Pre-Game Talk: Wild vs Jets - Fri Oct 20 7:00pm CT

[ijuka]

I guess with the line shuffling it's good to look at xGF%(expected goals %) for all the lines that have played at least 10 minutes together. Yeah, that's a low number, but there's been so much line shuffling that otherwise there'd be just a few lines. This is for 5v5.

Ehlers-Scheifele-Wheeler - 39.82%
Perreault-Little-Laine - 57.45%
Matthias-Lowry-Tanev - 62.34%
Matthias-Copp-Tanev - 56.52%
Matthias-Lowry-Petan - 38.33%
Ehlers-Little-Laine - 64.63%

Well, only the first two have a decent amount of ice time(66 and 45 minutes respectively). But we can still see a couple of signals - For example, Ehlers-Scheifele-Wheeler doesn't work at all. When it comes to goals for and against, Laine and Little combine for 5 for 1 against which is the best by far. All the third lines have zero goal actions for or against and in fact, while the GF%s are good the actual expected goals are REALLY low. Real low event stuff going on with these versions of the third line.

Wheeler-Scheifele-Ehlers are at 3 goals for 4 against.


So if we were to assume that the third line variations don't have enough of a sample size, it might at least be good to recognize that something would need to be done to Ehlers-Scheifele-Wheeler, which they are. So that's a good start.

 

[KingBogo]

Last ten games against the Wild 5-5-0 Not sure who struggles against who

I guess with the line shuffling it's good to look at xGF%(expected goals %) for all the lines that have played at least 10 minutes together. Yeah, that's a low number, but there's been so much line shuffling that otherwise there'd be just a few lines. This is for 5v5.

Ehlers-Scheifele-Wheeler - 39.82%
Perreault-Little-Laine - 57.45%
Matthias-Lowry-Tanev - 62.34%
Matthias-Copp-Tanev - 56.52%
Matthias-Lowry-Petan - 38.33%
Ehlers-Little-Laine - 64.63%

Well, only the first two have a decent amount of ice time(66 and 45 minutes respectively). But we can still see a couple of signals - For example, Ehlers-Scheifele-Wheeler doesn't work at all. When it comes to goals for and against, Laine and Little combine for 5 for 1 against which is the best by far. All the third lines have zero goal actions for or against and in fact, while the GF%s are good the actual expected goals are REALLY low. Real low event stuff going on with these versions of the third line.

Wheeler-Scheifele-Ehlers are at 3 goals for 4 against.


So if we were to assume that the third line variations don't have enough of a sample size, it might at least be good to recognize that something would need to be done to Ehlers-Scheifele-Wheeler, which they are. So that's a good start.

I'd like to see your source. The E-S-W line scored 4 goals and had 10 points in the Edmonton game alone.

 

[Jimby]

The xGF% is from Corsica. Ehlers, Scheif and Wheeler have sucked 5 on 5.

 

[jetsforever]

Awesome to see Lemieux get in

 

[Aavco Cup]

The xGF% is from Corsica. Ehlers, Scheif and Wheeler have sucked 5 on 5.

Played a lot against McDavid and had a terrible game against Calgary and Columbus

 

[ijuka]

I'd like to see your source. The E-S-W line scored 4 goals and had 10 points in the Edmonton game alone.

Corsica is the source.

The game against Oilers:

Goal 1: Matthias Scheifele Wheeler
Goal 2: Ehlers Scheifele Wheeler
Goal 3: Ehlers Lowry Tanev
Goal 4: Ehlers Scheifele Wheeler

So that's 2 goals, not 4. Feel free to look for the third, I generally trust Corsica as a source. Regardless, it was more about the awful xGF% here. The actual goals for and against was more of a bonus.

 

[KingBogo]

The xGF% is from Corsica. Ehlers, Scheif and Wheeler have sucked 5 on 5.

Are you comfortable with the sample size? If you extend the sample size to a reasonable amount going back to last season they are pretty dominant.

 

[Jimby]

xGF% looked better last year but they still gave up more goals than they scored.

Are you comfortable with the sample size? If you extend the sample size to a reasonable amount going back to last season they are pretty dominant.

 

[Ducky10]

Last ten games against the Wild 5-5-0 Not sure who struggles against who

And coincidentally enough, 11-11 in 22 games since the team came back to Winnipeg.

I assume some Wild fans likely figure they struggle against us as well.

 

[Eyeseeing]

Why is this in a Pre Game Thread?

Because there was conversation involving Maurice in this thread
Mods can move the or delete.
Like it or not Maurice is going to pop up in every thread in some way shape or form
Have a great evening

 

[ijuka]

Are you comfortable with the sample size? If you extend the sample size to a reasonable amount going back to last season they are pretty dominant.

Well, it's about how they've been doing this season. Using this logic, no line's been good this season and no line's been bad. Every line is unknown and it doesn't matter that Kucherov and Stamkos have scored 12 points for example.

I don't agree with this line of thinking. Them being bad until now doesn't mean they're bad for the whole season, but that doesn't change that they've been bad.

 

[KingBogo]

xGF% looked better last year but they still gave up more goals than they scored.

Well for now they are pretty much all that scores for the Jets

 

[KingBogo]

Well, it's about how they've been doing this season. Using this logic, no line's been good this season and no line's been bad. Every line is unknown and it doesn't matter that Kucherov and Stamkos have scored 12 points for example.

I don't agree with this line of thinking. Them being bad until now doesn't mean they're bad for the whole season, but that doesn't change that they've been bad.

It might have been the language you used "Ehlers-Scheifele-Wheeler doesn't work at all" where we know that they have produced a lot together reaching into last season. A small sample size this season doesn't change that IMO. And the 3 of them have gained the lion's share of points this season even if one was an empty net and another was on the PP.

 

[ijuka]

It might have been the language you used "Ehlers-Scheifele-Wheeler doesn't work at all" where we know that they have produced a lot together reaching into last season. A small sample size this season doesn't change that IMO. And the 3 of them have gained the lion's share of points this season even if one was an empty net and another was on the PP.

Hmm, that could be it. Well, I looked at last season's stats and their xGF% certainly was better. However...

Goals for: 9 Goals against: 11.

So this sample size of the line being negative overall now has a sample size of over a season. It was not a positive GF% line last season either like you were implying.

 

[ERYX]

Matthias-Lowry-Tanev - 62.34%
...
Ehlers-Little-Laine - 64.63%

For all the hate that these two line combos get, and specifically the amount of hate directed individually at Matthias, Lowry, and Tanev, they have some good numbers and have been effective.

 

[KingBogo]

Hmm, that could be it. Well, I looked at last season's stats and their xGF% certainly was better. However...

Goals for: 9 Goals against: 11.

So this sample size of the line being negative overall now has a sample size of over a season. It was not a positive GF% line last season either like you were implying.

E-S-W only scored 9 goals on the ice together 5-5 last season?

I was questioning on sample size because IMO small sample size leads to bad predictions. For example this season's ixGF/60 the Jets rankings are:

1) MP
2) Ehlers
3) Copp
4) Scheifele
5) Laine

Which according to Corsica 5v5 ixG60 is a better predictor of future 5v5 G60 than G60 itself (0.152 for forwards and 0.128 for defense vs. 0.140 and 0.076 respectively). Do we really believe Copp will out score Laine 5-5 this season?

 

[ijuka]

E-S-W only scored 9 goals on the ice together 5-5 last season?

I was questioning on sample size because IMO small sample size leads to bad predictions. For example this season's ixGF/60 the Jets rankings are:

1) MP
2) Ehlers
3) Copp
4) Scheifele
5) Laine

Which according to Corsica 5v5 ixG60 is a better predictor of future 5v5 G60 than G60 itself (0.152 for forwards and 0.128 for defense vs. 0.140 and 0.076 respectively). Do we really believe Copp will out score Laine 5-5 this season?

That is what surprised me too, certainly. 240 minutes of ice time, only 9 goals scored. Expected goals was about 11 and expected goals against was like 8 so they kind of flip flopped on those. Maybe it's bad luck? In any case, in the last and this season combined the line is 12-15 in 5v5 goals, which isn't where you want your first line to be.

For comparison, Ehlers-Scheifele-Laine was at about 11 xGF and actually scored 23.

 

[KingBogo]

That is what surprised me too, certainly. 240 minutes of ice time, only 9 goals scored. Expected goals was about 11 and expected goals against was like 8 so they kind of flip flopped on those. Maybe it's bad luck? In any case, in the last and this season combined the line is 12-15 in 5v5 goals, which isn't where you want your first line to be.

For comparison, Ehlers-Scheifele-Laine was at about 11 xGF and actually scored 23.

ixGF/60 seems a little sketchy to me. Looking at last season Liane was ranked 13th on the Jets right behind Matthias and right ahead of Tanev. This is suppose to be a better predictor of future scoring than actual past 5-5 GF/60

http://corsica.hockey/skater-stats/

 

[Howard Chuck]

Ahhh, the joys of slicing and dicing stats....

 

[Neuf]

Genuinely excited to see Lemieux's debut. It's he's a M Tkachuk lite, that would be great. Just don't take penalties unless you're drawing more or getting points.

 

[Agent Zub]

Is Connor going to start?

 

[WPGChief]

Why is everyone giving little snippets? Here's the full rap on the Jets line combos last year at 5v5 which are also adjusted.

I'd pay particular attention to the actual numbers of xGF and xGA so we're not entirely fooled by the percentages. Ehlers-Scheifele-Wheeler had the 4th highest xGF%, which came with 10.82 xGF (3rd highest). If you look at Matthias-Lowry-Armia, they had 13.49 xGF for top in that category, but that came because of their 367 minutes of TOI at 5v5. ESW accomplished that in less TOI, and were expected to give up fewer goals as well.

Additionally, remember that Corsica's xG model does not consider "shooter talent" (DTM's considered previous shooting percentage, based on the assumption that some shooters are better than others (i.e. Steven Stamkos's shot likely has a higher chance of scoring in certain shots than Tanner Glass's)). From his blog post from the Corsica 1.0 days:

Mine is a shot quality model. xG stats are by-products of assigning goal expectancy to shots. In my mind, this ability to assess shot quality is most important, though supplying information with which to devise better evaluative metrics is a welcome benefit. The model is similar in nature to that of @DTMAboutHeart,3 with some important distinctions. The most important difference is his inclusion of regressed shooting talent. I chose to exclude shooter talent not because it isn’t an important factor, but rather because I fear players may unfairly benefit or suffer from their linemates’ aptitude. Here’s what my model does account for:

• Shot type (Wrist shot, slap shot, deflection, etc.)
• Shot distance (Adjusted4 distance from net)
• Shot angle (Angle in absolute degrees from the central line normal to the goal line)
• Rebounds (Boolean – Whether or not the shot was a rebound)
• Rush shots (Boolean – Whether or not the shot was a rush shot)
• Strength state (Boolean – Whether or not the shot was taken on the powerplay)5

 

[ocdaddy]

The Jets are up against another well coached team who are well aquainted with their system.

That's an advantage to the visitors before the puck even drops.

 

[Gm0ney]

And coincidentally enough, 11-11 in 22 games since the team came back to Winnipeg.

I assume some Wild fans likely figure they struggle against us as well.

The struggle is fans trying to stay awake during games against the Wild.

 

[Zhamnov5GoalGame]

Hoping the 3a / b- lines can chip in some offense.
Let Laine and Connor pop off a couple of goals.
I'll be at the game so I'd like to see a decent match.