As a late edition to the Winnipeg Jets lineup, Keaton Ellerby was simply asked to fill holes created by injury. Injuries persisted and Ellerby became close to a full time player. Now he is a restricted free agent and General Manager Kevin Cheveldayoff has a decision to make. Does he qualify him?
The Basics
| GP | G | A | P | TOI/GP | S/60 | Sh% | PenD |
| 51 | 2 | 4 | 6 | 13.7 | 2.8 | 6.1% | +8 |
Numbers are for all minutes.
Keaton Ellerby isn't blowing the doors off offensively, but his six points is still more than a couple Jets forwards picked up this year. He average time on ice was the lowest on the team, so he wasn't being asked to do a ton. There really isn't much to see here. The +8 penalty differential does provide some nice hidden value. That differential on average would account for a +1 or so in the team's goal differential.
Usage
Player usage chart from extraskater.com.
| EV | Percentage of Team’s Available TOI | 5 on 5 | ||||
| ZS% | ZS%rel | EVTOI% | PPTOI% | SHTOI% | QoC TOI% | QoT TOI% |
| 53.4% | +2.4% | 27.3% | 1.9% | 3.3% | 27.9% | 26.6% |
| 3rd | 4th | 8th | 7th | 9th | 7th | 7th |
Rankings are given for team out of 9 defensemen to play 20 or more games.
In addition to being the Winnipeg Jets least used defender, Keaton Ellerby was heavily sheltered and didn't get much time on ice with the special teams units. In an ideal world the team would get a little bit more special teams value out of their depth defenders.
Underlying Numbers
The y-axis is given to show the typical 40-60 sustainable spread seen for NHL calibre players. Players are compared to particular line/pair depending on 5v5 TOI per game. Population mean (average), median (50th percentile), and other percentiles are relative to player performance between 2007-12.
| 5v5 TOI | Corsi% | relCorsi% | dCF/20 | dCA/20 | dCorsi% |
| 13.35 | 48.0% | -2.3% | -1.251 | +0.595 | -2.587% |
RelCorsi is team's percentage of shot attempts with player on ice minus shot attempts with player off ice. Delta Corsi values are team's shot attempt rates with player on ice, but relative to league norm given similar usage (quality of teammates, quality opponents, zone starts, and TOI). dCF/20 is shot attempts for -where a positive number is above average-, dCA/20 is shot attempts against -where a negative number is above average-, and dCorsi% is percentage of shot attempts.
These numbers are ugly. The fact that Ellerby was sheltered, but still put up poor raw possession stats doesn't speak well to his effectiveness. The dCorsi takes the numbers one step further. Throwing Ellerby over the boards seemingly neuters the Jets offense without adding anything in-terms of shot suppression.
Worth noting, Ellerby has some limited success playing in a supremely sheltered role along side Adam Pardy (Corsi for percentage of 53, but he really struggled with other partners. In tougher minutes along side Tobias Enstrom, his Corsi for percentage was 49.4 (Toby remains awesome). Along side Mark Stuart – likely middling minutes (can’t remember) – his Corsi for percentage dropped to 40.7.
Conclusion
The numbers suggest that Keaton Ellerby isn’t very good, but it is worth noting that he is relatively young – younger than Zach Redmond for example. Unfortunately for Ellerby, he doesn’t excel defensively and he doesn’t bring any offensive skills of note to the table. While he is young, he isn’t a player that screams upside.
Keep or lose?
This is a tough question. Cost controlled assets are always a plus, but in this case it might be worth parting ways with the RFA. Players like Zach Redmond and Paul Postma bring more special teams value and they post better numbers (possession and counting). The Jets would be better served adding a quality UFA to the roster or giving ice time to someone who has a little more upside.
Bonus:
Keaton Ellerby doesn't really present anything interesting cases from a stats perspective, so let's use the bonus section to give him the award for best hockey hair.
All numbers are from Extra Skater, Behind the Net and Hockey Analysis
