Quick Thought: Stat Tracking

Jeff Veillette (Jeffler)
August 28 2013 12:25PM

Michael Traikos had an interesting piece in the National Post about Jay McClement, his big year last year, and his thoughts on the Leafs moving forward. That's cool, and deserved - McClement was a massive help to the penalty kill last year, and produced above everybody's expectations. But a couple of things stated are a bit peculiar and are worth noting/counter-pointing:

The Leafs, who averaged 26.3 shots per game and allowed 32.3 shots per game, had one of the worst shot differentials in the league last season. It is a reason why some believe the team was lucky to win as many games as it did.

 Valid. Go on...

But McClement, who had an even plus-minus rating, said the numbers are misleading.

Using plus/minus to show off prowess here is interesting, considering the issues with using the stat and the fact he's a specialist at playing in a part of the game where plus/minus isn't used, but admittedly, that's me nitpicking and not the point at hand...

The Leafs were content giving up shots as long as they were from bad angles or far away from the net, he said,

 Here's the funny thing about the Leafs bleeding shots against. He's a little right about the Leafs having the bulk of their "disadvantage" in shots against coming from far away (30+ feet away). But is that the team "being content" or just not being up to task? More than half of the average NHL team's shot attempts come from long distance. Many of those leads to rebounds or secondary attempts. Goaltenders aren't immune to these shots (screens, deflections, or just bad misplays). Even still, the Leafs weren't exactly shutting down other teams close range either. SkinnyFish talked a bit more about shot distance and the 2013 Leafs a while back.

Maybe it's not a glaring problem, but I feel that like, combining with management comments this offseason, this organization seems to be brushing off the shot issue rather easily. Could they know something we don't? Sure. But it's still a concern. Moving on.

coaches tracked advanced stats on everything from turnover-to-takeaway differential

Turnover-to-Takeway differential is a great statistic for people who are new to statistics. When I was 15, I ran a site that had one purpose, and that was to run Bryan McCabe out of town. (It actually got embarassing amounts of media attention). The site is now gone (because it was stupid and McCabe was run out of town). One of the major arguments I used was TO-to-TA ratio/differential. Why? Well, he keeps losing the puck, and doesn't exactly gain it back, and that's bad, right?

Not really. Giving away the puck a lot requires two things: lots of ice time, and lots of time with the puck. You need to have lots of time to posess what you're going to give away, leading to that event being an inevitability in good players. I used 2006/07 as my year to really dig into McCabe, but looking at the leaders in giveaways that year, you had the likes of Zdeno Chara, Sergei Zubov, Dan Boyle, Andrei Markov, Rob Blake, Sheldon Souray, Nicklas Lidstrom, Dion Phaneuf, Brian Rafalski, and Brian Campbell in the top 15. In hind sight, he was in elite company.

To give a slightly more modern twist? Dion Phaneuf lead his position in the stat, but was slightly trailed by Dustin Byfuglien, Andrei Markov, Dan Boyle, and Norris Winner PK Subban, and that's only going into the top 8. Takeaways are nice, but really are awarded for things like intercepted passes, and forwards are more cautious about throwing a cross-ice pass when an elite defender is in the way, and will go for the dump, which can still cause a posession change but further away from the opponent's offensive zone, and without a takeaway, which is why Jeff Petry has a better ratio than your favourite superstar defenceman. 

This evens out a bit more with forwards, but still has the same issue - the better a player is, the more cautious his opponents are in making a pass in front of him. The better he is, the more opportunity he has to posess the puck and give it away at times. Next.. 

to where players were finishing their bodychecks that provided an accurate indicator of how well the team was playing.

This is simpler and treats on similar concepts. My first theory to support this making sense was "oh, well if you look at where they're finishing hits, you can see where the opponent has the puck and judge how the team is doing from there". This fell apart when I remembered that the same thing can be tracked by, well, looking at where the other team is holding the puck, seeing as that includes times where they haven't forced themselves into a wall or made Dion Phaneuf taste blood in open ice. Odds are, the Leafs actually care where their players are throwing hits.

But why? If you're hitting another player, you don't have the puck. If you're finishing a check, you're usually ignoring the puck, and it's a crapshoot as to which team gets another guy in on time to continue the play, usually leaning towards the team of the hit player. So a player who throws a hit starts and ends the play without posession of the puck. Hitting is entertaining, sets the tone of the game, and all of that fun intangible stuff, but it's not the most effective thing in the world and certainly not an important metric to decide how well your team is playing.

Again, I could be wrong. Maybe the Toronto Maple Leafs have devised a system unlike anybody elses that works, and the people who are raising their eyebrows may have to give credit where credit is due. Maybe they do look at the same stuff many of us (and other NHL teams) do, and McClement just happened to mention the two outliers. Plus, I don't have an issue with most of McClement's other statements in the article, particularly that the team, like it or not, has an identity now and a roster that is moulded to better suit that identity, and may that may lead to better night-by-night play for now. But if the above is an indication of what the Leafs are focusing on when breaking down what happens on the ice, they seem to be focusing effort into the wrong spaces.

7cb905bdffc4d09e93770ff4a1889462
I bring news about the Toronto Marlies, opinions about the Toronto Maple Leafs, and a bunch of ridiculous thoughts about everything else.
Avatar
#1 leafer4life
August 28 2013, 01:25PM
Trash it!
2
trashes
+1
2
props

Just like +_ is misleading for some players (Fraser), I find that Corsi and hits can be misleading stat when trying to guage the talent of certain players.

So sure Mclemment Corsi (shot differential) is terrible but there is a reason and you need to consider context of how he used, his zone starts, who he plays against and who his teammates are rather then blindly focussing on shot differentials when accessing a players contribution.

Avatar
#2 leafer4life
August 28 2013, 07:22PM
Trash it!
2
trashes
+1
1
props
Cam Charron wrote:

*Searches post*

*Finds that the only mention of 'Corsi' comes from the comment section*

Yeah, that's about par for the course here.

Nice response but in case you are being candid then it is time brush up on your advanced stats as shot differential attempts is essentially Corsi.

Avatar
#3 SkinnyFish
August 28 2013, 12:34PM
Trash it!
1
trashes
+1
0
props

"The Toronto Maple Leafs have devised a system unlike anybody elses that works, and the people who are raising their eyebrows may have to give credit where credit is due."

They have? We do? You sure about that?

Avatar
#4 Cam Charron
August 28 2013, 03:55PM
Trash it!
1
trashes
+1
4
props

*Searches post*

*Finds that the only mention of 'Corsi' comes from the comment section*

Yeah, that's about par for the course here.

Avatar
#5 Justin Fisher
August 28 2013, 07:33PM
Trash it!
1
trashes
+1
1
props

In all fairness, maybe when the Leafs look at takeaway/giveaway ratios, they're looking for huge giveaway numbers to identify elite defencemen? Huh? Maybe? Huh huh? I tried.

Avatar
#7 josh
August 28 2013, 01:08PM
Trash it!
0
trashes
+1
0
props

You realize the actual quote was "on everything from..." meaning not only the two extreme cases you mention, but everything in between as well. Blown out of proportion much...

Avatar
#10 Justin Fisher
August 28 2013, 07:32PM
Trash it!
0
trashes
+1
2
props

Not to slag my boss, but I do wish @CamCharron would brush up on his advance stats.

/lols

Avatar
#11 Lance Tsunami
August 30 2013, 03:34PM
Trash it!
0
trashes
+1
1
props

Jeffler, you do lots of good stuff...but I think there are some gross assumptions here that many people seem to be missing.

You are making assumptions that:

1) The Leafs are using the NHL turnover-to-takeaway "definition", which yes, ignores the issues mentioned. 2) The Leafs are using the stat without putting it in context (e.g. where the turnover was made and what the score of the game was at the time).

The Leafs may define turnovers as "being stripped of the puck via a pass or while stickhandling when in a certain area of the ice when tied or defending a lead".

The Leafs could be monitoring this as a team, to try and keep them under "x" turnovers a game as a target for improvement...not necessarily to "rank" players.

The Leafs may also be looking at this individually, but recognizing the differences between Frazer McLaren (who should never ever have these turnovers and plays 5 minutes a game) vs. Kessel (who needs to take chances at appropriate time to create offence).

Jeffler said"But if the above is an indication of what the Leafs are focusing on when breaking down what happens on the ice, they seem to be focusing effort into the wrong spaces."

Agreed. But we don't really know what the Leafs are tracking and how they are using the data.

Comments are closed for this article.