By The Numbers: Is Petan on the power play a big enough payoff to place over Thorburn?

There has been plenty of discussion surrounding Paul Maurice’s recent usage of Nic Petan, and the recent decisions to play 33-year-old grinder, Chris Thorburn, over the 21-year-old skilled forward. Maurice gave some reasoning behind his decisions on March 3rd, when he said that Petan was “not a big enough payoff on the power play” per Scott Billeck.

We know for a fact that Petan has been pretty solid producer on the power play (carrying the Winnipeg Jets second highest 5-on-4 point production per hour of ice time) and that Petan has been struggling at even strength (although a large chunk of this may come from usage and archaic role usage).

However, only with numbers can we compare the pros and cons to make an informed decision on whether or not the tradeoffs truly work in Petan’s or Thorburn’s favor.

In my last post, I broke down what WAR is and why it matters. Essentially, WAR combines multiple statistics into one currency so we can get a decent approximation of a player’s overall contributions towards wins in multiple areas.

We could just end the analysis here and point out Maurice is wrong, but let’s delve in deeper.

What we see here is that Nic Petan and Chris Thorburn have had a similar overall impact on the Jets at even strength (EV_OV). Both of them have been slightly below replacement level. Although, an additional 23 per cent of even strength ice time means that Petan is actually the superior option with being relatively less negative per minute. Petan has been superior offensively (EVO), while Thorburn has ben superior defensively (EVD).

What about power play?

We see here that Petan has a 1.3 goal above replacement value with his offensive contributions on the power play (PPO). Overall, Petan gives the Jets their third highest overall power play impact, and second highest relative to ice time. This impact is about 13 times in magnitude to the impact either player carries at even strength.

For a team that is struggling on special teams, this is a huge boost. Petan’s power play impact boost is so large, it pushes him by 217 positions upwards in ranking when comparing all skaters in even strength impact (EV_OV) versus their overall impact when excluding penalty differentials (PURE_OV).

Of course, penalty differentials matter as well. Thorburn has quite the penchant for taking penalties and this caries a stark negative impact (TAKE). Thorburn’s penalty impact ends up nearly as negative as Petan’s positive power play impact.

So, while we have two players that are fairly similar in overall even strength impact over the season, there is a huge difference in their overall impact on the team (OVERALL). In fact, over 250 skaters sit between the two when looking at every NHL defender’s and forward’s impact on the game.

As I noted in my last post, WAR is never the ending point of discussion, but it should be the starting point.

We already did some further investigating in seeing why Petan had such a high PPO rating with his team second highest power play production per minute. We also investigated into why Thorburn has such a negative value despite only being slightly negative in even strength, due to his massive penchant for taking penalties (which may undervalue his true impact as the Jets are a below average penalty killing team).

Another thing I want to look at is their similar even strength ratings (EV_OV), aside from Petan actually being about 18 percent better per minute. So, we then break down even strength offense into its components.

The three components are a player’s individual offensive contributions through boxscore stats (OBPM), their ability to improve linemate offense through improving team expected goal generation (OXPM), and their ability to hurt the opposition’s offense through reducing expected goals against (DXPM).

We see here that Petan is worse defensively than Thorburn, but neither are below replacement level (DXPM). In terms of offense, Petan is better than Thorburn, both in generating scoring himself (OBPM) and improving his linemates (OXPM). Both are below replacement level in terms of the individual offense, but only Thorburn is below replacement level in impacting expected goal generation.

Most of this makes sense, like Thorburn being similar to his historical norm. However, Petan’s offensive values do not seem to rest well relative to what we should expect with Petan’s skill.

So, we then break things down even further.

Petan is a skilled forward, but he hasn’t been producing much 5-on-5 offense, and 5-on-5 point production is typically what drives OBPM value. In addition, XPM uses priors to assume who contributed to on-ice results. Essentially, if something good happens and it does not know how to divvy up the responsibility, it will default to the player that was superior the season prior.

It expects you to play well with strong players and worse with weaker players.

However, I’ve discussed before how Petan’s performance is not as simple as that:

 

If you wanted context to these stat lines, the top one looks like a bad enforcer, while the top one looks like Blake Wheeler last season. This is quite the split. Overall, we see the Jets do better than expected with Petan playing with good shot differential players, but even worse than expected with Petan playing with bad shot differential players. It’s not just Petan being better with the top performers and worse with the bottom performers, but Petan actually makes the top performers even better while making the bottom performers even worse.

In other words, the results we see are not as simple as being with better is optimal. I have hypothesized that this is a chemistry like effect. It makes sense though that a small, skilled, playmaker would require a player to pass to. It makes sense that when with the right players, a skilled playmaker creates a lot, but without those players things fall apart.

While I expect Petan’s true talent to be no where nearly as high as what we see with those numbers away from Thorburn and Drew Stafford, it suggests that Petan’s OBPM and OXPM are probably falsely identified by the WAR model due to it not accounting for chemistry effects. I expect Petan to be a much stronger OBPM performer in the future, even if nothing else changes aside from Petan’s usage.

Maurice’s assumption that Petan’s offense on the power play is not enough to pull him ahead of Thorburn is completely off the mark and there is no real evidence behind it.

While both players have been slightly similar in even strength performance, Petan still is the better player in those minutes. Not only that, but we expect these models may be undervaluing Petan’s offensive abilities due to his peculiar usage in 5-on-5 situations. Even prior to accounting for Petan’s extreme usefulness in special teams, Thorburn should never play over Petan.

While Thorburn does have superior defensive value, overall the Jets will win more games with Petan than Thorburn, all else held constant.