HFBoards

Go Back   HFBoards > NHL Western Conference > Pacific Division > Edmonton Oilers
Notices

Captain Obvious and the Oilers powerplay ...

Closed Thread
 
Thread Tools
Old
02-17-2004, 03:50 PM
  #26
dawgbone
Registered User
 
Join Date: Jun 2002
Country: Canada
Posts: 21,104
vCash: 500
Send a message via ICQ to dawgbone Send a message via MSN to dawgbone
Quote:
Originally Posted by Digger12
Not sure what's going on then, 'cause when I look at the top 50 PP point getters I only count 12 Dmen in the top 50.

Looking at the 02/03 season, and there's only 10 Dmen in the top 50 PP scoring, and only 2 Dmen in the top 20.

Maybe not so pointless?
Sorry, I had meant defencemen, typo on my part.

dawgbone is offline  
Old
02-17-2004, 03:51 PM
  #27
igor*
 
Join Date: Mar 2002
Posts: 3,276
vCash: 500
Quote:
Originally Posted by dawgbone
We aren't going to start comparing baseball to hockey again are we?

Didn't we beat that to death last time and come to the conclusion that while stats have some merit, a hockey game cannot be broken down into small finite conclusions like baseball can.

Not only that, but the Oilers were a low budget team that made the playoffs as well... and just like they A's, they couldn't get over the hump of the first round.
That's not correct. Baseball seems more intuitively obvious to apply stats too, simply because an 'at bat' is an easier unit to grasp than a 'minute of playing time'. And the quality of defence (pitcher and fielders) is roughly the same for each player ... so it is easier to apply.

When you look at hockey ---> it is a more talent-driven, high-event game (in terms of shots, scoring chances, action, etc.) and the data is now available to separate the wheat from the chaff, and presumeably the NHL has spent millions accumulating this data (published or unpublished) for a reason. NHL hockey is for that reason a much better sport to apply stats to than is MLB IMO.

One could really clarify the value of different players if you spent the time. Separating out game situations and linemates and opposition in a quantifiable way. I'm not about to volunteer to do this ... but surely dozens of other people are already doing just this for NHL teams. (There are dozens of data analysts employed by NHL teams, I have to wonder what they have been doing, day in and day out, year after year )

If you've ever read the extent to which Roger Nielsen and others have pushed this, even 10 or 15 years ago ... you'd be absolutely shocked dawgbone. And guys like Keenan and Bowman are his biggest fans. I shudder to think how far ahead teams like OTT are in this regard. Obviously its not the be-all and end-all ... but its an important part of the overall picture, and its not 'an edge the Oilers should grab' ... sadly its an edge they've already lost IMO.

Every Saturday night Don Cherry sells stupid to an audience that already has cupboards full ... but don't be fooled, there are some smart fokkers in this league.

My 2 cents.


Last edited by igor*: 02-17-2004 at 03:55 PM.
igor* is offline  
Old
02-17-2004, 04:04 PM
  #28
dawgbone
Registered User
 
Join Date: Jun 2002
Country: Canada
Posts: 21,104
vCash: 500
Send a message via ICQ to dawgbone Send a message via MSN to dawgbone
Quote:
Originally Posted by mudcrutch79
Just out of curiousity, how did we come to this conclusion? I think that, even in baseball, there is nothing certain because of what the stats say, but if, if Igor is onto something here, isn't that stats showing the way. I've never really seen any logical writing on the impact of stats in hockey, the way that there has been in baseball, but I think that the traditional hockey stats are so useless that there is nothing that can be gleaned from them. If you go back a ways, I've got a thread on Ray Whitney that I wrote in late September, suggesting that teh Wings spent 10 million bucks over 3 years on a guy who would be a 50 point scorer for them. At the time, some argued that he had always been a 60-70 point scorer, and that he was going to a better team etc etc, but I argued that it was teh situation that you had to put the guy in that would make the difference. Whitney has been hurt a bit, but if he had played 82 games, he'd be on pace for about 50 points. I know that the Wings don't really have a budget, but if they did, and had a system in place to do statistical analysis, they might have questioned that acquisition a bit more. I know that we don't have the statistical breakdown that we do in baseball, but hockey has, over the past couple years, started to produce stats that can be used for meaningful analysis. THere are still utterly useless stats-PIM for example, but on the whole, there is a lot that can be learned.

At the very least, taking an in-depth look at what can be quantified allows for the qualitative discussion to be within certain parameters. Let's say that prior to acquiring Izzy, the Oilers had info about players with a similar career arc, and could look at those who became 35-40 goal scorers, and tried to identify what the factors were that enabled them to take the next step. There would still be no guarantees that he would do it, but you're taking more of an informed gamble.

Along the same vein, look at this thread. We've realized that the Oilers tend to be a lot better with that fourth forward on teh ice during the PP. Now we can try and figure out qualitatively just why that is, looking at who they were playing and where they were playing. I've kind of floated the theory that maybe we can explain away some of that difference by the fact that it was favourable matchups at home, and we would see drastically different results if this were common practice. Of course, I can't be sure, having yet to really get down with the numbers.

To suggest that hockey can't be broken down into smaller pieces that allows for meaningful statistical evaluation is asinine. This is the mindset that drafts guys like Bonsignore because they look good, and puts Cory Cross and Steve Staios on teh ice during the PP because you have to have two d on the ice. In order to win with a budget like the Oilers, you have to be creative and innovative, and that means challenging every belief that you have. Otherwise, you fall into errors that are avoidable, if only you'd question the action.
Granted, I understand what you are saying...

I had meant that the game itself can't be broken down like baseball can and has been. You can breakdown statiscial performances and approximate things (much like the comparing of Salo to Johnson in one of the other threads). I'll go on the assumption that the wings weren't expecting 70+ points from Whitney because they weren't going to give him the same top line minutes he got in both Columbus and Florida.

Creativity and innovation are important in a small budget, and for the most part the Oilers have shown that. Despite limited resources, compared to the rest of the league, they have continually ended up in the top half of the league. As for drafting Bonsignore... your whole argument was that stats are important, and Bonsignores stats were never an issue, it was his head that was and I don't know of too many stats that calculate that.

dawgbone is offline  
Old
02-17-2004, 04:18 PM
  #29
dawgbone
Registered User
 
Join Date: Jun 2002
Country: Canada
Posts: 21,104
vCash: 500
Send a message via ICQ to dawgbone Send a message via MSN to dawgbone
Quote:
Originally Posted by igor
That's not correct. Baseball seems more intuitively obvious to apply stats too, simply because an 'at bat' is an easier unit to grasp than a 'minute of playing time'. And the quality of defence (pitcher and fielders) is roughly the same for each player ... so it is easier to apply.

When you look at hockey ---> it is a more talent-driven, high-event game (in terms of shots, scoring chances, action, etc.) and the data is now available to separate the wheat from the chaff, and presumeably the NHL has spent millions accumulating this data (published or unpublished) for a reason. NHL hockey is for that reason a much better sport to apply stats to than is MLB IMO.

One could really clarify the value of different players if you spent the time. Separating out game situations and linemates and opposition in a quantifiable way. I'm not about to volunteer to do this ... but surely dozens of other people are already doing just this for NHL teams. (There are dozens of data analysts employed by NHL teams, I have to wonder what they have been doing, day in and day out, year after year )

If you've ever read the extent to which Roger Nielsen and others have pushed this, even 10 or 15 years ago ... you'd be absolutely shocked dawgbone. And guys like Keenan and Bowman are his biggest fans. I shudder to think how far ahead teams like OTT are in this regard. Obviously its not the be-all and end-all ... but its an important part of the overall picture, and its not 'an edge the Oilers should grab' ... sadly its an edge they've already lost IMO.

Every Saturday night Don Cherry sells stupid to an audience that already has cupboards full ... but don't be fooled, there are some smart fokkers in this league.

My 2 cents.
Wouldn't the fact that baseball has already done it be a reason why the NHL is spending millions crunching numbers? I mean baseball has been abusing stats for decades in order to provide an obscure detail to benefit a player, coach, fan, etc...

Baseball has proven to be the ultimate game for statistics, and how to apply their usage to get the one up on your team. Of course other leagues are going to copy this, and attempt to use it to get an edge in one aspect or another.

Unfortunately, baseball has one huge distinct advantage over hockey in terms of applying statistical data.

Time.

Baseball has the advantage everytime of more or less stopping, analyzing the situation, and making a decision before the start of every play. If you don't like your matchup, you can stop and change it at any point.

This is why hockey will always be inferior to baseball when it comes to statistical analysis. When a player has a puck on a 2 on 1, he can't pull the stat about his success rate when he passes to team mate X whily coming in on oponent Y and Goalie Z. Baseball is full of that, which is why you see pitchers coming out of the bullpen and not throwing a single pitch before they are replaced.

Heck, even football and basketball are significantly easier to break down with statistics than Hockey is. Yes you can break down which of your defensemen is on the ice for the most ppg against/min, but does that really compare to Jason Giambi batting .171 on sunday afternoon games on natural grass in domes against left handed pitchers whose last name begins with X?

Baseball comes down to situational events that have happened over and over again, and have been analyzed over and over again, and the manager has the time to act on that accordingly. There is not that option in hockey. Yes there have been hundreds of 2 on 1's but there is no statistical breakdown for them that can be analyzed and transmitted in time to assist the player in any meaningful way.

So while there is certainly room for statistical analysis in Hockey, there is no way that it can be applied in the same magnitude or regularity that it can in Baseball.

dawgbone is offline  
Old
02-17-2004, 04:46 PM
  #30
mudcrutch79
Registered User
 
Join Date: Jul 2003
Location: The Big Smoke
Posts: 3,903
vCash: 500
The thing is, hockey does provide many stops when data can be used to make better decisions, as well as oppportunities before hand to determine what should be done. With the stats that the league likely has available, the Oilers could, for example, determine which of their defensive tandems has the most success against the Canucks big line. This is an obvious example that doesn't involve telling a player to do something, but rather to try and maximize the odds in your favour.

I think that we're kind of confusing two different things here. THrough video analysis and whatnot, coaches can determine tendencies of certain players, something that I don't think you can necessarily do with statistical analysis. This can enable you to realize what flaws a player has and how they can be exploited. However, what statistical analysis can do is tell you, the manager or coach, how to exploit the situation. There are three obvious situations that can be capitalized on in hockey-even strength, power play, and shorthanded situations. Every situation is then further sub-categorized based on what players a team puts on the ice to encounter them. Statistical analysis can help with better decision making there. Say that you find a team is susceptible to having goals scored on them when you put out 4F-1D on the PP. That doesn't necessarily tell you that you should do exactly that, but it maybe identifies a flaw in their strategy and tells you what you should try and do.

It is a fallacy that you can't take advantage of the situations in hockey. Its a bit messier in baseball, but by searching for overall tendencies, I still think you can come out further ahead than talent alone would seem to justify.

mudcrutch79 is offline  
Old
02-17-2004, 05:03 PM
  #31
dawgbone
Registered User
 
Join Date: Jun 2002
Country: Canada
Posts: 21,104
vCash: 500
Send a message via ICQ to dawgbone Send a message via MSN to dawgbone
Quote:
Originally Posted by mudcrutch79
The thing is, hockey does provide many stops when data can be used to make better decisions, as well as oppportunities before hand to determine what should be done. With the stats that the league likely has available, the Oilers could, for example, determine which of their defensive tandems has the most success against the Canucks big line. This is an obvious example that doesn't involve telling a player to do something, but rather to try and maximize the odds in your favour.

I think that we're kind of confusing two different things here. THrough video analysis and whatnot, coaches can determine tendencies of certain players, something that I don't think you can necessarily do with statistical analysis. This can enable you to realize what flaws a player has and how they can be exploited. However, what statistical analysis can do is tell you, the manager or coach, how to exploit the situation. There are three obvious situations that can be capitalized on in hockey-even strength, power play, and shorthanded situations. Every situation is then further sub-categorized based on what players a team puts on the ice to encounter them. Statistical analysis can help with better decision making there. Say that you find a team is susceptible to having goals scored on them when you put out 4F-1D on the PP. That doesn't necessarily tell you that you should do exactly that, but it maybe identifies a flaw in their strategy and tells you what you should try and do.

It is a fallacy that you can't take advantage of the situations in hockey. Its a bit messier in baseball, but by searching for overall tendencies, I still think you can come out further ahead than talent alone would seem to justify.
I agree 100% that in some instances you can statistically analyze hockey and make a decision on that... but we are talking about 15 seconds (unless there is a T.V. or a regular timeout) in order to try and figure this out, as opposed to baseball where everything is predicated and laid out in front of you (i.e. pitching around one guy so you can get to another). Also in hockey, if your team doesn't have last change, which you don't in 50% of your games, or the simple fact that the other team makes a quick change before you are able to can throw everything out of whack.

For instance, you may have a lot of success when you put 4 forwards on the ice for a pp, but what is Detroit's success rate at not only killing penalties, but scoring short handed goals agaisnt 4 forwards, and is it worth the risk? That I can understand using statistics for, but that is a pre-determined circumstance, which is rare in NHL hockey, but common place in baseball.

Has there really been proof brought forward that statistical analysis in hockey will get you farther than talent? We've seen system play dominate talent, but has statistical analysis altered the game anywhere near as much as say the trap? The trap uses a very basic statistic as it's base, and that is if you give up less goals than you opponent, you will win more than you lose... it was a simple twist on the old score more goals than your opponent. And honestly, that's about a big a change as stats have made. Instead of trying to score more goals than the other team, everyone is simply trying to give up less goals than the other team.

I have gone way off and I apologize, but this is a very intriguing conversation...

dawgbone is offline  
Old
02-17-2004, 05:58 PM
  #32
Funkymoses
Registered User
 
Join Date: Mar 2002
Location: Ann Arbor, MI
Country: United States
Posts: 840
vCash: 500
i just read this article which might be of interest....

http://www.mlive.com/wolverines/aane...3281673460.xml

it's basically about Michigan's video system and the stats they collect. to summarize: they've got up to date stats on everything, almost immediately, to the point where they can use it *during intermission* to get favorable matchups.

I basically agree with the stat-heads in this conversation. It's ridiculous that the Oilers haven't seen these stats or, worse, have seen them and ignored them. that's a very bad sign.

Funkymoses is offline  
Old
02-17-2004, 06:12 PM
  #33
igor*
 
Join Date: Mar 2002
Posts: 3,276
vCash: 500
Quote:
Originally Posted by dawgbone
Wouldn't the fact that baseball has already done it be a reason why the NHL is spending millions crunching numbers? I mean baseball has been abusing stats for decades in order to provide an obscure detail to benefit a player, coach, fan, etc...
I think there is a lot of sense in the argument that what has happened in baseball recently has opened the eyes of some hockey owners. But the fact is that the NHL in general and some teams in particular were accruing an astonishing amount of detailed (and really relevant stuff IMO) statistical information long before baseball fell in love with stats as an administration tool. In fact I'd argue that the NHL (Nielsen in particular) has been ahead of baseball in this regard until recently. And the only reason that baseball moved forward is because a couple owners were innovative, and hired a couple of true businessmen into GM roles ... and they have had success ... and its snowballed. The same will happen in hockey. (if it hasn't already).

'oilswell' fished up several academic articles on stats in hockey, and their application. Some of it is pure academic drivel IMO, some of these guys have never even seen a game methinks. Other stuff is hardcore, and downright 'big picture' sensible. Google the names of these better authors and you'll find that they have been on either the NHL or NHLPA payroll at some point. And this is years-old material.

I know 'mudcrutch' envisions bringing the 'Billy Ball' thinking to hockey ... my point is that he is 15 years too late (not for the Oilers, but for the NHL in general).[/QUOTE]

igor* is offline  
Old
02-17-2004, 08:55 PM
  #34
Matts
Registered User
 
Join Date: Nov 2002
Posts: 2,911
vCash: 500
Db

I think your biggest problem is you have a hard time faulting anything that the Oilers do. Which means it's been a bad year for you and as long as Lowe continues to play ostrich regarding netminding and centre, it's not gonna get any better for you.

BTW, eight teams make the playoffs in MLB and 16 in the NHL

Matts is offline  
Old
02-17-2004, 09:51 PM
  #35
ZIM
Guest
 
Country:
Posts: n/a
vCash:
Quote:
Originally Posted by Funkymoses
i just read this article which might be of interest....

http://www.mlive.com/wolverines/aanews/hockey/index.ssf?/base/sports-0/107703281673460.xml

it's basically about Michigan's video system and the stats they collect. to summarize: they've got up to date stats on everything, almost immediately, to the point where they can use it *during intermission* to get favorable matchups.

I basically agree with the stat-heads in this conversation. It's ridiculous that the Oilers haven't seen these stats or, worse, have seen them and ignored them. that's a very bad sign.
I couldn't resist. Remember this MacT quote.
Quote:
I don't rely on game-to-game stats that much.
Can be found in this Oilers' MacTavish being outcoached? That's ridiculous article.

 
Old
02-18-2004, 12:00 AM
  #36
igor*
 
Join Date: Mar 2002
Posts: 3,276
vCash: 500
Quote:
Originally Posted by IceDragoon
... Last year I thought Reasoner would've looked good as a #1 PP pointman, but then he would've been on the ice over 30 mins some games. And, of course I like York (healthy) on the point too. (sigh)
Stoll is a smart old man in a young man's body. Although some may say he's too green, I think he could be a good choice.
Hmmm 3 smart, adaptive centres (I know York isn't...). Stoll only got to show his stuff because the other 2 got injured, but he would be my pick of what's left of the litter.
I like York on the point too. Reasoner is an interesting idea, he's certainly responsible enough defensively ... not sure what his slapshot is like though. Hemsky would still be my first choice.

Heck, if I was coach I'd be tempted to play 5 forwards

Quote:
...
As far as assessing a business...

I would need to make a more in-depth assessment before I could; "recommend dismissal of all management involved with the administration of this aspect (powerplay)." On the whole I have found that; a good 'attitude-adjustment/shake-up' will usually give you better results than a complete turnover of management. I've seen far too many 'adjusted' managers go on to become top producers, to think otherwise.

imho - KLo needs to assess and make adjustments accordingly. iow - I concur. It's broke. Fix it.
...
Yeah, you are right of course. I'm losing patience with the Oilers mishandling of the powerplay though ... so I reserve the right to be rash

I really wish that the Oilers would more clearly define the roles of the assistants though. Someone has to take ownership for both the successes and failures of the powerplay IMO. Same goes for the PK.

igor* is offline  
Old
02-18-2004, 01:16 AM
  #37
igor*
 
Join Date: Mar 2002
Posts: 3,276
vCash: 500
Quote:
Originally Posted by mudcrutch79
...
1. Does anyone know where we could get some accurate stats about how much time the Oilers have spent on the 4-3PP, or on the PP with the goalie out?
Well you could modify that macro I posted earlier to grab the PP times for each off of the NHL scoresheets ... that's pretty easy to do.

Then you could scour all the NHL.com superstat sheets and grab the goalie PP ice-times ... that would give all the raw data you need to figure out the amount of different types of ice-time. Or at least I think it would.

Quote:
... Not quite the five times better that Igor is suggesting, but still almost 2.5 times better.
You're right, my bad on the arithmetic.

Quote:
Had the team gone 4F-1D at all times, could we really be looking at a PP that was +36 instead of one that was +21? For those who are somewhat familiar with my obsession with goal differential, this would put the Oilers at +12 instead of -3, or somewhere between Calgary and LA. (By the way, my theory seems to be taking a beating this year, as three of the teams ahead of the Oilers have worse goal differentials, which is kind of strange)
I don't think you can consider your theory on goal differential a theory anymore. I think it is just fact. And as someone else here suggested ... apply Poisson's equation and watch it fly .

The way I see it ... goal differential gives you a probability of garnering X number of points. So for example (and these are just made-up numbers) --- with a goal differential of +2 there would be a 5% chance that you would get 95 points or more, a 60% chance that you would garner 87 points or more, and a 20% that you would garner 80 points or less.

BTW: Using goal differential alone for two teams about to play ... you can calculate the bookies odds with surprising accuracy. Much more imporatnat than any other stat (wins, losses, streaks, etc) in forecasting the outcome of a game.

My 2 cents.

igor* is offline  
Old
02-18-2004, 08:41 AM
  #38
dawgbone
Registered User
 
Join Date: Jun 2002
Country: Canada
Posts: 21,104
vCash: 500
Send a message via ICQ to dawgbone Send a message via MSN to dawgbone
Quote:
Originally Posted by Matts
I think your biggest problem is you have a hard time faulting anything that the Oilers do. Which means it's been a bad year for you and as long as Lowe continues to play ostrich regarding netminding and centre, it's not gonna get any better for you.

BTW, eight teams make the playoffs in MLB and 16 in the NHL
Really?

Talk about captain obvious. That is pretty much what I said in an earlier post about why I continually defend Oilers management about things.... wow, I hope it didn't take you long to come to that conclusion.

And yes, 8 teams make the playoffs in MLB, but they have a greater discrepancy in the haves/have nots. Would the Oakland A's have made the playoffs much if they were in the American League East? I doubt it, because they aren't in a division with all the haves.

dawgbone is offline  
Old
02-18-2004, 09:14 AM
  #39
dawgbone
Registered User
 
Join Date: Jun 2002
Country: Canada
Posts: 21,104
vCash: 500
Send a message via ICQ to dawgbone Send a message via MSN to dawgbone
Quote:
Originally Posted by Funkymoses
i just read this article which might be of interest....

http://www.mlive.com/wolverines/aane...3281673460.xml

it's basically about Michigan's video system and the stats they collect. to summarize: they've got up to date stats on everything, almost immediately, to the point where they can use it *during intermission* to get favorable matchups.

I basically agree with the stat-heads in this conversation. It's ridiculous that the Oilers haven't seen these stats or, worse, have seen them and ignored them. that's a very bad sign.
Do you really think that it's they've ignored them (or haven't seen them, whatever the case)?

For instance, in the L.A. game, the Oilers like to have lefties taking draws on the left side of the ice in their own end and righties on the right side. Your stats may say that your right hander may be better than your left hander, but do you want the puck drawn towards the front of your net? Sure you may win the face-off, but if the puck ends up in your net, does it matter? Also, do we know that these stats aren't taken into consideration? Is this why the Oilers like left handed draws on the left side in their own end?

Maybe it's just me, and the stats I have available, but everytime I try to apply a stat to a situation in hockey, there is a hole, or a question that stat doesn't answer, and to me that is a huge problem.

dawgbone is offline  
Old
02-18-2004, 09:23 AM
  #40
Mr Sakich
Registered User
 
Mr Sakich's Avatar
 
Join Date: Mar 2002
Location: Motel 35
Posts: 8,015
vCash: 500
great posts guys. A few things to consider:

1 in baseball, all the stats are used to give a team a small edge. If things were absolutely guarantteed to provide a desired outcome, then the team would use a computer instead of a manager. This edge is outweighed by the fluid nature of hockey. No point in trying to figure out a 1% edge if the margin of error is 10%.

2 In hockey, every player plays injured and it is up to the coach to figure out who will give him the best performance on a given night. In baseball, players are usually 100% or they don't play.

3 In baseball, it is one player vs the other team. In hockey, it is the combined output of three players plus 2 dmen vs an ever changing team. A good coach makes 1+1+1 = more than 3.

4 The real application of Billy Beane is in drafting and scouting. He determined that slugging percentage is a far greater indicator of major league potential than batting average and drafted accordingly. This is the area of greatest potential application for hockey. Perhaps shooting percentage or the ratio of goals vs assists or some other obscure measurement holds a key to effective drafting.


Matts

As for who should be hired, the Oilers don't hire to win or else the org would have away different look

I have read this 4 times now and I can't figure out what you mean by this?

Mr Sakich is offline  
Old
02-18-2004, 09:35 AM
  #41
igor*
 
Join Date: Mar 2002
Posts: 3,276
vCash: 500
Quote:
Originally Posted by dawgbone
Do you really think that it's they've ignored them (or haven't seen them, whatever the case)?
Personally, I think the Oilers have the data. Heck I'm sure they have pretty sophisticated software written specifically for the NHL that gathers this stuff. As well as captures video clips from games to file against the data ... and enables the coaching staff to tally the scoring chances and all the players involved in it. A lot of the decisions they make seem to indicate that this is the case.

But they are hockey guys, ex-teachers and ex-players ... my guess is that when facts fly in the face of what they believe in their hearts and minds ... they write it off as coincidence. Bias comes into play. If even one of these guys was completely unbiased, or came from a business background ... I'm sure they'd step back and ask questions like "exactly why is team X so much more successful than us in this aspect of the game" or "what are the chances of this result being a coincidence?". They'd play the odds in each situation (the tricky bit is figuring out what the odds really are). But they aren't, so they don't, at least not across the board.

Quote:
Maybe it's just me, and the stats I have available, but everytime I try to apply a stat to a situation in hockey, there is a hole, or a question that stat doesn't answer, and to me that is a huge problem.
You're probably using the wrong stats. As mudcrutch says ... most of the stats posted in the paper are kind of muddy and often worthless. Important to sit back and just think logically methinks ... if you forget what you already know about hockey stats and just apply common sense you WILL find surprising, undeniable correlations. (or, alternatively, track down old Roger Neilsen interviews). You'll also find that the majority of NHL teams are already applying these things. It really is NOT rocket science.

My two cents.

igor* is offline  
Old
02-18-2004, 10:03 AM
  #42
dawgbone
Registered User
 
Join Date: Jun 2002
Country: Canada
Posts: 21,104
vCash: 500
Send a message via ICQ to dawgbone Send a message via MSN to dawgbone
Fair enough... good food for thought anyways.

dawgbone is offline  
Old
02-18-2004, 10:51 AM
  #43
Matts
Registered User
 
Join Date: Nov 2002
Posts: 2,911
vCash: 500
Yeah, DB

The Mariners won over 100 games one year the A's were the WC and the next season the Angels and A's came out of the west and Ana won the series so yeah that division is certainly trash. The A's haven't won a playoff round yet but they aren't competiting in a weak division so you can drop that right now.

Whomever said that they'd use computers instead of managers, well some orgs believe in stats and some don't and the ones that do have GM's and managers working on the same page to the extend that maybe a guy who kills lefthanded pitching but does little else will be picked and will be used accordingly.

Teams that use stats have managers that respect stats and that's why Buck was let go in favor of Tosca.

Matts is offline  
Old
02-18-2004, 11:13 AM
  #44
dawgbone
Registered User
 
Join Date: Jun 2002
Country: Canada
Posts: 21,104
vCash: 500
Send a message via ICQ to dawgbone Send a message via MSN to dawgbone
Quote:
Originally Posted by Matts
The Mariners won over 100 games one year the A's were the WC and the next season the Angels and A's came out of the west and Ana won the series so yeah that division is certainly trash. The A's haven't won a playoff round yet but they aren't competiting in a weak division so you can drop that right now.
They certainly aren't in the hardest division. Granted they aren't in the central where you need only a pulse to make the playoffs, but that Western division certainly hasn't been much to brag about, especially considering they only have 4 teams, and get to play against the central division more times than say any team in the east division meaning an easier schedule.

The proof is in the pudding... despite dominating their division 3 out of the last 4 years, they haven't accomplished anything in the playoffs. Yes, once in that time a team in their division accomplished something, but is it any different than any other cinderella story?

And the year the A's won over 100 games, did they not still have players like Giambi in their lineup? Over the past 3 years, the team that won the west had less wins than the team that won it the previous year...

dawgbone is offline  
Old
02-19-2004, 02:26 PM
  #45
mudcrutch79
Registered User
 
Join Date: Jul 2003
Location: The Big Smoke
Posts: 3,903
vCash: 500
Quote:
Originally Posted by igor
Personally, I think the Oilers have the data. Heck I'm sure they have pretty sophisticated software written specifically for the NHL that gathers this stuff. As well as captures video clips from games to file against the data ... and enables the coaching staff to tally the scoring chances and all the players involved in it. A lot of the decisions they make seem to indicate that this is the case.

But they are hockey guys, ex-teachers and ex-players ... my guess is that when facts fly in the face of what they believe in their hearts and minds ... they write it off as coincidence. Bias comes into play. If even one of these guys was completely unbiased, or came from a business background ... I'm sure they'd step back and ask questions like "exactly why is team X so much more successful than us in this aspect of the game" or "what are the chances of this result being a coincidence?". They'd play the odds in each situation (the tricky bit is figuring out what the odds really are). But they aren't, so they don't, at least not across the board.


You're probably using the wrong stats. As mudcrutch says ... most of the stats posted in the paper are kind of muddy and often worthless. Important to sit back and just think logically methinks ... if you forget what you already know about hockey stats and just apply common sense you WILL find surprising, undeniable correlations. (or, alternatively, track down old Roger Neilsen interviews). You'll also find that the majority of NHL teams are already applying these things. It really is NOT rocket science.

My two cents.
I think you kind of make my case here Igor. You said up above that you figured I was 15 years too late to push this sort of thinking. The problem is that, although some teams may have cottoned onto it, if you peruse the lists of NHL GM's and coaches, what you find are a bunch of guys who played in the NHL, and as a result, have all of the biases and what not that comes along with that. I always think its kind of interesting that the teams who are really good seem to feature as GM's guys who were, at best, marginal players, if they were at all. When you think of people reknowned as great GM's, you think of people like Sam Pollock, Glen Sather, Pierre Lacroix, and Bill Torrey. Even today, two of the perennial contenders in Colorado and Detroit are run by guys with a limited playing background. I find what Feaster has done in TB to be interesting as well. Here's a guy who apparently can't even skate, grew up in the US, and he's turned TB into one of the top teams in hockey. I find his refusal to trade Lecavelier particularly interesting, because I have a belief that generally, you're going nowhere without a great centre, and I think that they are a hell of a lot harder to find then great wingers. I haven't quantified that one yet, but it's on the to-do list.

As for the Oilers PP, the obvious thing that they should be doing is seeking a balance between the rate at which they achieve a + at 4F-1D, and a + at 3F-2D. Right now, they should be analyzing the situations where they've had success at 4F-1D, and pushing it forward to situations where they think that the strategy would be less likely to work. Simple operations management.

Just because teams have the data, doesn't mean that it's been applied, or applied properly. You'll remember that Bill James was publishing his findings on baseball since the late 70's, and it took until the mid to late 90's for someone to jump on it. I suspect that the same is probably true in hockey, if not more so, given the reverence people attach to heart, guts, etc etc.

mudcrutch79 is offline  
Closed Thread

Forum Jump


Bookmarks

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off



All times are GMT -5. The time now is 05:04 PM.

monitoring_string = "e4251c93e2ba248d29da988d93bf5144"
Contact Us - HFBoards - Archive - Privacy Statement - Terms of Use - Advertise - Top - AdChoices

vBulletin Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
HFBoards.com is a property of CraveOnline Media, LLC, an Evolve Media, LLC company. 2014 All Rights Reserved.