Nylander signs 6-year contract

[Zee]

Like I said, Nylander is the reverse Samson, much better with short hair

[Guilt Trip]

Shave the beard, cut the hair, whatever it takes to turn things around.

[Highlander]

you mean you want him to play for the Islanders

[herman]

Like I said, Nylander is the reverse Samson, much better with short hair

William Nylander said his aunt who cuts his hair is in town, hence the new look. Maybe will change his scoring luck a little too. Said he’s not happy with where game is at yet, but working as hard as he can to get it back up.

— Kristen Shilton (@kristen_shilton) December 31, 2018

Turns out it wasn’t due to any visit from Lou, just his stylist/aunt wasn’t around until now. Playing with Johnsson is sure to get it back up.

[Zee]

Like I said, Nylander is the reverse Samson, much better with short hair

William Nylander said his aunt who cuts his hair is in town, hence the new look. Maybe will change his scoring luck a little too. Said he’s not happy with where game is at yet, but working as hard as he can to get it back up.

— Kristen Shilton (@kristen_shilton) December 31, 2018

Turns out it wasn’t due to any visit from Lou, just his stylist/aunt wasn’t around until now. Playing with Johnsson is sure to get it back up.

Guy is a millionaire but relies on his aunt who's been cutting his hair since he was a kid

[herman]

William Nylander's December
My assessment using the latest SKATR update.https://t.co/99aA86pWeq #LeafsForever #Leafs #LeafsNation #MapleLeafs

— 🇨🇦 Bill Comeau Crush the Curve 📉 (@Billius27) January 1, 2019

The top bunch of bars being where they are is why most fans are disappointed with Nylander so far.


All the big blue bars where they are even better than before is why I'm not worried because that second block of bars is a far more stable indicator of the player's controllable effect on the game. The top chunk is scoreboard results.

Check out the SKATR comparison tool for yourself to see how Willy stacks up.

If you look at Matthews' you can see how he has struggled outside of actual scoring (which is great) without Willy on his wing. Look to see that rebound as we run through the back half of the season with Willy and Matthews reunited.

[Zanzibar Buck-Buck McFate]

I am not trolling you herman but I absolutely don't get the line of reasoning that the least controllable effects a player has on a game are the only results that really matter.  In philosophy they call that a "category mistake" (or a basic ontological error).

[herman]

I am not trolling you herman but I absolutely don't get the line of reasoning that the least controllable effects a player has on a game are the only results that really matter.  In philosophy they call that a "category mistake" (or a basic ontological error).

Happy to discuss!

First off, let’s look at what I actually wrote:

that second block of bars is a far more stable indicator of the player’s controllable effect on the game.

I’m certainly not saying that goal results do not matter; what I’m saying is that puck possession metrics in this sample of games is a better indicator of the player’s performance. Implicit in shot metrics are how much the team has the puck when this player is on the ice. We can probably pick out why Nylander is such a good play driver with some video, as outside of the first 4 games he’s been forechecking well, stealing pucks at a higher per 60 rate than anyone else on the team right now, and exiting/entering zones with impunity. The numbers say when Nylander Ian on the ice, the puck ends up in the OZ for shots far more often than not.

If that’s happening so often, why isn’t the puck in the net? He’s attacking from good parts of the ice, his teammates aren’t garbage, what gives? Puck luck, whether people recognize it or not, is a thing. A player can do everything right to get into position and take a good pass in his wheelhouse and something prevents the goal (broken stick, rut in the ice causes a dman to stumble into just the edge of the puck and deflects it, whatever). Goals are rare in this game for a reason. On the other hand you also get Ron Hainsey goals where he blindly slaps it 4 ft wide but it clips the dman’s back skate as he’s turning to crosscheck someone and the puck finds the net. In a small sample (i.e. a handful of games), those results are highlighted. After a quarter of a season (and sometimes it takes a whole season — see William Karlsson or David Clarkson) their actual level of play is revealed (how many has Hainsey scored since his week of two goals?).

The Leafs as a team, by shot metrics, has climbed significantly since Nylander’s return. Not all of it is specifically because of him but his presence adds to our depth and our depth is overtaking their opponents. It’s setting up our top lines for cushy offensive starts and generally getting them into better position to do damage.

[IJustLurkHere]

I am not trolling you herman but I absolutely don't get the line of reasoning that the least controllable effects a player has on a game are the only results that really matter.  In philosophy they call that a "category mistake" (or a basic ontological error).

Isn't Herman saying that the second group of bars are the more controllable effects, and the ones that matter? i.e. shots share is more controllable than points?

Or am I missing something?

[Zanzibar Buck-Buck McFate]

Thanks herman, appreciate the thoughtful answer.

We have already disagreed before over how big a role randomness plays in goal scoring so I won't replay that.  (To digress, my own view is that the innate low-scoringness of hockey is primarily a function of how difficult it is to achieve the object, which is fundamentally determined by the ratio of the net size to the goaltender combined with the size of the gamepiece (the puck) and the speed with which it can be propelled.  Compare soccer, also low-scoring, where all these factors are different but utlimately in combination still give you a low-scoring game. In other words, scoring goals in hockey is hard mostly because the designers of the game made it that way.)

I totally understand that players can make significant contributions to scoring without scoring themselves.  But at the end of the day if those top bars aren't blue, then a player like Nylander is failing.  He's expected to score, not just support scoring indirectly.  I think it's fine to use these stats to point out the things he is doing right.  And I don't expect him to stay mired in a scoring slump all year. 

But I do think it is legitimate for fans to now begin to question how much longer non-scoring from him can be explained away.  That was the gist of something I posted last week, to the effect that he'd better start putting points up real real soon after the Xmas break or else face criticism. 

He's had enough time now to begin producing.  People can disagree with that but I don't think they can go on to characterize people like me who do hold that opinion as being unreasonable.

And again, for the record, I am a big fan of 29. 

[Zanzibar Buck-Buck McFate]

I am not trolling you herman but I absolutely don't get the line of reasoning that the least controllable effects a player has on a game are the only results that really matter.  In philosophy they call that a "category mistake" (or a basic ontological error).

Isn't Herman saying that the second group of bars are the more controllable effects, and the ones that matter? i.e. shots share is more controllable than points?

Or am I missing something?

I'll let herman answer that, but part of what I am arguing is that you cannot say the first set of bars don't matter.

[herman]

IJustLurkHere was merely pointing out that I didn't say that first set of bars don't matter. My elaboration clarifies that I meant those results are not indicative of Nylander's true performance.

You're welcome to describe it as anything other than luck, but would it be fair to say there are certain elements to a goal's probability equation (which I'm out of my depth to generate) which are directly under the player's control, under other players' control, and subject to the whims of the environment?

The on-ice metrics measure the effect of the elements of the game most directly under said player's control. Everything else has a bigger spread of result possibilities that take many more samples to even out due to all the external variables (regress to their mean -- peak of the normal/bell curve).

As for when even I will start to be worried/disappointed in his lack of results, again I will refer to Johnny Gaudreau's rough first two months after missing training camp as that's really the closest comparable to this situation (and I wouldn't count players with pre-season injuries because that's a different issue). Scale accordingly for Nylander who missed training camp and two regular season months; it won't be a linear adjustment either as mentioned previously because all the players around him are in mid-season form. Personally, I think he'll be fine come mid-January and I wouldn't be truly worried until March.

[Coco-puffs]

Lets not forget the reason why Corsi became popular was because it was the most PREDICTIVE measure analysts could come up with.  ie.  If we are going to use past performance to determine probabilities of future success, Corsi currently stands as the best measure.

So, herman is not incorrect in saying he expects Nylander to turn it around.  I think we all do, considering he has begun to look better and the underlying stats say its PROBABLE that things will improve.

That said, I don't think using Corsi is the best way to measure past success.  I would not agree that "those results are not indicative of Nylander's true performance".  Yes herman, they are.  He hasn't provided the offense we expect.  Some of it is certainly luck, but you make your own luck in lots of cases.

[herman]

That said, I don't think using Corsi is the best way to measure past success.  I would not agree that "those results are not indicative of Nylander's true performance".  Yes herman, they are.  He hasn't provided the offense we expect.  Some of it is certainly luck, but you make your own luck in lots of cases.

I don't think I was measuring past success here, merely that as he regresses to the mean, it won't be these dismal results. By true performance, I mean his most probable outcomes (i.e. getting points), which his current season's totals are clearly not reflective of. If you look at how often Nylander is getting his team into the OZ and what his shooting locations are, that's precisely how an individual player creates his own luck.

[princedpw]

One thing I’d add is that it looks from those charts like Nylander has been very heavily sheltered — bad competition and heavy zone starts whereas last year he probably had the opposite (at least in terms of competition).

Also, there’s corsi and there’s luck but there are other things that may well be under a player’s control but just harder for us to measure and/or not public data (eg: generation of pre-shot movement and/or proximity of other players).  Such things may be factoring into our “eye test” when evaluating Nylander.