Goalie Performance Relative to Opposition
This analysis tries to answer the question of how goalies do compare to the offence they are facing.
I collected the stats for all 30 NHL teams in the 16–17 season, as well as the stats for all goalies with minimum 20 games played. For each goalie, I looked at how their save percentage compared to the offense they were facing (1-sh%) for the past 5, 10, and season. I only looked at games in which the goalie played at least a period.
The relative performance for each category is averaged for each goalie, and the season relative performance is also summed to create a metric to rank the goalies analyzed.
Some goalies were found to perform just as well as expected given the opposition, though the average relative performance changed over the season for some goalies. Below is Thomas Greiss, whose performance noticeably degrades in all 3 relative performance metric averages over the season.
The metric SrP* (sum of the relative performance of the goalie compared to the average of all of the opponent’s previous games played in 16–17) was used to rank all the goalies investigated.
With a sample size of 54 goalies, there is a pretty good distribution with a mean of 0.05.
Save percentage doesn’t explain SrP* scores completely, meaning there is some value to analyzing the SrP* scores of goalies.
As you can see from the ranking, there is an abundance of backups ranked higher than their team’s starter. This analysis may not be as valid for goalies with between 20–30 games played.
SrP* is an okay metric to see how goalies compare to expected performance based on their competition. Having the Vezina winner as the top tier goalie is a good sign, as well as 31.6% that isn’t explained by save percentage alone. The SrP* score needs to be taken with a grain of salt for back ups, as Dell probably would not do better than Jones if given a starter’s workload.