Management GM Pierre Dorion/Front Office Thread - Part IX [Mod Warning in post 1)

[Ice-Tray]

Another heartbreaking statement from PD: he LOVES analytics.

Said he goes the through the reports after every game. Says sometimes they come in at 3am and he’s up reading them.

 

[Hale The Villain]

I think Markstrom going from one of the best goalies in the league statistically to one of the worst, while having a backup with sub .900 save% is their biggest issue.

Sutter close behind them.

Huberdeau having his worst season as a pro is 3rd.

Yeah Markstrom having a bad year sunk them.

But Huberdeau and Kadri going from a combined 202 points last season to 101 points this year screams Sutter's system stifling offense.

 

[BonHoonLayneCornell]

Another heartbreaking statement from PD: he LOVES analytics.

Said he goes the through the reports after every game. Says sometimes they come in at 3am and he’s up reading them.

I'm curious, why is this heartbreaking?

Yeah Markstrom having a bad year sunk them.

But Huberdeau and Kadri going from a combined 202 points last season to 101 points this year screams Sutter's system stifling offense.

Ya that's why I'd run it back one more time with a different coach and see what happens. They can start the shedding at the deadline if they collapse.

They were one of the best only a year ago and while Hubs, Kadri & Weegar's deals may look bad on the back half, they are by no means over the hill yet.

Markstrom probably pulls an Anderson too and comes back strong after a bad year.

 

[Burrowsaurus]

Also mentions something about extending Sanderson.. $ are limited. I expect DBC to be traded

I think somethings are going to have to happen. Dead money off books. Then giroux retire. And then Joseph traded.

 

[Knave]

I'm curious, why is this heartbreaking?

Garbage in is garbage out

 

[Tnuoc Alucard]

Treliving was let go. Hope we can get him.

 

[BonHoonLayneCornell]

Garbage in is garbage out

 

[Ice-Tray]

I'm curious, why is this heartbreaking?


Ya that's why I'd run it back one more time with a different coach and see what happens. They can start the shedding at the deadline if they collapse.

They were one of the best only a year ago and while Hubs, Kadri & Weegar's deals may look bad on the back half, they are by no means over the hill yet.

Markstrom probably pulls an Anderson too and comes back strong after a bad year.

Heartbreaking to the PD haters who have argued for ages that we don’t use or like analytics, based on zero information beyond personal dislike.

It just another weapon removed from the arsenal.

 

[BonHoonLayneCornell]

Heartbreaking to the PD haters who have argued for ages that we don’t use or like analytics, based on zero information beyond personal dislike.

It just another weapon removed from the arsenal.

Gotcha. Definitely not clear lol, I thought you meant because he was still up at 3am.

 

[Knave]

View attachment 690059

At the end of the day the data needs to be collected. The more complex or ambiguous you make the data you are collecting... the more room there is for interpretation by the people collecting it.

So you have garbage data because it's not collected fairly, impartially and there is not one size fits all definitions for the variables being collected.

So when you put flawed data (garbage data) into your model thinking you're going to produce great results... you don't and instead you get garbage out.

What makes it even worse is players are people, they age and most will play fewer than 82 games a year. These aren't huge sample sizes and they're influenced by tons of other factors that none of the advanced stats will be able to account for now or in the near future.

Furthermore this isn't baseball or football. This isn't to slam those sports but they are stop and go. Hockey like Basketball or Soccer is fluid. Baseball has much easier metrics and variables to define. A batter walks up to the plate. The pitcher is pitching to them. A maximum of 3 other players on base might try to steal. Same thing with football - one team is defending, one team is on offense. The quarterback will have the ball snapped to him in almost all situations. Plays are predesigned, routes players run are predesigned and it's easy to understand if it succeeds, if it fails and if there is a scramble play that is improvised.

I feel like people read Moneyball and learned all the wrong lessons. First the As didn't win anything - kind of like Kyle Dubas - and second the As relied heavily on traditional scouting for their bullpen outside of one example in Moneyball. And that one pitcher was a relief pitcher. It would be like if we were playing up our #7D as some sort of proof or savior. It's ridiculous. Some of their more important batters were again... not particularly efficient according to sabermetrics. But they scored runs for wins.

I think sabermetrics and analytics more generally has been played up to extremes just like AI is being played up right now. I've worked with AI for many years now. It's trash. There's potential but right now? It's not reliable. And the same should be said for analytics instead of the nauseating praise it receives.


By the way - Pierre Dorion isn't alone in this, most of the league including Kyle Dubas is on this train and it's sad.

Anyways those are my disorganized thoughts. It feels like a cult for geeks who like sports.

 

[Tufted Titmouse]

At the end of the day the data needs to be collected. The more complex or ambiguous you make the data you are collecting... the more room there is for interpretation by the people collecting it.

So you have garbage data because it's not collected fairly, impartially and there is not one size fits all definitions for the variables being collected.

So when you put flawed data (garbage data) into your model thinking you're going to produce great results... you don't and instead you get garbage out.

What makes it even worse is players are people, they age and most will play fewer than 82 games a year. These aren't huge sample sizes and they're influenced by tons of other factors that none of the advanced stats will be able to account for now or in the near future.

Furthermore this isn't baseball or football. This isn't to slam those sports but they are stop and go. Hockey like Basketball or Soccer is fluid. Baseball has much easier metrics and variables to define. A batter walks up to the plate. The pitcher is pitching to them. A maximum of 3 other players on base might try to steal. Same thing with football - one team is defending, one team is on offense. The quarterback will have the ball snapped to him in almost all situations. Plays are predesigned, routes players run are predesigned and it's easy to understand if it succeeds, if it fails and if there is a scramble play that is improvised.

I feel like people read Moneyball and learned all the wrong lessons. First the As didn't win anything - kind of like Kyle Dubas - and second the As relied heavily on traditional scouting for their bullpen outside of one example in Moneyball. And that one pitcher was a relief pitcher. It would be like if we were playing up our #7D as some sort of proof or savior. It's ridiculous. Some of their more important batters were again... not particularly efficient according to sabermetrics. But they scored runs for wins.

I think sabermetrics and analytics more generally has been played up to extremes just like AI is being played up right now. I've worked with AI for many years now. It's trash. There's potential but right now? It's not reliable. And the same should be said for analytics instead of the nauseating praise it receives.


By the way - Pierre Dorion isn't alone in this, most of the league including Kyle Dubas is on this train and it's sad.

Anyways those are my disorganized thoughts. It feels like a cult for geeks who like sports.

This is a lot of words to basically say you don't really understand analytics or how they are used at higher levels.

Pierre built a pretty good team, there was just a lack of execution and the goaltending was not good. The contract situation is good outside of DeBrincat.

In terms of being like Dubas - Dubas traded away 2 analytics darlings for some guys that have charts full of red. Weird comparison when there are GMs like the Carolina group which are much more public with their relationship with analytics.

Not to mention the stats Dorion is using are nothing like what you have access to publically and for free. People like dom and JFresh are more about creating visuals with free to use stats, professional clubs are no doubt far more advanced.

 

[Knave]

This is a lot of words to basically say you don't really understand analytics or how they are used at higher levels.

Pierre built a pretty good team, there was just a lack of execution and the goaltending was not good. The contract situation is good outside of DeBrincat.

In terms of being like Dubas - Dubas traded away 2 analytics darlings for some guys that have charts full of red. Weird comparison when there are GMs like the Carolina group which are much more public with their relationship with analytics.

Not to mention the stats Dorion is using are nothing like what you have access to publically and for free. People like dom and JFresh are more about creating visuals with free to use stats, professional clubs are no doubt far more advanced.

You're essentially saying "trust me, bro". And no, I will not.

I am aware that the Senators likely contract analytics firms that attempt to collect data and create models. That changes nothing I have said.

 

[bicboi64]

I'm sure PD looks at analytics reports as in he visually looks at the paper. Not sure his ability to understand what they mean is there. You didn't need analytics to tell you Gus for Talbot is bad, or signing Murray to $6 million aav isn't a good idea.

Sucks that he feels like he needs to look at them at 3am, get a good nights rest Dorion. You've made some good contract extensions and need you fully rested if you're in charge of anymore.

I'm sure every GM in the league looks at analytics, but ultimately their final decision isn't solely influenced by them. I'm sure Dubas likes analytics, doesn't change the fact that he sucks at contract signings, the same way Dorion has poor pro eval.

 

[Ouroboros]

You're essentially saying "trust me, bro". And no, I will not.

I am aware that the Senators likely contract analytics firms that attempt to collect data and create models. That changes nothing I have said.

'You're too stupid to get it" is always the immediate go-to for devotees of Our Lady of The Expected Goals.

This offers insight into why they have embraced it so willingly.

 

[Golden_Jet]

You're essentially saying "trust me, bro". And no, I will not.

I am aware that the Senators likely contract analytics firms that attempt to collect data and create models. That changes nothing I have said.

No the Sens like every team in the league pay 200k/ year to use sportlogiq.
There is no model.
AHL , junior , college , broadcasters also use sportlogiq, the more you pay the more data you get,

 

[Snowwy]

I feel like people read Moneyball and learned all the wrong lessons. First the As didn't win anything - kind of like Kyle Dubas - and second the As relied heavily on traditional scouting for their bullpen outside of one example in Moneyball. And that one pitcher was a relief pitcher. It would be like if we were playing up our #7D as some sort of proof or savior. It's ridiculous. Some of their more important batters were again... not particularly efficient according to sabermetrics. But they scored runs for wins.

Ooof.. I think it is pretty clear that you did not understand Moneyball..

 

[Neil Patrick Harris]

I'm curious, why is this heartbreaking?

It means PD is probably not getting a proper 8 hours of sleep every night

 

[Knave]

Ooof.. I think it is pretty clear that you did not understand Moneyball..

No I very much got it. It's like a Malcolm Gladwell piece.

 

[BonHoonLayneCornell]

It means PD is probably not getting a proper 8 hours of sleep every night

Well then to Dorion I say...

 

[Tufted Titmouse]

'You're too stupid to get it" is always the immediate go-to for devotees of Our Lady of The Expected Goals.

This offers insight into why they have embraced it so willingly.

The difference I guess is that people who are interested see it as a tool and are open to using as many tools as possible to form their opinions. Then there are people who use terms like "Our Lady of Expected Goals" or "spreadsheet nerd" who think they see it all with their eyeballs and don't need data.

Most of us prefer to use all the tools we can, but you are welcome to make that decision on your own. I think a lot of GMs around the league have a higher opinion of your team than some of the fans here.

Here's where Ottawa finished 5v5 (natstattrick which can differ from hockey ref) - they just couldn't score 5v5. Norris will bump that up, and no one expect DeBrincat to go from a 40g scorer to struggling to hit 27.

 

[PlayItAgain]

No I very much got it. It's like a Malcolm Gladwell piece.

No I don’t think you do get it. Your statement earlier that garbage in = garbage out makes that clear. The data coming in is based off official nhl statistics so unless you’re implying they completely make the numbers up, I don’t get what you’re trying to say.

 

[Tragedy]

Mommmm! The eye test guys and number crunchers are arguing again!

 

[Knave]

Mommmm! The eye test guys and number crunchers are arguing again!

I'm too cool to care about something

 

[FunkySeeFunkyDoo]

Heartbreaking to the PD haters who have argued for ages that we don’t use or like analytics, based on zero information beyond personal dislike.

It just another weapon removed from the arsenal.

Which of Dorions moves in the past 12 months have been driven by analytics?

 

[Relapsing]

Heartbreaking to the PD haters who have argued for ages that we don’t use or like analytics, based on zero information beyond personal dislike.

It just another weapon removed from the arsenal.

Nah, they'll just say he doesn't understand what the analytics are showing... Unlike them, of course.