Management Thread | Blurst of Times

[pitseleh]

In Hronek's case, his zone exit numbers cratered when paired with Hughes, because of course he defers to the best zone exit guy in the NHL. Does that mean that Hronek is worse than when he was the primary puck mover on his pairing in Detroit? Absolutely not, but these graphs will now say he's shit at it.

In fairness, Hronek’s exit numbers were poor in Detroit too. I thought the same as you a while ago but Burke’s Evil Spirit posted a good article showing that wasn’t the case. He was like fourth or fifth on the Red Wings in controlled exits when he played there. Strong defensive impacts, strong work in the offensive zone, but average-ish in getting the puck out with control. I think that’s in part why he doesn’t do well with Soucy.

If the team is ok continuing to play Myers with Hughes, it’s why I think him and Pettersson could make a good pairing.

 

[pitseleh]

I am curious what a JFresh card would have said about the Sedins back in the day when they were running 55-60% xGF rates together but 40-45% xGF rates when apart.

 

[Bleach Clean]

I mostly just think once you’re slicing and dicing numbers below 20 game samples it starts to get pretty noisy and I would lean on the eye test more heavily. He’s got great numbers outside of a handful of games when he was coming back from injury and playing with arguably the worst defender on the team. I’m inclined to chalk that up to health/partner.

Fair enough, I would prefer the bigger samples (31 and 81 games from last year) and not controlling for his worst partner(s).

The eye test shows me he's not generating enough offensive events that would suggest he's a strong #3 Dman carrying a weaker partner. What I see is Forbort has stabilized the bottom end of his play, unlike Soucy, which has allowed him to get close to even on the 2nd pair, but still not a net positive.

What I didn't account for is non-standard partners like Brannstrom and Myers (primarily) buoying his xGF% in the 1/3rd block of minutes that does not include Hughes, Soucy and Forbort.

The problem with last year is that he was clearly injured down the stretch. I don’t know exactly when he got hurt but picking an arbitrary date he posted a 49% xGF rate in 251 minutes without Hughes before January 31 and a 41% xGF rate in 111 minutes without Hughes after January 31.

Then factor in he got 31% ozone starts when he wasn’t playing with Hughes vs 65% when he was. I think zone starts tend to be an overrated factor but when you have big splits like that it is going to knock down your percentages by a couple percentage points.

Is that noise, a drop off in performance, something else? Personally, watching him I didn’t think he was great last year, and really wished we could have seen more with him away from Hughes to make a better judgment on him. I think the JFresh chart is a bit tough on him due to the injury impact. But I’d agree it is directionally correct.

But this season I’ve thought he’s largely been very good and don’t have the same concerns I did last year.

A #3 Dman is my bar for him away from Hughes. In that regard, he hasn't met expectations last year or this year. When he's looked good with Forbort, he's sawing off minutes with him (44.83 xGF%), not leading that pair forward. That's my concern.

Even if that JFresh chart is hard on him due to injury, it lightens the critique on him too by incorporating his minutes with Hughes... 4/5ths of his minutes buoyed by the 65% Ozone starts you reference, still 58% WAR rating and below the 50% EV threshold in Offense and Defense... Not good. (But thank you for at least acknowledging the chart)

This year and last year look pretty similar in terms of non-Hughes minute xGF% (incorporating Soucy for both). His counting stat rate is dropped to the lowest in 5 years too. And so, the eye test is the last best argument for him... And I don't think that's enough.

We should all expect his play to bear out in the numbers across the two years.

 

[ExarKun80]

The heavy lifting on D is . . . mission accomplished

The heavy lifting on top6 C has . . . just begun

There is also a need for a serious winger. What we have now is not going to cut it. Brock and DeBrusk are too streaky. You can't rely on guys who score in bunches for 4-5 games and then go 10-15 without a goal.

Using last year as any sort of evaluation is a mistake since it was an aberation just like Jimbo's bubble. Management fell for it and downgraded the defence thinking that our offence+Demko will make up for it and here we are.

 

[pitseleh]

Fair enough, I would prefer the bigger samples (31 and 81 games from last year) and not controlling for his worst partner(s).

The eye test shows me he's not generating enough offensive events that would suggest he's a strong #3 Dman carrying a weaker partner. What I see is Forbort has stabilized the bottom end of his play, unlike Soucy, which has allowed him to get close to even on the 2nd pair, but still not a net positive.

Interestingly the microdata JFresh posted showed someone who is providing strong offensive creation impacts in the offensive zone, just poor transition impacts.

A #3 Dman is my bar for him away from Hughes. In that regard, he hasn't met expectations last year or this year. When he's looked good with Forbort, he's sawing off minutes with him (44.83 xGF%), not leading that pair forward. That's my concern.

Even if that JFresh chart is hard on him due to injury, it lightens the critique on him too by incorporating his minutes with Hughes... 4/5ths of his minutes buoyed by the 65% Ozone starts you reference, still 58% WAR rating and below the 50% EV threshold in Offense and Defense... Not good. (But thank you for at least acknowledging the chart)

I don’t follow your reasoning. As far as I know we don’t know how JFresh “lightens the critique by incorporating his minutes with Hughes”. My experience with is that in these types of situations the model JFresh relies on weighs those minutes away from Hughes very heavily and does not give much credit for the minutes with Hughes.

I think that’s pretty clear from the chart when you consider that he has the trendline for Hronek as being around the 10th percentile for defence in 2023/24. You can’t get there from his numbers unless your model assumes the results with Hughes were almost entirely due to Hughes.

This year and last year look pretty similar in terms of non-Hughes minute xGF% (incorporating Soucy for both). His counting stat rate is dropped to the lowest in 5 years too. And so, the eye test is the last best argument for him... And I don't think that's enough.

We should all expect his play to bear out in the numbers across the two years.

But which numbers? Dom Lucyzyzyn’s model at the Athletic had Hronek as the 30th best defender in the league last year. Even this year with the time missed he’s tied with a bunch of players in the 35-45 range. That seems pretty consistent with how people view him.

Even under the model JFresh uses Hronek would be in the 80-95 range for the league (based on the 58th percentile and assuming he has 6-7 defenders per team meeting his minutes cutoff), which puts Hronek as a low end #3 or high end #4.

So you have a strong #2 on one end and a high end #4 on the other, which seems to be around the range of debate here.

 

[vancityluongo]

tangentially related but i once talked to the owner of a dairy queen in the neighbourhood that laurence gilman lived in, and he said his theory was that the canucks lost to the bruins because they started having kids and they couldn't focus during the playoffs. i thought he was insane at the time, but maybe he's right.

this rules hahaha

laurence gilman's neighbourhood dq owner >>>> joe sakic's bbq salesman

 

[Ernie]

I just don't think you can evaluate a defenceman who is shuffling around the lineup like he does when he's not with Hughes. He has to play an entirely different game when he's with Hughes vs when he's running his own pairing.

He was re-signed to caddy Hughes after he helped elevate Hughes to the Norris trophy. Can a cheaper player replicate most of what he does with Hughes, and can he be the play driving half of a 2nd pairing? Perhaps it's worth finding out, but it's more than likely that the best way for him to move the needle is to play the role that he has been successful in on this team.

 

[theguardianII]

Hughes still out. Many top players suffered through a sports hernia and kept playing. They aren't always need a stretcher situation.
You know there must be something because his injury is being downplayed by Dhali.

Could be just preparation to sit out the 4 Nations thing but it could also be him making sure he can play with his brother(s) in the 4 Nations. His first chance really

 

[Bleach Clean]

Interestingly the microdata JFresh posted showed someone who is providing strong offensive creation impacts in the offensive zone, just poor transition impacts.



I don’t follow your reasoning. As far as I know we don’t know how JFresh “lightens the critique by incorporating his minutes with Hughes”. My experience with is that in these types of situations the model JFresh relies on weighs those minutes away from Hughes very heavily and does not give much credit for the minutes with Hughes.

I think that’s pretty clear from the chart when you consider that he has the trendline for Hronek as being around the 10th percentile for defence in 2023/24. You can’t get there from his numbers unless your model assumes the results with Hughes were almost entirely due to Hughes.

But which numbers? Dom Lucyzyzyn’s model at the Athletic had Hronek as the 30th best defender in the league last year. Even this year with the time missed he’s tied with a bunch of players in the 35-45 range. That seems pretty consistent with how people view him.

Even under the model JFresh uses Hronek would be in the 80-95 range for the league (based on the 58th percentile and assuming he has 6-7 defenders per team meeting his minutes cutoff), which puts Hronek as a low end #3 or high end #4.

So you have a strong #2 on one end and a high end #4 on the other, which seems to be around the range of debate here.

With JFresh's model, it would be 6 defenders per team (192), and a 58th percentile rank of that would be 111th no? And that's with Hughes. In other words, only 57% of scores would be lower than his score, despite playing with Hughes in 4/5ths of his minutes in 2023-24. Correct?

(Edit: Is the min cut off you propose the same as what JFresh is using?)

I saw those microstats. Fair point about poor transition impacts (I've been saying he defers too much). Mixed results in DET too. His DZone impacts in VAN (denials, retrievals and preventions) don't favour him either, but he was very good once in the Ozone. That aligns with how I see him play.

On gauging him as a strong #3 dman: Last year his impact sans Hughes would have been around 181st in terms of xGF%. This year it's 166th. All Hughes min last year at 70th (Strong #2 Dman, as you said), and Hughes mind this year 52nd (#1 Dman). That's a #1 Dman to almost a bottom end #6, all turning on Hughes. That's such a wide range. I'm looking for 64-96th ranked impacts without Hughes. Is that unfair?

If your read is that he was a high end #4 last year, why did his play not seem very good to you? (eye test?)

 

[Bleach Clean]

I just don't think you can evaluate a defenceman who is shuffling around the lineup like he does when he's not with Hughes. He has to play an entirely different game when he's with Hughes vs when he's running his own pairing.

He was re-signed to caddy Hughes after he helped elevate Hughes to the Norris trophy. Can a cheaper player replicate most of what he does with Hughes, and can he be the play driving half of a 2nd pairing? Perhaps it's worth finding out, but it's more than likely that the best way for him to move the needle is to play the role that he has been successful in on this team.

No doubt, I agree it's best for him to be with Hughes. For the team, however, he's best played away from Hughes to run his own pair if Myers can mimic his impact. So far, Myers has done this:

In 344 min this year with Hughes, Myers is at 58.98 cGF%
In 402 min this year with Hughes, Hronek is at 57.12 xGF%

Last year Myers edged out Hronek's performance again with Hughes, but it was only across 159 min (too disparate).

Him driving the 2nd pair, even though he hasn't really done it, is worth exploring. The potential payoff could be immense.

 

[Jyrki21]

Other than Pittsburgh which was obviously a non-option given the mess there, Buffalo is the closest NHL market to his home town and are a reasonably promising young team where he would have come in as 'the guy' which might have appealed to him. It's still probably a no but it isn't something I would have ruled out entirely and if I were Buffalo I would have been moving mountains to make it happen and convince him. And that certainly didn't seem to happen.

Columbus is actually the next-closest market to Pittsburgh, and they'd have the added bonus of being able to advertise him as an Ohio native. And in spite of everything, they seem to be headed in the right direction -- it seems to me they really could have used him too.

 

[pitseleh]

With JFresh's model, it would 6 defenders per team (192), and a 58th percentile rank of that would be 111th no? And that's with Hughes. In other words, only 57% of scores would be lower than his score, despite playing with Hughes in 4/5ths of his minutes in 2023-24. Correct?

You have the math backwards. If there are 192 defenders, 57% (109) would be below him and at the same level, making him 81-83 in the league (since 2-3 players will be at the same percentile).

On gauging him as a strong #3 dman: Last year his impact sans Hughes would have been around 181st in terms of xGF%. This year it's 166th. All Hughes min last year at 70th (Strong #2 Dman, as you said), and Hughes mind this year 52nd (#1 Dman). That's a #1 Dman to almost a bottom end #6, all turning on Hughes. That's such a wide range. I'm looking for 64-96th ranked impacts without Hughes. Is that unfair?

I don’t think it’s an unreasonable expectation, but recognizing:

- in doing so you are only measuring the quality of defender by even strength play.

- the data has not been adjusted for zone starts and quality of teammates/competition, which may have a small but material impact. I wouldn’t be surprised if adjusting to neutral zone starts alone would add 1-2 percentage points to Hronek’s xGF ratio based on the numbers I have seen on Corsi impacts from zone starts.

- that in the sample sizes you’re looking at the data may not with a high level of confidence show an absolute ranking.

If your read is that he was a high end #4 last year, why did his play not seem very good to you? (eye test?)

My not very good comment was relative to expectations - to me he looked more like a 3-4 defender who had a hot few months with Hughes and should be paid $1.5-2 million less, not a $7.25 million top pairing defender.