Welcome back, Hank. GVT style.

[IKnowPhysics]

It's been almost two months since the last thread, so let's try this again. I think (and hope) Hank Tallinder's going to help out quite a bit this season.

 

Not just because he's had good chemistry with guys on our roster, and we're hoping that it relights fires or positively influences them. But there's a case that he's now one of our better defenders. Warning: advanced stats involved.

 

Goals versus threshold (GVT) is a measure of how much an individual player contributes to his team's goal differential for a given season. It has limitations, like it doesn't measure talent (only the actual contributions to the goal differential, ie, a goalie that doesn't face any shots doesn't contribute much, regardless of talent), it doesn't measure intangibles, it doesn't factor in quality of team, and a few other limits. Players that play on teams with enormously different goal differentials are hard to compare. Players that have missed games from injury or benching, end up with a lower GVT because they didn't play. Again, simply put, GVT measures contribution to your team's goal differential.

 

For example, you can look at the Sabres with higher GVT, and things sort of make sense:

Miller 9.7

Vanek 8.7

Pominville (over the whole season) was 5.1

Hodsgon 4.5

Ott 4.3 (good defensive contributions here as a forward)

Ehrhoff 3.8

Flynn 3.3

Sekera 2.5

 

Specifically, you can line up last year's Sabre defenders like this:

Ehrhoff 3.8

Sekera 2.5

Weber 1.8

Sulzer 1.8

Pardy 1.5

Leopold 1.1

Pysyk 0.2 (in 19 games)

Myers -0.1

Ruwedhel -0.2 (in 7 games, not really enough stats here)

Regehr -0.3

 

New Jersey finished with a goal differential this season of -8, nearly identical to Buffalo's -7, so you can compare the GVTs of players from the two teams without too much trouble. NJ defenders:

Andy Greene 8.8 (enormous D contribution to goal differential, not necessarily a measure of talent)

Mark Fayne 4.3

Marek Zidlicky 4.2

Adam Larsson 3.7

Anton Volchenkov 3.3

Henrik Tallinder 2.0 (25 games)

Bryce Salvador 0.7

Peter Harrold 0.3 (23 games)

 

Henrik Tallinder on the NJ squad, the way he ended up being used in 25 games, was the sixth defenseman on a very defensively-oriented Devils roster. Even just playing those 25 games for the Sabres last season, he could've been the defender with the 3rd best contribution to the Sabres' goal differential. If he played the full 48 games, it's possible that he could have been our second best with a GVT of something close to 3.8, close to Ehrhoff's GVT (I'm not sure if that's cheating or not).

 

You can point at the way the Devils play, or you can point at our horrible defense last year, in these comparisons, etc. But we're getting a defenseman that should be able to effectively contribute to the team in exchange for the low low price of Riley Boychuk. And he's a familiar face. Welcome back, Hank.

[TrueBlueGED]

Not sure I agree with the interpretation here. Partially based on sample size, but more so based on roster. NJ had some good possession forwards who also play defense...the Sabres, not so much. I expect him to he asked (whether on purpose or not) to "do more" here, and I'm not sure I expect him to handle it well at this stage of his career.

 

That said, I think we should just start throwing advanced stats into every thread the way "Pegula bad" seems to infect every thread :P

[IKnowPhysics]

Not sure I agree ... more so based on roster. NJ had some good possession forwards who also play defense...the Sabres, not so much.

 

Understandable, but BEHOLD THE POWER OF STATS: :w00t:

 

Tallinder's RCQoT was 0.952, only really Bryce Salvador's was higher at 1.498, but that's middle of the road, sort of, for Sabres defensemen:

Sulzer 4.006 (what?)

Leopold 1.591

Myers 1.428

Pysyk 1.409

Ehrhoff 0.973

Weber 0.903

Pardy 0.501

Regehr 0.325

Ruwedhel -1.137 (ugh)

 

So he did play with NJ's high-posession guys, as much as any but Salvador, but in a team-wide context, you gotta remember that while NJ's top lines were good two-way possession lines, their bottom lines were total junk.

 

The real knock on Tallinder is his RCQoC: he was the lowest of the NJ defensemen at -0.509, getting the easiest minutes competition-wise, and nobody on the Sabres D corps had that easy of a time; Sulzer was closest at -0.213, everyone else was pretty much even or higher. So if he fails hard, stats point towards that it's because the competition is harder for him this time around, not that his new (old?) teammates suck a whole lot more.

[TrueBlueGED]

Understandable, but BEHOLD THE POWER OF STATS: :w00t:

 

Tallinder's RCQoT was 0.952, only really Bryce Salvador's was higher at 1.498, but that's middle of the road, sort of, for Sabres defensemen:

Sulzer 4.006 (what?)

Leopold 1.591

Myers 1.428

Pysyk 1.409

Ehrhoff 0.973

Weber 0.903

Pardy 0.501

Regehr 0.325

Ruwedhel -1.137 (ugh)

 

So he did play with NJ's high-posession guys, as much as any but Salvador, but in a team-wide context, you gotta remember that while NJ's top lines were good two-way possession lines, their bottom lines were total junk.

 

The real knock on Tallinder is his RCQoC: he was the lowest of the NJ defensemen at -0.509, getting the easiest minutes competition-wise, and nobody on the Sabres D corps had that easy of a time; Sulzer was closest at -0.213, everyone else was pretty much even or higher. So if he fails hard, stats point towards that it's because the competition is harder for him this time around, not that his new (old?) teammates suck a whole lot more.

 

Do you even watch the games? ;)

[IKnowPhysics]

Haha, I do.

 

In reality, if Myers squares his ###### away, like I hear he's trying hard to do this offseason, Tallinder could be a three-four guy for us with Ehrhoff, Myers, and Pysyk, and, with any luck, McBain and Weber floating around for us.

[Ghost of Dwight Drane]

Let's do this Madlibs style....

 

Select my response:

 

1) I would have figured Tallinder's differential had been 3.0 to 1

 

2) This isn't the first time the words spread and sheets have been used in describing Tallinder's attributes

 

3) I was assuming GVT Style stood for Gynecological Violating Terrorist

 

I'm not married to any of them. Sorry, this is your baby. Don't break my arm...multiple times....

 

OK. I'm done.

[IKnowPhysics]

Just like the 3rd jersey, the Sabres will unveil their latest acquisition on twitter:

 

[...]

 

 

Do you even watch the games? ;)

 

What's the point? Number crunching is obviously way more interesting. Also, if you WATCH the game, you get the completely wrong idea. It's only when you pour over the stats do you have any idea what's going on in the game.

 

 

[rakish]

What's the point? Number crunching is obviously way more interesting. Also, if you WATCH the game, you get the completely wrong idea. It's only when you pour over the stats do you have any idea what's going on in the game.

 

QFT

[Andrew Amerk]

Craig Rivet.

[TrueBlueGED]

What's the point? Number crunching is obviously way more interesting. Also, if you WATCH the game, you get the completely wrong idea. It's only when you pour over the stats do you have any idea what's going on in the game.

 

You can. Unless you've invented a way to remove your bias when watching a game and have photographic memory of hundreds of games and every moment within them, in which case, you should probably patent that and get rich.

 

(Yes, I have willfully broken my sarcasm detector).

[...]

 

 

You can. Unless you've invented a way to remove your bias when watching a game and have photographic memory of hundreds of games and every moment within them, in which case, you should probably patent that and get rich.

 

(Yes, I have willfully broken my sarcasm detector).

 

Why? Why watch a game of hockey, then? Or a movie?

[TrueBlueGED]

Why? Why watch a game of hockey, then? Or a movie?

 

We may be talking past each other. I'm not talking about no longer being a fan of the team. But when you're judging player value, especially discussing it on a message board, why wouldn't you want to be as objective as possible and use the tools at your disposal?

[nfreeman]

We may be talking past each other. I'm not talking about no longer being a fan of the team. But when you're judging player value, especially discussing it on a message board, why wouldn't you want to be as objective as possible and use the tools at your disposal?

 

You're kinda testy lately.

[...]

 

 

We may be talking past each other. I'm not talking about no longer being a fan of the team. But when you're judging player value, especially discussing it on a message board, why wouldn't you want to be as objective as possible and use the tools at your disposal?

 

I'm just taking your broken sarcasm meter and running with it. I understand stats and their/its use(s), and appreciate the usefulness of them/they/it. It is delicious, though, that, according to stats-wielders, watching/experiencing the game in real time leads to wrong/inaccurate impressions/ideas/conclusions about the game.

[Weave]

You're kinda testy lately.

 

Another Angry. We have AngryTrueBluePhD to go with AngryEleven.

[TrueBlueGED]

You're kinda testy lately.

 

Football is starting, what Bills fan isn't? :P

 

(I know, a lot of stuff going on, will try to soften the tone a bit)

 

I'm just taking your broken sarcasm meter and running with it. I understand stats and their/its use(s), and appreciate the usefulness of them/they/it. It is delicious, though, that, according to stats-wielders, watching/experiencing the game in real time leads to wrong/inaccurate impressions/ideas/conclusions about the game.

 

Are you completely excluding this possibility? In the GDTs people will jump on their favorite whipping boy when something goes wrong (Myers, for instance) only to have another poster evaluate the play a little more deeply and maybe see that a forward lost his man, not Myers. Or when Leopold got turned into a pylon and Miller let in a weak goal--the Miller haters piled on, while others pointed to Leopold's failures. I know you're in the GDTs enough to see stuff like this happen.

 

Besides, the eye misses things and the mind forgets (or misremembers, as Roger Clemens would say) things. Stats can help fill in some of these gaps. Obviously stats aren't perfect and you can't learn everything from a spreadsheet, as visuals help fill in gaps that stats can't capture (for example, a 10 foot shot is considered to be pretty high quality...but a 10 foot shot from out wide isn't as good as from the slot, and there's no state for where that shot came from on the ice). The stats guys, at least the ones on this board, tend to be pretty reasonable in accepting stats don't tell you the whole story...but the "eye test" crew seems to be highly insulted or blow off a stats-related counterpoint whenever one of us points out that his/her interpretation may not be accurate because of the stats.

[...]

Of course I am not "completely excluding this possibility". Sometimes phenomenon are observed first, and then verified with numbers/science. Sometimes numbers/science reveal phenomenon missed by observation.

 

The stats guys, at least the ones on this board, tend to be pretty reasonable in accepting stats don't tell you the whole story...

 

This, however, I can not agree with. I think the divide on HOW to "fix" the Sabres is even greater now thanks to the number crunchers than it was when I first started posting here. Perhaps it's the nature of the number crunchers, but they sometimes take things way too literally.

 

I submit the thread on where deluca and I (and others, fer shure) were trying to argue that the Sabres need bigger and tougher players with more "heart". You, with all due respect and no intent to disparage, were in that thread taking what we were saying and simply not getting the gist.

 

If anything, I think there is a decently balanced curve ranging from stats <--> observers, with most folk in the middle somewhere (like me).

 

Yes, some observers get upset when faced with stats, but the opposite is true as well.

 

In the middle of the action, though, while watching, or playing, for me, no way do I care what the stats are saying. I play sports, have played a little hockey, and it is, ultimately, a game of feeling, instincts, and cunning. We have seen too many times what happens when a player can not absorb the information they are presented and are able to react to it instinctively - their thoughts bog them down, like having a wet blanket tightly wrapped around them on the ice.

 

It's fun and entertaining...number crunching to me is not fun and entertaining, as useful as it is. Now, I can completely and utterly dig that there are folks who get off on number crunching and doing so IS part of the entertainment factor of following sports.

 

Personally, I get off on observing the performance in the moment. It can be torn apart later with whatever tools/methods seem prudent.

 

 

 

 

[Sabres Fan in NS]

You're kinda testy lately.

 

I first read this as tasty ... :wub:

[bunomatic]

Problem is that the people at the game that are paid to compile the stats have biases and probably misinterpret the play or actually miss things as well. Thats hit # 12 for Ott and oh lawdy momma lookie that hot dress walking up the aisle in section...

[IKnowPhysics]

Problem is that the people at the game that are paid to compile the stats have biases and probably misinterpret the play or actually miss things as well. Thats hit # 12 for Ott and oh lawdy momma lookie that hot dress walking up the aisle in section...

 

It's true, "real time" stats like hits and takeaways varying slightly in their methods from arena to arena. With a little analysis, it's possible to see which ones. It's also possible to see which stats don't have that much bias in them, like shots. Some don't have any bias, like goals, shots, minutes of ice time, zone starts, etc.

[TrueBlueGED]

It's true, "real time" stats like hits and takeaways varying slightly in their methods from arena to arena. With a little analysis, it's possible to see which ones. It's also possible to see which stats don't have that much bias in them, like shots. Some don't have any bias, like goals, shots, minutes of ice time, zone starts, etc.

 

Depending on what we're talking about, some of the bias will also wash out with a large enough sample. If you're only looking at the Rangers, you might be led to believe hits correlates to possession, but when looking at all the teams in the league that Rangers home bias becomes irrelevant. Also why home/away splits are great.

[bunomatic]

It's true, "real time" stats like hits and takeaways varying slightly in their methods from arena to arena. With a little analysis, it's possible to see which ones. It's also possible to see which stats don't have that much bias in them, like shots. Some don't have any bias, like goals, shots, minutes of ice time, zone starts, etc.

 

Shots is a funny one. I remember watching Ray Bourque and Boston and Bourque would fire the puck in from centre ice at the goalie probably 10 times a game and that would count as a shot on goal but wasn't really a scoring chance. If you're at the game you know that and understand the situation that led to that lopsided shot count. Someone reads the paper the following day and sees shots listed as 42-28 for Boston and assumes the game was maybe dominated a little by Boston. Another that is misleading is second assist. I think many times teams are guilty of giving second assists too freely.

[Weave]

Shots is a funny one. I remember watching Ray Bourque and Boston and Bourque would fire the puck in from centre ice at the goalie probably 10 times a game and that would count as a shot on goal but wasn't really a scoring chance. If you're at the game you know that and understand the situation that led to that lopsided shot count. Someone reads the paper the following day and sees shots listed as 42-28 for Boston and assumes the game was maybe dominated a little by Boston. Another that is misleading is second assist. I think many times teams are guilty of giving second assists too freely.

 

Whoever used to count shots in the Aud used to have a reputation as being very "liberal" with what was counted as a shot for Buffalo. It was well known enough that it would sometimes show up in commentary in places like The Hockey News.

[TrueBlueGED]

Whoever used to count shots in the Aud used to have a reputation as being very "liberal" with what was counted as a shot for Buffalo. It was well known enough that it would sometimes show up in commentary in places like The Hockey News.

 

On the flip side, MSG is currently known for over counting shots against to pump up Lundqvist's SV%.