Introduce yourself!

[GKJ]

I joined the dark side about last summer. I rejected it because I was bad at math. Then I said 'well, they can just do the math for me.' Now it works better.

 

[TheDevilMadeMe]

I'm the only moderator of the History Board who isn't also a mod here (and there's still time for that ).

I think advanced statistics are an invaluable tool, but I think a lot of stats people get so bogged down in numbers that they sometimes loose track of the essence of the game itself. I find instances where the statistician overreaches and claims his stats how much more than they actually do to be unhelpful.

Two specific criticisms of mainstream hockey analytics:

1) The assumption that "conventional hockey wisdom" is worthless if it can't be statistically proven. Countless times, I have seen conventional wisdom tossed aside by the latest man with "The Answer" only to see later statistical work indicate that yes, the conventional wisdom had something to it. Off the top of my head, "goaltenders have no effect on shots against" and "skaters have no effect on save percentage" to be mindbogglingly ignorant statements (especially the second one), yet for a time were (and in same cases still are) accepting as truisms by some in the hockey analytics community.

I think the responsible thing to do would be to start with the assumption that conventional wisdom has a grain of truth to it, and should only be thrown out in the face of convincing evidence to the contrary (which we do have in quite a few cases). The current assumption seems to be that conventional wisdom should be dismissed off the bat unless convincing (statistical) evidence can be found in favor of it.

Taken to the extreme, the collective opinions of paid NHL GMs and coaches are dismissed as those of a bunch of meatheads stuck a past without the newfangled stats.

2) The tendency to dismiss every effect that can't be easily explained statistically as "luck." The easiest example I can think of is the commonly used blanket statement that any increase or decrease in playoff performance is due to random variation. This would make sense of players were simply machines driven by probability engines, but completely ignores the psychological difference between the playoffs and the regular season both in terms of pressure and in terms of playing the same opponent over and over again.

This might be a specific example of #1 (dismissing conventional wisdom as luck out of hand).

Back to me now: In case you haven't noticed, I'm much better at criticizing studies than coming up with my own

 

[Chainshot]

I've poked around with some of the stats to validate my opinion on certain players for a few years now and now it's starting to gain traction. I find them very useful tools and a good way of finding more information (but not ALL information about) how a player plays.

 

[Czech Your Math]

I think advanced statistics are an invaluable tool, but I think a lot of stats people get so bogged down in numbers that they sometimes loose track of the essence of the game itself. I find instances where the statistician overreaches and claims his stats how much more than they actually do to be unhelpful.

I agree. Some of the best studies and metrics are the simplest in principle. For instance, Overpass' adjusted plus-minus is very simple in principle. Even the math used to estimate SH/PP GF/GA on ice is relatively simple. It's recognizing how to use the available data properly is not always so simple. In contrast, some "all-in-one" metrics that utilize more complicated math without reasoned support for the methodology are of little use and can be misleading.

Two specific criticisms of mainstream hockey analytics:

1) The assumption that "conventional hockey wisdom" is worthless if it can't be statistically proven. Countless times, I have seen conventional wisdom tossed aside by the latest man with "The Answer" only to see later statistical work indicate that yes, the conventional wisdom had something to it. Off the top of my head, "goaltenders have no effect on shots against" and "skaters have no effect on save percentage" to be mindbogglingly ignorant statements (especially the second one), yet for a time were (and in same cases still are) accepting as truisms by some in the hockey analytics community.

I think the responsible thing to do would be to start with the assumption that conventional wisdom has a grain of truth to it, and should only be thrown out in the face of convincing evidence to the contrary (which we do have in quite a few cases). The current assumption seems to be that conventional wisdom should be dismissed off the bat unless convincing (statistical) evidence can be found in favor of it.

Taken to the extreme, the collective opinions of paid NHL GMs and coaches are dismissed as those of a bunch of meatheads stuck a past without the newfangled stats.

I think the opinions of knowledgeable hockey people should be given some respect. However, it should not stop there, but rather be an impetus to further support or (at least partially) disprove that "conventional wisdom."

2) The tendency to dismiss every effect that can't be easily explained statistically as "luck." The easiest example I can think of is the commonly used blanket statement that any increase or decrease in playoff performance is due to random variation. This would make sense of players were simply machines driven by probability engines, but completely ignores the psychological difference between the playoffs and the regular season both in terms of pressure and in terms of playing the same opponent over and over again.

This might be a specific example of #1 (dismissing conventional wisdom as luck out of hand).

Back to me now: In case you haven't noticed, I'm much better at criticizing studies than coming up with my own

While it may be rash to just dismiss conventional wisdom, it's also unwise to blindly accept it. I think what's most important is to use logic when analyzing and assessing data. For instance, playoff performance is difficult to study, because the conditions are unequal. One could say "Messier's playoff PPG is similar to his regular season PPG, which is unusual, so he must be one of the most clutch playoff players." This overlooks some important factors:

- Messier's regular season PPG was lower than a lot of other great players, and it's generally easier to maintain a lower PPG than a higher one.

- Messier was not in the playoffs his last several seasons, so his regular season PPG is decreased by these lesser seasons, while his playoff PPG isn't. Also, a player's playoff games each season varies a lot more than his regular season games, and since league avg. scoring varies each season, this also has an effect.

- Messier was often on strong regular season teams, which were often stronger than their opposition in the playoffs. Such strong teams will tend to outperform their opposition, so the players on strong teams will tend to outperform players on weaker teams.

It shouldn't be expected that everyone perform or cite a study to support their position. What's frustrating is that many abandon simple logic when assessing the data available. I can understand when someone doesn't believe they have the math skills to perform or understand a study. I can't understand when they refuse to use logic.

 

[Greg02]

Hi, I'm Greg. I', a CS/Math major primarily interested in theoretical computer science. I'm fairly skeptical of the ability to draw meaningful conclusions about hockey from sabermetrics.

I'm going into my senior year of college majoring in Math and Computer Science. Id definitely consider myself more of a computer science student, I love programming and algorithms, but have also recently begun really delving into Machine Learning, and, by extension Stats! Id love to poke around here and learn a few things!

Hmm... Islanders, Haskell, Machine Learning...

We must be mortal enemies or something (Rangers/Scheme/PLT)

 

[puckguy11]

I'm Andy. Big on the hockey (Not so much on the math), but the idea of sabremetrics in hockey intrigues me, considering the lengthy contracts given out the past few seasons especially for smaller markets that may not have the money nor the appeal to bring in free agents but still want to win.

 

[team_alex]

I always though HF should start a board for this stuff. There should be enough of us quant-types around to make things interesting. I have a Bsc in Econ and the CFA designation. My math isn't the best, but I've had a few levels of econometrics & calculus. So I hope I can have useful input somewhere along the way.

 

[Iain Fyffe]

2) The tendency to dismiss every effect that can't be easily explained statistically as "luck." The easiest example I can think of is the commonly used blanket statement that any increase or decrease in playoff performance is due to random variation. This would make sense of players were simply machines driven by probability engines, but completely ignores the psychological difference between the playoffs and the regular season both in terms of pressure and in terms of playing the same opponent over and over again.

The problem here is that you have made a blanket statement yourself (by claiming that analysts make that blanket statement), and also rely upon a truism (the playoffs are "different" in a meaningful way). What you're probably missing is the amount of analysis that has gone into a question like that, which reveals the strong effect that variance has on the playoffs.

If there are real differences in these players, it should be persistent and repeatable. And yet, clutch players one year often disappear the next, and "playoff" teams fail to repeat their performance.

Luck is never the only factor. But it's a biggie, the fewer games you play. That some teams perform differently in the playoffs being used as evidence that the playoffs are different is circular reasoning, and normal variance explains it just as well without begging the question.

 

[Church Hill]

This is something that has always interested me, I've just never really sat down and thought at length about quantifying concretely hockey performance with pencil-in-hand. Maybe this forum can serve as a catalyst. Also, I have a degree in Math & Economics. I know a lot about statistics too, as the job I'm pursuing uses it extensively. I'm excited to see where this board goes and maybe I'll try to contribute some ideas.

 

[Mint Berry Crunch]

My username is my real name, and you can find my stuff online. My blog is here, and I wrote for Hockey Prospectus for some time but not really anymore. And my old site is still around. I've done a lot of statistical analysis on modern players and teams, but generally focus on old stuff now, pre-1927.

Your name is awesome. (I seriously mean that if it comes across as sarcasm at all.)

And back on track --

Hey all. Developmental psych major here. Love hockey; Rangers fan. Huge love for soccer as well (Arsenal). All-around lovable nerd, really.

 

[CanadianHockey]

I'm CH, minimal background in mathematics and statistics beyond high school level calc and my intro to statistics courses for an undergrad degree in political science. Looking forward to seeing some of the great minds of HF at work. Hopefully I'll be able to pick up a thing or two.

 

[xtra]

Ill be honest i don't have a fancy degree like some of you or have i been published in at any MIT Sloan conventions but im here to learn as this seems like the wave of the future and i would love to know what i am talking about before everyone else jumps on the bandwagon

P.s. I'm a canucks fan and i know they use this stuff so hopefully learning this cna help me see the value in some of their moves

 

[Coconuts]

I have a degree in Mathematics, but now am working and continuing my education in the Computer Science field.

I've always been heavily into hockey statistics and trivia, but in the last few years I've become more interested in using all available data to be able to predict future results better than anyone has been able to so far.

About a year and a half ago my interest was at its peak and I wrote a program to retrieve all event and shift data from NHL.com so I could have all available raw data available for mining. It worked but was not very robust, and the algorithms I was using in my data mining were inefficient.

These days I have a number of other projects that must take precedence, so I am sure I won't be able to contribute as much in terms of substance as I would like, but I will definitely be reading. Maybe this will help spark some motivation again.

 

[King Woodballs]

Hi I am KW
My math now a days begins and ends with 1+1=3

I don't have time to come up with funky numbers and spread sheets... but I do enjoy reading others.

 

[Yurog]

Hi guys
I'm 21 years old. I'm from Russia. I'm intersting History of hockey. I missed prelockout NHL. To catch up, I began to collect classic games and sistemize merit of players numerically.
My background is industrial automation, studied the numerical methods,mathematical models of technological processes, optimization techniques , statistical processing of data, betting, like programming on C++.

I develop goalies rankings since 50s

 

[Sorge Georos]

I'm Mike, at 18 years old. Loved baseball stats since I was 6 and learning Sabremetrics has given me a greater understanding and appreciation of the game. Have some statistical background because my father teaches Biostatistics at a med school and I'm majoring in Accounting for college.

Honestly I think hockey's where baseball was in the 1980s, in terms of condescending attitudes towards analytics and the reliance on narratives like clutchness and intangibles.

Arguments are just not as fun when a fan can just say "you don't watch the games" if you say anything's off about their player's game. Scouting has it's place as does watching games, but so do stats.

 

[LyricalLyricist]

Great subforum. Not sure how much I'll contribute but I'll be passing by and looking around for sure.

I'm a mental math nut and my educational background is pretty much being in the process of an industrial engineering bachelor's degree.

 

[Hivemind]

Mechanical Engineer with a few undergrad prob & stat courses under his belt. Became interested in "SABR"-type stats through baseball, and followed that interest to hockey. Due to my workload and other commitments, I doubt I'll have the time or energy to invest in running or coordinating any of my own studies (would have done so well before now if I could), but I'm more than happy to be involved in the discussions and provide whatever constructive criticism I can.


I'm also not afraid to yell at you if you abuse statistics, especially if you wind up giving them a bad name. I'm look at you, Neil Greenberg (ESPN/Washington Post/ex-RMNB).

 

[Thirty One]

Very excited to see this sub-board.

I'm Jared, 25, Accountant. Very interested in performance metrics. The less we use viewing a player in a small sample and writing down how that player makes us feel as a method of evaluation, the better.

 

[Frank Gallagher]

I'm Patrick. 19, History Ed. Major. I love stats and know way too many useless ones. I however probably shall contribute very little in the way of work... Regardless, I will definitely be watching to see what you number-crunchers can do.

Godspeed gentlemen!

 

[garret9]

Hello all. My math background is modest (limited to the introductory calc and stats classes needed for my B.Sc), but I've always been fascinated by stats and their application in sports. I was especially drawn in by the work Ken Pomeroy, a college hoops stats guru.

Hi Hank!!! Haha

Everyone else:
Hi my name is Garret.
I'm a long time hockey fan, coming from a family that is deeply involved in many facets at many levles, but is new to online community (brought by the return of my childhood team).

My mathematical/statistical background predominately comes from my undergrad experience (analytical chemistry, biostatics, multivariable calc, differential equations, etc)... but the statistical side of hockey was introduced to me by the lovely site of Arctic Ice Hockey.

 

[headsigh]

I like hockey and I failed math class almost as hard as the Thrashers failed.

But I'm here anyway.

 

[AlienWorkShop]

Hellooooo (la la la)

Just noticed this subforum, cool stuff.

I've always loved stats and I'm the type that probably enjoys reading about baseball stats more than actually watching the game haha. I've come to realize part of my joy for hockey is given it's free-flowing nature and almost impossible to control environment, I can simply watch the game and not have freaking numbers running through my head the entire time haha

My background... did economics for my undergrad and I'll be starting a master's in economics overseas in the fall, so any actual contribution I can make will probably be pretty limited, but I'll be sure to drop by occasionally and perhaps make some contributions/suggestions/etc.

Random thoughts:
1) I'm hardly the first to say so, but I'm very skeptical of most hockey stats given how uncontrollable the environment is compared to baseball. I haven't really done much analysis on my own as I tend to get frustrated by biased numbers and I'm not much of an excel jockey. However, if optical tracking is properly implement in hockey one day... oooooh boy.

2) I get a little annoyed by both sides of "regression to the mean" arguments. I'm generalizing, but adherents to regression may treat it as an actual law and try to shut down discussion accordingly, while detractors will characterize it as a law and then simply point to deviations to disprove it. It's generally just a rule of thumb that is heavily dependent on variation depending on the data at hand (this may be more of a baseball stats issue haha, but just saying)

3) If hockey stats could be controlled like baseball stats (maybe one day...), I'd be very tempted to say goodbye to my econ aspirations and take it up as my career haha

4) There seems to be a lot of debate over just how useful advanced hockey stats can be, with good reason. In general, I tend to think that while some advanced baseball stats may explain, say, 75% of a team/player's output, the very best hockey stats are probably in the 10-20% range with luck and intangibles playing a much larger part. Even if that is "low", I still think it's worthy exploring.

uh, ok, cool.

 

[SaskRinkRat]

Really excited about this new subforum.

I'm really interested in whether or not the new movement toward hockey metrics can identify higher volume, more frequently occurring improvement-type metrics. A lot of the focus right now seems to be on post-hoc evaluation (i.e., over the course of five seasons, which events counted most in the win-loss regression analysis). I'm curious about whether or not those events can be further broken down into smaller events that happen during a game so that player evaluation can happen in the moment. If we know that shots lead to goals, then what leads to shots? Then what leads to those things that lead to shots, and what skill development is necessary to ensure those things occur with more frequency.

I think analytics have the potential to be a huge coaching tool in addition to a player evaluation tool. I feel like most are currently geared toward, say, a GM who is concerned with a player's past performance, whether it will continue into the future, and whether he can get good value for that player if he brings him on his roster. I'd like to see another branch start up that is geared toward, say, a coach who wants to understand how his players are doing right now in the areas that most lead to net goal production.

 

[Shrimper]

Passed my maths at GCSE, failed at A-Level yet learning as an accountant. Will look to learn things here and participate.