Coach Discussion: All Purpose Coaching Thread Part III

[Fastfrde]

So, say you have a team with some very offensively talented players and possibly not a very good defense. This same team is trying to limit goals against by playing a conservative style and collapsing the forwards back to help the defense. Offensively, trying to be opportunistic on scoring chances, which they can convert based on talent level, at a fairly consistent rate. Would this not result in a bad xGF? Would this not be further exasperated if you were actually winning more games than losing, hence playing more conservative?

Not saying that it's sustainable or not, but it seems pretty explainable in the way the team is trying to play.

 

[Gm0ney]

I'm not sure about the value of "adjusted" xG. Why should the expected goal value of a shot depend on game situation?

I guess the idea is that some shots would never have happened had the game been tied, or would not have been in that position?

Just intuitively, if a team is leading by a couple of goals in the 3rd, their bottom 6 and 3rd pairing are getting more ice time, people aren't risking life and limb to block shots and "pay the price" - and the trailing team is shortening the bench, and being more aggressive in the offensive zone...these things all add up to more and higher-quality chances against for the team that's chasing.

 

[Gm0ney]

I dunno dude... when 8 of the top 11 xGF% teams are not in a playoff position, it just doesn't seem like an appropriate stat to hang your hat on. If you don't like my cut-off, how about I take the top two teams in each division and see where they rank:

12, 17, 18, 21, 23, 24, 25, and 27

Looks like expected goals is pretty bogus since it doesn't correlate to actual success at all. In fact, a bad xGF% is more telling of a good team, at least for this year!

If you're looking for strong correlations...

The Jets today sit at 44.11% xGF. Since the 2007-2008 season, only 2 teams have made the playoffs with an xGF% that low. The bottom 3 teams in the league in xGF each season have made the playoffs 4 times total since 2007-08 - and one of those teams was in the lockout-shortened 2012-13 season.

 

[Guffman]

If you're looking for strong correlations...

The Jets today sit at 44.11% xGF. Since the 2007-2008 season, only 2 teams have made the playoffs with an xGF% that low. The bottom 3 teams in the league in xGF each season have made the playoffs 4 times total since 2007-08 - and one of those teams was in the lockout-shortened 2012-13 season.

Can you explain why almost every quality team (1st or 2nd in their division) ranks in the bottom half of the league in the xGF%?

 

[Whileee]

I guess the idea is that some shots would never have happened had the game been tied, or would not have been in that position?

Just intuitively, if a team is leading by a couple of goals in the 3rd, their bottom 6 and 3rd pairing are getting more ice time, people aren't risking life and limb to block shots and "pay the price" - and the trailing team is shortening the bench, and being more aggressive in the offensive zone...these things all add up to more and higher-quality chances against for the team that's chasing.

I understand why you would adjust for score and venue for shot attempts (Corsi), because of the reasons you note. But if you are trying to assess which team actually had more expected goals in a game then adjustment for score makes no sense. You can say the Jets wouldn't have generated those chances if the game was tied, but it doesn't change the actual estimate of expected goals because that is based on shot type, location, etc.

My general point is a bit of a "devil's advocate" approach to the current trend of judging team performance through the xG lens. I understand it has shown some predictive properties, but I think those have been overstated in the general usage. I think the xG estimate for last night's game indicate that they can be misleading for the balance of play for an individual game, in addition to having limits on the predictive value for future team performance.

 

[KingBogo]

Thanks for that. I was under the impression we won xGF% clearly for some reason. That makes a lot more sense, especially considering score effects.

And unless I'm mistaking the team, the Canes have previously been a good Corsi team but not a good xGF (meaning they take a lot of low-percentage-shots).

The Canes seem different now. They have evolved exactly like Ehlers has. They can actually turn possession into goals much more effectively.

According to Natural Statstrick the Jets "won" the 5-5 xGF% in all 3 periods ending up with 2.44 xGF and 1.85 xGA for a 56.95% xGF% for the game. IMO this is a pretty clear xGF% "win" playing at 5-5.

Carolina Hurricanes @ Winnipeg Jets, 2019-12-17

I just saw Whileee's comments and agree it makes no sense to look at adjusted xGF% like it adds anything meaningful.

 

[Whileee]

According to Natural Statstrick the Jets "won" the 5-5 xGF% in all 3 periods ending up with 2.44 xGF and 1.85 xGA for a 56.95% xGF% for the game. IMO this is a pretty clear xGF% "win" playing at 5-5.

Carolina Hurricanes @ Winnipeg Jets, 2019-12-17

I just saw Whileee's comments and agree it makes no sense to look at adjusted xGF% like it adds anything meaningful.

I think adjusted xGF% might be useful as a data point for a predictive or analytical model, but for describing the game and balance of chances the adjustment obscures more than it reveals.

 

[Jetfaninflorida]

According to Natural Statstrick the Jets "won" the 5-5 xGF% in all 3 periods ending up with 2.44 xGF and 1.85 xGA for a 56.95% xGF% for the game. IMO this is a pretty clear xGF% "win" playing at 5-5.

Carolina Hurricanes @ Winnipeg Jets, 2019-12-17

I just saw Whileee's comments and agree it makes no sense to look at adjusted xGF% like it adds anything meaningful.

That game looked just like many Jets games this year. Helle puts up a 0.930 and we win that game.

 

[Gm0ney]

Can you explain why almost every quality team (1st or 2nd in their division) ranks in the bottom half of the league in the xGF%?

I'd say it's a bit of cherry picking the top 2 in each division instead of the top 8 in points% - xGF powerhouses Pittsburgh and Carolina are excluded because they're in the Metro with Washington and the Islanders - in exchange, xGF bottom-halfers Arizona and Buffalo are counted...that's 25% of the sample.

It's also a bit of making things sound worse than they really are by lumping a lot of teams into the "bottom half of the league in xGF%" - e.g. Boston is a division leader in the bottom half of the league in xGF% - they're 16th with a 49.7% xGF. So they're not really rubbing shoulders with the Jets here. Also using adjusted xGF% raises Colorado to 50.00% and Boston to 50.35%...so not awful xGF.

Plus there's more to the game than 5v5. Special teams and goaltending and outshooting expectations can all play big roles.

 

[ecolad]

Just a few quick thoughts about last night`s game.

- Was struck by how structured the Canes were, in transitioning out of their end and in entering our end. Yes, they kept everything at a high tempo, with excellent sticks, but they were also positionally very strong and moved with a clear purpose. Notice how well they attacked with the puck, with all 5 men in the unit rotating and trying to create back-door chances - impressive and something clearly practiced as set plays. This is significantly different than what our boys are doing, with our D hesitant to actively enter the zone and primarily remaining static at the points.
- Was also struck by the aggressive overload fore-check scheme used by the Canes to intensify pressure on our guys. This seems pretty basic to their overall approach, but as it happens, is truly kryptonite to our already struggling efforts to vacate the D zone. This again is very much different than our approach of recent where Coach has gotten very passive with offensive zone fore-check and instead has focussed on efforts in the neutral zone and our D zone to limit scoring attempts against.

 

[KingBogo]

I'd say it's a bit of cherry picking the top 2 in each division instead of the top 8 in points% - xGF powerhouses Pittsburgh and Carolina are excluded because they're in the Metro with Washington and the Islanders - in exchange, xGF bottom-halfers Arizona and Buffalo are counted...that's 25% of the sample.

It's also a bit of making things sound worse than they really are by lumping a lot of teams into the "bottom half of the league in xGF%" - e.g. Boston is a division leader in the bottom half of the league in xGF% - they're 16th with a 49.7% xGF. So they're not really rubbing shoulders with the Jets here. Also using adjusted xGF% raises Colorado to 50.00% and Boston to 50.35%...so not awful xGF.

Plus there's more to the game than 5v5. Special teams and goaltending and outshooting expectations can all play big roles.

Seems to me that @Guffman has been pretty straight forward and clean on his dividing lines while you are dancing around on the head of a pin trying to make your case with very shaky lingages and even then they really don't add that much support to your argument.

 

[Gm0ney]

Seems to me that @Guffman has been pretty straight forward and clean on his dividing lines while you are dancing around on the head of a pin trying to make your case with very shaky lingages and even then they really don't add that much support to your argument.

I took a look at the Top 11 teams based on xGF %.
...

Can you explain why almost every quality team (1st or 2nd in their division)...

Yes, those straight forward metrics everyone's always talking about: the old "Top 11" and the "1st or 2nd team in each division when teams have played a differing number of games" measuring stick of team quality...I mean what's the reason for selecting a Top 11 or just the top 2 in each division when teams haven't played the same number of games? Why not Top 10 or go on Points% to level the playing field? And why measure against the bottom half in xGF%? There's a big difference between 16th and 31st - it's larger than the difference between 16th and 1st.

So anyway, back to my dancing on the head of a pin: over the last 12 seasons, the bottom 3 teams in xGF% have managed to make the playoffs a total of 4 times (11%).

My reason for looking at the "bottom 3" is because the Jets are currently in the bottom 3. They actually have a worse xGF% than some last place teams in some seasons (5/12 = 41.7%). I'm looking at the extremes, because the Jets are currently at the extreme.

Of the teams with a worse xGF% over the last 12 seasons than the Jets current mark, only 2 have made the playoffs - and one was the 2012-13 Maple Leafs in the lockout shortened season (and everyone was expecting the wheels to fall off that team at the time - the season just wasn't long enough).

Looking at the Top 3 teams in xGF%, 5 of those teams have missed the playoffs in 12 seasons (14%).

Anyway, that's my case. The defense rests.

 

[Guffman]

Like, if it’s such a big deal to say Top 11, I’ll change it to Top 10...7 aren’t in the playoffs, 2 are WC teams and the last was a third place team.

If you didn’t like me picking the top 2 teams in each division, we can take the top 8 teams my point %. Still won’t change a thing.

Best xGF% teams are almost all out of the playoffs, while the best teams in the league have terrible xGF%.

I think we should be asking why great teams don’t have a high xGF%.

Edit: Maybe all the elite teams are afflicted by bad coaching but are being carried by great goaltending and/or elite snipers that transcend xGF% established by average hockey players!

 

[Saidin]

That game looked just like many Jets games this year. Helle puts up a 0.930 and we win that game.

I would say there's a in between point from .930 and what we got from Helly on Tuesday.

 

[Gm0ney]

Like, if it’s such a big deal to say Top 11, I’ll change it to Top 10...7 aren’t in the playoffs, 2 are WC teams and the last was a third place team.

If you didn’t like me picking the top 2 teams in each division, we can take the top 8 teams my point %. Still won’t change a thing.

Best xGF% teams are almost all out of the playoffs, while the best teams in the league have terrible xGF%.

I think we should be asking why great teams don’t have a high xGF%.

Edit: Maybe all the elite teams are afflicted by bad coaching but are being carried by great goaltending and/or elite snipers that transcend xGF% established by average hockey players!

I was just answering your question. Your parameters make things look more lopsided than they really are. And 5v5 xGF% isn't the whole game - elite goaltending, good shooting luck, and special teams can overcome a negative 5v5 xGF%. That's what's happened with the Jets...well the special teams haven't been amazing but taking fewer penalties has been beneficial.

And historically, a lot of stars have to align for a team with as bad an xGF% as the Jets to make the playoffs. The good news for us is that those stars have been aligning for the Jets.

As for how this relates to coaching - do you think it's by design that the Jets are losing the xGF battle most nights? Do you think 44% is all we can muster with this lineup? This team used to be dominant at 5v5 and it took us to the WCF. Now we're relying for miracles in net and hoping our shooting talent can continue to outscore expectations...you shouldn't even need to look at xGF to see that.

 

[Whileee]

I was just answering your question. Your parameters make things look more lopsided than they really are. And 5v5 xGF% isn't the whole game - elite goaltending, good shooting luck, and special teams can overcome a negative 5v5 xGF%. That's what's happened with the Jets...well the special teams haven't been amazing but taking fewer penalties has been beneficial.

And historically, a lot of stars have to align for a team with as bad an xGF% as the Jets to make the playoffs. The good news for us is that those stars have been aligning for the Jets.

As for how this relates to coaching - do you think it's by design that the Jets are losing the xGF battle most nights? Do you think 44% is all we can muster with this lineup? This team used to be dominant at 5v5 and it took us to the WCF. Now we're relying for miracles in net and hoping our shooting talent can continue to outscore expectations...you shouldn't even need to look at xGF to see that.

What proportion of teams with a 0.615+ points % on December 15 miss the playoffs?

Also, it looks like the correlation between xGF% at this point of the season and future points% really weakened in the past two seasons. Measurement error? Changing systems? Fluke?

 

[Saidin]

I think Mo's approach to be better 'defensively' is failing hard (just look at the buttery soft and inviting zone infront of our net) and in doing so, sacrifices all our Offense. Why not just let the hounds loose and see where our chip lie. Can't be any worse than it's been this year can it?

 

[Guffman]

I was just answering your question. Your parameters make things look more lopsided than they really are. And 5v5 xGF% isn't the whole game - elite goaltending, good shooting luck, and special teams can overcome a negative 5v5 xGF%. That's what's happened with the Jets...well the special teams haven't been amazing but taking fewer penalties has been beneficial.

And historically, a lot of stars have to align for a team with as bad an xGF% as the Jets to make the playoffs. The good news for us is that those stars have been aligning for the Jets.

As for how this relates to coaching - do you think it's by design that the Jets are losing the xGF battle most nights? Do you think 44% is all we can muster with this lineup? This team used to be dominant at 5v5 and it took us to the WCF. Now we're relying for miracles in net and hoping our shooting talent can continue to outscore expectations...you shouldn't even need to look at xGF to see that.

Agreed that 5 v 5 xGF% isn't the whole game. I wasn't really cherry picking. If 5 v 5 xGF% was a good part of the game, then you'd expect to see at least one contending team within the top 10 of xGF% rankings, right?

I think when you do have a quality team, the way you play may be quite a bit different than how a weaker team plays. Are contending teams trying to maximize xGF% based on the players they have? If not, why not? I haven't really thought about it but maybe that's worth exploring.

As for the Jets, the reality is that we have a decimated defense corp. It's ridiculous turnover. Based on having so many AHL/waiver wire players, I would expect a big degradation in results. Going into this season without Buff and the turnover we had, I didn't think this was a realistic playoff year. Not surprised we would have a terrible xGF%. Now, if you believe a poor xGF% correlates to bad coaching, then I guess all the coaches on contending teams should be fearing for their lives and professional hockey people shouldn't be suggesting Maurice should be considered for Coach of the Year.

If I thought teams tried to maximize xGF%, maybe I would share your opinion.

 

[Gm0ney]

Agreed that 5 v 5 xGF% isn't the whole game. I wasn't really cherry picking. If 5 v 5 xGF% was a good part of the game, then you'd expect to see at least one contending team within the top 10 of xGF% rankings, right?

Meh, not necessarily...stats and probabilities are funny like that sometimes and we're looking at just this one part of the game - which is pretty important, but not so important that its impact can't be overshadowed by the other stuff.

I think when you do have a quality team, the way you play may be quite a bit different than how a weaker team plays. Are contending teams trying to maximize xGF% based on the players they have? If not, why not? I haven't really thought about it but maybe that's worth exploring.

I think it's more likely that a team with lots of talent and a flawed plan can paper over bad xGF...like with a killer powerplay or great goaltending. I think ideally every team would like to consistently carry the play and generate more dangerous chances on offense, while also preventing dangerous chances against.

We've heard Maurice talk about zone-time - it seems to be something he considers important. And I can perhaps understand if his overall strategy is to stay out of the defensive zone as much as possible - although last year with basically the same lineup as the 2017-18 team that went to the conference finals I really thought that was a dumb plan and had a lot to do with the Jets' eye-test and fancy-stat woes...I wanted that team to pin its ears back and go. Anyway, this plan isn't working. The Jets have the lowest xGF/60 in the NHL and 3rd highest xGA/60...we're only low-event in the offensive zone.

And this plan wasn't working last year either. Even with Trouba and Buff (for half the season) and Myers and Chiarot and Tanev and Little the Jets weren't creating much offensively or preventing much defensively. Their xGF/60 last season was 7th lowest in the league. Their xGA/60 was 9th highest.

So yeah, we're worse this year. At both offense and defense:

	xGF/60	xGA/60
2018-19	2.17	2.38
2019-20	1.97	2.49

[TBODY] [/TBODY]

As for the Jets, the reality is that we have a decimated defense corp. It's ridiculous turnover. Based on having so many AHL/waiver wire players, I would expect a big degradation in results. Going into this season without Buff and the turnover we had, I didn't think this was a realistic playoff year. Not surprised we would have a terrible xGF%. Now, if you believe a poor xGF% correlates to bad coaching, then I guess all the coaches on contending teams should be fearing for their lives and professional hockey people shouldn't be suggesting Maurice should be considered for Coach of the Year.

If I thought teams tried to maximize xGF%, maybe I would share your opinion.

This team hasn't been good since the 2018 playoffs. There's been a dropoff this year but the team wasn't playing well to begin with. To go from 8th-worst xGF to 2nd-worst xGF with worse personnel...it's not an accomplishment. We're just burning more prime years of our skilled players, spinning our tires with Maurice.

 

[Gm0ney]

What proportion of teams with a 0.615+ points % on December 15 miss the playoffs?

Also, it looks like the correlation between xGF% at this point of the season and future points% really weakened in the past two seasons. Measurement error? Changing systems? Fluke?

Not many, I'd suspect...but I'm not saying the Jets are going to miss the playoffs...although I wouldn't be shocked if they did a massive swoon and fell out of the playoffs, either.

 

[KingBogo]

Yet another article by Murat in the Athletic that is very complimentary to a part of the systems the Jets run. There have been a few this year and every read offers another layer of insight on what the Jets are doing on the ice.

What is 'second quick'? A simple yet foundational part of...

 

[lablite47]

Just a few quick thoughts about last night`s game.

- Was struck by how structured the Canes were, in transitioning out of their end and in entering our end. Yes, they kept everything at a high tempo, with excellent sticks, but they were also positionally very strong and moved with a clear purpose. Notice how well they attacked with the puck, with all 5 men in the unit rotating and trying to create back-door chances - impressive and something clearly practiced as set plays. This is significantly different than what our boys are doing, with our D hesitant to actively enter the zone and primarily remaining static at the points.
- Was also struck by the aggressive overload fore-check scheme used by the Canes to intensify pressure on our guys. This seems pretty basic to their overall approach, but as it happens, is truly kryptonite to our already struggling efforts to vacate the D zone. This again is very much different than our approach of recent where Coach has gotten very passive with offensive zone fore-check and instead has focussed on efforts in the neutral zone and our D zone to limit scoring attempts against.

This has been the case since Maurice arrived. Decent coach, seems to be well liked by players, is very dedicated, works hard. Great communicator, both in the dressing room and in front of the mic. Unfortunately though, not able to bring this team over the top, mostly due to the style of play he commands IMO. Not very dynamic and/or creative offensively and way too passive on PK. Time to move on from both him and Huddy who's been here since their return in particular.
There has to be better options available to bring this team over the top. Hope they do well in Seattle.

 

[Jetfaninflorida]

I would say there's a in between point from .930 and what we got from Helly on Tuesday.

View attachment 294679

Sure. 0.875 is in between.

 

[Howard Chuck]

I think like any coach in the league, or any league, some coaches are more suited to certain things than they are to others.

At a very high level, Maurice seems to be great at helping and playing young growing players, and has just shown us that he can help a team that is seriously depleted through injuries.

What I'm not sure that he can do, is take a group of exceptionally talented players and make them better than the sum of their parts. He seems to be great at helping a team overcome adversity because of any number of reasons, but just seems to have trouble with that last push to team greatness.

Right now we are in that 'facing adversity' stage due to injuries, so I give him credit for our record, but I will reserve judgement about when we are at full strength or closer.

 

[Hunter368]

Yet another article by Murat in the Athletic that is very complimentary to a part of the systems the Jets run. There have been a few this year and every read offers another layer of insight on what the Jets are doing on the ice.

What is 'second quick'? A simple yet foundational part of...

Yup pretty positive overall.