Locker Clean Out Day

[Bartho]

What we're discussing is Pavelec's performance on whether or not the Jets made the playoffs despite him. Not on whether or not Pavelec is a good goalie. I'm not trying to convince you on that, although we seemingly agree on that discussion point if we were to make it.

Descriptive vs predictive.

My point is that while Pavelec stopped 92% of shots rather than his norm 90.6%, it was grouped in a manner that was least productive towards helping the Jets win%... especially given the Jets were an above average outshooting team with average shooting percentage.

Pavelec did stop a higher percentage of shots than the average goalie did.
This fact does not change if you adjust for shot location.
If he were someone who maintained this for long periods of time (which he failed to do), I would also conclude he is likely to continue stopping pucks.
He did however perform below average for most of his season, and only performed well near the end.
This is why I say despite Pavelec, even though Pavelec did perform exceptionally well for a period.

This is pretty much the only reason/way I'd use QS%.

Ok. So when the better stat makes Pavelec look good and the inferior stat makes him look mediocre, we go with the inferior stat. I get it now.

 

[garret9]

Ok. So when the better stat makes Pavelec look good and the inferior stat makes him look mediocre, we go with the inferior stat. I get it now.

Haha funny but no.

One is performance in how many shots a player saved. This stat (and its variants) is the best stay in showing true talent level in large samples (which Pavs has never done well in)

The other is distribution into performance showing how often the goaltender gave the team a level in which a league average team performance would garner a win (which Pavs has never done well in).

If you read it fully, it points out how Pavelec performed well in stopping the puck, but it his excellency was predominately confined to a very small part of his season. This is something everyone knew already, just the numbers show by how much.

However, neither stay paints Pavelec as good when you deal with acceptable sample sizes to use performance as a measure of talent...

 

[garret9]

Right.



I think glossing over the sv% is incorrect. During that shut out streak across 3 games, the Jets only scored 4 goals. So they didn't just need a quality start in those games. They needed an exceptional goaltending effort. Which is what they got. Had Pavs been "average" in those games, we might have received 0 points (or at least certainly far fewer) from them.

Was Pavs inconsistent between great and bad across the season? Yes. But so we're the Jets.

So, regarding Pavs' inconsistency, you could just as easily paint it the opposite way. Which is that Pavs was exceptional when he needed to be. There were of course other games the Jets won where Pavs was bad. But Pavs didn't need to be better in those games. We won anyway.

So things get pretty mushy when you argue QS over sv% for a specific set of games. Maybe if Pavs had been more consistent (but with the same sv%), we could have actuaĺly had a worse record last year. And missed the playoffs.

Whereas sv% gives us an inarguable fact. Pavs stopped more pucks than the average goaltender. Nothing mushy about that.

It is very difficult to argue that a different distribution of his sv% (which again, was good) across the season would have resulted in more or less points. For the specific season in question.

It's really simple:

1) Pavelec played amazing for a short period of time pushing us into he POs.
2) Pavelec was the largest reason why the Jets record wasn't better at that point (over much larger numbers of games) and needed the hot streak in the first place.
3) Had we had meh instead of terrible goaltending tinge Jets could have dropped the ball on those games and been fine.
4) The Jets made the playoffs despite Pavelec.

 

[raideralex99]

Pavelec is not a top 10 goalie end of story.
You can't use one good year or a hot streak as an indication ... maybe the shooters were having an off game.
Throw away Pavs best and worst seasons and then use the rest of his stats and that's his true rating.

That win against the Wild at the end of the season ... Pavs was terrible but had good stats. He was out of position so many times ... the Wild players missed the net or could not get their stick on the puck. So someone who did not see the game looks at the stats and It shows Pavs played great but it was really the Wild shooters having an off game.

 

[SensibleGuy]

Pavelec is not a top 10 goalie end of story.
You can't use one good year or a hot streak as an indication ... maybe the shooters were having an off game.
Throw away Pavs best and worst seasons and then use the rest of his stats and that's his true rating.

That win against the Wild at the end of the season ... Pavs was terrible but had good stats. He was out of position so many times ... the Wild players missed the net or could not get their stick on the puck. So someone who did not see the game looks at the stats and It shows Pavs played great but it was really the Wild shooters having an off game.

Pav is not a top 30 goalie.

 

[SensibleGuy]

So things get pretty mushy when you argue QS over sv% for a specific set of games. Maybe if Pavs had been more consistent (but with the same sv%), we could have actuaĺly had a worse record last year. And missed the playoffs.
.

By what reasonable means could you possibly make the argument that had Pavs given us .920 sv% consistently though out the season we could have been worse off? I mean everything else being the same, how does stopping more pucks leave us in a worse situation?

 

[Whileee]

It's really simple:

1) Pavelec played amazing for a short period of time pushing us into he POs.
2) Pavelec was the largest reason why the Jets record wasn't better at that point (over much larger numbers of games) and needed the hot streak in the first place.
3) Had we had meh instead of terrible goaltending tinge Jets could have dropped the ball on those games and been fine.
4) The Jets made the playoffs despite Pavelec.

QS is still a pretty crude statistic though, isn't it?

Wouldn't you want to adjust for scoring chances and/or shots 4v5? Jets probably led the lead in games with 5+ PP against. That's going to mess with the ability to have a quality start (2 goals against is almost always a quality start, 3 goals against is almost never a quality start).

It would actually be pretty easy to develop more sophisticated models, I would think. Dichotomizing individual game performances based on crude cutoffs like the QS without further adjustment doesn't seem like "advanced" stats to me.

Lest anyone get all frothy about this discussion, I still think Pavs is a crappy goalie. This is more of a conceptual discussion for me at this point.

 

[SensibleGuy]

QS is still a pretty crude statistic though, isn't it?

Wouldn't you want to adjust for scoring chances and/or shots 4v5? Jets probably led the lead in games with 5+ PP against. That's going to mess with the ability to have a quality start (2 goals against is almost always a quality start, 3 goals against is almost never a quality start).

It would actually be pretty easy to develop more sophisticated models, I would think. Dichotomizing individual game performances based on crude cutoffs like the QS without further adjustment doesn't seem like "advanced" stats to me...

I'm gonna go ahead and bet that the advanced stats guys have already thought that through...

 

[garret9]

QS is still a pretty crude statistic though, isn't it?

Wouldn't you want to adjust for scoring chances and/or shots 4v5? Jets probably led the lead in games with 5+ PP against. That's going to mess with the ability to have a quality start (2 goals against is almost always a quality start, 3 goals against is almost never a quality start).

It would actually be pretty easy to develop more sophisticated models, I would think. Dichotomizing individual game performances based on crude cutoffs like the QS without further adjustment doesn't seem like "advanced" stats to me.

Lest anyone get all frothy about this discussion, I still think Pavs is a crappy goalie. This is more of a conceptual discussion for me at this point.

We're talking about a team that also scored an above average numbers of goals for...

There are better models out there, there is AAA% and WT% which is being developed by Nick Mercadante of Hockey Graphs. He presented these at VanHAC.

I did mention these when I first started talking this point.

 

[Whileee]

It's really simple:

1) Pavelec played amazing for a short period of time pushing us into he POs.
2) Pavelec was the largest reason why the Jets record wasn't better at that point (over much larger numbers of games) and needed the hot streak in the first place.
3) Had we had meh instead of terrible goaltending tinge Jets could have dropped the ball on those games and been fine.
4) The Jets made the playoffs despite Pavelec.

I think that's too simple, G.

Games are discrete events, but QS% is a continuous variable. To make the claim that a team won "despite" a low percentage of quality starts, you need to analyze the game-by-game results as discrete events.

1) To succeed "despite" low quality starts, a team would win a higher than expected number of games where the goalie failed to provide a quality start, or have another goalie carry them.

2) Jets record in Pavs' "non-quality" starts was 4-14-5, which is certainly not above the expected during non-quality starts. So the Jets weren't winning very often "despite" Pavs.

3) Hutch only had 52.8% QS, so basically 2 more QS than Pavs.

4) The Jets actually won more than expected during Hutch's non-quality starts (record was 6-10-2), so if anything, the Jets were more likely to win despite a poor start by Hutch.

5) The Jets scored 4 or more goals 7 times during Hutch's 18 "non-quality" starts (39%). They scored 4 or more goals in 4 of Pavs' 23 non-quality starts (17%). This contributes to the Jets being more likely to win "despite" a non-quality start by Hutch.

Seems that it might be more appropriate to say that the Jets won "despite" Hutch.

What the Jets did remarkably well last year was win when they got a quality start. I think this is a combination of two factors: 1) their stellar 5v5 play meant that a quality start was usually converted into a win; 2) a quality start is often associated with not having PP goals against, and when the Jets stayed out of the box they won often. Conversely, the Jets penchant for taking penalties probably contributed to some "non quality" starts.

Going forward, I think we can be encouraged that if the Jets get a high number of quality starts and fewer penalties, they will be a very competitive team.

 

[Whileee]

We're talking about a team that also scored an above average numbers of goals for...

There are better models out there, there is AAA% and WT% which is being developed by Nick Mercadante of Hockey Graphs. He presented these at VanHAC.

I did mention these when I first started talking this point.

See above... seems that they scored more when Hutch was in net than when Pavs was, especially during non-quality starts.

 

[garret9]

You'll have to take it one step further, as sh% and shot differentials are impacted by goal states which would differ depending on goaltender performance.

But, as I said, more sophisticated models, like WinThreshold% and AboveAverageApperence% also suggest the same thing.

 

[Whileee]

You'll have to take it one step further, as sh% and shot differentials are impacted by goal states which would differ depending on goaltender performance.

But, as I said, more sophisticated models, like WinThreshold% and AboveAverageApperence% also suggest the same thing.

Well, Pavs is crappy, so I guess any way you look at it will conclude the same thing.

Hutch seems pretty mediocre too, which is a disappointment.

Bottom line is that if the Jets get average to above average goaltending, and reduce their PK time, I think they'll be a playoff contender.

 

[garret9]

Well, Pavs is crappy, so I guess any way you look at it will conclude the same thing.

Hutch seems pretty mediocre too, which is a disappointment.

Bottom line is that if the Jets get average to above average goaltending, and reduce their PK time, I think they'll be a playoff contender.

Ya... Hutch has been pretty disappointing.

Hope he bounces back, but not going to place any bets on it.

Had a good talk with a SJS scout at VanHAC. He was talking about how Hutchinson confuses him, being amazing one year and terrible the next, not even just at NHL level (stats wise though this was Hutch's only bad year).

 

[raideralex99]

Ya... Hutch has been pretty disappointing.

Hope he bounces back, but not going to place any bets on it.

Had a good talk with a SJS scout at VanHAC. He was talking about how Hutchinson confuses him, being amazing one year and terrible the next, not even just at NHL level (stats wise though this was Hutch's only bad year).

Either he had injuries, family problems or poor coaching. I think he needs a good coach.

 

[Whileee]

Either he had injuries, family problems or poor coaching. I think he needs a good coach.

Not sure what he needs, but I kept on feeling that he had the "yips", from the latter part of last season and through most of this season.

 

[Puckatron 3000]

By what reasonable means could you possibly make the argument that had Pavs given us .920 sv% consistently though out the season we could have been worse off? I mean everything else being the same, how does stopping more pucks leave us in a worse situation?

.920 is .920, no matter how you break it up. It's the exact same number of stopping pucks.

 

[SensibleGuy]

.920 is .920, no matter how you break it up. It's the exact same number of stopping pucks.

Yeah except that Pav got himself to 920 by posting 3 shutouts at the end of the season. His short hot streak brought his average up nicely. If he had a .920 sv% going into the last 8 or so games we probably would have been able to rest some bodies because we'd have sealed our PO spot early...which might have helped us to win a game instead of getting swept!

 

[Andy6]

Yeah except that Pav got himself to 920 by posting 3 shutouts at the end of the season. His short hot streak brought his average up nicely. If he had a .920 sv% going into the last 8 or so games we probably would have been able to rest some bodies because we'd have sealed our PO spot early...which might have helped us to win a game instead of getting swept!

That has to be the most contrived anti-Pavs argument ever! What a sneaky devil he was for using shutouts to artificially boost his save percentage!

 

[Whileee]

That has to be the most contrived anti-Pavs argument ever! What a sneaky devil he was for using shutouts to artificially boost his save percentage!

Jets scored a total of 3 goals in his last 3 shutout wins, but those wins shouldn't count as much as earlier wins.

 

[JetsFan815]

Really good Chevy interview http://www.tsn.ca/radio/winnipeg-12...ing-on-building-a-team-relationships-1.471354

 

[Evil Little]

Really good Chevy interview http://www.tsn.ca/radio/winnipeg-12...ing-on-building-a-team-relationships-1.471354

Let me guess:

"process"
boathouse reno talk
getting really mad when they talk about Pavelec

 

[SensibleGuy]

Jets scored a total of 3 goals in his last 3 shutout wins, but those wins shouldn't count as much as earlier wins.

they shouldn't count MORE than earlier wins. And Pav's hot streak at the end shouldn't count more than Hutch's work in the first half. That's all. They were wins. We needed them. But I'd trade those three shutouts at the end for 6 or 8 .935 games earlier in the year...

 

[SensibleGuy]

That has to be the most contrived anti-Pavs argument ever! What a sneaky devil he was for using shutouts to artificially boost his save percentage!

I'll tell you what's contrived - using Pav's hot streak last year as an excuse to absolve him of his awful performance for the other 95% of his career as a pro...

 

[YWGinYYZ]

I'll tell you what's contrived - using Pav's hot streak last year as an excuse to absolve him of his awful performance for the other 95% of his career as a pro...

Who's absolving him of his awful performances? The media?