Jump to content
The Official Site of the Vancouver Canucks
Canucks Community

Edmonton fired Tyler Dellow (summer of analytics hire)


Recommended Posts

Derek Van Diest @SUNdvandiest

While on analytics topic, sources say #Oilers guy Tyler Dellow returned to TO after Eakins firing and is unlikely to be used any more

My question is now that TO hasn't improved at all and Carlyle is gone, are all those TO analytics hires on notice as well?

Perhaps not, as Carlyle is made to look like the sole problem there.

The hirings of Kyle Dubas, Cam Charron, Darryl Metcalf and Rob Pettapiece signalled Torontos shift toward an analytics-first philsophy, and we all know the enduring hallmark of the Carlyle era was horrible analytics.

http://www.thehockeynews.com/blog/where-do-the-maple-leafs-go-after-carlyle-bylsma-babcock-or-beyond/

As opposed to the eras that preceded him? TO hasn't exactly been a force in any regard, 'advanced' or not, since they lost their top center to retirement. Not sure what Carlyle was supposed to do except work with what he's been given. And what did Nonis do but some questionable moves since those hires were made? Like the Gunnarsson for Polak trade. An 'anti-corsi' move if there ever was one.

I felt the Summer of Analytics was an overblown story to begin with, but to see Edmonton just toss their guy aside like yesterday's trash pretty much confirms it was just a knee-jerk move to perhaps instill some interest from media and fans before another god-awful season commences.

But TO is on a different level. They have an entire staff of bloggers to attribute some of their performance to now. So far it's being made to look like they represent a 'new era' there or something, where proxies to determine puck possession are somehow supposed to change what coaches have known how to coach for decades, but I think it's only a matter of time before the blame game comes to them and they get cast aside too.

Meanwhile, NHL fans lost a few of their own resources over the same period. Isn't ExtraSkater more valueable that whatever it's creator is doing in TO? (According to my quick analysis, he is doing nothing.) And with the disappearance of CapGeek, it makes you wonder if the NHL is going to move towards hoarding all statistical and cap-related content in the future. That would be a shame, as this content being made available makes the game more interesting for the fans. And what does the NHL have to hide anyway?

Link to comment
Share on other sites

The Oilers' management is such a joke. How do these people get the higher-end jobs in the first place?

Even if I were jobless and I were in the hockey industry, I'd never sign with the Oilers. Never. You could walk away looking worse off than when you started.

#AnywherebuttheOilers.

Such a poorly run organization. I honestly think people from CDC could do a better job of running it, at one tenth of the money that they're being paid.

MacTavish needs to be fired just for the Draisaitl handling fiasco alone.

Link to comment
Share on other sites

My question is now that TO hasn't improved at all and Carlyle is gone, are all those TO analytics hires on notice as well?

Perhaps not, as Carlyle is made to look like the sole problem there.

http://www.thehockeynews.com/blog/where-do-the-maple-leafs-go-after-carlyle-bylsma-babcock-or-beyond/

As opposed to the eras that preceded him? TO hasn't exactly been a force in any regard, 'advanced' or not, since they lost their top center to retirement. Not sure what Carlyle was supposed to do except work with what he's been given. And what did Nonis do but some questionable moves since those hires were made? Like the Gunnarsson for Polak trade. An 'anti-corsi' move if there ever was one.

I felt the Summer of Analytics was an overblown story to begin with, but to see Edmonton just toss their guy aside like yesterday's trash pretty much confirms it was just a knee-jerk move to perhaps instill some interest from media and fans before another god-awful season commences.

But TO is on a different level. They have an entire staff of bloggers to attribute some of their performance to now. So far it's being made to look like they represent a 'new era' there or something, where proxies to determine puck possession are somehow supposed to change what coaches have known how to coach for decades, but I think it's only a matter of time before the blame game comes to them and they get cast aside too.

Meanwhile, NHL fans lost a few of their own resources over the same period. Isn't ExtraSkater more valueable that whatever it's creator is doing in TO? (According to my quick analysis, he is doing nothing.) And with the disappearance of CapGeek, it makes you wonder if the NHL is going to move towards hoarding all statistical and cap-related content in the future. That would be a shame, as this content being made available makes the game more interesting for the fans. And what does the NHL have to hide anyway?

The guy from Capgeek has some undisclosed 'health issues'. Not sure why he'd need to lie about that, but unless he's being threatened at gunpoint, there's really no need to 'permanently' destroy something that is/was obviously useful. Gillis, during his reign, might have shot down the fact that capgeek's numbers are/were wrong, but Ottawa's GM (Murray) referred media to Capgeek. LOL.

If I were to take a guess, Capgeek's disappearance has nothing to do with what your last paragraph is speculating, but you don't need analytics to see that the people being fired aren't people that should've been fired first, if there was even merit to the firings.

Link to comment
Share on other sites

The Toronto hirings seemed like a Shanahan initiative, so I didn't think they were in trouble with the Carlyle firing. Then again, I didn't think Dellow would have anything to do with an Eakins firing either, but it may not be entirely related either.

Link to comment
Share on other sites

The thing I hate about analytics is not the concept, but how people use them.

If you do research, the only thing you know for certain is that your numbers are wrong, because measurement is always imperfect. Whatever you measure, you measured it imprecisely. The analytics gurus never take this into consideration (e.g., develop confidence intervals) let alone mention the limitation of their methods.

They all act like Corsi numbers, time of possession, success on zone entries are measured infallibly.

These are all stats being measured by some goon sitting on his/her couch or while they are half-paying attention in a press box as they try to juggle other responsibilities.

Hell, FO wins are not always entirely clear and it becomes a judgement call. If you go by official statistics, some teams conflate stats. Cal Clutterbuck would have ten hits a game back when he played in Minnesota. You can't trust the raw data, yet some people want to religiously follow the resulting statistics.

Lots of people are in a rush to be leaders in the analytics movement, but until measurement is improved, the approach to understanding the game will remain incredibly imprecise.

/rant

Link to comment
Share on other sites

The thing I hate about analytics is not the concept, but how people use them.

If you do research, the only thing you know for certain is that your numbers are wrong, because measurement is always imperfect. Whatever you measure, you measured it imprecisely. The analytics gurus never take this into consideration (e.g., develop confidence intervals) let alone mention the limitation of their methods.

They all act like Corsi numbers, time of possession, success on zone entries are measured infallibly.

These are all stats being measured by some goon sitting on his/her couch or while they are half-paying attention in a press box as they try to juggle other responsibilities.

Hell, FO wins are not always entirely clear and it becomes a judgement call. If you go by official statistics, some teams conflate stats. Cal Clutterbuck would have ten hits a game back when he played in Minnesota. You can't trust the raw data, yet some people want to religiously follow the resulting statistics.

Lots of people are in a rush to be leaders in the analytics movement, but until measurement is improved, the approach to understanding the game will remain incredibly imprecise.

/rant

Plus puck possession is not the only way to be a successful team.

But if you are baseball fan, you will know by now that having good sabremetric players does not mean you have a winning team.

Link to comment
Share on other sites

The Toronto hirings seemed like a Shanahan initiative, so I didn't think they were in trouble with the Carlyle firing. Then again, I didn't think Dellow would have anything to do with an Eakins firing either, but it may not be entirely related either.

If I recall correctly, Eakins was a pretty big fan of Dellow's work, and may have been pushing/requesting for his hire in the offseason.

Link to comment
Share on other sites

If "analytics" were behind decisions in Toronto and Edmonton, it wouldn't simply be the coaches being fired.

The GMs would have lost their jobs by now.

Analyics are for managers more so than coaches imo. Coaches work with what they already have. You fire a coach for gameplanning and how he utilizes his lineup, for whether he does or does not get the best out of players.

'Analytics' are more useful imo to the team-builder than the foreman.

In any event, 'analytics' in themselves have no value - they are all still relative. Crapty analytics are offered up all over the place. For example, in Toronto they'll try to sell you the 'analytical' truth that Bozak piggybacks on Kessel's talent. They conclude this because Bozak's underlying numbers aren't as good when he doesn't play with Kessel. The problem with such oversimplified bunk is that Bozak's situational use is far different when not playing with Kessel - far harder minutes actually - and without taking the comparative contexts into account, the isolated myth that Kessel makes Bozak better is really just a misleading oversiimplification, propped up by appearance of 'analytics'.

If one were to take "analysis" at face value as all being of equal value, they should simply listen to the "analysts" on panels around the sports world, and rethink the myth that "analytics" are generic.

"Analytics" always depend on "analysis". Some "analysts" could have a mountain of valuable information in front of them and make a pile of turd out of it. While some GMs simply have the gift of observation and come by their "analysis" with a relatively objective eye. But there are obvious limits to observation - no one has time to observe all the teams, all the games etc - so the composite objective outcomes of advanced statistics can certainly be useful.

"Analytics" ought to tell you not to devalue and let go of your best two ways players, guys like MacArthur or Grabovski - or you will become an even worse puck possession team. Toronto GMs seemed oblivious to these kinds of reality repeatedly.

They might tell you that exchanging Gunnarson for Roman Polak is not the answer to your defensive struggles - but do you really need 'analytics' to figure that out?

They should probably tell you that running with a third line of Draisaitl and Yakupov is a horrible idea - but a coach probably shouldn't need 'analytics' to figure that out, and a GM probably shouldn't be providing a roster where that happens.

They can tell you that Santorelli and Raymond are very good moneyball signings.

Or that Boyd Gordon can help a team like the Edmonton Oilers.

But a few gems of information are not enough to turn an outhouse into a cabin.

Anyhow, if Toronto were being ruled by competent analysis they wouldn't be mired in the odd contradictions they find themselves stuck in.

Link to comment
Share on other sites

The Oilers' management is such a joke. How do these people get the higher-end jobs in the first place?

I think the problem is the management's "status" of being part of the former Oilers team when Gretzky was around (ie. Lowe, MacTavish, now Messier's in management there too). It's been taken advantage of and abused .... a lot....

EDIT: Just as a funny extra, when asked what Messier's position was, he couldn't give a clear message as to what he actually does.

Link to comment
Share on other sites

What was supposed to happen in half a season of listening to what some blogger had to say about the team anyway?

The process of acquiring truly great players takes years and years and requires FAR more astute skills than mere stat watching.

Agreed that if they fired him already, then it was a joke to hire him in the first place.

Link to comment
Share on other sites

If "analytics" were behind decisions in Toronto and Edmonton, it wouldn't simply be the coaches being fired.

The GMs would have lost their jobs by now.

Analyics are for managers more so than coaches imo. Coaches work with what they already have. You fire a coach for gameplanning and how he utilizes his lineup, for whether he does or does not get the best out of players.

'Analytics' are more useful imo to the team-builder than the foreman.

In any event, 'analytics' in themselves have no value - they are all still relative. Crapty analytics are offered up all over the place. For example, in Toronto they'll try to sell you the 'analytical' truth that Bozak piggybacks on Kessel's talent. They conclude this because Bozak's underlying numbers aren't as good when he doesn't play with Kessel. The problem with such oversimplified bunk is that Bozak's situational use is far different when not playing with Kessel - far harder minutes actually - and without taking the comparative contexts into account, the isolated myth that Kessel makes Bozak better is really just a misleading oversiimplification, propped up by appearance of 'analytics'.

If one were to take "analysis" at face value as all being of equal value, they should simply listen to the "analysts" on panels around the sports world, and rethink the myth that "analytics" are generic.

"Analytics" always depend on "analysis". Some "analysts" could have a mountain of valuable information in front of them and make a pile of turd out of it. While some GMs simply have the gift of observation and come by their "analysis" with a relatively objective eye. But there are obvious limits to observation - no one has time to observe all the teams, all the games etc - so the composite objective outcomes of advanced statistics can certainly be useful.

"Analytics" ought to tell you not to devalue and let go of your best two ways players, guys like MacArthur or Grabovski - or you will become an even worse puck possession team. Toronto GMs seemed oblivious to these kinds of reality repeatedly.

They might tell you that exchanging Gunnarson for Roman Polak is not the answer to your defensive struggles - but do you really need 'analytics' to figure that out?

They should probably tell you that running with a third line of Draisaitl and Yakupov is a horrible idea - but a coach probably shouldn't need 'analytics' to figure that out, and a GM probably shouldn't be providing a roster where that happens.

They can tell you that Santorelli and Raymond are very good moneyball signings.

Or that Boyd Gordon can help a team like the Edmonton Oilers.

But a few gems of information are not enough to turn an outhouse into a cabin.

Anyhow, if Toronto were being ruled by competent analysis they wouldn't be mired in the odd contradictions they find themselves stuck in.

Are you sure you don't hate Toronto? :P

Link to comment
Share on other sites

:bigblush:

But really Surfer - it's just part of the game. If you go to the Trades, Rumours forum, you'll find Dreger and Cox threads from weeks ago, when the Leafs were winning and the Canucks were losing. It's what they do - troll us at any and all opportunies. It's what we do in turn.

Perhaps it's almost to the point where we should give the Leafs a break heh?

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...