Player Discussion Jake Sanderson (D) PART 3

[coladin]

Enjoying the bar scene I hear!

 

[dumbdick]

More documentation would be better, but it's still pretty clear what is going on in his model.

Nobody is treating statistical models as on par with direct observation, and let's not act like direct observation in this context is anything scientific. NHL teams directly observed Panarin not getting drafted. They also directly observed drafting Yakupov first overall.

Well keep me posted if I ever say any of those words you put in my mouth.

 

[ijif]

Well keep me posted if I ever say any of those words you put in my mouth.

That’s a solid dodge lmao

 

[OD99]

Enjoying the bar scene I hear!

Good for him!

 

[Cosmix]

To be fairer still, I don't think data scientist modelling geeks can sit in front of a computer and develop models with the ability to accurately project the future of teenagers.

I think "data scientist modelling geeks" can provide useful inputs to those responsible for making draft picks and trades. Scouts can do the "eye tests" and subjective forecasts too. I consider both useful inputs. It would be interesting to compare the results of each.

 

[JD1]

I think "data scientist modelling geeks" can provide useful inputs to those responsible for making draft picks and trades. Scouts can do the "eye tests" and subjective forecasts too. I consider both useful inputs. It would be interesting to compare the results of each.

Perhaps I'm jaded. I've worked with quite a few so called data scientist s, some educated to the PhD level, but I haven't worked with one that contextually understood what they were trying to analyze

 

[Cosmix]

Perhaps I'm jaded. I've worked with quite a few so called data scientist s, some educated to the PhD level, but I haven't worked with one that contextually understood what they were trying to analyze

So they were not hockey fans then.

 

[Cosmix]

I remember reading the Bill James baseball Abstract book, the 1986 version I think,. It was a very interesting read.

 

[JD1]

I remember reading the Bill James baseball Abstract book, the 1986 version I think,. It was a very interesting read.

Ya I read that too. Great book.

 

[SlapJack]

Perhaps I'm jaded. I've worked with quite a few so called data scientist s, some educated to the PhD level, but I haven't worked with one that contextually understood what they were trying to analyze

Not saying that understanding the subject is irrelevant, but if you're trying to show something through data you don't necessarily need to know the nuances. It could lead to bias.

For hockey, the data required isn't quite there to make solid predictions. Everything you see with your eyes could theoretically be quantified with data, but much of it is too complex to measure

 

[Masked]

I remember reading the Bill James baseball Abstract book, the 1986 version I think,. It was a very interesting read.

Yes but baseball is not hockey. Baseball is all independent events while hockey is all interdependent events. You can separate out the impact of your teammates in baseball. In hockey how do you make a statistical model for linemates who are always on the ice together?

Assen na yo!

 

[JD1]

Not saying that understanding the subject is irrelevant, but if you're trying to show something through data you don't necessarily need to know the nuances. It could lead to bias.

For hockey, the data required isn't quite there to make solid predictions. Everything you see with your eyes could theoretically be quantified with data, but much of it is too complex to measure

From a data science view, ok, maybe. But data is not much use if you don't understand what data to compile, whether it's useful, how to use it etc.

Do you watch the game last night. Did you see the short handed 2 on 1 that Brown had with Formenton I think. That's a clear cut 2 on 1 with no back pressure. That's a grade A opportunity to score. But Brown being Brown, he lost the puck trying to pass it, basically buried it into a sliding D man. And that play is representative of what's wrong with data science in a hockey world.

 

[AchtzehnBaby]

 

[Cosmix]

Yes but baseball is not hockey. Baseball is all independent events while hockey is all interdependent events. You can separate out the impact of your teammates in baseball. In hockey how do you make a statistical model for linemates who are always on the ice together?

Assen na yo!

I agree hockey is more complicated to track versus baseball. I expect hockey analytics to improve with time and adoption of computerized tracking systems.

 

[slamigo]

Honestly, I think that aside from skating (both speed and agility) the highest metric a player can possess is the ability to send/receive a pass. If you track a player's ability to pass, receive and manage the puck, you'd be about 90% there in terms of scouting. The game is so quick at the NHL level that any tiny bobble of the puck basically kills the play.

 

[Micklebot]

Yes but baseball is not hockey. Baseball is all independent events while hockey is all interdependent events. You can separate out the impact of your teammates in baseball. In hockey how do you make a statistical model for linemates who are always on the ice together?

Assen na yo!

I don't really follow it much, but apparently analytics in Basketball has been very effective. They do a much better job of capturing data though. Once Hockey analytics ditches the RTS that are poorly tracked and get the more robust tracking they do in basketball (obviously tougher to track a puck than a basketball) I think you'll see to big steps forward in how accurately things can be predicted.

 

[SensFan1010]

Grieg at 18 1 spot behind rossi is nice to see

 

[SpezDispenser]

Grieg at 18 1 spot behind rossi is nice to see

I'll take 18! This kid is so underrated around the league, it's nice to see Button giving him some love.

 

[Masked]

I don't really follow it much, but apparently analytics in Basketball has been very effective. They do a much better job of capturing data though. Once Hockey analytics ditches the RTS that are poorly tracked and get the more robust tracking they do in basketball (obviously tougher to track a puck than a basketball) I think you'll see to big steps forward in how accurately things can be predicted.

I don't follow basketball much either but I'm guessing that their use of analytics has led players to try for 3 pointers more often and things like that. That makes sense because sinking a basket is a much more independent event than trying to beat a goalie.

Assen na yo.

 

[Micklebot]

I don't follow basketball much either but I'm guessing that their use of analytics has led players to try for 3 pointers more often and things like that. That makes sense because sinking a basket is a much more independent event than trying to beat a goalie.

Assen na yo.

certainly possible, and that's one of the bigger impacts I'm aware of, but I think it goes well beyond just that. I know there's a lot of stuff on pass completion percentages and stuff like that that has apparently been beneficial, it breaks down whether the jump shot was contested or not, so you can adapt how much pressure you put on a given player based on their personal aptitude, whether they're more effective moving one way vs the other ect.

 

[NHL Dude 120]

No doubt. If they could, teams would hire them instead of scouting.

I'd say hire former hockey players who were data scientists. You need to combine the skillset with the expertise of hockey. Former CHL and NCAA players not going pro are perfect candidates to look at.

 

[Sens With Benefits]

Any news on if he suits up at all this year?

 

[cudi]

Any news on if he suits up at all this year?

I think last word was they're 'hopeful' so really no new news yet.

 

[SensFan1010]

How did Power look in his nhl debute?

 

[GCK]

How did Power look in his nhl debute?

Good. Very stable and confident.