silverfish
got perma'd
Matt Barlowe is pulling all of his public work down, and moving to NYC. There are some hints, but I'm not trying to get my hopes too high to have them come crashing down.
The Advanced Stats Thread Episode IX
- Thread starter Kakko
- Start date
-
Xenforo Cloud will be upgrading us to version 2.3.5 on March 3rd at 12 AM GMT. This version has increased stability and fixes several bugs. We expect downtime for the duration of the update. The admin team will continue to work on existing issues, templates and upgrade all necessary available addons to minimize impact of this new version. Click Here for Updates
Oscar Lindberg
Registered User
Why is it that these guys pull down their resources when they get hired? You think it is a condition by the team that hires them?
That would be the only thing that makes sense
A site like Corsica I could understand pulling down because Manny isn't going to be able to maintain it. But wiping visuals and twitter history off the map completely seems a bit unnecessary
That would be the only thing that makes sense
A site like Corsica I could understand pulling down because Manny isn't going to be able to maintain it. But wiping visuals and twitter history off the map completely seems a bit unnecessary
Manny also offline but might just be one of his short escapes, he does that from time to time, then he comes back trying to be professional with no shit posts 
100% part of the deal. teams aren't paying them for them to share their stuff with the world
Imagine NYR building a new analytics team spearheaded by Manny and Matt, i'd be so happy... Not only are they super smart, they're also great at shitposting.
Machinehead
HFNYR MVP
Machinehead
HFNYR MVP
It's not easy out here. The opinions are so toxic.Manny also offline but might just be one of his short escapes, he does that from time to time, then he comes back trying to be professional with no **** posts
silverfish
got perma'd
Blue Blooded
Most people rejected his message
Do the heatmaps look this good because Quinns system is good or is it a Mirage based on voodoo?
Blue Blooded
Most people rejected his message
I dunno, but we are score-adjusted 27th/27th (5v5/all) in CF% and 5th/3rd in xGF% during the streak. Runaway 1st in xPDO goes without saying; we have 101.98, 2nd has 101.14.
So I was thinking recently. If a team is playing poorly but are winning significantly more than you're losing, and have high PDO they're seen as a fluke, but all of the circumstances are the same but they're losing all of the games they lose in scenario 1 by more goals and the PDO is closer to 100 then that team is not a fluke. It doesn't add up. I know that this scenario is not super likely, but I remember how last year to start the season the Pens were winning close games and getting blown out when they lost. I don't know what their PDO was though.
- Oct 23, 2014
- 30,336
- 43,025
PDO is just one thing to look at. Goal differential is another.
If you are winning the close games but getting blown out otherwise, you are probably getting a tad 'lucky' in those wins. Thus, you question how sustainable that is moving forward. The 2017 Senators who were a shot away from the SCF, had a negative goal differential in that regular season and barely snuck into the playoffs. Were a lottery team next year.
- Oct 23, 2014
- 30,336
- 43,025
Roy, Hasek, Brodeur = untouchables?
Roy, Hasek, Brodeur = untouchables?
On the mainboards of all places, I found a pretty interesting discussion. Basically, someone is making the claim (first post) that SV% goes up as Shots Against go up and using it to support an agneda for Marty Brodeur in comparison to Dom Hasek and Patty Roy, as he faced fewer shots than those guys in the Devils trap years.
The poster recites a pretty large data set of games where a goalie faced <= 29 shots and >= shots and the SV% in games where the goalie faced 30 or more shots is noticeably higher for every single goalie.
Now the 2nd poster made a good point about selection bias and goalies being pulled. A goalie playing bad or on an off night will be pulled before he can eclipse 30 shots against so that skews the data, and also that the poster is using a rate stat (SV%) against a raw stat (shots against) and would be better off served using SA/60 as a measure of tendy workload.
Found this pretty interesting. The 2nd poster claims there is no legitimate corelation to SV% and shots faced and offered to better analyze the data if he was provided it. Figured I'd post it here and see if anyone else has any insights.
Roy, Hasek, Brodeur = untouchables?
On the mainboards of all places, I found a pretty interesting discussion. Basically, someone is making the claim (first post) that SV% goes up as Shots Against go up and using it to support an agneda for Marty Brodeur in comparison to Dom Hasek and Patty Roy, as he faced fewer shots than those guys in the Devils trap years.
The poster recites a pretty large data set of games where a goalie faced <= 29 shots and >= shots and the SV% in games where the goalie faced 30 or more shots is noticeably higher for every single goalie.
Now the 2nd poster made a good point about selection bias and goalies being pulled. A goalie playing bad or on an off night will be pulled before he can eclipse 30 shots against so that skews the data, and also that the poster is using a rate stat (SV%) against a raw stat (shots against) and would be better off served using SA/60 as a measure of tendy workload.
Found this pretty interesting. The 2nd poster claims there is no legitimate corelation to SV% and shots faced and offered to better analyze the data if he was provided it. Figured I'd post it here and see if anyone else has any insights.
silverfish
got perma'd
Let me give you guys the best career advice you ever will get, don't assume that everyone around you are idiots. I think you would be in for a pretty rude awakening if you walked into a coaching room trying to pitch the idea that you figured out that Pionk had bad Corsi...
1. They are of course aware.
2. They of course have much better cleaned numbers and other sources of data, come on, we have several people walking 50-60 hours a week with this.
3. Expect them to surely have a very clear picture of what happened during every single shift every game played. What caused a shift to play out as it did, how well did we handle the situations that occurred. Like a shift spent in your own end does not just "happen", there is a very obvious reason for it.
And as a result of this and many other factors they still choose the deployment that they do. To think there is a diagram that would change their opinion on this topic is pretty darn childish.
BTW I dropped in to ask if anyone have seen any work done on Philly lately?
Hackstol has some things going that could be pretty interesting. Did anyone notice how they have the far end winger turn back towards the board and to become a passing option? And Philly was constantly seeking that option with long hard cross ice passes. Drove our Ds nuts.
A common transition play is that you have like the LW head up ice along the board and if a pass does not come the LW curves into center ice and heads home facing his own net, so that the Ds can hit him with a pass that is redirected to either side of the ice. What Philly was doing was in principle the same, but instead of having a LW go from the board to center ice they reversed it. They had the LW cut in and then turn towards the boards and head home. Then they were constantly throwing pucks across the ice to that guy who could get it to the center or RW with another cross ice pass. And it was quite successful.
The big positive effect of Hackstol's play is the quality that is required on that first cross ice. If you hit the boards with it, the pass can be both high and really hard and be hard to pick up. Some teams have really struggled with Philly thanks to that play.
Hackstol has some things going that could be pretty interesting. Did anyone notice how they have the far end winger turn back towards the board and to become a passing option? And Philly was constantly seeking that option with long hard cross ice passes. Drove our Ds nuts.
A common transition play is that you have like the LW head up ice along the board and if a pass does not come the LW curves into center ice and heads home facing his own net, so that the Ds can hit him with a pass that is redirected to either side of the ice. What Philly was doing was in principle the same, but instead of having a LW go from the board to center ice they reversed it. They had the LW cut in and then turn towards the boards and head home. Then they were constantly throwing pucks across the ice to that guy who could get it to the center or RW with another cross ice pass. And it was quite successful.
The big positive effect of Hackstol's play is the quality that is required on that first cross ice. If you hit the boards with it, the pass can be both high and really hard and be hard to pick up. Some teams have really struggled with Philly thanks to that play.
- Oct 23, 2014
- 30,336
- 43,025
Machinehead
HFNYR MVP
I see we're back to the argument that the Rangers are the cutting edge of analytics and have stuff we haven't heard of yet.Let me give you guys the best career advice you ever will get, don't assume that everyone around you are idiots. I think you would be in for a pretty rude awakening if you walked into a coaching room trying to pitch the idea that you figured out that Pionk had bad Corsi...
1. They are of course aware.
2. They of course have much better cleaned numbers and other sources of data, come on, we have several people walking 50-60 hours a week with this.
3. Expect them to surely have a very clear picture of what happened during every single shift every game played. What caused a shift to play out as it did, how well did we handle the situations that occurred. Like a shift spent in your own end does not just "happen", there is a very obvious reason for it.
And as a result of this and many other factors they still choose the deployment that they do. To think there is a diagram that would change their opinion on this topic is pretty darn childish.
How long have we had an analytics department? A year?
silverfish
got perma'd
Machinehead
HFNYR MVP
Let's say the Rangers do have something cutting edge that that Pionk is good at.
He's still outshot, outchanced, and outscored. Does it matter what he's good at? It's a simple game. You're a positive on the ice, or you're not.
"If he's a good hitter, then why doesn't he hit good?"
He's still outshot, outchanced, and outscored. Does it matter what he's good at? It's a simple game. You're a positive on the ice, or you're not.
"If he's a good hitter, then why doesn't he hit good?"
Machinehead
HFNYR MVP
Let me give you guys the best career advice you ever will get, don't assume that everyone around you are idiots. I think you would be in for a pretty rude awakening if you walked into a coaching room trying to pitch the idea that you figured out that Pionk had bad Corsi...
1. They are of course aware.
2. They of course have much better cleaned numbers and other sources of data, come on, we have several people walking 50-60 hours a week with this.
3. Expect them to surely have a very clear picture of what happened during every single shift every game played. What caused a shift to play out as it did, how well did we handle the situations that occurred. Like a shift spent in your own end does not just "happen", there is a very obvious reason for it.
And as a result of this and many other factors they still choose the deployment that they do. To think there is a diagram that would change their opinion on this topic is pretty darn childish.
How does me posting i want Manny and Matt to join NYR make you write this post to me? ?!?!?! :s Truly confused.
