Proposal: TBL-MIN

[Bazeek]

View attachment 268111

Last year Sergachev was way better and he is a lot younger so it's hard to believe Brodin is any better. Sergachev has a false reputation as a bad defensive dman though so I can understand why you would think Brodin is better.

Do you know who Sergachev was typically played against?

Brodin-Spurgeon took a lot of the hardest minutes last year up until Dumba's injury. I can't remember how the usage changed after that, but from January on he was playing similarly difficult minutes with Pateryn, who's okay but not really a 2nd pairing guy.

 

[AKL]

Do you know who Sergachev was typically played against?

Brodin-Spurgeon took a lot of the hardest minutes last year up until Dumba's injury. I can't remember how the usage changed after that, but from January on he was playing similarly difficult minutes with Pateryn, who's okay but not really a 2nd pairing guy.

5th highest time on ice per game on the roster, 2nd highest powerplay time on ice, 8th highest penalty kill time on ice. I think we can assume he’s getting much easier minutes.

 

[Bazeek]

5th highest time on ice per game on the roster, 2nd highest powerplay time on ice, 8th highest penalty kill time on ice. I think we can assume he’s getting much easier minutes.

Probably, though the numbers in that chart appear to be even-strength only. I'm not even sure how much the usage/quality of competition stuff matters, I just know Brodin was playing pretty thankless minutes on a struggling team last year. His numbers this year look ugly too, but for the most part he's been very good.

I'm sure Sergachev generates more offense regardless though, which may have been the point he was making.

 

[AKL]

Probably, though the numbers in that chart appear to be even-strength only. I'm not even sure how much the usage/quality of competition stuff matters, I just know Brodin was playing pretty thankless minutes on a struggling team last year. His numbers this year look ugly too, but for the most part he's been very good.

I'm sure Sergachev generates more offense regardless though, which may have been the point he was making.

What I meant was, I think based on the minutes he gets, it’s not a huge leap to assume he’s not playing against other teams top lines. Brodin likely gets much tougher QoC.

 

[Bazeek]

What I meant was, I think based on the minutes he gets, it’s not a huge leap to assume he’s not playing against other teams top lines. Brodin likely gets much tougher QoC.

Fair enough, and you're probably right.

 

[AndreRoy]

some of us are NOT worried on cap cause simply some guys gonna move whether we win cup or not expect 2 guys forwards outta here next year plain and simple

If you’re not worried then you’re not paying attention.

Who are we going to move out? The only high-priced forward without a full NTC (aside from guys like Point and Kuch who we obviously aren’t going to trade) will be Killorn as of July 1st. He’ll have a 16-team trade list so he can eliminate half the league as potential options right off the bat; then there’s the difficult matter of finding a team on his list that a) has the cap space to take him, b) is willing to spend that cap space on a 31-year old middle six forward who is a third-liner on a good team, and c) is willing to do so with minimal salary coming back our way. That will not be easy considering most of the teams on his list will likely be contenders, which generally don’t have a ton of spare cap space or an interest in taking on a big contract for futures. Most of the teams on Killorn’s list will be in the same boat that we’re in: trying to ship out contracts like his.

But let’s assume that we are able to find a taker for Killorn. After all, he might view an opportunity to play in the top six as more important than playing for a contender, which if so would increase our chances of finding a suitable trade partner. We then have to somehow convince one of Palat, Gourde, or Johnson to waive his full NTC, and none of them has any reason to do so. Put yourself in their shoes for a moment and then tell me what any of them has to gain by agreeing to be traded from Tampa. Go ahead, I’ll wait.

 

[DistantThunderRep]

Do you know who Sergachev was typically played against?

Brodin-Spurgeon took a lot of the hardest minutes last year up until Dumba's injury. I can't remember how the usage changed after that, but from January on he was playing similarly difficult minutes with Pateryn, who's okay but not really a 2nd pairing guy.

Probably, though the numbers in that chart appear to be even-strength only. I'm not even sure how much the usage/quality of competition stuff matters, I just know Brodin was playing pretty thankless minutes on a struggling team last year. His numbers this year look ugly too, but for the most part he's been very good.

I'm sure Sergachev generates more offense regardless though, which may have been the point he was making.

What I meant was, I think based on the minutes he gets, it’s not a huge leap to assume he’s not playing against other teams top lines. Brodin likely gets much tougher QoC.

Welcome to the CCman no context chart. He is famous on TBL Forums for posting a chart with zero context and basing all facts around it. Brodin had harder minutes all season than Serge. Serge was mostly used on our third pairing last season with Girardi or Coburn, depending on injuries. He was the most successful paired with Coburn last year. He was on the 2nd PP unit which would roughly see about 30 seconds every PP we got. It's understandable that Brodin's stats were lower solely based on deployment.

 

[57special]

Brodin is a defensive Dman with exceptional skating skills, defend line rushes, and has a great ability to retrieve and move the puck out of his zone. He seldom plays on the PP, partly because of a lack of offensive imagination, and partly because MN is chock full of excellent offensive Dmen. Not a guy who is going to fill up the stats sheet, but he is the guy you put out to defend against McDavid.

His game does not seem to resemble Sergachev's very much. Any team trading for him would presumably know that they are not getting a Tyson Barrie type.

 

[VoluntaryDom]

Two completely different play styles. Brodin has been a top suppression player for awhile. If you think an 8 game sample size proves that Sergachev is better then that is your opinion.

:shrug:

 

[Bazeek]

View attachment 268309
:shrug:

I assume this was posted to back up the point he was making?

 

[VoluntaryDom]

if we average goals for impact and expected goals for impact to estimate offense, and use expected goals against impact to estimate defense, by RAPM, here is how sergachev and brodin shake out in the 2018-19 season.

brodin comes out to a -.198 offense per 60, with minus (minus being good on defense) .223 on defense. this means he is a +.025 player per 60 mins, or slightly above average.

sergachev is much better on offense. his xGF impact is less than his GF and Corsi impacts which hurts him a bit here but including GF shows his offensive talent. sergachev ends up being a +.0565 on offense by this method. his defense is not bad either. hes a -.063 there using that method, remember that minus is good. overall he comes out to be +.1195, which is noticeably above average and starting to reach top pair territory. both of them played well over 1000 minutes at ES, serg about 1200 and brodin over 1450. these stats are per 60, so i have to take that into account. by this measurement, sergachev added 2.4 goals above an average player (note that average is above replacement level so this isnt GAR), while brodin added 0.6. brodin has a larger sample of performing above average so its fine if you think its closer than that, but i dont really see how brodin is so clearly better than sergachev.

 

[VoluntaryDom]

I assume this was posted to back up the point he was making?

if we average goals for impact and expected goals for impact to estimate offense, and use expected goals against impact to estimate defense, by RAPM, here is how sergachev and brodin shake out in the 2018-19 season.

brodin comes out to a -.198 offense per 60, with minus (minus being good on defense) .223 on defense. this means he is a +.025 player per 60 mins, or slightly above average.

sergachev is much better on offense. his xGF impact is less than his GF and Corsi impacts which hurts him a bit here but including GF shows his offensive talent. sergachev ends up being a +.0565 on offense by this method. his defense is not bad either. hes a -.063 there using that method, remember that minus is good. overall he comes out to be +.1195, which is noticeably above average and starting to reach top pair territory. both of them played well over 1000 minutes at ES, serg about 1200 and brodin over 1450. these stats are per 60, so i have to take that into account. by this measurement, sergachev added 2.4 goals above an average player (note that average is above replacement level so this isnt GAR), while brodin added 0.6. brodin has a larger sample of performing above average so its fine if you think its closer than that, but i dont really see how brodin is so clearly better than sergachev.

when you consider the numbers that those bars represent it becomes clear that sergachev had the better season. a similar SD in goals for is a larger value than in expected goals against since there are many players who drive goals for at a rate that no one can drive expected goals against. this is why i value offense and defense equally but forwards and some dmen can impact offense more than anyone not named mikko koivu or nik hjalmarsson can impact defense.

 

[VoluntaryDom]

Do you know who Sergachev was typically played against?

Brodin-Spurgeon took a lot of the hardest minutes last year up until Dumba's injury. I can't remember how the usage changed after that, but from January on he was playing similarly difficult minutes with Pateryn, who's okay but not really a 2nd pairing guy.

RAPM adjusts for quality of competition

 

[Bazeek]

if we average goals for impact and expected goals for impact to estimate offense, and use expected goals against impact to estimate defense, by RAPM, here is how sergachev and brodin shake out in the 2018-19 season.

brodin comes out to a -.198 offense per 60, with minus (minus being good on defense) .223 on defense. this means he is a +.025 player per 60 mins, or slightly above average.

sergachev is much better on offense. his xGF impact is less than his GF and Corsi impacts which hurts him a bit here but including GF shows his offensive talent. sergachev ends up being a +.0565 on offense by this method. his defense is not bad either. hes a -.063 there using that method, remember that minus is good. overall he comes out to be +.1195, which is noticeably above average and starting to reach top pair territory. both of them played well over 1000 minutes at ES, serg about 1200 and brodin over 1450. these stats are per 60, so i have to take that into account. by this measurement, sergachev added 2.4 goals above an average player (note that average is above replacement level so this isnt GAR), while brodin added 0.6. brodin has a larger sample of performing above average so its fine if you think its closer than that, but i dont really see how brodin is so clearly better than sergachev.

I'll ask the dumb question then, because it's not clear to me from the chart itself: are GF/60, xGF/60, CF/60, xGA/60 and CA/60 individual numbers or "on ice" numbers?

 

[VoluntaryDom]

I'll ask the dumb question then, because it's not clear to me from the chart itself: are GF/60, xGF/60, CF/60, xGA/60 and CA/60 individual numbers or "on ice" numbers?

neither in this case. they are RAPM numbers, which use on-ice numbers and statistical regression to isolate the impact of an individual player. sort of similar to RPM in basketball if you are familiar with that.

 

[Bazeek]

neither in this case. they are RAPM numbers, which use on-ice numbers and statistical regression to isolate the impact of an individual player. sort of similar to RPM in basketball if you are familiar with that.

It may have been more effective to point that out than going straight to the mic-drop

My stats are rusty enough to not be able to critique their approach, but I do wonder how well it adjusts for the 2018-19 Lightning being in god mode for the regular season while the 2018-19 Wild were... not.

And just to be clear: I'm mostly agnostic on who's better right now. There were similar debates among Wild fans about Brodin vs Dumba for a few years and I tended to favor Dumba for the same reasons that you seem to favor Sergachev. I guess I lose patience with the "Bar Chart, deal with it!" posts.

 

[Merrrlin]

when you consider the numbers that those bars represent it becomes clear that sergachev had the better season. a similar SD in goals for is a larger value than in expected goals against since there are many players who drive goals for at a rate that no one can drive expected goals against. this is why i value offense and defense equally but forwards and some dmen can impact offense more than anyone not named mikko koivu or nik hjalmarsson can impact defense.

Do those charts adjust for Quality of Teammates?

I'd be really curious to see how a change of teams could impact a players stats. Maybe a guy like Duchene going from awful Ottawa to contender CLB would be a good case study one day.

 

[DistantThunderRep]

Do those charts adjust for Quality of Teammates?

I'd be really curious to see how a change of teams could impact a players stats. Maybe a guy like Duchene going from awful Ottawa to contender CLB would be a good case study one day.

Of course it doesn't. Meet our second resident chart dropper, not know what he's talking about talker, Dom. Basically, Dom sees the chart and will create reasons that end up having as many holes as Swiss cheese.

The charts don't take into account quality of teammates or deployment. It's just all data without context.

Serge is on a better team, deployed in a sheltered manner last season, and played less minutes in high danger situations. To notice most of this, you would need to watch the game and pay attention, not look at a chart and numbers.

 

[DistantThunderRep]

when you consider the numbers that those bars represent it becomes clear that sergachev had the better season. a similar SD in goals for is a larger value than in expected goals against since there are many players who drive goals for at a rate that no one can drive expected goals against. this is why i value offense and defense equally but forwards and some dmen can impact offense more than anyone not named mikko koivu or nik hjalmarsson can impact defense.

Are you seriously going to point out the most obvious thing like it's some massive discovery you made? Nothing you said here requires anymore than watching a hockey game and understanding the rules.

 

[Bazeek]

Do those charts adjust for Quality of Teammates?

I'd be really curious to see how a change of teams could impact a players stats. Maybe a guy like Duchene going from awful Ottawa to contender CLB would be a good case study one day.

The RAPM numbers attempt to take all sorts of things into account: strength of schedule, linemates, score effects, etc. The problem is that the model is such a cocktail of statistical methods that it's hard to critique on a message board. It's fun to try to get your head around it, but fans throw the numbers/charts out as authoritative arguments without being able to think critically about the underlying math. It "accounts" for quality of teammates, but how does it do that? Does it do it well? How do we know?

Someone once gave me the following rule for statistical models like this: before applying the model you should be able to name a.) 5 things the model is good at, b.) 5 things the model is bad at, and c.) 5 things the model is utterly useless at. That's probably overkill for a a hockey board, but 2 of each would be nice.

 

[Merrrlin]

The RAPM numbers attempt to take all sorts of things into account: strength of schedule, linemates, score effects, etc. The problem is that the model is such a cocktail of statistical methods that it's hard to critique on a message board. It's fun to try to get your head around it, but fans throw the numbers/charts out as authoritative arguments without being able to think critically about the underlying math. It "accounts" for quality of teammates, but how does it do that? Does it do it well? How do we know?

Someone once gave me the following rule for statistical models like this: before applying the model you should be able to name a.) 5 things the model is good at, b.) 5 things the model is bad at, and c.) 5 things the model is utterly useless at. That's probably overkill for a a hockey board, but 2 of each would be nice.

Well said!

That's why it's so hard to dig into statistical models around here. We're like a bunch of farmers discussing weather patterns...we know little bits about it, we know some fancy terminology, but it's still incredibly shallow and the devil is in the details. It's especially apparent when I see someone try to compare a guy like Point to Barzal...they are playing in environments that are completely impossible to normalize.

 

[TropicalFruitGirl2]

As both a Lightning and Wild fan, I would love for the two teams to be able to do some dealings that would be mutually beneficial.
In fact, I've often wondered why the teams never made deals together before.

Bolts seem to always have a surplus of forwards, while Minnesota doesn't, and the Wild seem to have reasonably sound defensive type players come up...an area where the Bolts seem to struggle historically: drafting and developing solid quality NHL defensive type players...most of the Bolts best defensemen have come from elsewhere, Victor Hedman being the exception of course.

Anyways, the way things are going for both teams, and as much as I would like to see a deal between the two clubs, I honestly don't feel Tampa Bay and Minnesota make particularly good trade partners.
Sure, there is PLENTY on the Lightning the Wild would want and could use, but sadly, I don't feel the Lightning would benefit nearly as much from what the Wild have to offer.

Minnesota currently has many average to slightly above average prospects, and no one at the NHL level with a reasonable contract, is worth acquiring at the cost of picks or players for the Lightning, IMO.
Brodin is a solid defender, but that's it.
I actually think he would fit in well with the Lightning, but his salary, as some have already pointed out, is too much for what his role would be.
He simply doesn't bring much else other than blocking shots and stick checking...and let's face it, that's his whole game.
Not much in the way of physicality, not much in the way of offense....I am pretty sure the Bolts can get from somewhere a player to do what Brodin does at half his cost.

I just think most trades involving the two clubs would be heavily in favor of the Wild and the Lightning would have no real reason to deal with them because it doesn't bring any great benefits to them.

Zucker? I like Zucker.
But the Bolts have about a dozen "Zuckers" between the NHL club and the AHL club, and probably a few in Europe or in Juniors.

Bolts need solid young two way defenders who project to be at least 2nd pairing guys...and the Wild don't have any to offer that fit that billing.
Brodin certainly doesn't fit that. Dumba might, but the Wild would be fools to part with him...and then of course, there is Dumba's current 6M price tag which the Bolts likely won't be able to fit and keep other players as well.

I don't see a deal to be made between the two.

 

[VoluntaryDom]

Of course it doesn't. Meet our second resident chart dropper, not know what he's talking about talker, Dom. Basically, Dom sees the chart and will create reasons that end up having as many holes as Swiss cheese.

The charts don't take into account quality of teammates or deployment. It's just all data without context.

Serge is on a better team, deployed in a sheltered manner last season, and played less minutes in high danger situations. To notice most of this, you would need to watch the game and pay attention, not look at a chart and numbers.

uhh are u unfamiliar with RAPM? it DOES adjust for context such as zone starts, teammates, and competition. it is an isolated impact metric not an on ice one

 

[VoluntaryDom]

The RAPM numbers attempt to take all sorts of things into account: strength of schedule, linemates, score effects, etc. The problem is that the model is such a cocktail of statistical methods that it's hard to critique on a message board. It's fun to try to get your head around it, but fans throw the numbers/charts out as authoritative arguments without being able to think critically about the underlying math. It "accounts" for quality of teammates, but how does it do that? Does it do it well? How do we know?

Someone once gave me the following rule for statistical models like this: before applying the model you should be able to name a.) 5 things the model is good at, b.) 5 things the model is bad at, and c.) 5 things the model is utterly useless at. That's probably overkill for a a hockey board, but 2 of each would be nice.

i’ve looked into multiple isolated impact models (such as EW GAR, Corsica GAR, and HockeyViz threat) and have come to the conclusion that RAPM is the best one. the main problems with it, if there are any, are in EW’s xG model itself. modern xG models are far from perfect. at least it isn’t like Corsica’s xG model which way overvalues the crease and undervalued the slot, corsica’s GAR which correlates more with PDO than anything else, and hockeyviz threat which thinks every forward is wayne gretzky and every dman is robert hagg.

 

[LTIR Trickery]

View attachment 268111

Last year Sergachev was way better and he is a lot younger so it's hard to believe Brodin is any better. Sergachev has a false reputation as a bad defensive dman though so I can understand why you would think Brodin is better.

Sergachev got some very primo, very sheltered minutes.