I’ve spent an inordinate amount of hours this summer digging into the passing data generated by Ryan Stimson and his team of volunteer trackers. The fundamental statistic resulting from the passing project is “primary shot contributions” or PSC. A PSC is a shot or a pass that leads to a shot. So if Victor Hedman carries the puck into the zone and passes to Nikita Kucherov who takes a shot, both players would record a PSC. Ryan’s background work on PSCs and how they predict future scoring is found here.
The passing project also tracks several different types of passes, and different types of passes lead to shots with different shooting percentages. For example, passes back to the point lead to shots that get past the goalie about 1% of the time while royal road passes lead to shots that convert about 15% of the time. Seeing that, it makes sense to weight each shot by the top of pass that preceded it. The specifics on the types of passes and the weights can be found here.
Once each shot is weighted by the type of pass preceding it, the resulting statistic is a “weighted primary shot contribution” or wPSC. That might seem a little wordy but the concept is simple. wPSCs are shots and primary shot assists weighted by the likelihood of the resulting shot to be a goal. Because this measure only looks at offensive contributions, it is not a measure of overall play. It’s more a measure of offensive skill and creativity. The piece linked above shows that wPSC is an improvement on traditional counting stats in predicting future scoring for forwards. However, the same is not true for defenders. So wPSC in its current form is only relevant for forwards.
With that context in mind, below is a chart that shows the Lightning forwards wPSC based on the games tracked in the project thus far. Included are players who have at least 120 minutes tracked at 5v5 in the project. The numbers next to the bar are percentile ranks among the included players for context. As usual, Nikita Kucherov and Steven Stamkos are the most dangerous offensive players on the team. Interestingly, Ryan Callahan looks pretty good in this sample of games, which are mostly from before he was injured in the spring. One name that stands out by its relatively low ranking is Vladislav Namestnikov. However, I would argue that spending so much time in the bottom six puts him in a role where he is expected to play with a defense first mentality.
For a version of the chart that can be filtered by team, click here.
If you’d like more detail, you can compare individual players using the chart below. I’ve shown an exaggerated case by default to illustrate how the chart can be used. JT Brown and Valtteri Filppula are both below average offensive contributors. But they get their by completely different paths. Brown is primarily a shooter while Filppula is primarily a passer.
A customizable version of the chart is available here.
As a last step in this look at the Lightning through the lens of wPSC, we can look at something new that hasn’t previously been available in hockey statistics. Which players combine to create the most shots? Because the NHL doesn’t offer shot assists, we can only learn this type of information through manual tracking. So for fun, let’s look at which Lightning duos combine to create the most shots as well as the most dangerous shots based on wPSC. Thanks to Muneeb Alam and Mike Gallimore for pulling the combo TOI for me. I would not have been able to do per 60 minute calculations without their help. And again, the chart includes duos with at least 120 minutes tracked together at 5v5 in the project.
A customizable version of the chart is available here. For teams with less games in the sample, 120 minutes might be too restrictive. For those teams, change the minimum to 60 minutes and you should be able to see more duos.
Not surprisingly, Kucherov and Stamkos are the most dangerous duo based on wPSC/60. In fact, either Kucherov or Stamkos are involved in seven of the top ten passing combinations for the Lighting. The chart below shows the quantity of shots generated by each pair on the x axis and the danger level of the shot on the y axis. From this view, we can see that Kucherov and Stamkos generate an absurd volume of shots while Kucherov and Killorn (the next most dangerous combination on the team) generate fewer shots but at a higher danger level.
A customizable version of the chart is available here.
Weighted primary shot contributions are an interesting way of examining creative offensive play. However, they ignore the defensive contributions of players. The next logical step in utilizing this data set is to attempt to quantify on-ice and relative wPSC measures for defenders to see if anything is to be learned from that exercise. So if you never hear from me again, it’s because I died trying to quantify defensive impact at the player level.
