Thursday, January 31, 2008

Soupy vs. Hank: Only One Can Survive

All right, kids, pull up a chair, turn on your brains, and bring a calculator. This post is going to be wordy - okay, they're all wordy but this one's gonna be way worse - and filled to the brim with numbers. Today we shall prove once and for all that Brian Campbell is not that much more valuable than Henrik Tallinder if he is indeed more valuable at all. If you get bored and/or confused, scroll down to the bottom where I will summarize our findings in few short, easy-to-read bullet points

Now most people compare two players at the same position in one way: point totals. Brian Campbell has a robust 33 points so far this season (4 goals, 29 assists) while Henrik Tallinder has a measly 6 points (0 goals, 6 assists). Campbell must be better, right? Not so fast. I really don't think simply looking at point totals works in this case and not just because it works against my favorite player (although that's definitely a strike against it). Here are my problems with just comparing points in this situation: One, it favors Campbell's strength as a player (offense) while disregarding Tallinder's strength (defense). Two, you're not looking at defense at all which is at least a little bit important when you're talking about defensemen. (I know, it's crazy.) So we're gonna bust out some other numbers up in here.

First a few disclaimers. All numbers were taken from BehindTheNet.ca which is an absolutely amazing website. It's a gold mine of statistics and it boggles my mind what you can find there. I also have to give a huge shout-out to reader, commenter, and fellow blogger Meg. She did all the foot work in parsing the following numbers. All I'm doing is writing it up in blog form. (I did ask for her permission which she kindly granted.) Finally all numbers are based on defensemen who have played at least 30 games this season and who play at least 15 minutes at even strength per 60 minutes of game time. That eliminates any outliers - guys who have great numbers but have only played 3 shifts all season due to injuries or whatever. Meg just looked at even strength numbers because she was laaaa-zy (just kidding, Meg!) but I think that's okay with these two players anyway. I would consider special teams kind of a wash. Most people aren't going to argue that Tallinder isn't more valuable on the penalty kill and most people aren't going to argue that Campbell isn't more valuable on the powerplay. And if you are going to argue either of those things, well, you'll have to do it somewhere else. Ha! So all that said, let's get ready to rumble!

The first stat we'll look at is Quality of Competition. This takes into consideration who the players in question are playing against. I think this is pretty important because to me, for a guy to be considered a top pairing defenseman, he needs to play his share of time against the opposing teams' top lines. Even against a team like Ottawa where the third line still has talent, the top line is well, the top line. No other line on the Sens team is as good as the Heatley-Alfredsson-Spezza line. And on teams that don't have as much talent as Ottawa, let's say, Tampa Bay for example, there's an even bigger difference between playing against the top line and playing against a lower line.

Tallinder's Quality of Competition rating is .15, the highest on the Sabres and good for 13th out of the defensemen who met our criteria. I won't bore you with all the names above him but it's a very solid group that includes Nick Lidstrom and Anton Volchenkov just as a couple of examples.

Campbell's Quality of Competition rating is .01 which puts him below Tallinder as well as in the lower half of the defensemen we're looking at here.

So that shows us that Tallinder is playing against tougher players but it doesn't tell us how he's handling them. To judge that we move on to a stat called Goals Against On/60 which tallies how many goals the opposing teams scores while the player in question is on the ice. Just to be clear, the lower the number, the better.

Tallinder's GA ON/60 is 1.82 good for 10th best in the league out of the players who met our criteria. Above Tallinder are players such as Nick Lidstrom (really, the man is not human), Mathieu Schneider, Chris Pronger, and Paul Martin.

Campbell's GA ON/60 is 3.03, 10th worst out of the group we're looking at. That means that Campbell is facing a lower level of competition than Tallinder - considerably lower, I'd say - but allowing more goals per 60 minutes. Out of all the Sabres defensemen, Tallinder is facing the hardest competition while also allowing the fewest goals per 60 minutes of ice time.

Great, Heather, but what about the offense? Hold on, I'm getting there. The next stat we'll look at is the opposite - Goals For On/60 or how many goals the Sabres score while the defenseman in question is on the ice. In this case, the higher the number, the better.

Tallinder's GF ON/60 is 2.82 which is right in the middle of the pack. Not great, not terrible. It is important to note however that, as average as it is, it is a good bit higher than his GA ON/60. For every 60 minutes Tallinder spends on the ice, the Sabres score one more goal than they allow.

Campbell does have Tallinder beat in this category. His GF ON/60 is 3.17, putting him at 12th best among the defensemen we're looking at. (I don't have a list of the guys ahead of him but I'm sure Lidstrom is one of them. Seriously, he's a freak of nature.) So you could definitely make a case that Campbell does bring a lot to the table offensively. However his GF ON/60 is negated somewhat by his high GA ON/60. The Sabres score .35 more goals when Campbell is on the ice than when he's not but they also allow 1.21 goals more when he's on the ice.

One thing Meg pointed out that is in Campbell's favor is that he's lower than Tallinder in Quality of Teammates, meaning Tallinder plays with either a higher rated defensive partner or with higher rated forwards or some mixture of both. Tallinder's rating is a .10 while Campbell's is a -.09. So that could account for some of the difference in their overall numbers but their rankings are close enough here that I don't think it accounts for everything.

So let's summarize:

- Campbell averages about a minute and a half more in ice-time than Tallinder per 60 minutes but Tallinder is facing a much higher level of competition, almost always taking on the opponent's first line.

- Despite playing a higher level of competition, Tallinder is on the ice for fewer goals against than Campbell. For every 60 minutes Tallinder is on the ice, the Sabres score one more goal than they allow.

- Campbell is on the ice for more goals for the Sabres but also on the ice for more goals against the Sabres. In fact, his goals for number (3.17) is almost completely wiped out by his goals against number (3.03).

So while you could say that Campbell is more valuable than Tallinder offensively, you could also say that Tallinder's defensive ability is, in the long-run, more valuable to the Sabres because of how often Campbell's defensive lapses hurt his offensive contributions and again, this is despite Campbell playing against lesser skilled lines. I know I'm really pounding that but these are defensemen. I think it's important to take defense into consideration. And let's not forget that this is all based on a season that most everyone would agree has been pretty uneven for Tallinder.

In other words, I win! Brian Campbell is NOT worth 5 million dollars plus per season and the Sabres are NOT going to fall apart without him. I win! Tell all your friends!

(Thanks again, Meg! You're awesome!)

Hank Tallinder FTW!

22 comments:

Katebits said...

See?! I can totally stomach statbits when I know they are eventually going to prove a point that I hold near and dear to my heart.

Excellent work Heather and Meg! This totally justifies my rage against Soupy! Now, could you gals please get to work on proving (with statbits) that Goose is the handsomest Sabre?

Meg said...

(I don't have a list of the guys ahead of him but I'm sure Lidstrom is one of them. Seriously, he's a freak of nature.)

Lidstrom's #1 among defensemen. Of course he spends the bulk of his ice time playing with Zetterberg, Datsyuk, Holmstrom, and Rafalski, so that's not exactly hurting him. ;) Still, the fact remains that he is indeed a freak of nature.

Unknown said...

But, but, Bucky says Soupy is worth 5 mil a year, so it must be true. Logic and stats aside.

Seriously though, both you and Meg did an awesome job of compiling all this information and presenting a thoughtful analysis of it.

Just a quick thought, if you've proven that Soupy is not worth 5/25; do you think Hank would be, if his current deal was up at the end of this season?

Meg said...

The analysis is all Heather. So is the logical organization and presentation. I basically just vomited a bunch of lists and numbers at her.

See?! I can totally stomach statbits when I know they are eventually going to prove a point that I hold near and dear to my heart.

I HATE statbits when they don't show me what I want to see. The kind that provide affirmation make me so much happier.

Heather B. said...

Just a quick thought, if you've proven that Soupy is not worth 5/25; do you think Hank would be, if his current deal was up at the end of this season?

Well, first of all, I admit that we probably didn't technically prove that Soupy isn't worth five million so much as we proved - IMO - that he isn't the most valuable d-man on the team and isn't a player who's going to necessarily make a huge difference from night to night. There's too much bad along with his good. (I don't think he's worth five million but the market is the market. It's crazy.)

But to answer your question, considering that Hank consistently does a pretty good job of limiting top lines even in what has been a bit of an off season, yeah, I think you probably could argue that he's worth 5/25. And that's not even taking into consideration that he's a good penalty killer too. If I were a GM I would certainly rather pay a d-man like Tallinder than a d-man like Soupy. I wouldn't take Tallinder to elite status because he's not particularly gifted offensively but I think 5/25 right now is pretty reasonable for him.

So I guess that's a yes? Should I be forwarding this to Hank's agent? ;-)

Gambler said...

Great job Meg and Heather! How far back do the stats on BehindTheNet go? I'd be interested in taking a look at Hank's numbers during a season that hasn't been quite so inconsistent. Also, I'm tempted to soothe my Yo-Yo-less heartache by looking at some of his numbers over there. This post has just shown me that statbits can be sexy!

I have to say, though, you left out the compelling category of "Time on Ice Spent Looking Like A Drowned Rat" or TOISLLAD, in which Hank has a solid advantage. The importance of this quality cannot be overlooked.

And I love that picture of smiley Hank SO MUCH! Mr. Tall Indian is quite fetching!

Anonymous said...

Great job, Heather and Meg and Hank! Bad job, Soupy!

Seriously, I agree with Katebits. When statbits are presented this clearly and to prove a point I agree with, they're like man's best friend.

I cannot tell you how happy it makes me to see that Soupy's GF are washed out be his GA. Not that I'm happy you guys are stuck with a bad player, but happy that my instincts on him were right. I never have statbits to back up my opinions so just saying, "But Soupy doesn't seem good to me..." hasn't been a very compellin argument.

Also, go Paulie! Woo-hoo!

Gambler said...

"Time on Ice Spent Looking Like A Drowned Rat" or TOISLLAD

Um, pretend there's an R on the end there, kay? I know letters. Promise.

Gambler said...

Oh my God! I'm sorry for comment spamming, Heather, but I think I just made it sound like Hank looks like a drowned rat. That is not what I meant at all! Soupy looks like a drowned rat, and it's gross! Hank looks like a tall, smiling Indian. And it's wonderful. That's the quality that cannot be overlooked.

I'd like to publicly apologize for that trainwreck of a comment I just left. I was drunk, or something. I'm going to go away now and stop embarrassing Top Shelf with my presence.

Heather B. said...

I think I just made it sound like Hank looks like a drowned rat. That is not what I meant at all!

Gambler, you did make it sound kind of like that but I knew what you meant. And while I'm a firm believer in TOISLLADR, I thought maybe we should leave Soupy with a shred of dignity, you know? Hockey is one thing but there is no debate about who is more attractive here. Soupy gets smoked by the Tall Indian. (That, btw, will never not be funny.)

Meg said...

How far back do the stats on BehindTheNet go?

They only go back to last season, unfortunately.

Gambler said...

(That, btw, will never not be funny.)

I'm glad to hear that, because I don't think I will ever stop saying it!

Mark B said...

All those stats made my brain hurt, but I gotta say - Hank at the top of his game (a la 2005-2006) trumps Soupy's best performances over that same time. He's more valuable when it really counts on the back end.

Anonymous said...

First, this is one of the better Sabres blogs in the universe. Thank you for heartfelt and thoughtful posts (as opposed to so many sites with ill-informed and mean spirited claptrap).

Second, great analysis. It's kind of like Phil Housley v Mike Ramsey. If you had to choose between them back in the day, who would you pick?

It's a no-brainer.

Heather B. said...

Anonymous:

Thanks for your kind words! Part of my goal in starting this blog was to offer something that was a little more reasonable and positive than what was out there and I'm glad to hear I'm meeting that goal at least some of the time.

Also, thanks for agreeing with me. You get bonus points for that :D

Housley and Ramsey were before my time here in Buffalo but for the record, I just ran that question by my husband and he said, "I'd take Ramsey easy and Housley was my favorite player. You would've loved Mike Ramsey." So I guess I'm going Ramsey on that one!

Rich said...
This comment has been removed by the author.
Anonymous said...

btw, Behindthenet stats only go back to last year because they use the NHL's TOI pages to determine who was on the ice together at all times.

The NHL has only published these in processable text form since the middle of the 2005-06 season. Last year was the first full season for the NHL and the first year I put the stats on the site.

Anonymous said...

I believe...especially after HT's goal tonight in the shootout!

S.A.M. said...
This comment has been removed by the author.
S.A.M. said...

Wow, I'm just getting around to reading this one (It's been a busy few days at my house) but I love the way you and Meg put all that together! (with the stats from Behindthenet too) rock on!

And I'm soooooo happy to see my ranting at Soupy being justified! The more I see him, the more I hate him as a defenceman. He just tends to forget the defence in front of the title and yeah. BUt that goals for/against thing is the most telling stat of all, I think. And I've been looking for something like that for a while, i'm glad you were able to provide it!

Hank wins, hands down. Absolutely no question in my book.

Have the Sabres traded Soupy yet? Please? Because I am SO SICK of seeing him every night. grrrr.

ANyway well done, and feel free to bust out the stats anytime you wish. I loooooove hockey numbers.

Anonymous said...

1. Brian Campbell as a top pairing defenseman DOES play against the opposition's top line.
2. You're not taking into account Campbell's linemates when saying "Campbell gets more goals scored on him per 60 minutes than...," and that is omitting an important fact since the forwards he plays with (Derek Roy, Pominville, etc) are for the most part young and inexperienced, and they like to play a transition game, so it's not necessarily Campbell's fault that more goals get scored against his line than against Buffalo's checking line.
3. Campbell's presence has completely transformed the SJ Sharks' offense (for the better), and he has been pretty defensively responsible (for the most part). Although he's no Mike Green, and his shot isn't as good as Liddy's, Neidermayer's, or Pronger's, he can skate and pass as well as Liddy, Neidermayer, Pronger and Green (although Green's puck-handling skills surpass them all), and he is DEFINITELY worth five million a year to the Sharks. And he's WAY better than the nobody Dman you mentioned, despite your prestigious argument.

Heather B. said...

Anonymous, I can't speak for his time in San Jose but in Buffalo, Brian Campbell did NOT play against opposing top lines unless he got caught on the ice with them. That's a fact. In fact, when you compare Campbell's level of competition against to other Sabres it would appear that he was somewhat hidden from tough assignments and he still had worse defensive numbers than many of his teammates.

I believe I did mention the Sabres on the ice might be affecting Campbell's numbers some but I don't think it's enough to explain the difference, especially since the two players you mentioned - Roy and Pominville - are two of our best two-way forwards with Pominville in particular being extremely responsible defensively. In fact, Pominville does play on the checking line so I'm going to assume that maybe you're not as familiar with the Buffalo team Campbell was on this year as I am.

This post was obviously about his time in Buffalo. If Campbell is more successful in SJ, good for him. SJ seems to have more defense first d-men to pair him with which I think helps and has one of the best players in the league for him to play with on the PP. That was simply not the case in Buffalo and Campbell's weaknesses were more obvious. If you think he's worth five million to the Sharks, I'll take your word for it although I'd be shocked if he signs for that "little." I don't think he was worth that to the Sabres and I'm glad they passed on him to focus on Miller and Pominville.

And finally, I would never argue that Tallinder or any other Buffalo d-man is stronger offensively as an individual than Campbell. Clearly, Campbell is a very good skater and puck handler and a good passer and he was definitely the best on Buffalo's squad. But I did think it was worth pointing out that the team's offensive/defensive differentials were slightly better when Tallinder was on the ice because Campbell (and Spacek's) defensive play is often so sloppy. Them's the stats and as someone who's watched Campbell play for 82 games the last three or so years, I think they're pretty accurate.

Again, I'm not saying Campbell doesn't have some amazing strengths and I'm not denying that he may be a better fit with San Jose's personnel. I think that's quite possible. This post was simply written to show that he's not the all-around elite d-man that some in the league were saying he is, he's not (IMO) worth the money he's going to be commanding, and he's not as strong defensively as the nobody d-man who really is the top defensive d-man in Buffalo.

Thanks for commenting!