When Craig MacTavish was hired as Oilers GM, he talked about advanced stats and analytics. Both David Perron and Marco Roy were what we might call ‘analytics acquisitions’ for Edmonton. A year later, Perron’s addition is universally regarded as MacT’s signature move.
- Craig MacTavish, summer 2013 on Perron trade: “It’ll be a good test for our analytics guys. They have him with some of the game’s elite. He’s right up there with controlling the play and shots for and against differential.” Source
Fair? If you’re Craig MacTavish, that’s got to be a good arrow for the math men, right? I would think so. In the week after the 2013 draft, I interviewed Michael Parkatti from the Boys on the Bus blog. Parkatti won the Oilers hackathon and did some numbers work for the Oilers before the draft, basically confirming what their scouts were viewing (or possibly raising red flags).
- Michael Parkatti: “The model really liked Marco Roy. There’s a lot of things to like about the player outwardly, and if you look at the statistics diagnostically it’s not just the point production but also the situation he found himself in. The team he played for this year was very good, but it was one of those situations where he wasn’t being dragged along by anybody. The way I like to look at it was he was the good player on the team, he played an integral part on that team. And you really saw that in the playoffs.”
We’re one season into the project, and—based on what we know—things are going pretty well. Perron, with the help of the analytics guys, was a home run. Roy, who started out very well, slipped some but still has a nice future. The model worked, Roy got hurt.
The player in the photo is Rourke Chartier, who to my eye fits what Parkatti was talking about above. Right? The draft is over now, this Chartier fellow is a San Jose prospect now, the die is cast. So, question: what does Liam Coughlin, who was taken No. 130 overall by the Oilers, have that Rourke Chartier doesn’t have, beyond size?
PRINCE ALBERT RAIDERS EXTRA SKATER, DRAFT ELIGIBLES
Rourke Chartier is 5.11, 180. Liam Coughlin is 6.03, 201. ALL other numbers that are available to us overwhelmingly suggest Rourke Chartier is the better man. The Edmonton Oilers use analytics, we proved that at the top. They acquired David Perron in part because of it, and correctly identified Marco Roy as a quality prospect in 2013 (he was ripping the cover off the ball before injuries).
What’s up? Do we have the right to expect scouts to stand aside while the math takes over? Maybe we don’t. Maybe hiring Liam Coughlin and his big fast body then hoping he’ll learn to score is the right call. Then again, Rourke Chartier and his 2.5 EV points-per-60 estimate are going to get a chance to prove math right and Oilers scouts wrong.
Should the Edmonton Oilers trust their scouts? Of course they should. Should the Edmonton Oilers make a selection as high and outside as Liam Coughlin? I’m not so sure. I wonder if Marco Roy’s injury cost the analytics department in the amateur procurement area this season. From a Jason Gregor interview with Craig MacTavish before the draft:
- Gregor: The Blue Jackets have a draft value chart. Do the Oilers have one?
- MacTavish: Yes we do. It is very beneficial on draft day and it gives us a value of what a draft pick’s value is based on past drafts.
- Gregor: When you’re looking at the history of the pick, I’m guessing there also cases where Stu MacGregor is talking to you and maybe saying, ‘Craig I really like this guy and I think that we should move up eights spots, or 10 spots for instance. How much do you still have to go on the scouting rather than just the analytics of that specific draft spot?
- MacTavish: Well, I think the value is in the amalgamation of both departments: your analytics department and your amateur scouting department. That’s really the sweet spot in terms of evaluation of players, there’s no question about that. Stu MacGregor is a very progressive guy, a very open-minded guy as is Dan. Analytics is used as it should be by Stu MacGregor as part of the overall evaluation process and as a tool.
SAW. HIM. GOOD.
It’ll be the death of us.
I believe the smart play here is to trust (your scouts) but verify (with the math). If the math says no, you better be right.