Q&A With Athletic Calgary's Kent Wilson on Flame's Analytics

[Toonage]

I can't excuse analytics entirely, but hockey is far too dynamic a sport to rely on them completely.

[Samonadreau]

Im pretty close to half and half eye vs analytics but if i have to chose one itll be analytics. More often than not it will show you who is tilting the ice both offensiively and defensively.

[evman150]

Originally Posted by Ashasx

All of the Athletics content is behind a paywall. Who is surprised?

I had never heard of "the Athletic" before this morning.

Originally Posted by undercoverbrother

Can't Bingo do whatever he wants on his site?

Of course. And like I said, I found the thread compelling enough to actually make a substantive post (which I rarely do anymore - post, that is). I just wanted to back the critique up in a small way, because I thought it was being unfairly dismissed.

[dammage79]

I've never been a fan of advanced stats. But I tend to think of them like Treliving does. They paint a small part of the picture. If they didn't, teams wouldn't use them.

So yes, they do have a function in the league but relying on them to tell your story is too binary of a logic process.

The game is far too fluid to rely on the picture they paint alone.

using them as a predictor is fool hardy as well. If player X was successful with his mates but when he gets paid as a UFA to perform like he did with those linemates and falters greatly on the new team with new players, then the stats were wrong, due to human dynamics that skew the results.

[Bingo]

Originally Posted by dammage79

I've never been a fan of advanced stats. But I tend to think of them like Treliving does. They paint a small part of the picture. If they didn't, teams wouldn't use them.

I thought Gulutzan had a great quote on analytics last year as well ... something to the effect of how he uses the analytics to test his biases and assumptions on each player.

[killer_carlson]

As a follow up to my question above, if the analytics group create a modifier or variable to account for guys who are playing purely for their own stats and not for the team's success, will you ask them to call it "The Hall Modification" or "discounting by the Hall factor".

[evman150]

Originally Posted by dammage79

using them as a predictor is fool hardy as well. If player X was successful with his mates but when he gets paid as a UFA to perform like he did with those linemates and falters greatly on the new team with new players, then the stats were wrong, due to human dynamics that skew the results.

I don't think you can say this. Perhaps the stats were interpreted poorly, or extrapolated improperly, but the stats themselves are not "wrong".

[Bingo]

Given what happened in Montreal 18 months ago (Subban trade), I wonder what kind of voice analytics departments have around the NHL.

If I remember correctly the guy disagreed with the trade, they made it anyway, he talked against it and was let go.

I wonder what the fight against traditional hockey views looks like from the inside. My only look at it was the movie Moneyball.

[Toonage]

Originally Posted by Bingo

I thought Gulutzan had a great quote on analytics last year as well ... something to the effect of how he uses the analytics to test his biases and assumptions on each player.

So use them to re-enforce or refute the eye test. The perfect tandem really.

[dammage79]

Originally Posted by Bingo

I thought Gulutzan had a great quote on analytics last year as well ... something to the effect of how he uses the analytics to test his biases and assumptions on each player.

Make sense. I think the stats may eventually be able to better aid coaches in putting line chemistry to paper once the data pool is deep enough.

Kind of like re-affiring hunches that the eye test was already telling you. Kind of like putting your gut feeling onto paper and into percentages and numbers.

[Fighting Banana Slug]

I interpret the poll question to read, "if you could only pick one way to evaluate a hockey player, do you watch the games or compare advanced stats". It has to be eye test.

[Caged Great]

Originally Posted by Bingo

I thought Gulutzan had a great quote on analytics last year as well ... something to the effect of how he uses the analytics to test his biases and assumptions on each player.

That's how I do it as well. Most of the time it correlates unless the player does a specific set of tasks that the stats don't account for like Hamonic or Glass.

[Enoch Root]

One thing I have never seen from any hockey stat is any measure of the confidence level for the stat in question (whether it be a t-stat or z score or whatever is the appropriate metric for the item in question).

People like to compare corsi between players for instance (player A has a 5on5 CF of 52 and player B is on 48.5!). The problem (one problem) is that these numbers are meaningless without knowing the degree of noise within the stat. Is 52% statistically different from 50%?

How do you address this and is there any information on this?

Thanks

(that is a question I would have if I were on the other side of the paywall)

[Metro Gnome]

Originally Posted by thymebalm

Feels like advertising to me...

If he wants the opinion of this forum on an article, we shouldn't have to pay to see it. I appreciate that The Athletic is trying something different, but to come to a free site and plug paywalled content is a little much.

Hey Metrognome, you release the article for free to CP and you'll have more chances to receive feedback and maybe even hook a few people into subscriptions...

I guess I'll deal with this first.

I'm actually not here for the forum's opinion on the article. As noted by Darcy, my interview with Snow was longer and more in-depth than what appeared at the Athletic. And because of the nature of the encounter, I thought it would be something worth sharing with this community. Essentially this is a way for me to share the content for free with CP.

In addition, I talked about this project over with Darcy before doing it specifically so it wouldn't seem like advertising. I made no quid pro quo request for being here. I didn't even ask for a link to the article to be included - Darcy has done that as a courtesy.

Of note for questions: Snow did not share with me particular projects or inner workings of what they do in his department (because they would never share that with the media). The back-and-forth was much more 50,000 foot level: where he came from, how he got here, his challenges in operating as the analytics guy, etc.

[Locke]

How highly regarded is +/- these days?

[Locke]

Originally Posted by Metro Gnome

I guess I'll deal with this first.

I'm actually not here for the forum's opinion on the article. As noted by Darcy, my interview with Snow was longer and more in-depth than what appeared at the Athletic. And because of the nature of the encounter, I thought it would be something worth sharing with this community. Essentially this is a way for me to share the content for free with CP.

In addition, I talked about this project over with Darcy before doing it specifically so it wouldn't seem like advertising. I made no quid pro quo request for being here. I didn't even ask for a link to the article to be included - Darcy has done that as a courtesy.

Of note for questions: Snow did not share with me particular projects or inner workings of what they do in his department (because they would never share that with the media). The back-and-forth was much more 50,000 foot level: where he came from, how he got here, his challenges in operating as the analytics guy, etc.

Oh....cool.

Seeing as how an 'Analytics Guy' is now managing a team in Arizona this seems pretty interesting.

[pepper24]

Where did Kent Wilson play junior and/or pro hockey? Don't see him on Elite Prospects.

[Metro Gnome]

Originally Posted by killer_carlson

We saw during the Eakins era in Edmonton that certain players looked like they were playing for Corsi rather than playing to win.

Do they adjust their modelling when the eye test is saying that a guy is out there to pad a stat?
Example, rather than flipping the puck deep for a line change or to enable a teammate to go fight for it Hall would float a lazy writer from outside the blueline.

Not helping the team succeed in any stretch

Chris stressed that analytics work is collaborative in the Flames org, such that he actively works with the coaching staff and the scouts. Meaning analysis is a back-and-forth between experience and math as much as possible.

As for gaming corsi in particular by taking long shots, I haven't really seen any player do this successfully. Guys trying to get free, poor quality shots will likely lead to turn overs, easy transition for the other team.

That said, analytics is moving in the direction of expected goals over pure outshooting measures. Expected goals tend to include other shot information like type and on-ice location. When combined with volume, that can better show and predict if a team (or player) is driving useful possession or not.

[Metro Gnome]

Originally Posted by Gaskal

How does he think of the Flames analytic package in comparison to the analytics departments of other teams around the league?

We can't really know this since all the teams are hush-hush about what they do. I have worked with people who are now in NHL head offices and they won't even give me any details on the down low. I'm guessing there is an NDA or two involved with conducting analytics work for NHL clubs.

[Bingo]

Did Snow hint at anything the Flames are doing that is different than what other clubs are doing?

I know the Oilers keep referring to their "own data" whenever they want to defend a player that is getting ripped in public analytics circles.