One More Time: A Deeper Look into The CHL and IPP

[youtube http://www.youtube.com/watch?v=lN1iQQS5E_A&w=420&h=315]

[NOTE: I’ve written an update on this project here related to empty net goals]

The other day, I published a piece looking at the Individual Points Percentage (IPP) of the top 24 draft eligible CHL forwards. This piece was occasioned by a new tool offered by McKeen’s scouting service.

The purpose of such an investigation to try and gain another piece of information regarding the on-ice play of the draft eligible players and their teams. IPP is simply the percentage of points earned by a player on even strength goals scored while a player is on the ice. So, let’s say a player is on the ice for 10 goals for and he accrues 5 points on those ten goals. In that case, his IPP is 50%.

This record can be used to give us a sense of the scoring burden players share in while on the ice. A player with a high IPP could be considered a vital catalyst for even strength goal scoring while on the ice. Additionally, a player with high IPP likely plays on a team that heavily relies on him for even strength scoring.

All the caveats from my previous post remain. IPP does not tell us who is the strongest even strength scorer, or who is the best player.

Building a Larger Sample

Following my last post, Hockey Twitter’s own Speeds asked an interesting question:

Basically, Speeds is asking whether Draisaitl’s 91% IPP (a number that would be a large outlier in the NHL) is so uncommon after all? The logic here is that because teams in the CHL vary widely in terms of team strength, perhaps Draisaitl’s situation (a kind of lone wolf as far as the forward group goes) is repeated elsewhere.

In order to investigate this hypothesis, I gathered a larger sample of CHL players. I looked at the top 20 forward scorers from each league in the CHL regardless of status (drafted, draft eligible, undrafted, over-age, etc.). This gave me a group of 60 of the CHL’s elite scorers. I’ve broken the group into 3rds. And, I’ve totalled the even strength points for each group in an effort to see if, among this group of high scorers, there is some correlation between IPP ranking and even strength scoring.

NB: In the case of a tie, I’ve ranked the highest even strength scorer first. I’ve removed two players who were traded and had split seasons and two defensemen and simply added the next 4 highest scorers.

Here’s the first 20:

Screen Shot 2014-06-09 at 8.34.49 PM

(click all images to embiggen)

Of the 60 top scorers compiled, only 5% share the range with Draisaitl’s IPP (2 players: SSM’s Tolchinsky and Portland’s Leipsic).

Totalling the column of even strength points, this group scored a total of 1168 ES points for an average of 58.4 ES points per player.

Here’s the 2nd 20:

Screen Shot 2014-06-09 at 8.43.37 PM

Totalling the column of even strength points, this group scored a total of 1132 ES points for an average of 56.6 ES points per player.

And, here’s the 3rd 20:

Screen Shot 2014-06-09 at 8.48.14 PM

Totalling the column of even strength points, this group scored a total of 969 ES points for an average of 48.45 ES points per player.

Analysis, Such as it is

The primary goal of this study was to see if Draisaitl’s 91% IPP was an outlier for premier scorers in the CHL. In the NHL, a season in the high 80s and up is typically an outlier and not repeatable averaged over a larger sample. What larger samples of elite NHL players show is that top scorers tend to have high IPP. Even the best players, as we saw in the last article, do not, however, have an IPP in the range of Draisaitl’s score. They may for a single season, but not over a larger sample (again, as noted in the last article, Taylor Hall had an extraordinary year from an IPP perspective).

As we discussed in the last article, a drop or gain in IPP from year to year is not necessarily an indicator that a player’s performance has improved or declined. Rather, it is an record of that player’s contribution to even strength scoring relative to his team. If a player’s 4 on-ice teammates improve or decline year over year, we should expect a correlating increase or decline in his IPP.

Returning to the question at hand: while an IPP as high as 91% over a season is a clear outlier in the NHL, is the same true of the CHL?

Based on this expanded study, though admittedly limited in a number of ways*, it appears that the CHL militates against exorbitant IPP scores just as much as the NHL. The take away from this expanded study, therefore, might simply be that Draisaitl’s situation this year really was fairly unique, even by CHL standards.

*I’d like to note a few of the limitations of this study for the record: I’ve only looked at a group of 60 players from a single season. I haven’t looked at previous seasons to see if a very high IPP was perhaps more common in prior years (perhaps this year was an outlier itself in having so few players score very high IPP). I haven’t looked at players’ year over year and averaged out IPP scores. In part, this is simply because I don’t have the data. But, I also would be concerned about such a study insofar as the development of junior players and the make up of junior teams varies so highly from year to year that such a study may produce unreliable results.
An Interesting Sidebar

With this larger sample to work with, I thought it might be interesting to see how strongly IPP scores correlate with even strength scoring. I haven’t done any intensive analysis here. I’ve simply blocked the larger set into three groups based on their IPP rankings.

Looking at the results in this rough manner, two things stand out.

First, the assumption is borne out by the data. The three groups ranked by their IPP scores correlate with their ranking in terms of even strength scoring.

Second, the difference in even strength scoring between group one and two is marginal. Whereas, the difference between the first two groups and the third is quite stark. Group one averaged 58.4 ES points per player. Group two 56.6. And, group three 48.45.

The threshold between group two and three is a 70% IPP score. I’m tempted to suggest that the baseline of 70% for IPP scores may be indicative of a real drop off in even strength scoring. At the very least, it is a record of a player’s even strength scoring burden being more evenly, or perhaps heavily, distributed among others.

Check Back Later

This new tool from McKeen’s gives us a lot to play with. In the coming days, I plan to have a look at even strength on-ice goal differential in relation to even strength team goal differential for the top 2014 CHL draft eligible players. This is something Lowetide has already started playing around with in earnest. An expanded look at the 2014 draft class seems in order.

To Top