Roles and Stats Part VII: Can roles predict future shooting and save percentage?

In the first six parts of this series I have looked at the relationship between roles and player statistics and, particularly for forwards, there is a clear relationship between how a player is used and their resulting statistics. Now this isn’t an absolute relationship as we saw with Brandon Sutter in part six of this series but it does exist. Thus far I have only presented an “in-sample” relationship (relationship between current season role and current season stats) but can roles predict future statistics as well? This is what I intend to show in this post.

Leading and Trailing TOI%

I am going to focus on forwards as they showed a much more significant relationship between roles and performance. Let’s first look how leading and trailing TOI% in years 1 and 2 relate to shooting percentage in years 3 and 4.


Ok, so if a player is getting a lot of time in an offensive role now (i.e. when playing catch up hockey) he will post a higher shooting percentage in the future. This is no surprise as we are talking top offensive players and top offensive players now will likely be top offensive players in the near future. Conversely if you are playing defensive roles one the third line this year you likely will be next year too. Current role can predict future on-ice shooting percentages.

How about save percentages?


The relationship is not that strong here but it certainly does exist. Given an offensive role now you are more likely to have a negative impact on save percentage in the future. This is predictive in every sense of the word. On average, offensive players negatively influence their goalies save percentage while 3rd and 4th liners generally boost it. On average and for most players in may not be overly important however for some it certainly could be

Defensive and Offensive Face Offs

Let’s look at this again but using defensive and offensive zone face offs to define roles.


Those players who get more offensive zone face offs will continue to post higher shooting percentages in the future.


The relationship doesn’t hold quite as strongly for save percentage but generally speaking those players that get more offensive zone starts now will continue to have negative impacts on save percentage in the future.


Current roles do have a predictive relationship with future shooting and save percentage which means that shooting and save percentages is either role driven or talent driven or some combination of the two. Either way, players can and do maintain elevated on-ice shooting and save percentages and we have to be aware of this.

A players ability to influence shooting percentage is quite a bit greater than it is to influence save percentage which makes considering shooting percentage in player evaluation far more important however, that doesn’t necessarily mean that save percentage can be ignored for all players.