Post by Montreal Canadiens on Aug 13, 2021 11:12:28 GMT -5
I have another question on how the ratings are made, i was just perusing things and noticed that Jeff skinner has a 76 def rating and Joel Armia has a 72, one was -16 and im sure has never seen 10seconds of PK time and the other was plus 10 and kills penalties at an elite level?
Post by Minnesota Wild on Aug 13, 2021 11:23:19 GMT -5
Former 40G scorer Jeff Skinner is elite at all facets of the game, he simple chooses to only exhibit 1 facet per season. Without Skinner on that line his linemates would've been twice as negative.
Kyle Connor - JT Miller - Nikita Kucherov Bryan Rust - Nazem Kadri - Valeri Nichuskin Rickard Rakell - Yanni Gourde - Trevor Moore Parker Kelly - Teddy Blueger - Nic Dowd
Drew Doughty - John Marino Ryan Pulock - Jakub Chychrun Trevor Van Riemsdyk - Nick Perbix Darren Raddysh
I have another question on how the ratings are made, i was just perusing things and noticed that Jeff skinner has a 76 def rating and Joel Armia has a 72, one was -16 and im sure has never seen 10seconds of PK time and the other was plus 10 and kills penalties at an elite level?
Hey Rick, thanks for the inquisition. For the DF the breakdown is as follows:
If we look at RAPM xGA/60 which takes a look at 12 different variables to account for a player's ability to prevent goals against and remove "luck" from it, Jeff Skinner was actually a pretty good defensive player despite his terrible contract. If we compare Jeff Skinner and Joel Armia in this metric you can see that Armia was actually a below-average player in the NHL this past year.
If we look at PK TOI/GP, Joel Armia averaged 1:31min/gp while Skinner didn't kill penalties so you are correct that Armia was a more useful PK guy than Skinner in that he killed penalties. But my question for you is how do you know Armia is an elite PK guy? Is it because he is used on the PK a lot of actually more effective than the average PK guy? If it is the latter, then can you point me to a metric that supports you statement because I'd love to use that in the ratings package moving forward. If it's the prior, we have to be careful about correlation and not causation, just because someone plays a lot of minutes on the penalty kill, doesn't necessarily mean they are good at that position (even though I am somewhat assuming that in the ratings package because I didn't know what else to use to add a PK weight into the DF rating).
Third, Defensive Zone Start % and this most recent year Skinner had a 44.4% while Armia was 48.6% which means they were deployed more as attacking forwards as they both started more than half their shifts in the O zone.
Lastly, blocks. Despite being used on the PK a lot, Armia only has 13 blocks all season. In comparison Skinner had 12 so this metric is a wash between the two.
In conclusion, Skinner has a better DF rating than Armia because he was a much more effective goal suppressor at even strength when we look at numerous variables to remove the luck element that would otherwise skew the results. The only metrics that Armia was superior than Skinner was at was being deployed in a PK position, but is that an actual skill, or a disposition by the coaches? Both players had similar zone start % and blocked shots.
Cheer up buddy, a good leader knows how to delegate. And you're delegating the puck science to someone really, really good at it while continuing to facilitate league operations. It's a sign the league is growing and improving and able to embrace change, and I think that's awesome. Plus, without your bearracuda connections literally no league business would ever get done.
2024-25 Toronto Maple Leafs Brady Tkachuk - Brayden Point - Tyler Toffoli Carter Verhaeghe - Nick Suzuki - Jordan Eberle Jake Neighbours - Blake Coleman - Luke Evangelista Jordan Martinook - Michael Amadio - Michael Eyssimont
Gustav Forsling - Noah Dobson Adam Larsson - Alex Pietrangelo Justin Schultz - Jackson LaCombe
Cheer up buddy, a good leader knows how to delegate. And you're delegating the puck science to someone really, really good at it while continuing to facilitate league operations. It's a sign the league is growing and improving and able to embrace change, and I think that's awesome. Plus, without your bearracuda connections literally no league business would ever get done.
Bearracuda has become a great location for the league meetings
I have another question on how the ratings are made, i was just perusing things and noticed that Jeff skinner has a 76 def rating and Joel Armia has a 72, one was -16 and im sure has never seen 10seconds of PK time and the other was plus 10 and kills penalties at an elite level?
Hey Rick, thanks for the inquisition. For the DF the breakdown is as follows:
If we look at RAPM xGA/60 which takes a look at 12 different variables to account for a player's ability to prevent goals against and remove "luck" from it, Jeff Skinner was actually a pretty good defensive player despite his terrible contract. If we compare Jeff Skinner and Joel Armia in this metric you can see that Armia was actually a below-average player in the NHL this past year.
If we look at PK TOI/GP, Joel Armia averaged 1:31min/gp while Skinner didn't kill penalties so you are correct that Armia was a more useful PK guy than Skinner in that he killed penalties. But my question for you is how do you know Armia is an elite PK guy? Is it because he is used on the PK a lot of actually more effective than the average PK guy? If it is the latter, then can you point me to a metric that supports you statement because I'd love to use that in the ratings package moving forward. If it's the prior, we have to be careful about correlation and not causation, just because someone plays a lot of minutes on the penalty kill, doesn't necessarily mean they are good at that position (even though I am somewhat assuming that in the ratings package because I didn't know what else to use to add a PK weight into the DF rating).
Third, Defensive Zone Start % and this most recent year Skinner had a 44.4% while Armia was 48.6% which means they were deployed more as attacking forwards as they both started more than half their shifts in the O zone.
Lastly, blocks. Despite being used on the PK a lot, Armia only has 13 blocks all season. In comparison Skinner had 12 so this metric is a wash between the two.
In conclusion, Skinner has a better DF rating than Armia because he was a much more effective goal suppressor at even strength when we look at numerous variables to remove the luck element that would otherwise skew the results. The only metrics that Armia was superior than Skinner was at was being deployed in a PK position, but is that an actual skill, or a disposition by the coaches? Both players had similar zone start % and blocked shots.
I have another question on how the ratings are made, i was just perusing things and noticed that Jeff skinner has a 76 def rating and Joel Armia has a 72, one was -16 and im sure has never seen 10seconds of PK time and the other was plus 10 and kills penalties at an elite level?
Hey Rick, thanks for the inquisition. For the DF the breakdown is as follows:
If we look at RAPM xGA/60 which takes a look at 12 different variables to account for a player's ability to prevent goals against and remove "luck" from it, Jeff Skinner was actually a pretty good defensive player despite his terrible contract. If we compare Jeff Skinner and Joel Armia in this metric you can see that Armia was actually a below-average player in the NHL this past year.
If we look at PK TOI/GP, Joel Armia averaged 1:31min/gp while Skinner didn't kill penalties so you are correct that Armia was a more useful PK guy than Skinner in that he killed penalties. But my question for you is how do you know Armia is an elite PK guy? Is it because he is used on the PK a lot of actually more effective than the average PK guy? If it is the latter, then can you point me to a metric that supports you statement because I'd love to use that in the ratings package moving forward. If it's the prior, we have to be careful about correlation and not causation, just because someone plays a lot of minutes on the penalty kill, doesn't necessarily mean they are good at that position (even though I am somewhat assuming that in the ratings package because I didn't know what else to use to add a PK weight into the DF rating).
Third, Defensive Zone Start % and this most recent year Skinner had a 44.4% while Armia was 48.6% which means they were deployed more as attacking forwards as they both started more than half their shifts in the O zone.
Lastly, blocks. Despite being used on the PK a lot, Armia only has 13 blocks all season. In comparison Skinner had 12 so this metric is a wash between the two.
In conclusion, Skinner has a better DF rating than Armia because he was a much more effective goal suppressor at even strength when we look at numerous variables to remove the luck element that would otherwise skew the results. The only metrics that Armia was superior than Skinner was at was being deployed in a PK position, but is that an actual skill, or a disposition by the coaches? Both players had similar zone start % and blocked shots.
Thanks for the explanation, I still don’t agree and you won’t find anyone in the real world who is taking Jeff skinner as a puck possession shut down guy over armia. He blocked less shots in 10 less games, scored the same pts in 10 less games and was plus 10, and killed penalties regularly. So skinner gets 0 out of 15 for Pk, he has almost 10% less DZ starts then Armia, he blocks less shots per game and he is still higher? This is where stats don’t tell the story. I’m ok with it, I’m sure I have another player with skewed DF ratings to the good but I think there might be something to revisit here moving forward.
If skinner starts at 85% because he never kills penalties and only lost 9 more pts on the remaining metrics his contract may not be as bad, that would put him ahead of Bergeron on selke nominations.
Thanks for the explanation, I still don’t agree and you won’t find anyone in the real world who is taking Jeff skinner as a puck possession shut down guy over armia. He blocked less shots in 10 less games, scored the same pts in 10 less games and was plus 10, and killed penalties regularly. So skinner gets 0 out of 15 for Pk, he has almost 10% less DZ starts then Armia, he blocks less shots per game and he is still higher? This is where stats don’t tell the story. I’m ok with it, I’m sure I have another player with skewed DF ratings to the good but I think there might be something to revisit here moving forward.
If skinner starts at 85% because he never kills penalties and only lost 9 more pts on the remaining metrics his contract may not be as bad, that would put him ahead of Bergeron on selke nominations.
I agree with you in that nobody takes Skinner as a shutdown guy, but last year he proved that at even strength, when accounting for things like zone start, quality of competition, whether he was playing at home or away, trailing or in the lead, etc. he was an above average forward at preventing goals being scored while he was on the ice; Armia was poor in this respect.
I think we may be arguing two different points. From what I can tell, you are saying that since Armia kills more penalties he's automatically a better defensive player than Skinner, which I mentioned before is correlation, not causation. We could've been put on the ice on a Habs PK unit for a few games and we would've logged more PK time than Skinner. Would it have made us better defensive players than Skinner? Absolutely not.
I will say that data has a silver-lining, the plus is that it is absolute, there is no subjectiveness to this. It's purely objective and objectively Skinner was the more defensively responsibly player than Armia last season at even strength. We could throw data out and just use the eyeball test and individually rate 1300 players and their 14 different stat catgeories, but that would be time consuming and open up a whole new argument of "you don't know what you're talking about"s.
The negative, and I think you and I are in agreement here is that there could be some data points that the analytics community aren't collecting that could make the data even more accurate than before. Like I said before, if there is a data set where you can prove that Armia was not only a penalty killer, but a penalty killer that was more effective than half the penalty killer's in the league at prevent power play goals, please send it my way because I would love to use it in the ratings package moving foward and this would objectively prove that Armia is an efficient player at one aspect of the DF rating.
Calgary Flames: that unassigned list has been huge
Aug 7, 2023 9:14:49 GMT -5
Dallas Stars: Go Dallas
Dec 10, 2023 20:47:04 GMT -5
Boston Bruins: Did we lose the discord?
Jan 23, 2024 7:15:42 GMT -5
Philadelphia Flyers: Announcement coming soon......
Jan 23, 2024 22:32:24 GMT -5
Anaheim Ducks: I got logged off Discord and I don’t know how to get back on. That’s frustrating.
Mar 5, 2024 16:05:08 GMT -5
Dallas Stars: Go Stars
Mar 12, 2024 12:48:29 GMT -5
Montreal Canadiens: not sure what teh point of this is anymore, when your 3rd line is better then the top line on 4 teams ahead of you in the standings simulation obbioulsy has zero reality to it.
Apr 11, 2024 13:16:33 GMT -5
Boston Bruins: @montreal... I feel your pain. Got spanked in a must win game by a team I hadn't lost to all season... Just another reason for hockey to make me drink!!! lol
Apr 15, 2024 11:01:12 GMT -5
Los Angeles Kings: 3rd line better ummm scratching my head
Apr 23, 2024 20:37:13 GMT -5
Dallas Stars: Stars winning the cup
May 10, 2024 19:30:28 GMT -5
Montreal Canadiens: Dakota joshua at 6 million, eeeeesh
Aug 17, 2024 21:03:54 GMT -5
Los Angeles Kings: yup one of those contracts you look 2-3 years down the road and go, why did I sign that deal
Aug 20, 2024 10:20:39 GMT -5
Dallas Stars: I had the money to spend but exactly why i didn't spend it. Just because you have the ability to, doesn't mean you should! Unless you plan on retaining for trade value after the 50 game no trade limit as a rebuilding team imo
Aug 20, 2024 16:31:28 GMT -5