Our complete data set of shots---the one broken down by the individuals involved and the game state---is not ready yet. But using the shot locations data, I have put together some estimates of goalkeeper value from the 2013 season. To be clear, these are estimates, and our own Will Reno's Goals Saved Above Replacement (GSAR) is likely to be more accurate when it is complete. These ratings are simply the differences and ratios between actual goals allowed by a keeper and expected goals allowed by a keeper. They are based on our shot locations, and they take into account which games keepers actually played. Here are the top-ten:
Goals above average (GAA) is that aforementioned difference, and numbers less than zero indicate a better keeper. Goal Ratio (Gratio) is the ratio, or fraction, of goals allowed to expected goals allowed. GAA1 and Gratio1 come from calculations based only on shots on target, while GAA2 and Gratio2 are based on total attempts. As an example of how to interpret these stats, one might say that Nick Rimando allowed 7.39 fewer goals than expected by an average keeper. One could also say that he allowed just 79.1 percent of the goals an average keeper would have allowed in his position. These are, of course, rough estimates, and one can assume that keepers have some influence themselves over the shots they face (thus, the inclusion of the GAA2 and Gratio2).
The complete Excel file can be found here.