Friday, January 26, 2007

Another Look At The West Conference Scoring Leaders

After nearly a week without any NHL-calibre games it is time to play hockey again. One of the biggest often unmentioned developments in the league is the discrepancy in talent between the East and West Conference. The West Conference has the better record Tom Benjamin addresses this. The West Conference has a winning record in head-to-head games. They have 56 wins and 37 losses (with 7 losses in overtime and shootout. This is clear evidence that the West Conference is superior this year.

This season, the East Conference is higher scoring. 6.2 goals per game are scored in East Conference and 5.7 in West Conference games. Thus the top scorers in the league come from the weaker East Conference. This is not surprising. Weaker leagues are always higher scoring. The AHL is always higher scoring than the NHL. The ECHL is always higher scoring still.

Why is the West Conference better? It's not clear how that happened, though here is my theory. The West Conference has always been disadvantaged in terms of travel. It is sprawled over four time zones in North America while the East Conference is all in some portion of the eastern time zone. The west teams lose time that could have been used for practice, rest etc. It is hard on the western teams. It has been shown to cost them points each season. Over time, the East Conference teams thought they were a bit better than they were and the West Conference teams thought they were a bit worse. This led to the west working to improve themselves while the east was complacent and didn't do the same. At any rate, the West Conference is currently the better one.

In the beginning of December, I listed the top ten scorers in the West Conference. Things have changed as time has passed. Chris Pronger, who was previously the second highest west scorer has fallen off the pace with his injury.

Here are the current top 10 scorers in the better West Conference

Top 10 Scorers In The West Conference (as of all star break)
Name Team Goals Assists Points Standing in League
Joe ThorntonSJ1347609
Teemu SelanneAna30295910
Jarome IginlaCal23305313
Paul KariyaNas16375315
Patrick MarleauSJ22305219
Joe SakicCol20325220
Daymond LangkowCal19315024
Alexander FrolovLA23244729
Andy McDonaldAna12354731
Steve SullivanNas15314633

There are 23 east players with as many or more points than the top 10 west players. It is easier to score in the east. Goals are more plentiful because the opposition is not as good.

I believe that these ten players are likely having better offensive seasons than the top 10 scorers in the east (a group that can be much more easily seen with a look at the NHL top scorers)

The east still has more offensively talented players. Crosby, Ovechkin, Kovalchuk, Spezza, Heatley, Alfredsson, Sundin, Jagr, Savard, Malkin, etc. would all still put up great offesnive numbers in the west but in general the west has better defenses and goaltending.
The east doesn't hjave more offensively talented players. They have more offence. You look at the higher numbers from, the higher offence and assume higher offensive talent.

Sure the names you listed would put up good offensive numbers in either conference, but the numbers would be lower in the west.
What I am sure about is that neither of us can be completely sure of our claims without doing more research.
David. Define research. In my mind, it means watching hockey games. I think in your mind it means crunching numbers.

I think it might be something that could be shown with a little number crunching. However, I would argue that merely watching some west conference games should be enough to convince you that Iginla, Thornton, Sakic, Frolov, Datsyuk, Havlat, Selanne, Kariya, Marleau and Smyth are ten west players who are at least equal offensive talents to the 10 east players you named.
Have you watched every western conference team and every eastern conference team play 5-10 games to get an accurate read of their offensive and defensive ability and can be completely objective about it? How many games have you seen Steve Sullivan play? What about Ray Whitney? Have you honestly seen enough of both of these guys play this year to accurately say that Sullivan's 47 points is better than Whitney's 57? How about Olli Jokinen and his 52 points? If not, then some objective analysis of data would certainly be useful for backing up your claim.

I watch a lot of hockey and try real hard to be independent thinking but I also realize that it is next to impossible for any single person to fairly evaluate fairly the hundreds of players who play in the NHL. Teams employ many scouts, not just one, for this very reason.

Besides, humans aren't normally good at evaluating players (or anything) objectively. We tend to focus on the extremes and not the norms. We focus on the one bad blunder a player makes and label him a bad defensive player or we see another player make a spectacular play and he is suddenly a star offensive player.

How many people think that Samsonov is a very good offensive player? A lot do. But he isn't. He's actually pretty mediocre. Sure he is a good skater and has some skill with the puck but too many people focus on what he can do or what he does every now and again but fails to recognize he is actually not very good. Statistics tell you this because statistics don't focus on the one or two brilliant plays he makes every now and again but on what the player is as a whole. Humans often struggle with that distinction. If you read my blog I spent oodles of time prior to the season trying to convince Montreal Canadiens fans that Samsonov was a horrific signing, and well, he has been.

So you can talk all you want about research being about watching hockey games but ultimately the numbers don't lie and someones perception based on a small number of games watched can certainly be wrong.
So you can talk all you want about research being about watching hockey games but ultimately the numbers don't lie and someones perception based on a small number of games watched can certainly be wrong.

Its entirely true that watching hockey is an imperfect science. However, you overrate the importance of numerical analysis of statistics.

Numbers do lie. They are often wrong. They are often inaccurate.

Whenever you watch hockey to evaluate players or do a statistical study its important to ask yourself what is missing, overemphasized or underemphasized.

I am going to quote Bill James. He is talking about baseball, but it applies here.

Baseball statistics are simplifications of much more complex realities. It may be unneccessary to say this because, of course, all human understanding is based on simplifications of more complex realities. Economic theories are simplified images of how an economy works, replacing billions of complicated facts with a few broad generalizations. The same is true of psychological and sociological theories, it is true in medicine and astronomy. The search for understanding, wherever it roams, is a search for better simplifications-simplifications which explain more and distort less. Even the understanding gained from experience is, of course, a simplification of experience into the generalizations whivh are distilled from many experiences.

The difference between a good statistical analyst and a poor statistical analyst is that a good statsitcal analyst, like Pete Palmer or Craig Wright, understands this, and a bad one implicitly denies it.

A good statistical analyst, in studying the statistcal record of a baseball season, asks three or four essential questions:

1) What is missing from the picture?
2) What is distored here, and what is accurately portrayed?
3) How can we include what has been left out?
4) How can we correct what has been distorted?

We all know many things and many different types of things which are not reflected in the statistical record. Acknowledging this, a good statistical analyst is sometimes able to reach out and draw areas of the game which were previously undocumented inside the tent, inside the focus of the statistical record. Sabremetrics is sometimes able to invent a way to correct for one or another distortion of the statistical picture.

The bad statistical analyst, onm the other hand, will assume that what the statistical record tells him must be true and complete- and by making that assumption, will forfeit his ability to add anything significant to the record.

Baseball statistics are interesting not because they answer questions for us, but because they may be used to study issues. The value of baseball statistics in identifying the greatest players is not that they answer all of the questions involved, but that they provide answers to some of the questions involved, which enables us to focus on the others.

My point is the numbers often do lie. And it is the bad statistical analyst in Bill James's quote that denies this fact.
I agree with everything in that Bill James quote. Here is what is happening.

1. You see that the majority of the top scorers in the NHL are form the eastern conference.

2. You ask yourself part of what Bill James says every good statistician should ask: What, if anyhting, is missing here? What is distorted and what is accurately portrayed?

3. You answer that by surmising that the players in the west are in fact just as good as the players in the east offensively, it's just that the players in the west are better and thus that reduces their point producing potential.

So far everything you have done is valid. You may have a valid point. Problem is, you have only applied half of Bill James's theory of what makes a good statistician. The other half, which I said needed to be done, was to answer the questions of how can we include what has been out and how can we correct what has been distorted. (though to some extent you still have to prove that the points statistics are in fact distorted)

So, thank-you for posting that Bill James quote because I agree with it 100% and everything that I have ever done on my website has been to apply this theory. My power rankings are designed to eliminate the distortion caused by the difference in schedule. My player rankings are designed to eliminate the distortion in a players statistics that are a result of who that player plays with and against. My adjusted hits, giveaways and takeaways statistics are designed to eliminate the human caused differences in how each NHL game observer defines what is a hit, a giveaway and a takeaway.

And that Bill James quite also backs up my point when I said: "What I am sure about is that neither of us can be completely sure of our claims without doing more research."

Oh, and not anywhere in that Bill James quote did he say that human observation was better or ever should supercede (quality) statistical analysis. I just don't see how watching hockey games can accurately tell you with any degree of confidence whether Steve Sullivan is a better or worse offensive player than Whitney or Jokinen?
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?