One of the odd things about the debate over Jordan Eberle's underlying numbers is the sentiment expressed by dawgbone in his recap of Eberle's standing in our Top 25 Under 25, namely, that Eberle is now being "punished" by for being overly lucky last season. While part of what db says is correct - Eberle is in for a bit of a bounce from his current shooting numbers - but it's not as if Eberle is terribly unlucky right now.
After Eberle played in his 160th game, I broke his career out into 10 separate 16-game segments. His shooting percentage in each segment is reflected below. The last segment is the current one and totals only 9 games.
|Jordan Eberle S%|
My takeaway from the table above centers around two things:
- First, Eberle's "slump" to start the season wasn't the worst run of luck in his career. In games 49-65, he shot just 7.7%. This segment correlates with the Oilers post-Hall broken ankle swoon during Eberle's rookie season.
- Second, Eberle's incredibly lucky run last season was marked by his two highest shooting percentage segments of his career. For 32 games last season, he shot 23%. Those two segments also happened to be back-to-back.
Below is a line graph displaying the moving average of Eberle's career shooting percentage in game-by-game format. Note the arc created by the 32 game heater.
I put all of this together because of the sentiment that is bubbling up around Eberle's season. The sentiment that this run is an aberration and he'll be back to the Eberle we know in no time. "This is just bad luck." In the table below I've stolen a concept that Scott Reynolds loves: you can't toss out the exceptionally bad streaks without tossing out the exceptionally good streaks. It's sound logic. If you're looking for the predictive performance, the heart of the bell curve, you toss out the whale tails, but understand that they can happen at any time.
The segments detailed below are his two best and two worst 16-game segments. A quick, non-mathy analysis of the segments leads us to a predictive shooting percentage for Eberle:
Eberle's career shooting percentage currently stands at 14.2%. He has 58 goals on 408 shots. Toss out the worst segment, games 49-64 and the best, games 113-128 and Eberle has 47 goals on 332 shots, or 14.2% - dead on his career average. Toss out the next worst segment, games 145-160 and the next best, games 97-112 and Eberle is left with 33 goals on 247 shots, or 13.4% - slightly lower than his career average.
For those of you less statistically-inclined, the exercise performed above, eliminating the tails in search of Eberle's true talent, is an extremely simple, though flawed, method of performing regression analysis without all of that complicated math mumbo-jumbo.
Accepting that Eberle's true talent for shooting is somewhere between 13.4% - 14.2%, the probability that he would score 7 goals or less on 84 shots (his totals in his two worst segments) is somewhere between 7.5% and 11%. It's not a likely occurrence, but it's not a rare occurrence, either. On the flip side of the coin, the probability that he would score 18 goals or more on 77 shots (his totals in his two best segments) is somewhere between .5% and 1%. His production through that 32 segment was a rare occurrence.
Focusing on the 10-game segments themselves, the probability of 11 goals or more on 46 shots (games 113-128) is between 1.5% - 2.5% while the probability of 4 goals or less on 45 shots (games 145-160, which includes his recent slump) is between 21% - 26%. The probability of 6 goals on 59 shots (his 2013 season) is 25% - 30%.
What's the takeaway? Eberle's 18 goals on 77 shots last season was a very rare occurrence that he will be hard-pressed to ever duplicate. His slump this season isn't the most likely outcome, but it's far from rare. In fact, he's likely to go through similar slumps in the future. Reset your expectations for Eberle to shoot between 13.4% and 14.2% and you'll have a much more realistic view of the player and the production you can expect from him for seasons to come.