Player Discussion Jake DeBrusk IV: The debate rages forth
- Thread starter Mr. Make-Believe
- Start date
ODAAT
Registered User
while I don`t understand most things about analytics, I do accept that there`s a place for some of it in the game but this analytic is one I would put under the "why bother" headline
Any way we cut if folks, Jake ain`t paid to simply look engaged, he`s paid to put the puck in the net, he`s not doing it and fans here have hammered far less prospects/players in the past for the same thing, can`t have double standards
Jake needs a helluva playoff or his days might be numbered
Look at expected goals as the next step to measuring shot attempts.
The old joke about Dougie Hamilton was he'd throw anything at the net for that shot attempt measurement. Expected goals take it a step further and says "okay, what are quality of those shots? If we look at hundreds of thousands of shots across years and years of hockey, what is the likelihood that shot A is going in vs. shot B?"
So expected goals will discount a Chris Pronger type shot on net from center ice and give a boost to someone getting a shot on net in the slot, where it's more likely that a goal will be scored.
Obviously more than goes into it but at a very high level that's what the measurement is attempting to get at.
The old joke about Dougie Hamilton was he'd throw anything at the net for that shot attempt measurement. Expected goals take it a step further and says "okay, what are quality of those shots? If we look at hundreds of thousands of shots across years and years of hockey, what is the likelihood that shot A is going in vs. shot B?"
So expected goals will discount a Chris Pronger type shot on net from center ice and give a boost to someone getting a shot on net in the slot, where it's more likely that a goal will be scored.
Obviously more than goes into it but at a very high level that's what the measurement is attempting to get at.
I'm not opposed to analytics either. I am absolultely opposed to it being used as the sole measure of a player's performance. A guy could ace analytics and still not be all that productive.
I think more people should chime in and tell us they don't understand "analytics" but give their opinion on "analytics" anyway.
Don’t like Jake moving back to RW tonight after he had a very good game last night
- May 28, 2011
- 1,299
- 2,769
Maybe this is a more intuitive way to look at it. 74 was credited with four shots last night, all of them high danger shots in the area around the net. A common refrain you hear around here is that "he's not contributing in other ways when he's not scoring". Well, last night I think he was. This is exactly what he needs to be doing in order to start potting some goals.
The key will be to continue doing this night in and night out but it's a step in the right direction.
The key will be to continue doing this night in and night out but it's a step in the right direction.
Definitely very interesting, I agree. You'd think they'd go DeBrusk-Frederic-Lazar to get all three guys in spots they're pretty comfortable...but just my guess....they're hoping they can use this time between now and playoffs to get #74 as much experience on the right side as possible, get him going, and be able to bump him back up and Kuraly down come playoff time.
Not a bad thing to have options I guess.
BruinDust
Registered User
- Aug 2, 2005
- 24,646
- 22,471
I have a background in stats (not to the degree that the guys who run these advanced stats website, but a background none the less) and have an understanding of xGF and personally I think the stat is complete bull-shit. 30+ years on the ice playing hockey is enough to tell me some (not all) of these variables that are plugged into this particular algorithmic formulas are garbage when evaluating scoring chances/shots.
Are they measuring shot velocity? How about quickness or deceptiveness of the players release? What about accuracy? Was the shooter trying to pick corner? And if so, did he hit his spot or hit the goalie in the chest?
Expected Goals (xG) Models Explained
See, to me, xGF takes a bunch of measures that are far less impactful on actual scoring chances than measures that are ignored (velocity, release, etc.) Granted some like deceptiveness of release are hard to measure. But I can't put much stock into a scoring chance statistics that ignores what I would call the prime factors in determining a good shot from a bad one.
I'll be honest, I don't know jack about Corsi or Fenwick. And I don't want to.
He looked engaged and battled for the puck the other night. Good to see.
What I don't like to see is the erosion of his talent and on ice impact. Jake is better than just a 4th line energy guy.
What I don't like to see is the erosion of his talent and on ice impact. Jake is better than just a 4th line energy guy.
We're aware.
Forgive me for being blunt here but the rest of your post makes me incredibly incredibly doubtful about your "background in stats"
Even a freshman stats 101 class would teach you that it's about probabilities, not perfection. No one with a "background in stats" would say "well it doesn't measure if the shooter was trying to pick a corner so it's bullshit" and side with anecdotal, biased evidence of "30+ players playing hockey is enough for me". Not a single person with a real "background in stats."
"Background in stats" should tell you that all of those variables you mentioned are factored in some form when you're looking at a data set of hundreds of thousands of shots.
So I definitely appreciate your opinion, but like I said before, all you're saying is exactly the quoted piece above. "I don't understand it so it's all stupid"..and that's fine. Seems there was some confusion around what that number represented so I tried to give an explanation of what expected goals means.
RussellmaniaKW
Registered User
- Sep 15, 2004
- 19,701
- 21,810
imagine claiming you have a background in stats and then concluding your post with "I don't want to understand this better"
BruinDust
Registered User
- Aug 2, 2005
- 24,646
- 22,471
I don't. I have a background in stats, but I don't enjoy it or like it and certainly don't want to worry about it when discussing or watching something I enjoy like hockey.
RussellmaniaKW
Registered User
- Sep 15, 2004
- 19,701
- 21,810
the xG number for Jake just says that he's shooting more from dangerous areas. why is that so f***ing hard to understand or appreciate?
you could google it and see the value in it yourself for a fraction of the effort it takes to post about how stupid you think it is (while clearly not understanding it).
you could google it and see the value in it yourself for a fraction of the effort it takes to post about how stupid you think it is (while clearly not understanding it).
RussellmaniaKW
Registered User
- Sep 15, 2004
- 19,701
- 21,810
Last night was a clear improvement in his game, don’t really need advanced stats to see it but they help back it up. He had down moments but he kept them to a minimum.
He needs to build on this, it’s definitely a step in the right direction.
He needs to build on this, it’s definitely a step in the right direction.
ON3M4N
Ignores/60 = Elite
- May 28, 2011
- 1,299
- 2,769
Smiles/60 are still replacement level. Going to need to crunch some numbers on his expected Oreo intake to determine whether he's waiver fodder or worth maybe a 7th round pick...
ON3M4N
Ignores/60 = Elite
Thought Jake was better to start the game when the team was buzzing. Seemed like he faded a bit towards the end of the game, but it was a better all around effort than we've seen previously. Really hope he's watching Hall and how he attacks because that's the mentality Jake needs to play with.
neelynugs
Registered User
- Feb 27, 2002
- 35,480
- 10,043
i'd like to see this combo before the season ends. seems that the bottom 6 makeup will probably
be determined by which opponent we draw. you get the pens and are probably looking for more
speed and pop. you get the caps and you want more of a physical bent.
BruinDust
Registered User
- Aug 2, 2005
- 24,646
- 22,471
So don't and let those that want to, do so. Seem pretty simple to me
Well considering I was replying originally to a poster who wasn't all hardcore advanced stats, I thought that was what I was doing.
And that being said.....because I disagree with the validity of advanced stats in a free-flowing sport like hockey, why should I have to keep that opinion to myself? I'm confused don't we have free speech here? I didn't realize my posts about advanced stats were breaking some sort of rule around here.
Feel free to ignore any of my posts regarding the lack of validity in advanced stats as we move forward.
Seems like that's going to be the combo tonight but all in different spots.
11-20-74
Number8
Registered User
- Oct 31, 2007
- 18,243
- 17,609
To be fair, as a fan, I have very little interest in advanced stats. I've tried, but at the end of the day, they all line up pretty much with the "eye test". And after more than 50 years of watching a ridiculous amount of hockey, my eye test is pretty good. I'm not a scout, nor an evaluator of talent, nor a guy who's going to come up with magic trade proposals. Easilly 85% of the hockey I watch is Bruins biased as all hell. And I like it that way.
Using the eye test I can tell you that for the past 10-15 games Charlie Coyle has looked better and I stopped worrying about him a while ago.
Using the eye test tells me that Taylor Hall is dangerous every time he steps on the ice and he back checks like a tenacious SOB.
Using the eye test tells me Pasta is beginning to feel much better.
Using the eye test I can tell you that back in the day, when Kevan Miller was a massive whipping boy around here, I had zero tolerance for that. He's no different today than he was in the past. A tough, rugged stay at home dman who can move the puck surprisingly well and steadies things when he's on the ice.
Same with Andrew Ference. People liked to whip on him too. Missing the difference in leadership, the intensity, and the frigging win/loss record when he was in or out of lineup.
I don't know whether their Corsification relation between their expected performance matrix and how it relates to the heat map in sections of the game when the temperature on the ice was suboptimal. But I don't need to -- I'm a fan.
I know these stats have value in the right hands. And can help players improve thier games if they use them in the right way.
And for fans that enjoy them? Fill your boots. This should be fun first and foremost and fun alone.
For now, and as far as Jake DeBrusk goes? I could give two shits about his advanced stats. It's effort and tenacity -- 100% everytime he's on the ice. He needs to mimic Curtis Lazar and Craig Smith. NOT their advanced stat metrics. Their EFFORT on the ice.
Jake has skills. If he puts in the effort, those skills will start to appear again. If not, he's going to get moved for, sadly, pennies on the dollar. What he does if he goes somewhere else is irrelevant to me. If he does well, it will only piss me off that he wasted the very long rope he got here.
Using the eye test I can tell you that for the past 10-15 games Charlie Coyle has looked better and I stopped worrying about him a while ago.
Using the eye test tells me that Taylor Hall is dangerous every time he steps on the ice and he back checks like a tenacious SOB.
Using the eye test tells me Pasta is beginning to feel much better.
Using the eye test I can tell you that back in the day, when Kevan Miller was a massive whipping boy around here, I had zero tolerance for that. He's no different today than he was in the past. A tough, rugged stay at home dman who can move the puck surprisingly well and steadies things when he's on the ice.
Same with Andrew Ference. People liked to whip on him too. Missing the difference in leadership, the intensity, and the frigging win/loss record when he was in or out of lineup.
I don't know whether their Corsification relation between their expected performance matrix and how it relates to the heat map in sections of the game when the temperature on the ice was suboptimal. But I don't need to -- I'm a fan.
I know these stats have value in the right hands. And can help players improve thier games if they use them in the right way.
And for fans that enjoy them? Fill your boots. This should be fun first and foremost and fun alone.
For now, and as far as Jake DeBrusk goes? I could give two shits about his advanced stats. It's effort and tenacity -- 100% everytime he's on the ice. He needs to mimic Curtis Lazar and Craig Smith. NOT their advanced stat metrics. Their EFFORT on the ice.
Jake has skills. If he puts in the effort, those skills will start to appear again. If not, he's going to get moved for, sadly, pennies on the dollar. What he does if he goes somewhere else is irrelevant to me. If he does well, it will only piss me off that he wasted the very long rope he got here.
LSCII
Cup driven
This is the main issue right here whenever we talk about advanced stats versus the old eye ball test. While most people understand and accept that it's all about a probability and playing the odds, people take that information and turn it into an absolute to help them prove their point or argument. It's not meant to be used or interpreted that way at all, but most advanced stat nerds do use it in that manner, which is infuriating.
BruinDust
Registered User
- Aug 2, 2005
- 24,646
- 22,471
See that is the issue right there. You take that stat and break it down into "he's shooting more from dangerous areas" like there is no question that all those variables plugged into a formula tell you exactly that, and there is no subjectivity at all.
It was like another stat I saw when Marchand was out that the stat was saying "Bergeron and Pastrnak are awful together". Which we all know is crap, taking Marchand away from those two didn't make them bad players. Is there an impact, likely negative? Sure. But just because the number the formula spits out is low (or bad or negative) doesn't the make underlying assumption (in this case 37 and 88 w/o 63 aren't good) true. Does that make sense?
