Player Discussion Lars Eller

[Mr Jackpot]

so what do you guys think of Larry?

 

[Hivemind]

so what do you guys think of Larry?

Hasn't been a particularly notable player so far. Very little production. But aside from some early penalties, he's not been bad either. He's mostly doing the little things right (particularly in terms of shot suppression), but hasn't found line mates that he can generate with offensively. Can quietly eat minutes and keep the puck on the right half of the ice, but isn't doing much to help with the depth scoring issues in terms of generating for himself or for his line mates.

I do wish he was winning more of his faceoffs. 47.9% total on the year, and only 36.6% shorthanded. Other than the faceoff circle, he's probably been the team's best PKing forward in terms of limiting goals against.

 

[Calicaps]

Hasn't been a particularly notable player so far. Very little production. But aside from some early penalties, he's not been bad either. He's mostly doing the little things right (particularly in terms of shot suppression), but hasn't found line mates that he can generate with offensively. Can quietly eat minutes and keep the puck on the right half of the ice, but isn't doing much to help with the depth scoring issues in terms of generating for himself or for his line mates.

I do wish he was winning more of his faceoffs. 47.9% total on the year, and only 36.6% shorthanded. Other than the faceoff circle, he's probably been the team's best PKing forward in terms of limiting goals against.

Beagle says hi!

 

[Hivemind]

Beagle says hi!

Beagle has given up 7.47 GA/60 while shorthanded. Eller has given up only 5.72 GA/60.

It's still early in the season for special teams rates (as they're a few minutes per night job), so aren't completely stable yet and single goals can still nudge them pretty far. However three year samples for both players also favor Eller when compared against Beagle and when compared against Eller's former Montreal teammates. From 2013-16, Eller gave up only 4.58 GA/60 while shorthanded. The next best Habs were Pacioretty (5.57) and Plekanec (5.80). Among players with at least 300 PK minutes since 2013, Eller has the 2nd lowest GA/60 among forwards in the entire NHL, behind only Zubrus. He's an elite penalty killer.

 

[trick9]

Hasn't been a particularly notable player so far. Very little production. But aside from some early penalties, he's not been bad either. He's mostly doing the little things right (particularly in terms of shot suppression), but hasn't found line mates that he can generate with offensively. Can quietly eat minutes and keep the puck on the right half of the ice, but isn't doing much to help with the depth scoring issues in terms of generating for himself or for his line mates.

I do wish he was winning more of his faceoffs. 47.9% total on the year, and only 36.6% shorthanded. Other than the faceoff circle, he's probably been the team's best PKing forward in terms of limiting goals against.

I don't know. Beagle takes most of the toughest PK minutes because of the FO ability you were talking about. He takes twice as many FO's on the PK than Eller who often comes on when the puck has already been cleared out of the Capitals D zone. Eller is a great PK'er but face-offs have huge impact on the PK. Jay Beagle is 2nd in the NHL in FO's on the PK with 59.1%, only behind Antoine Vermette (65%) who takes way less draws.

Not sure what's going on with Backstrom there. He has only won 3 of the 21 FO's on the PK for 14.3%. Trotz doesn't really even trust him there (FO's, not PK in general) anymore unless either Eller or Beagle is on the box.

 

[HecticGlow]

Beagle has given up 7.47 GA/60 while shorthanded. Eller has given up only 5.72 GA/60.

It's still early in the season for special teams rates (as they're a few minutes per night job), so aren't completely stable yet and single goals can still nudge them pretty far. However three year samples for both players also favor Eller when compared against Beagle and when compared against Eller's former Montreal teammates. From 2013-16, Eller gave up only 4.58 GA/60 while shorthanded. The next best Habs were Pacioretty (5.57) and Plekanec (5.80). Among players with at least 300 PK minutes since 2013, Eller has the 2nd lowest GA/60 among forwards in the entire NHL, behind only Zubrus. He's an elite penalty killer.

He must play fewer short handed minutes than Beags though, both because Beags is always on the ice first, and because Eller has so often been in the penalty box...

Presumably, though, Beagle is facing more of teams' first power play units than Eller, which might also make a difference. Either way Beagle + Winnik/Oshie and Eller + Wilson make strong PK units.

 

[artilector]

Eller seems like a serviceable, responsible player who does not drive play. Nothing to write home about so far, given his position/salary.

 

[Hivemind]

He must play fewer short handed minutes than Beags though, both because Beags is always on the ice first, and because Eller has so often been in the penalty box...

Presumably, though, Beagle is facing more of teams' first power play units than Eller, which might also make a difference. Either way Beagle + Winnik/Oshie and Eller + Wilson make strong PK units.

I don't know. Beagle takes most of the toughest PK minutes because of the FO ability you were talking about. He takes twice as many FO's on the PK than Eller who often comes on when the puck has already been cleared out of the Capitals D zone. Eller is a great PK'er but face-offs have huge impact on the PK. Jay Beagle is 2nd in the NHL in FO's on the PK with 59.1%, only behind Antoine Vermette (65%) who takes way less draws.

I'd buy those explanations a lot more if Eller didn't have such a strong track record of PKing in Montreal, as well. And there's no reason Eller couldn't be deployed in a left/right combo that allowed him to take draws on his strong side (he's 8.8% better against righties on the dot this season).

And while faceoffs are important, just a reminder that we're talking about a difference of an event that happens only 3.6 times per game (for Beagle) and 1.8 times per game (for Eller) while shorthanded. A 20% difference in an event that happens fewer than 4 times per game is a relatively minor shift in the actual outcome. The difference comes out to be less than 1 faceoff difference per game. The remainder of play can more than make-up for that.

 

[txpd]

Eller is 4th among forwards in pk/toi average. Beagle averages nearly a full minute more per game on the pk. 30-40% more. That doesn't sounds a lot like the best Pk'r on the team to me.

Eller was brought in to push Beagle to 4 and improve the production in that 3C hole. So far Eller as failed at that in almost every phase of the game.

I can write that off to being on a new team that is quite a change from where he came from. But he is not close to better than Beagle right now at anything

 

[HecticGlow]

I'd buy those explanations a lot more if Eller didn't have such a strong track record of PKing in Montreal, as well. And there's no reason Eller couldn't be deployed in a left/right combo that allowed him to take draws on his strong side (he's 8.8% better against righties on the dot this season).

And while faceoffs are important, just a reminder that we're talking about a difference of an event that happens only 3.6 times per game (for Beagle) and 1.8 times per game (for Eller) while shorthanded. A 20% difference in an event that happens fewer than 4 times per game is a relatively minor shift in the actual outcome. The difference comes out to be less than 1 faceoff difference per game. The remainder of play can more than make-up for that.

Oh I'm not doubting he and Wilson have made a strong PKing duo - not at all. Just pointing out he's probably not that much better than our other PKers, who come at considerably less cost. His 5-on-5 with the Caps is considerably worse than it was with Montreal.

 

[Hivemind]

Eller is 4th among forwards in pk/toi average. Beagle averages nearly a full minute more per game on the pk. 30-40% more. That doesn't sounds a lot like the best Pk'r on the team to me.

Eller was brought in to push Beagle to 4 and improve the production in that 3C hole. So far Eller as failed at that in almost every phase of the game.

I can write that off to being on a new team that is quite a change from where he came from. But he is not close to better than Beagle right now at anything

Playing more minutes doesn't mean you are better, it means you are used more. That can be for any variety of reasons, from coach familiarity, to coach biases, to saving a player for other minutes, to comfort with partners.

Eller has been, without a doubt, the most effective penalty killing forward on the team. That's not a subject for debate. The Caps give up fewer goals against while he's out there than any other forward. This matches up with his excellence at PKing in Montreal. Lars Eller is an elite penalty killer, and is better than Jay Beagle at preventing the opposition from scoring goals while the Caps are shorthanded.

 

[txpd]

I can buy that. Trotz chooses Beagle over Eller because he is either biased for Beagle or misuses his roster because he doesn't know better.

Bad coaching.

 

[Hivemind]

Biases don't have to be conscious (and most often they are not). Beagle is a higher visibility player than Eller, that's for sure. We all remember Beagle's impressive forechecking on the PK, driving opponents back into their own zone. Plays like that could cause a coach to favor Beagle. Alternatively, Trotz could be highly valuing Beagle's faceoff ability. There are numerous reasons why Trotz may value Beagle more. It doesn't mean that Eller is less effective than Beagle, though.

 

[SpinningEdge]

Elmer was brought in to help scoring depth. He's failed terribly at it.

It's great he's doing well on the PK but given his salary we don't need another Brooks Laich and that's what Eller is acting like. There's a reason Laich was shipped out. No production for the money he's getting paid.

Eller needs to improve offensively and I hope he does. A lot of players on this team on the bottom 6 have been extremely disappointing offensively.

 

[AlexBrovechkin8]

He's been more noticeable in the offensive zone as of late. I think the points will come.

 

[Hivemind]

I view Eller as more of Joel Ward in his early tenure with the Capitals than late tenure Brooks Laich. Eller is doing a lot of things right, but he's just not on the scoreboard. I think the points will come eventually if they're patient with him.

 

[Jags]

Biases don't have to be conscious (and most often they are not). Beagle is a higher visibility player than Eller, that's for sure. We all remember Beagle's impressive forechecking on the PK, driving opponents back into their own zone. Plays like that could cause a coach to favor Beagle. Alternatively, Trotz could be highly valuing Beagle's faceoff ability. There are numerous reasons why Trotz may value Beagle more. It doesn't mean that Eller is less effective than Beagle, though.

You're still glossing over the fact that Eller is deployed on the 2nd PK unit, meaning he's way more often deployed versus our opponent's 2nd PP unit. The degree of difficulty is significantly easier. The same was true in Montreal -- 2nd unit minutes, sub-50% on faceoffs.

Beagle is a first-unit PKer that went 56% in faceoffs last year versus our opponents' best. Most faceoffs won lead to clears, and most clears take 15+ seconds off the clock. And again, that's versus the other team's best power play unit. Eller offers none of that. Beagle is also superior on the forecheck, as you point out.

Eller being an "elite" PKer is arguable. He finished 4th or 5th in PK time in Montreal over the last 3 years. His faceoff numbers are disappointing. His forechecking is solid, but nothing to write home about. This is not the pedigree of an elite penalty killer.

Jay Beagle being favored by a great coach on the PK over Eller isn't a head-scratcher. He's a better penalty killer across the board. That he's on for more goals against is *entirely* explained by the fact that he's on versus a vastly superior caliber of opponent.

None of this diminishes Eller's value. He IS a standout PKer, and in that capacity we're lucky to have him. But no, he's not a better PKer than Beagle.

That's been true at even strength this year, too...

Despite more TOI per game, with 32% of the season over, Eller is currently on pace for 7 goals and 10 points. Beagle's looking more like 16 goals and 35 points.

Beagle's won 62% of his faceoffs to Eller's 48%.

He's a +6 to Eller's -5.

You want your elite PKers out of the box, so his 20 PIMs to Beagle's 6 aren't doing him any favors.

Beagle has 2 shorthanded points and 2 game-winning goals. Eller has none of either.

Beagle has over twice as many blocks.

Eller has almost three times as many missed shots.

Eller's defensive-zone faceoff percentage is 37.8%. Beagle's is 61.3%. Eller's significantly behind Beagle in every zone, but this is a glaring fail for a defensive forward -- worst of our 4 starting centers; even Kuznetsov.

When statistics are properly viewed in context, there is literally no metric to suggest that Eller is significantly better than Beagle at *anything*.

Here again I'll stress that this doesn't mean that Eller is horrible. He is playing with a new team in new systems. He has been banged up a bit. He hasn't had much of a chance to gel with linemates because of all the shuffling.

But the fact remains that, as of right now, his underperformance is pretty staggering for a guy with a $3.5m cap hit that cost us two 2nd-round picks.

 

[Hivemind]

Despite your giant wall of text, Eller has allowed fewer goals against on the PK consistently through-out his career. Both in Washington and Montreal. He's outperformed Beagle and every other forward in Washington. He's out performed his teammates in Montreal. So, yes, there is a stat that suggest Eller is superior to Beagle at *something,* and that *something* is penalty killing. Eller has now and in the past a lower goals against/60 than Beagle does on the PK, by a significant margin. Eller is, simply put, better at stopping the opponent from scoring while killing penalties.

You attempt to paint the 2nd PK unit as a unit that gets lesser competition. That's simply not true. PK units typically change every time the puck is cleared from the zone, while PP units will stay on for 60-90 seconds of a PP. Each PK unit gets a heavy dose of the opponent's first unit. Plus, regardless of what personnel the opponent has on the ice, all PK assignments are a difficult one. There's no such thing as an sheltered deployment while killing penalties.

And once again, minutes doesn't mean better or worse. It just mean the coach selects a different player.

You can get all hung up on face-off% you want, but that's both easy to mask (have someone else take draws or platoon the draws to Eller's favorable RH match-up) and a relatively rare event. As I pointed out earlier, the difference between 62% and 48% faceoff percentage is only less than once faceoff per game when counting Beagle's and Eller's shorthanded faceoffs combined. For Eller's 1.8 PK faceoffs per game, improving that % would only mean one more faceoff won every four games. The percentages seem like a big difference, but in terms of actual results, it's minuscule.

Also, +/- is a (flawed) stat that only tracks even strength play. I'm not sure why you're bringing it up in a discussion about penalty killing. Same with your rant about game winning goals (how is that even remotely relevant to short-handed play?).

This isn't taking any credit away from Beagle, who's had a terrific start to the season. This is simply stating that Eller is the superior penalty killing player. Because that's the objective truth.

 

[Eskobar]

Despite your giant wall of text, Eller has allowed fewer goals against on the PK consistently through-out his career. Both in Washington and Montreal. He's outperformed Beagle and every other forward in Washington. He's out performed his teammates in Montreal. So, yes, there is a stat that suggest Eller is superior to Beagle at *something,* and that *something* is penalty killing. Eller has now and in the past a lower goals against/60 than Beagle does on the PK, by a significant margin. Eller is, simply put, better at stopping the opponent from scoring while killing penalties.

You attempt to paint the 2nd PK unit as a unit that gets lesser competition. That's simply not true. PK units typically change every time the puck is cleared from the zone, while PP units will stay on for 60-90 seconds of a PP. Each PK unit gets a heavy dose of the opponent's first unit. Plus, regardless of what personnel the opponent has on the ice, all PK assignments are a difficult one. There's no such thing as an sheltered deployment while killing penalties.

And once again, minutes doesn't mean better or worse. It just mean the coach selects a different player.

You can get all hung up on face-off% you want, but that's both easy to mask (have someone else take draws or platoon the draws to Eller's favorable RH match-up) and a relatively rare event. As I pointed out earlier, the difference between 62% and 48% faceoff percentage is only less than once faceoff per game when counting Beagle's and Eller's shorthanded faceoffs combined. For Eller's 1.8 PK faceoffs per game, improving that % would only mean one more faceoff won every four games. The percentages seem like a big difference, but in terms of actual results, it's minuscule.

Also, +/- is a (flawed) stat that only tracks even strength play. I'm not sure why you're bringing it up in a discussion about penalty killing. Same with your rant about game winning goals (how is that even remotely relevant to short-handed play?).

This isn't taking any credit away from Beagle, who's had a terrific start to the season. This is simply stating that Eller is the superior penalty killing player. Because that's the objective truth.

 

[Jags]

Eller has allowed fewer goals against on the PK consistently through-out his career.

Except last year, right? Beagle's SHGA60 was over 30% better than Eller's last year, despite Eller being the Habs' best PKer overall.

If we can agree that Holtby's play and Price's injury were big factors in that -- and we'd be stupid not to -- then you pretty much have to concede that Eller playing in front of Price went a long way toward helping him achieve the low SHGA60 he did.

So no, Eller has clearly not been better throughout his career. He was 30% "worse" for a full year just last year.

This year, sure, Eller's number is lower. But he's also logged 36% fewer minutes. "Per 60 minutes" stats aren't that simple. More minutes per game adds up. That Beagle has played 22% more PK time than any other Caps forward matters.

Which brings us to the faceoffs -- another thing you like to dismiss out of hand. The only reason Beagle's playing so many more minutes than everyone else on the PK is that he's killing it on draws. If our elite penalty-killing center wasn't such a glaring liability in the faceoff circle, Beagle wouldn't be out there for those lopsided minutes in the most disadvantageous situations (shorthanded defensive zone draws). You think that doesn't pull his stats down? That you think faceoffs don't matter is staggering to me.

It also matters that Eller's always on with the same partner while Beagle has split time with 4 different guys. Consistency and continuity matter.

And once again, minutes doesn't mean better or worse. It just mean the coach selects a different player.

So it doesn't mean anything that Eller logged less PK time for Montreal last year than two forwards that played way fewer games than he did? That makes sense to you? It means nothing that though he's their best PKer -- far and away by your estimation -- he doesn't log the most time? It says nothing that the 8th winningest coach in NHL history has him logging 36% less time on the PK than another guy?

You don't attribute that at all to performance? Faceoffs? Forecheck? The fact that he's been invisible at even strength, and THAT might make coaches less apt to lean on him when the chips are down?

None of that tracks for you at all?

Minutes don't matter, faceoffs don't matter, an entire year of stats that don't support your argument don't matter...

Yeah, I can see why you've reached this "objective" conclusion. When you're this apt to completely discard huge, weighty factors that anyone would concede profoundly affect a player's deployment on the PK, it's easy to reach pretty much ANY "objective" conclusion you feel like tossing out there.

Of course minutes, faceoffs, pairings, ES performance, compete level, and the other guys on the ice matter. And no, OF COURSE our 1st and 2nd units don't see equal time versus their opponents' best power play unit. OF COURSE the guy logging 36% more time sees those units more -- that guy sees 36% more of EVERYTHING. OF COURSE the disadvantageous nature of Beagle's "extra" minutes figure in big-time.

And just like I did twice in my last post, let me reiterate again that, yes, Eller is a great PKer. I'm not saying he isn't. We're lucky to have him in that regard. But saying it's an objective truth that he's clearly a better PKer than Beagle is nuts. It's easily arguable, which we know because -- HI! -- I'm here arguing it. And I'm doing a halfway decent job.

Using SHGA60 as your only metric is just silly. A great goalie offsets any meaningful differential entirely. So does a bad one.

In other news, I agree that +/- is a crap statistic. But when you're comparing two guys that play the same position for the same team, rotating pretty evenly between the same two lines with same personnel around them, versus the same opponents and in front of uniformly great goaltending, yes, a -11 differential is very telling. There are some perspectives where +/- has meaning, and this is one of them.

And lastly, your argument about the faceoffs boiling down to one per game, the same can be said of Eller and Beagle's SHGA60 differential. When you factor in all the mountains of things you discard that you really shouldn't, that gap closes a bit, and whatever's left amounts to very little when spread out over 30% of a season. Given that between them they've averaged about 60 PK minutes this year, we're talking about 2 goals. That's a couple bounces, a couple blown saves, a couple lapses by ANY of the other guys on the ice at the time... Doesn't have to all be on Beagle.

All things being equal, if Eller were perceived as clearly that much better, he'd be getting the time. He's not, so unless you think you know more than Barry Trotz, you're probably best served taking a more well-rounded look at the stats that matter to find out why Eller has ALWAYS been deployed significantly less on the PK than 3 or 4 other forwards on whichever team he plays...

 

[billcook]

Some career numbers.

4v5, relative to teammates:

Player | TOI | CA60 | xGA60 | SCA60

Jay Beagle|532.41|4.35|0.24|1.00
Lars Eller|608.32|-4.30|0.19|1.17

TOI - Time On Ice
CA60 - Corsi Against per 60 minutes [shot attempts]
xGA60 - Expected Goals Against per 60 minutes [basically 'quality' of shots]
SCA60 - Scoring Chances Against per 60 minutes [high danger shots]

Beagle is a little better at preventing scoring chances

Eller is a lot better at suppressing and reducing quality of shot attempts



Raw numbers:

Player | GA60 | Sv%

Jay Beagle|4.73|90.37
Lars Eller|4.44|88.89

GA60 - Goals Against per 60 minutes
Sv% - Save Percentage while player was on ice [basically goaltending 'influence']

 

[japhi]

Habs fan here that has watched Eller play 300 or so games. Eller does two things very well; he looks like a hockey player, and he is elite at keep away. Everything else he is a marginal NHL'r. He simply can not read the play, and create offense. I found him painful to watch - he has all the tools but can not pull it together.

At some point he will get hot, put up some points and the cycle will begin again where people will think he can do more. But he can't, he is a soft, poor in the dot, offense black hole 4th liner that has been fooling people since his draft year.

 

[Hivemind]

Except last year, right? Beagle's SHGA60 was over 30% better than Eller's last year, despite Eller being the Habs' best PKer overall.

If we can agree that Holtby's play and Price's injury were big factors in that -- and we'd be stupid not to -- then you pretty much have to concede that Eller playing in front of Price went a long way toward helping him achieve the low SHGA60 he did.

So no, Eller has clearly not been better throughout his career. He was 30% "worse" for a full year just last year.

From 2013-2017, Eller gave up 4.67 GA/60, 2nd in the NHL among all forwards. During that same span, Beagle gave up 5.96 GA/60. That's as far back as stats.hockeyanalysis.com will let you take it if you still want to include 2017.

Beagle had a exceptionally good last year, but that year is an outlier compared to the remainder of Beagle's career (he was at a team-worst 7.43 GA/60 the season before that). Eller's goal limitation has been much more consistent. The fact he was typically outperforming his teammates suggests it's not just the Price factor.

This year, sure, Eller's number is lower. But he's also logged 36% fewer minutes. "Per 60 minutes" stats aren't that simple. More minutes per game adds up. That Beagle has played 22% more PK time than any other Caps forward matters.

Once again, ice time is a reflection of who's played more, not who's been more effective in their ice time. I'm not pro-rating points here and claiming a 3rd liner will put up first line numbers based on point/60. This isn't about sheltered match-ups or deployments that can inflate or deflate numbers like a depth scorer. This is about actual goal limitation. And Eller has done a much better job at limiting goals against, both in total figures (4 goals against to Beagle's 9) and per ice-time terms.

Which brings us to the faceoffs -- another thing you like to dismiss out of hand. The only reason Beagle's playing so many more minutes than everyone else on the PK is that he's killing it on draws. If our elite penalty-killing center wasn't such a glaring liability in the faceoff circle, Beagle wouldn't be out there for those lopsided minutes in the most disadvantageous situations (shorthanded defensive zone draws). You think that doesn't pull his stats down? That you think faceoffs don't matter is staggering to me.

Once again, the difference in their faceoff number is less than one draw per game combined. That's not a hard concept to grasp.

So it doesn't mean anything that Eller logged less PK time for Montreal last year than two forwards that played way fewer games than he did? That makes sense to you? It means nothing that though he's their best PKer -- far and away by your estimation -- he doesn't log the most time? It says nothing that the 8th winningest coach in NHL history has him logging 36% less time on the PK than another guy?

We've been over this numerous times already. Your appeal to authority argument is weak. Alexander Semin was also an exceptional penalty killer (to the point where he was actual even between short-handed goals and goals against over a two year span), but he didn't see much PK time because the coach preferred other players and saved Semin for other ice time. Coaches have favorites. It happens.

You don't attribute that at all to performance? Faceoffs? Forecheck? The fact that he's been invisible at even strength, and THAT might make coaches less apt to lean on him when the chips are down?

None of that tracks for you at all?

What doesn't track for you at all seems to be that coaches can have biases and/or mistakes.

Minutes don't matter, faceoffs don't matter, an entire year of stats that don't support your argument don't matter...

Show me WHY they matter rather than just claiming they do. I've laid out my argument repeatedly, and you keep going back to the same things. I've shown with larger sample sizes that Eller outperforms Beagle. I've shown that the difference in faceoff percentage is relatively minor (less than one per game).

Yeah, I can see why you've reached this "objective" conclusion. When you're this apt to completely discard huge, weighty factors that anyone would concede profoundly affect a player's deployment on the PK, it's easy to reach pretty much ANY "objective" conclusion you feel like tossing out there.

Give me the evidence that the difference is profound, rather than just stating it. Show me that PK performance declines with minutes played. Otherwise you're just mouthing off without any supporting evidence.

Of course minutes, faceoffs, pairings, ES performance, compete level, and the other guys on the ice matter. And no, OF COURSE our 1st and 2nd units don't see equal time versus their opponents' best power play unit. OF COURSE the guy logging 36% more time sees those units more -- that guy sees 36% more of EVERYTHING. OF COURSE the disadvantageous nature of Beagle's "extra" minutes figure in big-time.

Show me the evidence.

Using SHGA60 as your only metric is just silly. A great goalie offsets any meaningful differential entirely. So does a bad one.

Good thing they're playing in front of the same goalie then.

In other news, I agree that +/- is a crap statistic. But when you're comparing two guys that play the same position for the same team, rotating pretty evenly between the same two lines with same personnel around them, versus the same opponents and in front of uniformly great goaltending, yes, a -11 differential is very telling. There are some perspectives where +/- has meaning, and this is one of them.

It's a crap statistic that doesn't even describe short handed time, yet you're now using it to back up that one player is a better penalty killer? LOL

And lastly, your argument about the faceoffs boiling down to one per game, the same can be said of Eller and Beagle's SHGA60 differential. When you factor in all the mountains of things you discard that you really shouldn't, that gap closes a bit, and whatever's left amounts to very little when spread out over 30% of a season. Given that between them they've averaged about 60 PK minutes this year, we're talking about 2 goals. That's a couple bounces, a couple blown saves, a couple lapses by ANY of the other guys on the ice at the time... Doesn't have to all be on Beagle.

That's why I also included much larger sample sizes, and they still reflect better on Eller than Beagle. The difference is also larger than 2 goals (Beagle has 9 against, Eller has 4).



I removed several of your paragraphs when you kept repeating the same thing over and over and over again. Typing a wall of text doesn't mean you're bringing any new points to the table. Keep your arguments more concise.

 

[Jags]

From 2013-2017, Eller gave up 4.67 GA/60, 2nd in the NHL among all forwards. During that same span, Beagle gave up 5.96 GA/60. That's as far back as stats.hockeyanalysis.com will let you take it if you still want to include 2017.

Fundamentally, our disagreement is a matter of perspective. You prefer this macro version that makes it nearly impossible to parse for context because of all the many things that make the comparison apples and oranges -- different personnel (partners, D, goaltending), coaches, systems, the much poorer zone starts a good faceoff guy gets versus a bad faceoff guy, minutes played, and on and on and on. There's a staggering amount of necessary analysis to consider here that you're simply ignoring.

I prefer the smaller samples because they're easier to parse for that context to make the comparison more worthwhile, and the more recent samples because what happened in 2013 has so little bearing on what happens today. By looking at last year, I'm attempting to demonstrate how clearly those other factors affect the very narrow criterion you're using to determine the overall quality of a PK forward (SHGA60 ONLY).

In the context of last year, you'd be stupid to argue that Beagle had a great year and Eller had a bad one. Holtby had his best year ever and Price was out all year. That accounts for most of the disparity. This very simple example clearly demonstrates how staggeringly context matters.

In the context of this year, now that they're playing for the same team and can most easily be fully compared, yes, sussing out WHY Beagle gets so much more PK time than Eller is a worthwhile and necessary conversation to have. But you only want to argue SHGA60 and claim that nothing else matters, when that is obviously not the case.

So let's have it your way. Let's remove context entirely, remove the one year they can most reasonably be compared, and view the biggest sample we can get.

In the 8-year period between 2008 and 2016, Beagle's SHGA60 is 4.68 and Eller's is 4.86.

Context doesn't matter, right? Very large samples aren't misleading, right? SHGA60 is the only metric worth using, right?

Prior to this season, in the 8 years that encapsulate their entire careers, among forwards playing at least 400 shorthanded minutes, Beagle is #4 and Eller is #9.

These are the sample sizes and metrics YOU prefer. How this convinces you that Eller is "objectively" a better penalty-killer than Beagle is beyond me.

This is about actual goal limitation.

No, it's really not. Tons of other factors come into play. Plekanec got more time because he's a more complete player, Byron got more time because he's much more of a threat to disrupt and score, Eller's faceoff struggles limit his defensive zone deployments, and so on. So many other factors come into play when determining the OVERALL value of a PK forward.

Once again, the difference in their faceoff number is less than one draw per game combined. That's not a hard concept to grasp.

Pick a lane. You prefer the macro perspective on SHGA60 despite it making that stat less telling, simply because it suits you. And here with faceoffs you prefer the micro perspective despite it making that stat less telling, simply because it suits you.

For defensive forwards that take a lot of draws, defensive zone faceoffs are VERY important. Lost draws in that zone lead directly to chances against. So a 61% vs. 38% disparity IS a big deal. Whether at full strength or shorthanded, if you suck at something so profoundly that your coach doesn't even want you on the ice when it's happening, yeah, that's kind of important.

We've been over this numerous times already. Your appeal to authority argument is weak. Alexander Semin was also an exceptional penalty killer (to the point where he was actual even between short-handed goals and goals against over a two year span), but he didn't see much PK time because the coach preferred other players and saved Semin for other ice time. Coaches have favorites. It happens.

Again with the apples and oranges. Eller is nothing like Alex Semin. The ONE thing Eller is exceptional at is penalty killing. That EVERY coach he's EVER played for has deployed him on the PK less than other players matters, Hive. Some of the reason for that is what those other players do better than he does, and some is about the things he does poorly. It's NOT just a random, meaningless "coach has his favorites" situation. Eller is a PK specialist. That he has routinely, consistently been deployed significantly less than other guys on his team speaks directly to the parts of his game that don't measure up to those other guys.

What doesn't track for you at all seems to be that coaches can have biases and/or mistakes.

And what doesn't track for you at all is that ALL of those coaches know WAY more about this than you do. Forgive me for thinking that your "I have a stat site bookmarked and narrow view of what's important" perspective MIGHT be laughably less reliable than those of the professional NHL coaches that have consistently deployed this guy the same exact way for his entire career...

Show me that PK performance declines with minutes played.

You said yourself that playing on the PK in the NHL is uniformly difficult, period. You also seem to like baseless, apples-and-oranges comparisons, so...

If you drink 36% more alcohol than the guy next to you, you're going to get drunker, faster.

If you bash your head against the wall 36% more than the guy next to you, you're going to hurt yourself more.

If you spend 36% more time in the bitter cold than the guy next to you, you're going to be colder and die sooner.

MoJo, Backstrom, and Sanford have all played at least 10 SH minutes this year, and their SHGA60 are all less than the mighty Eller's. So they're better PKers, right?

Minutes matter, Hive.

It's a crap statistic that doesn't even describe short handed time, yet you're now using it to back up that one player is a better penalty killer? LOL

Now you're just making shlt up. When I originally brought up +/-, it was CLEARLY in the context of even strength. And in my last post I CLEARLY explained how that -11 differential IS very telling in-context.

So you lied in asserting that I made a PK argument using +/- and then refused to address the (admittedly rare) valid use of +/- to point out a pretty big disparity between two guys that have played remarkably similar roles and time this year.

And this thread is about Lars Eller overall. It's not the "Hivemind's Narrow View of Lars Eller's PK Stats" thread.

Keep your arguments more concise.

Sorry, but it takes space to counter abject stubbornness.

 

[Hivemind]

You still haven't actually given any evidence to support your numerous assertions. You just say "minutes matter," without showing any actual evidence that more minutes leads to decreased efficacy on the PK. There's a difference between "minutes matter" and "playing more minutes directly correlates with more goals against while shorthanded." You're making a giant leap between the two to try and explain Beagle's worse rates. You can apply "context," but when you attempt to apply context, the burden of proof is on you to demonstrate how much that context matters. Think of all the time people spent using zone starts as a context tool for possession metrics, only for it to turn out that zone starts largely don't impact those metrics. And guess what, faceoffs are in a similar boat, where it takes a massive difference in faceoffs to translate into a win. Show me the evidence that the context you're attempting to bring into this conversation accounts for the 44% difference in their goals against rates while shorthanded.

And go back and read my arguments again. You're jumping on this "micro" vs "macro" train. But I was referring to longer scale trends to simply show that this year is not a fluke. Now instead you're including years where neither player was a PK regular (and Eller had been on three different teams) in order to make your case. If you want to take a "macro" perspective, you can see that Beagle's performance varies wildly from year to year, both in terms of rate stats and his position relative to teammates, while Eller has been one of the best PKing forwards on his team in each of the past four seasons.

Once now your coaching argument is fully jumping into the "appeal to authority" fallacy. Coaches can have implicit or explicit biases. They can have other motivations aside of purely what player is the most effective on the PK. They can be luddites who hate statistical analysis. They can be many things, with many motivations, and sometimes they even just plain make mistakes. Using minutes as a substitute for performance is just plain wrong.