The top 10 Dman who played at least 1000 minutes 5on5 in relative xGF% and the bottom 10

[Siignal]

Brad Treliving waived Valimaki. His 2022 with the Flames was just incredible blunders. Good riddance

 

[PatriceBergeronFan]

As typical, advanced stats can be manipulated to tell any narrative. Carlo at the bottom? Brutal statistic.

 

[bambamcam4ever]

Those are sort of interconnected, aren't they? If Bouchard-Ekholm is on the ice while all the goals are being scored, some other pairing - let's say Nurse-Ceci - needs to be bleeding goals against. Or the other way around. So each end of the list is sort of reserved for defenders on not so well-balanced teams.

Also Bouchard-Ekholm play with McDavid the vast, vast majority of the time. When you put all the best players on your team in the ice together, obviously their relative numbers will look better. Relative numbers are misleading more often than not, only in certain situations are they actually illuminating.

#8 Olli played on the literal most sheltered pairing in the NHL this past season.

View attachment 894384

Because Gostisbehere is probably the worst Dman defensively in the league

 

[Vinh]

You have to consider deployment, system, competition, overall team quality, etc. Looking at these numbers in a vacuum provides no insight whatsoever.

 

[tarheelhockey]

I will never understand people who do things like post a “top 10” list with a run of Gudkas-Dumoulin-Maata smack in the middle, and not take a step back to ask whether maybe the list is meaningless.

 

[Three On Zero]

As typical, advanced stats can be manipulated to tell any narrative. Carlo at the bottom? Brutal statistic.

Heavily skewed towards offensive players who have good linemates. Hurts actual good defenceman

 

[JPeeper]

I will never understand people who do things like post a “top 10” list with a run of Gudkas-Dumoulin-Maata smack in the middle, and not take a step back to ask whether maybe the list is meaningless.

Bro, Jusso Valimaki is one of the top d-men in the league. Respect the stats.

 

[Machinehead]

135-Ryan Lindgren... Why im not surprised?

Yeah, everyone is shitting on this stat, but that f***ing tracks lol

 

[WhatTheDuck]

Top 10 are Dman who are well above their peers in positive impact and bottom 10 are the biggest liability for their team.


I mean compared to the other Dman on the Ducks, he definitely was a positive veteran force. The others were either bad or inexperienced rookies.

I don't think anyone from the Anaheim side would argue against the assertion, Gudas definitely made a positive impact on the backend.

 

[DJJones]

Relative numbers have always been useless to me. It's usually some random third pairing guy that they use in an offensive role.

For example Miromanov had the same relative expected goals as Bouchard, just didn't hit the minute cut off. 427 minutes isn't nothing though.

Tanev and Hanifin were both negative. Which I think we all know the conclusion we should come to here. Miro > Hanifin and Tanev combined

This chart is just showing Bouchard/Ekholm and likely McDavid are a lot better than whoever Nurse player with.

 

[McPoyle]

Relative xGF%? A pretty useless stat.

A player with a 45% GF on a 40% GF team would be considered better than a 51% GF player on a 58% team.

 

[Fourier]

Relative xGF%? A pretty useless stat.

A player with a 45% GF on a 40% GF team would be considered better than a 51% GF player on a 58% team.

While rel xGF% has big issues your example does not really support your concern. It is very possible that a player with a 45%GF on a 40% GF team is much better than a 51% GF player on a 58%GF team.

 

[McPoyle]

While rel xGF% has big issues your example does not really support your concern. It is very possible that a player with a 45%GF on a 40% GF team is much better than a 51% GF player on a 58%GF team.

It is possible, but its not a guarantee. Its just a silly stat that doesn't show much of anything wihout context. Whats the QOC for example? TOI? On its own rel GF% says nothing.

 

[Fourier]

It is possible, but its not a guarantee. Its just a silly stat that doesn't show much of anything wihout context. Whats the QOC for example? TOI? On its own rel GF% says nothing.

I fully agree context matters. And I also do not like rel GF%.

 

[MCB]

Wow! The ducks have two in the top 10?! It's time to start planning the parade!

 

[LetsGoBLUES91]

Goes to show once again how useless advanced stats are in hockey.