Friday, May 25, 2007

Bless you Ales/Shawn, Damn you Marty/Raffi?

I've been looking at NHL payrolls a lot lately with the intent of finding a relationship between how a team spends its money and its level of success. It sounds simple enough and this gave me the impression that I would find out a lot of things that I already know and that would be the end of it.

I started out by taking a look at the payroll structures of the top 8 regular season teams in the NHL this year - that is: Buffalo, Detroit, Anaheim, Nashville, San Jose, Dallas, New Jersey, and Ottawa - and comparing them to those of the NHL's 8 worst: Philadelphia, Phoenix, Los Angeles, Washington, Chicago, Edmonton, Columbus, and Boston. I included the 20 best skaters and two best goalies from each team, used actual as opposed to cap dollars, included active players regardless of health, and ignored inactive players (Alex Mogilny) who still get listed because of payroll/cap implications. I also added up the total contract value of each team's 3 most expensive forwards, two most expensive defensemen, and single most expensive goalie to form a separate, "top six players" dollar amount.

The basics:
Top 8 average payroll: $42,918,786.13
Bottom 8 average payroll: $34,152,804.38

Top 8 average "top six" payroll: $24,355,240.38
Bottom 8 average "top six" payroll: $19,271,989.00

Quite clearly the more successful teams spent more money, both at the top end and throughout their rosters. The actual distributions varied quite a bit in terms of which positions were getting the most money, from New Jersey going forward heavy with ~$16M tied up in Elias, Gomez, and Gionta to Anaheim's $13M Niedermayer/Pronger duo on defense to Dallas leading the way with $6,000,000 on Marty Turco in net. Payrolls ranged from very top heavy with cheap depth (Anaheim) to quite evenly spread out (Buffalo). Basically with my limited scope I was completely unable to identify a spending pattern that correlates with success other than the very, very obvious the more you spend, the more you get.

From there I decided to put the data into a chart to see if Excel could tell me something I missed. I made salary ranges of <$1M, $1-1.99M, $2-2.99M... all the way to $7+M, counted (by hand) how many players fit into each range and of course kept it divided into top 8 and bottom 8 categories. Here is the chart:

The differences may be slight but the top 8 teams clearly have more players making between the $2M (range) and the $5M ranges where the bottom 8 teams clearly have more players making in the $1M range or less. This is obviously to be expected - if a team spends more total money it should have more players in the higher ranges. The one (caution: small sample size) thing of note is that the best and worst teams in this case each had a very similar number of players who earned more than $6M this season. Could a reasonable conclusion be then that the best teams have star players backed up with very good depth where the bad teams may have stars but they're backed up by crap?

To further explore the idea I decided to see if I could attribute some measure of value to each player. This way I could see if perhaps the top 8 teams were getting more value for their dollars in addition to spending more of them in the first place. To do this, I used the very rudimentary tool of total points scored in each salary range (eliminating goaltenders of course) and dividing by the total number of players in that range to acquire a "points per player in X salary range figure". Here is that information in chart form:

Results here are mixed so I used Excel's "trendline" feature to get a rough line of best fit for each. With so few players in the top salary ranges and without separating forwards from defensemen there are huge sample size issues to be concerned with but to me the very interesting aspect of this comparison is that the difference in pts/player between successful/unsucessful teams appears to increase as you go higher in salary - that is to say both top and bottom teams received similar production from their cheapest players but the top teams generally received more production from their mid-range salaried players and much more from their very highest.

While you can see that I am obviously generalizing, there is intuitive reasoning to support such a trend. If a team is currently awful, it was probably at least in the neighbourhood of bad last season and the one before that. If this were the case, it would have probably have had access to high-end draft picks at some point in its recent history and would therefore be likely to have high production in its low salary ranges. In this way the production of Alexander Ovechkin props up his whole salary range. It is also intuitive that a team experiencing a great deal of success would have lots of production from its higher paid players and that a team doing poorly may have experienced some disappointing seasons from its highest paid players. Compound these disappointments with lower total payrolls and it's quite apparent that a team's season could be tanked.

As a final note, here is the chart the points/player data came from with the Oilers' own information thrown in. It's interesting to see that Edmonton competes quite well with the top 8 teams' production levels in the $2-2.99 and $3-3.99 salary levels but is disgustingly worse than either top/bottom 8 levels when it comes to 0-$0.99M and $1-1.99M levels.

There are a lot of flaws in the way that I have collected and analyzed the data here but hopefully it gives you something to think about. I look forward to reading people's impressions/responses/suggestions.


Blogger MikeP said...

I'm not surprised.

I posted something along these lines some time ago on hfboards, with no takers:

It takes money to win, pure and simple.

5/25/2007 7:51 pm  
Blogger Lowetide said...

Last summer, had Lowe signed even one additional ufa defender the Oilers imo would have been a much better club. You can say "well Tjarnqvist got hurt" but God Almighty they had nothing behind Smith, Staios, Tjarnqvist they could count on (except MAB who is effective in a defined role).

It takes money, but the Oilers roster makeup was curious to say the least.

5/26/2007 7:47 am  
Blogger Oil Stain said...

The Oilers had neither the great young bargain scorers or the 5+ million dollar men that make up over half the production of the top 8 teams.

No wonder they were last in the NHL in goals scored.

They are kind of stuck in limbo right now and need to aquire some elite players at all costs.

5/27/2007 8:41 pm  
Blogger Showerhead said...

MikeP: well look at that, you certainly did. I'll take credit for doing a whole lot more potentially useless work to come up with almost the same conclusion, but yep, you certainly said more $ means more chance of success. I wonder why nobody replied.

Lowetide: I tend to agree with you and so do a ton of Oil fans. Defense was an obvious hole from July of last summer onwards and it was never adequately filled. When you look at Lowe quotes harping on "five assets" for Pronger though, you have to imagine he was content with tanking the season from trade day onward. While I do think that he expected better than he got I would not be surprised at all if he had the idea of trading Smyth from last July as well. Could that not explain a lot?

Regardless of this hypothesizing, what I obviously want to know at the end of the day is when Lowe plans on turning his assets and his quantity into a quality hockey club. I'm personally feeling good about 08/09 with steps taken between now and then but that's so disappointing I can't imagine what happens if he waits longer.

Did someone say this will be his contract year? Maybe he will go all San Fernando on us in his own summer version of the playoffs.

Oil Stain: at all costs? I'm not sure. That could involve Briere at $7M or somesuch insanity. You're right though, Edmonton is sorely lacking in top end talent and its youth isn't good enough to produce unless playing behind such talent.

5/28/2007 8:16 am  
Blogger RiversQ said...

When you look at Lowe quotes harping on "five assets" for Pronger though, you have to imagine he was content with tanking the season from trade day onward.

SH: I think we should be clear on something. Harping on five assets is tantamount to tanking the franchise.

The impact of preferring quantity over quality is not limited to just this season. The Oilers didn't just get poor assets for now, they got poor ones for later as well.

5/28/2007 2:35 pm  
Blogger Showerhead said...

RiversQ: I don't know if I feel the same doom and gloom that you do. I have limited faith in Lupul and I hope his current contract is his last. I think Smid is a good candidate to follow the Eric Brewer route - IE his press clippings and general opinion will always paint him as a very good player with lots of potential - the only thing that's left to be seen is whether he gets traded in a few years for someone more established or develops into that good player for Edmonton. Seems like a good asset to me. This year's 1st round pick? No indication that there's a player there. Next year's? A little more likely based on depth + Anaheim can't win the Cup two years in a row can they? Still, not great assets. Combine them though with Edmonton's own 1sts this year and next and NYI's pick and you get much better odds of getting 1-2 good NHL players.

I realize that the obvious rub is that we gave away more in Pronger and Smyth than we got back. I'm not going to debate that even for a second. What I will say is that from today forward, Edmonton's very bad hockey team does indeed look to me as one that can be very good before Hemsky's contract is up. Whether picks/prospects get traded or they get packaged for legitimate players, there are enough pieces IMO for a smart GM to build a winner with.

I realize there are a lot of things that have to go right to make up for what's happened (preferably with someone new at the helm) but to me it's a winnable hand with the right person holding the cards.

5/29/2007 9:27 am  

Post a Comment

<< Home