All Purpose Analytics and Extended Stats Discussion

[Calicaps]

From a fancy stats standpoint in a vacuum it seems to be a wash, or close to it. Some models prefer Mantha more, some Vrana. But they're close. What is pretty universal between the models is that Mantha provides quite a bit more defensively than Vrana.

This will make Washington better this year, if for no other reason than I have to imagine Laviolette is going to trust Mantha more than Vrana. For argument's sake if you call the two players a wash talent-wise, then Mantha likely getting somewhere in the neighborhood of 16-18 minutes a night compared to Vrana likely only getting 12-14 minutes a night seems like a lot of additional value added for Washington. Is that value worth an additional first round pick? I don't know, probably? Could Vrana have been utilized better? I don't know, probably?

This seems like one of those scenarios where it probably just comes down to fit, and if Washington thinks Mantha is a better fit then I understand the trade.

Do these models account for quality of linemates and competition? In Det, Mantha had mostly so so companions and faced the top D every night. In DC, he'll likely play with 19 and face 2nd pair D. Is that factored into all of this?

 

[twabby]

Do these models account for quality of linemates and competition? In Det, Mantha had mostly so so companions and faced the top D every night. In DC, he'll likely play with 19 and face 2nd pair D. Is that factored into all of this?

Yes I think all of these models account for context such as competition, teammates, zone starts, score effects, etc.

 

[kicksavedave]

Yes I think all of these models account for context such as competition, teammates, zone starts, score effects, etc.

I was thinking about this but didn't know exactly how to quantify it in stats, but I thought you might. I would expect the Caps to have the puck a lot more than the Wings would, juts because of all around talent. So I would then expect that, if playing time were equal and deployments remained roughly the same as before, Mantha should see more time with the puck in our hands with the Caps vs what he saw with the Wings, and Vrana should see less time with his team holding the puck and more time on D, in Detroit vs what he's used to. Therefore we should see an uptick in Mantha's offense and downgrade to Vrana's points.

Not sure if this is true but wondering if there is a fancy that might show this. Have to think playing with Backi and Oshie would power Mantha to deliver more than he was in Detroit.

Also, Detroit may just be forced to play Vrana more than he was getting here, so that could blow this up anyway.

 

[g00n]

Yes I think all of these models account for context such as competition, teammates, zone starts, score effects, etc.

But they don't account for how shitty your team is and whether or not you're happy where you are. That's the old "change of scenery" factor.

To me it's impressive that Mantha's defensive metrics are apparently pretty good even when playing for a perpetually moribund team. That requires some effort. If you're just an offensive talent and you're on a shit squad the tendency is to ignore the coach and look for ways to cherrypick and boost your personal stats. Maybe AM gave in to this temptation at times and that's the "inconsistency" DET fans talk about.

But what's Vrana's excuse been? He seems to be a one-dimensional forward in every sense of the word, and based on the complaints from 5 consecutive coaches he has work ethic and attitude problems.

If not for the Cup win every fan here would probably be applauding the talent swap, draft pick price notwithstanding.

 

[twabby]

I was thinking about this but didn't know exactly how to quantify it in stats, but I thought you might. I would expect the Caps to have the puck a lot more than the Wings would, juts because of all around talent. So I would then expect that, if playing time were equal and deployments remained roughly the same as before, Mantha should see more time with the puck in our hands with the Caps vs what he saw with the Wings, and Vrana should see less time with his team holding the puck and more time on D, in Detroit vs what he's used to. Therefore we should see an uptick in Mantha's offense and downgrade to Vrana's points.

Not sure if this is true but wondering if there is a fancy that might show this. Have to think playing with Backi and Oshie would power Mantha to deliver more than he was in Detroit.

Also, Detroit may just be forced to play Vrana more than he was getting here, so that could blow this up anyway.

I don't know of one specific measure to quantify what you're asking for, but I'd guess Mantha's rate stats would be better in DC than in Detroit mainly because of the talent disparity, while Vrana's would drop off for the same reason. But I'd of course caution and say this doesn't mean Vrana is worse, or Manthat is better. Points are largely fleeting except for all-star level players.

Micah's model seems to think that the two things that impact a player's results most are individual talent level and a player's teammates.

Model: Magnus 4 (EV)

I don't actually think Backstrom or Kuznetsov are better than Larkin at this point, but I do think Washington's D is noticeably better than Detroit's D so I imagine Mantha's results will be much better in terms of goal differential. Whether it translates to points? I don't know, I would guess a lot of the offense would flow through Mantha so he'll pick up his share of points but I'm really more interested in the overall goal differential impact he'll have compared to Vrana's impact during his time in DC.

 

[twabby]

But they don't account for how shitty your team is and whether or not you're happy where you are. That's the old "change of scenery" factor.

To me it's impressive that Mantha's defensive metrics are apparently pretty good even when playing for a perpetually moribund team. That requires some effort. If you're just an offensive talent and you're on a shit squad the tendency is to ignore the coach and look for ways to cherrypick and boost your personal stats. Maybe AM gave in to this temptation at times and that's the "inconsistency" DET fans talk about.

But what's Vrana's excuse been? He seems to be a one-dimensional forward in every sense of the word, and based on the complaints from 5 consecutive coaches he has work ethic and attitude problems.

If not for the Cup win every fan here would probably be applauding the talent swap, draft pick price notwithstanding.

They account for how shitty your team is, but you're right models don't capture why players are performing the way they are outside of purely on-ice results. So yeah if Mantha is just unhappy in Detroit, it's reasonable to expect he'd be better in a place like Washington where they expect to contend.

On the other hand maybe Vrana will perform better in Detroit than he has in DC. He's clearly not the type of player Laviolette trusted, and I imagine that weighed on him quite a bit. Given that Detroit isn't contending I would expect Vrana to get a ton of ice time for Detroit team to evaluate what they have going forward before they negotiate his next contract. Maybe being a trusted top-line LW will help Vrana become a better overall player. I don't expect him to become some defensive stud, but I could see a more complete game developing.

 

[tenken00]

They account for how shitty your team is, but you're right models don't capture why players are performing the way they are outside of purely on-ice results. So yeah if Mantha is just unhappy in Detroit, it's reasonable to expect he'd be better in a place like Washington where they expect to contend.

On the other hand maybe Vrana will perform better in Detroit than he has in DC. He's clearly not the type of player Laviolette trusted, and I imagine that weighed on him quite a bit. Given that Detroit isn't contending I would expect Vrana to get a ton of ice time for Detroit team to evaluate what they have going forward before they negotiate his next contract. Maybe being a trusted top-line LW will help Vrana become a better overall player. I don't expect him to become some defensive stud, but I could see a more complete game developing.

That would be the best outcome. A win win for both teams.

 

[kicksavedave]

That would be the best outcome. A win win for both teams.

That's not how the world works. We need a winner and a loser to feel right.

 

[tenken00]

That's not how the world works. We need a winner and a loser to feel right.

Me good. You bad.

 

[g00n]

They account for how shitty your team is, but you're right models don't capture why players are performing the way they are outside of purely on-ice results. So yeah if Mantha is just unhappy in Detroit, it's reasonable to expect he'd be better in a place like Washington where they expect to contend.

On the other hand maybe Vrana will perform better in Detroit than he has in DC. He's clearly not the type of player Laviolette trusted, and I imagine that weighed on him quite a bit. Given that Detroit isn't contending I would expect Vrana to get a ton of ice time for Detroit team to evaluate what they have going forward before they negotiate his next contract. Maybe being a trusted top-line LW will help Vrana become a better overall player. I don't expect him to become some defensive stud, but I could see a more complete game developing.

Yes I've been wondering about Vrana in Detroit. With less pressure maybe he'll get away with some things that got him benched here, and find ways to really pad the stats. Hard to say. I don't know what kind of system they're running there.

 

[g00n]

Reading a post in another thread, I think Abro mentioned our goaltending is worst in the East.

This has me wondering if it's a team or player problem. We've looked in depth at the skaters but have we looked at VV and Sammy under the analytics microscope?

Specifically, I'd start with high-danger chances and breakaways vs the rest of the Conference or League. IMO you have to figure out if we're giving up too many goals due to defensive breakdowns (and the expected higher shooting % that goes along with breakaways, for example), or if the netminding is responsible.

Any stat-heads have a take?

 

[traparatus]

Reading a post in another thread, I think Abro mentioned our goaltending is worst in the East.

This has me wondering if it's a team or player problem. We've looked in depth at the skaters but have we looked at VV and Sammy under the analytics microscope?

Specifically, I'd start with high-danger chances and breakaways vs the rest of the Conference or League. IMO you have to figure out if we're giving up too many goals due to defensive breakdowns (and the expected higher shooting % that goes along with breakaways, for example), or if the netminding is responsible.

Any stat-heads have a take?

I was looking into this a bit on NaturalStatTrick.

Data seems to point to Caps not getting the goaltending performance. Caps rank 8th in High-Danger Chances Against at 9.26HDCA/60. That's good enough for 3rd in the division. It compares reasonably well to Islanders(8.53), Bruins(9.2), Pens(10.16).

Caps High Danger Goals Against tells a different story. At 1.4 HDGA/60, Caps are 16th in the league, though still 3rd in the division. Compare to Islanders at 1.08, Bruins at 1.27, Pens at 1.47.

Looking at GSAA numbers from the same source, results are not surprising. Among goaltenders who have played minimum of 600min this season, Vanecek ranks 27th(1.0 GSAA) out of 55. Samsonov is 45th(-4.99 GSAA). For comparision Varlamov is 5th(9.13 GSAA), Jarry is 18th (3.7 GSAA), Rask is 24th(1.99 GSAA).

 

[twabby]

Not that the numbers would necessarily paint a different story, but I’d rather look at GSAx rather than GSAA. Goals saved above expected (GSAx) takes into account the quality of shot the goaltender is facing, while goals saved above average (GSAA) does not take into account shot quality against.

 

[kicksavedave]

I was looking into this a bit on NaturalStatTrick.

Data seems to point to Caps not getting the goaltending performance. Caps rank 8th in High-Danger Chances Against at 9.26HDCA/60. That's good enough for 3rd in the division. It compares reasonably well to Islanders(8.53), Bruins(9.2), Pens(10.16).

Caps High Danger Goals Against tells a different story. At 1.4 HDGA/60, Caps are 16th in the league, though still 3rd in the division. Compare to Islanders at 1.08, Bruins at 1.27, Pens at 1.47.

Looking at GSAA numbers from the same source, results are not surprising. Among goaltenders who have played minimum of 600min this season, Vanecek ranks 27th(1.0 GSAA) out of 55. Samsonov is 45th(-4.99 GSAA). For comparision Varlamov is 5th(9.13 GSAA), Jarry is 18th (3.7 GSAA), Rask is 24th(1.99 GSAA).

Those GSAA numbers are fugly.

I've said in the past that goaltending is about two things: Not letting in the softies (making all the saves you are supposed to make), and making the occasional spectacular save at the critical times (the ones where Doc says "Oh and it didn't go").

Its not so important that a goalie makes a lot of spectacular saves vs high danger shots but also lets in three softies, you're still losing that game no matter what the defense does. On the other hand, if a goalie stops all the ones he should stop, then its on the defense to stop the impossible shots from happening. In the playoffs people focus on a goalie "stealing a game" or a series" but thats rarely the whole story, its often the case that the goalie makes all the routine stops and the defense plays solid. Tampa Games 6 and 7 were perfect examples - Shutouts, both of them, but Holtby was rarely tested in those two games.

So my point is, the goalie needs to stop the softies and let the defense do its job, and they will win most of their games. The Caps goalies have been inconsistent in this aspect but have shown the ability to do this. If the Caps D leaves the goalies out to dry, we won't win shit. If they play their game right, then its just up to the goalies to not be butter soft. I hope the are up to that task!

 

[g00n]

I think one factor is both goaltenders are young butterfly style (not uncommon) and they seem to be getting beat up high from fairly close range a lot.

"He goes down too early" is a criticism I've leveled as have some of the talking heads during the broadcasts, but maybe there's more to it.

If there isn't a basic mechanics or "skill" issue there (like just not being good with the glove) then the focus should be on forcing shots to come from farther away (duh) so the pads can do their job and the minders have more time to cover the upper areas. Or, they might have to play out a bit further to take away some of that shooting space/angle/time and use the stick more to stop the shot when possible.

We see a lot of side-to-side movement that looks like "swimming" in the crease but do we see many aggressive, confident challenges? Seems like they're hugging the post a lot.

Admittedly this might be bullshit coming from a faulty memory or just some bias of perception.

 

[traparatus]

Not that the numbers would necessarily paint a different story, but I’d rather look at GSAx rather than GSAA. Goals saved above expected (GSAx) takes into account the quality of shot the goaltender is facing, while goals saved above average (GSAA) does not take into account shot quality against.

Sure. Even of the top of my head, I know that number is not going to be pretty, considering that as a team Caps Expected Goals Against are a lot lower than actual goals against.

Caps are 10th in the league, 3rd in the division in xGA at 2.06 xGA/60. It's very close with Caps in a virtual tie with the Islanders, Boston slightly better. Caps are 17th in Goals Against at 2.48 GA/60. Not only is that 4th in the division but it places Caps a whooping 0.5 GA/60 behind the Islanders.

In a twisted kind of way, I still don't think goaltending is really the issue. Vanecek and Samsonov are performing about as well as I can reasonably expect two inexperienced goaltenders to perform. They are exhibiting all the typical issues of inexperience. Poor rebound control, excessive movement, dropping down into butterfly too early, as @g00n has pointed out.

The issue to me, if there even is one, is a coach's choice. Some coaches would look at the state of Caps' goaltending and formulate a game plan that focuses on making life as easy as possible for these netminders. Laviolette's plan appears to focus on winning games despite crappy goaltending. I said it before in the general thread, I think playing goalie for the Caps is really hard. It's not about the amount or quality of chances. IMO, it's about unpredictability and lack of defensive zone structure. It's a difficult game for a goaltender to read.

 

[kicksavedave]

Its also true that at least a non trivial amount of the goals against have been 100% directly caused by faulty puck handling of the goalie themselves, like Sammy's pass right to a Flyer that created the Mantha Zapruder Austopsy thread. I believe Sammy also cost himself a shutout earlier in the year with one dreadful mishandle leading right to the opponents only goal.

Point is, not all the GAA are on bad defense, bad overall goaltending technique, or "softies"... some of them are on easily correctable stuff, or stuff that simply a goalie should not do. Wonder how many goals like this there have been and how that would effect their numbers.

 

[g00n]

Its also true that at least a non trivial amount of the goals against have been 100% directly caused by faulty puck handling of the goalie themselves, like Sammy's pass right to a Flyer that created the Mantha Zapruder Austopsy thread. I believe Sammy also cost himself a shutout earlier in the year with one dreadful mishandle leading right to the opponents only goal.

Point is, not all the GAA are on bad defense, bad overall goaltending technique, or "softies"... some of them are on easily correctable stuff, or stuff that simply a goalie should not do. Wonder how many goals like this there have been and how that would effect their numbers.

But it's hard to tell how much of that is going on across the league. I'm sure other goalies mishandle the puck and make bad outlet passes from time to time, too.

 

[kicksavedave]

But it's hard to tell how much of that is going on across the league. I'm sure other goalies mishandle the puck and make bad outlet passes from time to time, too.

I mean Holtby did that in the SCF and cost himself a shutout vs Vegas, so yeah it happens. Feels like Sammy and Vanny has done it quite a bit though, but I don't have numbers to support my crap memory or compare to other teams.

Speaking of Holtby, what made his Cup run effective was not "The Save" which was awesome. It was that he gave up basically zero softies all playoffs long (the one above being the only one I can recall). He made the other teams earn every goal. And I have to think that helped inspire the team in front of him.

Something to be said for everyone just "doing their job" as one grumpy old NFL coach likes to say.

 

[twabby]

Various all-around stats for Justin Schultz this season:

Hockeyviz:

Evolving-Hockey:

JFresh:




Schultz is a below replacement level player being paid like a pretty good second pairing defenseman and is being played like a #3 defenseman. For the sake of the team BMac needs to do something about this. Rolling with him in the top 4 next season won't cut it, and on a capped out team $4 million is way too much to pay for someone who could do the same job on an ELC.

 

[Ovie's Neighbor]

Various all-around stats for Justin Schultz this season:

Hockeyviz:

View attachment 439240

Evolving-Hockey:

View attachment 439241

JFresh:




Schultz is a below replacement level player being paid like a pretty good second pairing defenseman and is being played like a #3 defenseman. For the sake of the team BMac needs to do something about this. Rolling with him in the top 4 next season won't cut it, and on a capped out team $4 million is way too much to pay for someone who could do the same job on an ELC.

We don’t need advanced stats to know that about Schultz. Just look at how the team performed when TVR was in the lineup. I’m sure that directly contributes to Schultz WAR.

 

[twabby]

We don’t need advanced stats to know that about Schultz. Just look at how the team performed when TVR was in the lineup. I’m sure that directly contributes to Schultz WAR.

A lot of people think Schultz had a good year and was worth playing over TVR, including the head coach of the Washington Capitals.

It’s not at all obvious to a large group of people exactly how bad Schultz was this year.

 

[g00n]

A lot of people think Schultz had a good year and was worth playing over TVR, including the head coach of the Washington Capitals.

It’s not at all obvious to a large group of people exactly how bad Schultz was this year.

He sucked late in the year and in the playoffs. I think everyone knows that. There's like one guy who's been defending him in general and most here wanted to see TVR vs Boston. He was the 3rd highest vote getter for our "on the hot seat this offseason" poll behind Kuzy and Chara. So "a lot of people" is false or misleading.

As for the coach's decision, you have to ask him. We're not in the locker room or war room so we don't know the Xs and Os they discuss, the little injuries and nuances, etc.

What about earlier in the season? What kind of numbers did he have then?

 

[txpd]

He sucked late in the year and in the playoffs. I think everyone knows that. There's like one guy who's been defending him in general and most here wanted to see TVR vs Boston. He was the 3rd highest vote getter for our "on the hot seat this offseason" poll behind Kuzy and Chara. So "a lot of people" is false or misleading.

As for the coach's decision, you have to ask him. We're not in the locker room or war room so we don't know the Xs and Os they discuss, the little injuries and nuances, etc.

What about earlier in the season? What kind of numbers did he have then?

I am guessing I am the one guy. I dont think I ever defended Schultz. I simply said that Laviolette's basic system called for the defense to be all in creating offense. Schultz is an offensive player and TvR had 3 points. In his final 10 regular season games Schultz had 9 points. They weren't going to scratch that.

My feeling is that his injury was as serious as Carlson's was. We were never told post playoffs what that injury was.

 

[twabby]

He sucked late in the year and in the playoffs. I think everyone knows that. There's like one guy who's been defending him in general and most here wanted to see TVR vs Boston. He was the 3rd highest vote getter for our "on the hot seat this offseason" poll behind Kuzy and Chara. So "a lot of people" is false or misleading.

As for the coach's decision, you have to ask him. We're not in the locker room or war room so we don't know the Xs and Os they discuss, the little injuries and nuances, etc.

What about earlier in the season? What kind of numbers did he have then?

I wasn't just talking about people here, I'm talking about Laviolette, I'm talking about the beat writers (Tarik El Bashir in particular who noted that Schultz had a "solid" first season in Washington), I'm talking about fans that I interact with elsewhere who thought Schultz had a good season overall. So yes, I think a lot of people is accurate.

Of course I do not have access to Peter Laviolette or have him in my rolodex, so unfortunately I cannot ping him and see what his thinking was. Given the information I have, in my opinion he made a pretty big mistake by either not limiting Schultz's minutes, or not scratching him completely given a perfectly good option on the bench to replace him.

Here's a rolling 10-game xGD graph:

So he did start out pretty well, but tailed off quite substantially toward the end. Of course a lot of his uptick is playing alongside Dmitry Orlov, who is an excellent playdriver.

More concerning than his first year in DC is his multiyear trend and his isolated impacts over the years, which have been trending down:

Maybe it's because of injuries, but do we really expect this guy to have a healthy season next year?