While discussing defensive pairings for Winnipeg Jets for later this evening, coach Claude Noel mentioned that he liked Dustin Byfuglien with Grant Clitsome last season. It's likely that Noel liked the +/- of the pair, since he's made line combinations based on +/- before. However, they benefited greatly from an overly inflated PDO (combination of on-ice shooting and save percentage). It's already well known fact that traditional +/- is a poor predictor of future success, since these percentages tend to regress to a natural mean. So what do the underlying numbers say for the Jets most effective pairings?
I have long been a supporter of keeping "Big Buff" with little Tobias Enstrom and I will show you why.
Who are their best partners?
Due to sampling biases and other natural variances, goal differentials are usually not indicative of future success. The exception is when you have a huge multi-year sample, but by then other variables such as system, partners, other linemates and developmental curves may have vastly changed throughout. Corsi differentials (all shot attempts including goals, shots on goal, missed shots and blocked shots) however have been shown to correlate strongly to all of time in attacking zone, puck possession, and scoring chances while also being a strong future predictor of future goal differentials and ultimately wins.
Essentially you can simplify all this by saying CF% is the best predictor of both future success and GF% (goal differential), and determiner of which pairing was best.
There are some variables beyond skill that effect the results but by restricting to situations where the opposing teams are within two points and ignoring the immediate 10 seconds after faceoffs, you limit the variables of the sample to be predominately controlled by skill, chemistry and effectiveness.
2011-13 Enstrom's Zone-Start Adjusted, Close-Game Situations WOWY
2011-13 Clitsome's Zone-Start Adjusted, Close-Game Situations WOWY
Patterns to note:
* Both left handed defensemen performed best when with Byfuglien. This makes sense sense as Dustin Byfuglien is a very potent (and sometimes underrated) defensemen. He has his faults, but overall he tends to overcome these faults with his positives.
* Clitsome is not as strong as Enstrom, duh, and you can see this with him being below Tobias in all three cases.
* You can also see how Clitsome lucked out nicely with some nice shooting percentages improving his goal differential in all three partners. This is unlikely to be the case in the future, but lightning can strike twice.
* Enstrom-Byfuglien pair is the most successful by a large margin. The increased possession in placing Byfuglien with Enstrom is larger than the loss of separating Bogosian from Enstrom.
* Some have argued that Zach Bogosian's sample size is small; however, by adding Hainsey, you can see that the overall general trend is correct, even if sample size may make the gap slightly more severe than in actuality.
Overall, Dustin Byfuglien and Tobias Enstrom have played some of their best hockey of their career when together. They have huge chemistry and compliment each other quite well, contrary to what some seem to believe. Statistically it makes sense, since Byfuglien is the best Jet blue liner when it comes to generating scoring chances for and Enstrom -- who is commonly miss-casted by fans as an offensive only defensemen -- is the best Jet blue liner when it comes to suppressing scoring attempts against.
So, we know that Byfuglien being paired Enstrom is really the best option for the Jets' top 5v5 pair and forms the most efficient top 4 grouping for the Jets... but...
How does Enstrom-Byfuglien stack up with the best in the NHL
We will take a look how the pair compares in both 5v5 ZS Adjusted CorsiClose possession and 5v5 scoring for the top pairs in the NHL. Now teams lines tend to be fluid with injuries and hot/cold streaks. Also, some teams may have two pairs that are essentially equal for 5v5 icetime... so because of this, some pairs may not always be the "number one pair"...
However, I think this gives a pretty good general idea:
As you can see here, when it comes to both shot differentials and offensive creation there are not many who out-perform the Jets duo. While it is true that linemates and team effects do influence the final results (see Shea Weber), the Jets top blue-line pair accomplished this on a very average possession team... which is quite the feat.
I do find the Calgary results curious as they were a terrible team last season but I didn't really follow them.
Noel shouldn't be blamed for trying and experimenting with combinations, as that is part of the job description. However, if he does end up trying to force something continuously throughout the season that fails to work, it would be appropriate to look back at this and start asking questions.
Contrary to common thoughts: Tobi + Buff = high end results.