Frozen Tools Forensics: Even-strength IPP and regression

Chris Kane

2020-12-11

With the NHL start date of January 13th announced let's turn our attention to projections. Obviously, make sure you pick up the Dobber Guide, but if you want to do some of your own there are a few stats that might help in that journey. For the next few articles, we are going to take a look at stats from 2019-20 and try and figure out what they can tell us about a player's performance and what they might mean for 2021.

This week: Even Strength IPP

A player's IPP, or Individual Point Percentage, is the number of goals that a player got a point on that were scored when he was on the ice. So, an IPP of 75 percent means that in he got a point on three of every four goals that were scored while on the ice.

How is this number useful? Well in general players fall into predictable ranges. These ranges are defined by their role and their own personal history. For example, elite top-line players can have an IPP consistently in the 75-80 percent range. Connor McDavid is pretty consistently around 80 percent while Sidney Crosby is usually around 75 percent. Less elite players, or players with more well-rounded, slightly more responsible games might have a slightly lower IPP. Jonathan Toews and Bo Horvat are typically between 65 and 70 percent. Power-play defensemen might be a bit lower still. Guys like Kris Letang, and Shea Weber usually hover around the 45-50 percent range, while a strong even strength string defensemen like Charlie MacAvoy might be closer to 35-40 percent.

If we look at a player's personal history and have a general idea of some of the ranges of sustainable IPP, and we see a change over a short period of time it is a clue that either something has changed in their game (usually deployment), or that their current scoring pace is likely unsustainable.

For today we are going to focus on even-strength IPP. Five-on-five play is by far the largest sample size we can look at, and most players score the majority of their points at even strength. It serves as a good baseline for a player's performance. Power-Play IPP is also important, particularly for elite players who score a lot of points on the power play, and the same rules apply as described above.

To get started today we are going to use the Advanced Stats report from Frozen Tools. The report gives a IPP in several states (total, even strength, and power play), points-per-60, five-on-five shooting percentage etc. and we will get to some of them later in this series.

Looking at one year's worth of data is only marginally helpful. If you have a really good grasp on a player's personal tendencies or expectations based on their role scanning through the list might flag some players whose numbers look off, but to really get into it we need additional years of data to establish a baseline. To do that I pulled three seasons' worth of player data and combined them into a three-year average. We can then compare a player's 2019-20 performance to their personal average to see who had the most consistent or inconsistent performance.

The table below shows the pl