Overall a solid post and I definitely am of the opinion that advanced stats and analytics are huge parts of building a successful team, but I also kind of think you're making a mountain out of a molehill here. I didn't see the whole interview or whatever but from what I saw in Pachla's post Nolan never once says anything about analytics being pointless or anything even close to that. Your "paraphrasing" of his comment on goals is extremely liberal, bordering on ridiculous. He said that scoring one more goal than the opposition is his number one analytic. Not that goals are the only analytic that matters. Nolan is and always has been an old school coach who preaches hard work and effort above all else and he's confident that he can judge those attributes with his own eyes. Just because he doesn't go out of his way to incorporate advanced stats into his coaching doesn't mean he finds them pointless. But regardless of how Nolan personally feels about analytics and advanced stats, I'm legitimately curious to see how people feel about how important it is for a head coach in general to be knowledgeable of and apply analytics. I guess it seems to me like it's far more important for the GM and the rest of the people building the roster to use them rather than the coach. As long as there are people on the staff (which clearly in the Sabres case there are plenty) who can use and interpret the numbers and get the information to the coach it shouldn't be that big of a deal that the coach personally doesn't know much about them.

Shots are a round peg trying to be hammered into to the square hole filled by bases in baseball. Shot-based stats are useful but they are inherently biased toward certain players and certain teams, namely players who habitually take weak shots and teams that are relatively unskilled. Also shots don't uniformly send a team 90 feet closer to a score; sometimes shots just kill offensive momentum.

nice work, jt...and thanx for reading

nice work, jt...and thanx for reading
- Michael Pachla

Thanks. And I like to follow the Sabres, so anytime.

Overall a solid post and I definitely am of the opinion that advanced stats and analytics are huge parts of building a successful team, but I also kind of think you're making a mountain out of a molehill here. I didn't see the whole interview or whatever but from what I saw in Pachla's post Nolan never once says anything about analytics being pointless or anything even close to that. Your "paraphrasing" of his comment on goals is extremely liberal, bordering on ridiculous. He said that scoring one more goal than the opposition is his number one analytic. Not that goals are the only analytic that matters. Nolan is and always has been an old school coach who preaches hard work and effort above all else and he's confident that he can judge those attributes with his own eyes. Just because he doesn't go out of his way to incorporate advanced stats into his coaching doesn't mean he finds them pointless. But regardless of how Nolan personally feels about analytics and advanced stats, I'm legitimately curious to see how people feel about how important it is for a head coach in general to be knowledgeable of and apply analytics. I guess it seems to me like it's far more important for the GM and the rest of the people building the roster to use them rather than the coach. As long as there are people on the staff (which clearly in the Sabres case there are plenty) who can use and interpret the numbers and get the information to the coach it shouldn't be that big of a deal that the coach personally doesn't know much about them.
- daaaahawks

Please locate the "Enter" key on your keyboard before posting in the future.

Please locate the "Enter" key on your keyboard before posting in the future.
- BINGO!

Thanks
for
the
tip!
Hope
this
is
easier
to
read!
I'm
sorry
I
didn't
know
the
protocol
for
the
comments
section
of
a
blog.

Anybody could clearly see the leafs were going to spiral out of the playoffs (besides leaf fans) just from watching the leafs play.

Shots are a round peg trying to be hammered into to the square hole filled by bases in baseball. Shot-based stats are useful but they are inherently biased toward certain players and certain teams, namely players who habitually take weak shots and teams that are relatively unskilled. Also shots don't uniformly send a team 90 feet closer to a score; sometimes shots just kill offensive momentum.
- Snowblind

Shots are also recorded differently in every arena.

Shots are also recorded differently in every arena.
- rangerdanger94

Yup, so are hits, and saves.

Some deflections off defensemen end up getting labeled a save.

At the critical point...when almost every statistical tally is registered...there is a human making a decision. A human observer.

What's the margin of 'human' error in these numbers?

This is a good post but there does seem to be a 'manage/coach by the numbers' mentality rising among some people that might be just as dangerous as not looking at the numbers at all.

In the end...just another tool in the toolbox. Doesn't mean you will be cup champion because you followed Dean Lombardi's formula.

The eye test is still valid. Bad and good can still be determined...largely...without consulting specialized stats. Borderline players...and maybe vets who are losing (have lost) a step...these stats probably can help confirm or refute what you think you see. What you do after that...well I guess that's up to the coach or GM. Do you move a player because of that? If you have a team that struggles with possession across the board, do you change your systems/approaches?

At the critical point...when almost every statistical tally is registered...there is a human making a decision. A human observer.

What's the margin of 'human' error in these numbers?
- icedog97

Especially true when it is a stat that simply does not matter as shots, hits, etc. really don't in the grand scheme of an NHL game. That is the Achille's Heel of fancy stats for hockey. Even QB passing ratings, as distorted as they are, are at least based on something (yardage) that is thoroughly vetted by refs and replays.

The team I watch the most is the Isles and they have one poster boy each for how little shots or hits, as a stat, actually matter. i couldn't give a sweet goddamn how solid Josh Bailey's relative corsi is or how Matt Martin is perennially among the league leaders in hits as on most given nights, they are the least important and most ineffectual players in the lineup.

Thanks
for
the
tip!
Hope
this
is
easier
to
read!
I'm
sorry
I
didn't
know
the
protocol
for
the
comments
section
of
a
blog.
- daaaahawks

I will refrain from trying to be helpful in the future.

Great blog James, you are quickly becoming my first read on HB every morning. I appreciate your interaction with other bloggers as well and I love the way you stand up for yourself. Keep up the good work buddy and go Yotes!

"The problem is, that over the course of a season, this could continue and you might not notice,... Now, over time, we tend to think luck balances out, and it does, eventually. 82 games is a long season, but even that is not really long enough to truly balance everything out - You could easily be a 20 goal scorer who got lucky enough to pot 40, or vise versa. "

The odds of that happening are so slim it's not even worth talking about. Find me more than 5, 40 goal scorers throughout one full season, that put up 20 or less in the prior(full) season, and then revert back to 20 or less after. You will find a handfull in a pool of hundreds of players....it's call odds and it's also why Casinos exist and usually thrive.

Furthermore, i find it somewhat dishonest of you to now make it seem like you're all about stats when in most of your arguments against your blog comments, you usually base it on "gut", opinion,and that "often, people are over reliant on stats". To prove my point if you were truly such a devout follower of stats, then what empirical evidence do you have that Gagner will hit 70 pts this season when all the stats point to <50?

The problem I have with analytics is that they really are based on modified traditional stats, as others have posted. Also, I don't need advanced stats to tell me that Doughty and Chara are beasts and that Jake Muzzin is better when he plays next to Doughty.

I feel like advanced stats are the captain obvious statements that usually only predict success. They are terrible at predicting failure. The Leafs are the only good example of when they were right. What did the entire analytics community predict about Anaheim, Colorado. Minnesota, or Montreal?

While advanced stats can be useful in certain situations, like line matchups, I don't believe there is anything there that can't be derived using traditional stats. The cream will always rise and that goes for teams as well as individuals. Toronto can hire all the experts they want, but they still won't be able to compete with the likes of Boston, Tampa, Detroit and Montreal for a few years. Their analytics team will probably admit this in the near future, but they will encircle it with the caveat that they need time to build a better possession team.

Analytics should be used like a lamp post, for support not illumination.

The problem I have with analytics is that they really are based on modified traditional stats, as others have posted. Also, I don't need advanced stats to tell me that Doughty and Chara are beasts and that Jake Muzzin is better when he plays next to Doughty.

I feel like advanced stats are the captain obvious statements that usually only predict success. They are terrible at predicting failure. The Leafs are the only good example of when they were right. What did the entire analytics community predict about Anaheim, Colorado. Minnesota, or Montreal?

While advanced stats can be useful in certain situations, like line matchups, I don't believe there is anything there that can't be derived using traditional stats. The cream will always rise and that goes for teams as well as individuals. Toronto can hire all the experts they want, but they still won't be able to compete with the likes of Boston, Tampa, Detroit and Montreal for a few years. Their analytics team will probably admit this in the near future, but they will encircle it with the caveat that they need time to build a better possession team.
- sniper11

Analytics help you decide who you want on the ice in certain situations and help your GM in picking talent.

Analytics help you decide who you want on the ice in certain situations and help your GM in picking talent.
- powerhouse

The key word is help. It should never be "analytics decide" anything.

Analytics help you decide who you want on the ice in certain situations and help your GM in picking talent.
- powerhouse

That's interesting, traditional stats do the exact same thing. Except that traditional stats don't tell you Joel Ward is a better player than Ovechkin because of zone adjusted, 5 on 5, points/60. I can tell you, with zero depth of analysis, that Ovechkin should be on the ice for an offensive zone faceoff and Ward for a defensive zone faceoff. Advanced stats, again, are very obvious. If you look at trend compilations for the future, the analytics are most of the time wrong, and are constantly adjusted AFTER each game, while traditional stats projections for the most part don't change much and don't miss the mark by very much. (injuries not withstanding for either).

Metrics are and always have been a business tool to feed stakeholders. From an operations standpoint, they don't leave the conference room, but they are reported for each period, whether its daily or whatever. They help to create action plans, but have zero say in the implementation of those plans. That part is always left to the managers and supervisors. Sports are no different. It is still up the managers, coaches, and players to win championships, the metrics won't do it for you. The human element is so important. Why do you think the last 2 mins of a close game is always so different from the previous 58 mins?

"The problem is, that over the course of a season, this could continue and you might not notice,... Now, over time, we tend to think luck balances out, and it does, eventually. 82 games is a long season, but even that is not really long enough to truly balance everything out - You could easily be a 20 goal scorer who got lucky enough to pot 40, or vise versa. "

The odds of that happening are so slim it's not even worth talking about. Find me more than 5, 40 goal scorers throughout one full season, that put up 20 or less in the prior(full) season, and then revert back to 20 or less after. You will find a handfull in a pool of hundreds of players....it's call odds and it's also why Casinos exist and usually thrive.

Furthermore, i find it somewhat dishonest of you to now make it seem like you're all about stats when in most of your arguments against your blog comments, you usually base it on "gut", opinion,and that "often, people are over reliant on stats". To prove my point if you were truly such a devout follower of stats, then what empirical evidence do you have that Gagner will hit 70 pts this season when all the stats point to <50?
- nikel

I go for a balance, not a either/or approach
,gagner has a career ppg avg that translates into an avg of 50 pts per year. Its logical to expect him to increase this in his age 25 season. I think so, anyways.

That's interesting, traditional stats do the exact same thing. Except that traditional stats don't tell you Joel Ward is a better player than Ovechkin because of zone adjusted, 5 on 5, points/60. I can tell you, with zero depth of analysis, that Ovechkin should be on the ice for an offensive zone faceoff and Ward for a defensive zone faceoff. Advanced stats, again, are very obvious. If you look at trend compilations for the future, the analytics are most of the time wrong, and are constantly adjusted AFTER each game, while traditional stats projections for the most part don't change much and don't miss the mark by very much. (injuries not withstanding for either).

Metrics are and always have been a business tool to feed stakeholders. From an operations standpoint, they don't leave the conference room, but they are reported for each period, whether its daily or whatever. They help to create action plans, but have zero say in the implementation of those plans. That part is always left to the managers and supervisors. Sports are no different. It is still up the managers, coaches, and players to win championships, the metrics won't do it for you. The human element is so important. Why do you think the last 2 mins of a close game is always so different from the previous 58 mins?
- sniper11

No matter what the numbers tell you, Yotes won't be good enough for the playoffs this year. I say they, and Edmonton, will be right around that 80 point mark, well behind the Canucks.

There are times when your OEL or Domi boner are so raging, that I can't even finish the article, and rush to post some asinine comment about you needing to rub one out on the nhl.com website.
Then there are times when you demonstrate that you can put together a blog that is an excellent read, and it makes me wonder why you dont expand into writing league wide articles, and moving away from writing strictly about the Yotes.

I don't think I need to explain which side of the fence Im sitting with this one JT.

I met Dubas at fan fest. Overheard him talking to a few fans. He is well spoken, and seems as if he is the type of guy that could change the thinking of a few dinosaurs.

That's interesting, traditional stats do the exact same thing. Except that traditional stats don't tell you Joel Ward is a better player than Ovechkin because of zone adjusted, 5 on 5, points/60. I can tell you, with zero depth of analysis, that Ovechkin should be on the ice for an offensive zone faceoff and Ward for a defensive zone faceoff. Advanced stats, again, are very obvious. If you look at trend compilations for the future, the analytics are most of the time wrong, and are constantly adjusted AFTER each game, while traditional stats projections for the most part don't change much and don't miss the mark by very much. (injuries not withstanding for either).

Metrics are and always have been a business tool to feed stakeholders. From an operations standpoint, they don't leave the conference room, but they are reported for each period, whether its daily or whatever. They help to create action plans, but have zero say in the implementation of those plans. That part is always left to the managers and supervisors. Sports are no different. It is still up the managers, coaches, and players to win championships, the metrics won't do it for you. The human element is so important. Why do you think the last 2 mins of a close game is always so different from the previous 58 mins?
- sniper11

1) Fenclose.

2) No.