Sportsnet's top 20 defenseman

[Frank Drebin]

Jeff Petry people


This is your brain on Corsi

 

[Bank Shot]

If advanced stats tell you that Jeff Petry is basically equal to Hedman and Letang, then your advanced stats are pretty garbage.

Yeah. Advanced stats work much better for forwards, as they are a reasonably good proxy for offensive contribution.

The fact that you get lists like this when analyzing defencemen just shows how off they are when measuring defensive contributions.

 

[93gilmour93]

Hedman at 12 is beyond stupid

 

[Zegs2sendhelp]

Yeah, I don't agree with everything on it either. Some of the measures are slightly subjective.

But it has some value, and I will concede that's he's obviously a better player than I thought.

Im so happy I can cry

Granted fowler shouldn't be in the top 10... not even sure id have him in my top 20... and def not ahead of lindholm... and I think manson could very easily be argued over fowler(but manson didn't even make top 40).

Depending on what stats you choose to focus on tho, fowler can look really good.

 

[Zegs2sendhelp]

This list is extremely bad as well. Morgan Reilly and John Carlson should both be top 40 (if not top 20)

Rielly was like 23rd I think...

the list is based on 3 seasons... Rielly was pretty middle of the pack during 2 of those 3 seasons.

 

[Albus Dumbledore]

Rielly was like 23rd I think...

the list is based on 3 seasons... Rielly was pretty middle of the pack during 2 of those 3 seasons.

He was 21st. I agree if anything rielly should have been closer to 30-35. He didn't put up points 2 years ago but had hard matchups and the year before that he put up okay numbers but wasnt facing top competition. He's really only had the one season. Easy to see and I'm a leafs fan.

 

[Zegs2sendhelp]

He was 21st. I agree if anything rielly should have been closer to 30-35. He didn't put up points 2 years ago but had hard matchups and the year before that he put up okay numbers but wasnt facing top competition. He's really only had the one season. Easy to see and I'm a leafs fan.

Ah ya, knew he was in the start of the last list.

If it was just last year then ya id agree he prob makes the top 20 pretty easily, but because its based on last 3 seasons I understand why he is further down the list then 1 might expect.... if they do the list again in 3 years, I expect Rielly will be in the 10-15 range pretty comfortably(assuming he continues playing the way he has been)

 

[Albus Dumbledore]

Ah ya, knew he was in the start of the last list.

If it was just last year then ya id agree he prob makes the top 20 pretty easily, but because its based on last 3 seasons I understand why he is further down the list then 1 might expect.... if they do the list again in 3 years, I expect Rielly will be in the 10-15 range pretty comfortably(assuming he continues playing the way he has been)

Yeah I don't disagree with anything you posted. Cheers!

 

[Chips]

Tyson Barrie and Petry are both better than John Carlson?

I’m a Caps fan and I absolutely believe John Carlson was overhyped overall despite having a great season, but it looks like this list is basing these on last few years for some defenseman, and just last year for other defenseman.

Even just looking at last season, were they better than Carlson? Just looking over HF comments I always thought Petry wasn’t supposed to be that good, and that Barrie was one dimensional and injury prone?

Obviously I don’t watch either of their teams lol

 

[Dessert Nights]

List might be bad but it is definitive ranking when you use those stats. If the results are wrong / misleading then their model needs more work. I'm sure they set the rules how they are comparing players and went with it. Top 20 is just an output of their results.

 

[Paul4587]

Suter at 23, Hedman and Petry basically ranked the same. Fowler at 6 (as a Ducks fan I can say that’s an absolute joke, he should be around 25-35, not top 10!)

Whatever statistical weighting’s are being used are wrong.

 

[CCM19]

Wait a minute you mean to tell me Morgan Rielly didn't even make the top 20. I thought Sportsnet always puts a Maple Leafs player in the top 10 of these lists so they can suck up to Leafs fans.

This was pretty funny the first time, now you’re just looking for Leafs attention.

Carry on

 

[CCM19]

This list is horrific, starting right at number 1.

Technically not, according to the advanced stats. Is what it is budz

 

[Raccoon Jesus]

List might be bad but it is definitive ranking when you use those stats. If the results are wrong / misleading then their model needs more work. I'm sure they set the rules how they are comparing players and went with it. Top 20 is just an output of their results.

Yeah, I'll give them credit for having a model and using that to describe the results. I think that's all they're trying to do and can appreciate that.

However, I've always felt like Berkshire has a bit of confirmation bias, and actively seeks out stats that prove his theories rather than questioning the model and stats themselves.

It's a good conversation starter though.

 

[Jeti]

People realize this ranking is made from advance stats and not personal opinions, right?

But their weighting of the various advanced stats is completely subjective and arbitrary. The advanced stats themselves are also arbitrary and subjectively chosen until they're proven to have some predictive power for future success. Right now they're merely proxies for certain events which may or may not be useful in measuring the usefulness of a player. For instance...

Their only knock on Byfuglien is that he doesn't rush the puck enough? That's absurd. D should focus on making good breakout passes rather than trying to rush the puck. Byfuglien is elite at breakout passes so saying he's bad (relative to top 20) at transition is ridiculous.

Byfuglien should be top 10. Klingberg and Petry should not even be top 20.

 

[Avs44]

People realize this ranking is made from advance stats and not personal opinions, right?

You realize that the "advanced stats" we have access to currently in the NHL are not a wholly predictive, sure science and not at the level of tracking in a league like the MLB, right? (Hopefully will improve when player and puck tracking is implemented, I am excited for that). You realize that for a system like this the weighting put on the various advanced stats used, and the very choices of which advanced stats to use, are 100% subjective and determined by the author's opinion, right? For example why is defense only given a weighting of 35? Why not 40? Why not 50? Why is "difficulty" given a +/- of 30%, why not make it 20? Why not make it 40? What are the weightings put on each stat within each category, e.g. are they all weighed the same? Why? These things are all subjective and can be bickered about. They are entirely personal opinion.

The weight put on the categories of "defense", "offense", and "transition" are completely subjective, as are the stats used within those categories, as are the weights each year, etc. Rankings like this are still created by people and their opinions, who just happen to be using stats.

That's not to mention problems with the stats themselves. The problems are sort of inherent until we get better tracking and better methods, but just as an example I notice the author is relying on corsi heavily throughout, e.g. even corsi QOC as part of their test to determine the "difficulty" each defensman faces, but it's hardly a science. You can read articles here, here, and here that point out some of the issues with that.

I certainly don't have the time to go do it, but I'm going to hazard a guess that if someone went through to make this list again, changed all of the weightings (for year and category), and pulled / added a few stats here and there based on what they think is useful, they could come up with a list that could look fairly different from this one -- and how would you go about determining which one is better and which one is "objective"....unless you use your personal opinion?

I do think lists like this are really interesting and do have good value, but we should not start pretending advanced stats in the NHL are at the point where a list like this is an agreed upon, objective science, and not something that still heavily involves opinion - in putting together, critiquing, and using.

 

[Flair Hay]

People realize this ranking is made from advance stats and not personal opinions, right?

All the more reason to rip into it

 

[Vlad The Impaler]

People realize this ranking is made from advance stats and not personal opinions, right?

Then they should call it the not-necessarily-good-but-kinda-good-on-paper-if-you-are-a-stat-wanker list of defensemen.

Petry is awful. He's not even a proper first pairing D.

Kinda glad to see the overrated Hedman a little down on the list, though.

 

[member 151739]

I'm not saying the formula to determine this list is perfect (it's not), but some people are so quick to disparage advanced stats if they go against any preconceived notions of who are the best (or simply good) players. Perhaps you should consider your biases and that maybe, just maybe, you are wrong about a certain player.

What I'm saying goes far beyond just this thread.

 

[KaseMeOutside]

LMAOOO where are all the stat guys who have sworn forever how awful Cam Fowler is, this article uses advanced stats and he's 6th

You either have to finally admit he's good or your advanced stats are junk science

LMAO

 

[Raccoon Jesus]

I'm not saying the formula to determine this list is perfect (it's not), but some people are so quick to disparage advanced stats if they go against any preconceived notions of who are the best (or simply good) players. Perhaps you should consider your biases and that maybe, just maybe, you are wrong about a certain player.

What I'm saying goes far beyond just this thread.

The criticism, though, to turn it around, is that the author doesn't do that.

However, the irony is, as you are saying, some folks fall into the same trap.

I'm big on advanced stats but I just don't think enough people question them in general particularly when a list like this is put together. An exhaustive use of stats that are manipulated by subjective weighting is more ripe for criticism than a completely objective presentation, yet some folks will take this list as gospel, ignoring the weighting biases. Guarantee we see people copy-and-paste these charts, contextless, all over the forums all year in arguments.

But as I mentioned earlier I think it's a fantastic jumping-off point for a conversation. I just also think the author calling it the "definitive" ranking is arrogant and smug as f*** even if I appreciate the scope of the work.

 

[member 151739]

The criticism, though, to turn it around, is that the author doesn't do that.

However, the irony is, as you are saying, some folks fall into the same trap.

I'm big on advanced stats but I just don't think enough people question them in general particularly when a list like this is put together. An exhaustive use of stats that are manipulated by subjective weighting is more ripe for criticism than a completely objective presentation, yet some folks will take this list as gospel, ignoring the weighting biases. Guarantee we see people copy-and-paste these charts, contextless, all over the forums all year in arguments.

But as I mentioned earlier I think it's a fantastic jumping-off point for a conversation. I just also think the author calling it the "definitive" ranking is arrogant and smug as **** even if I appreciate the scope of the work.

Oh, I'm not defending the author or the formula used in weighting this list. This problem goes far beyond this thread, as I said.

 

[Ducks in a row]

18.
Hampus Lindholm
Difficulty adjustment: +2.46%
Offence: 11.14/25 | Transition: 27.27/40 | Defence: 21.07/35
Total: 59.48/100

One of the toughest players to play against in the league because he’s so smart, Lindholm is an excellent shutdown defenceman who specializes in winning puck battles, blocking passes, and cutting players off in the neutral zone. Only a couple players in the league allow fewer controlled entries.
Lindholm doesn’t directly bring a lot of offence to the table, but his ability to move the puck up the ice provides far more and better chances for his teammates than you might expect by just looking at his point totals. Lindholm is heavily involved in directing neutral zone traffic and makes passes to open up lanes for teammates to attack more off the rush. He’s an incredibly important cog in the machine for an Anaheim team that struggles to play with pace.


6.
Cam Fowler
Difficulty adjustment: +3.08%
Offence: 11.69/25 | Transition: 28.98/40 | Defence: 22.05/35
Total: 62.72/100

I’m always a little surprised by how well Fowler shows in these breakdowns, but he is a smooth operator. On a team with a lot of slower, physical forwards, Fowler and Lindholm are the keys to the Ducks’ transition game. While Lindholm is more of a director in the neutral zone, Fowler is a versatile puck rusher and distributor in the defensive and neutral zones who breaks down defensive schemes with his playmaking and skating.
He is among the NHL’s most impressive defencemen with the puck at both blue lines, and he rarely ever makes missteps that result in turnovers. Though Fowler is 6-foot-2, he doesn’t strike anyone as a big player, though he is excellent at winning contested puck battles. He’s no Lindholm at defending the blue line, but Fowler is above average in that department, too.

Fowler at 6 with Lindholm at 18 this is crazy. Am I living in the Twilight Zone?

 

[glovesave_35]

I'm actually relieved there are no Toronto D in the top 20 so I don't have to read page after page of supposed Toronto biased media. Now will people accept that the other rankings (Matthews, Tavares, Marner and Nylander) were reasonable, unbiased and based on stats? Probably not, worth a shot though.

Three mentions of this by you guys so far. How about just letting it go and not mentioning the Leafs.

 

[traparatus]

While each individual stat measured for these lists is objective, how much weight you give them is entirely up to you.

Well, up to the GM, anyway. Replacing a lot of these players with higher ranked once would only serve to make their respective teams worse.

Hell, the only player on that list to win a Stanley Cup in the last 3 years is Letang. If you ask me, Pens would have been much better off replacing him with a much lower ranked but more risk averse defenseman during last playoffs.

Point is, the goal is to win the Cup and been ranked high up on a list ain't gonna get you there.