move this pawn, should I trade this piece, etc).  This ability to evaluate positions as good, bad, or equal is where their advantage lies, so naturally they want things to be as complicated as possible, positionally speaking.  At the lower ratings, the stronger player wants things to as complicated as possible, tactically speaking.
In the following game, Black plays the opening phase well enough, but misplaces one piece which causes him to lose the fight over a key square, which ultmately gives White a winning position, which is finished off with a tactic missed by Black.

So how does one improve his ability to evaluate positions?  One method that I find useful is to play over master games that are annotated by a master, preferably the master who won the game.  In the annotations, he will relate what he was thinking and why he played as he did.  Playing over master games that use the openings that you use is even better.  You will get a feel for what is right or wrong in certain opening positions, as well as the resulting middlegames.  In the game above, I had never seen the maneuver 10...Ba6, 12...Nxa6, 13...Nc7 in any of the dozens of games I had played over from books, as well as games I played myself over the board.  This does not necessarily constitute proof that a move or sequence is bad, but it certainly makes you think a little longer about what might be wrong with it and how you can exploit it.  In other words, you are in a position to EVALUATE the move as wrong and the position as better for you and know why.  In this game, I evaluated the position of the knight at c7 to be wrong because it did not participate in the struggle for e5 as it normally does from the usual squares d7 or c6.  Also, never having seen ...f6 played to kick the knight off of e5, I evaluated that move as an error due to the weakening of e6.  If Black had made the same evaluations BEFORE playing those moves, he probably would not have made them at all.  But his calculations seemed to be of the nature "My white-squared bishop is bad. Get rid of it."  "The knight on e5 is strong. Kick it out with f6."  While this is an improvement over much lower rated players, who often do not even have any concept of a piece being "bad" or "strong",  as the rating of one's opponents goes up, so does the depth and subtlety of the evaluations. For example, A GM may dismiss a certain pawn capture on move 6 in the opening with a comment like "This is bad because it gives Black a lost endgame."  And one might wonder how he can make such a comment on the endgame when it only move 6.  What he means is that now Black is compelled to win the game in the middlegame, since he has eliminated any recourse to an equal endgame by having damaged his pawn structure on move 6.

Improve your evaluation skills by playing over master games and your rating will go up. Trust me.