Venture back with me, to a time long ago, when temps were hot, humidity levels high. We were at a loss of things to talk about, when suddenly, the Wild signed Mikko Koivu to a seven year contract extension. The heat, coupled with the meltdown across the fan bases and media of the NHL that followed caused ice rinks to melt in the Yukon.
Back on July 20, we wrote a bit about one Mr. Peter Keating's blog, buried behind the ESPN Insider barrier, about Koivu's deal and how Mr. Keating feels it is an overpayment. If you remember, or go back and re-read, we came to the conclusion that using Mr. Keating's numbers, Koivu ranked as the 19th best player in the NHL, with the 21st highest contract in the league.
In his current blog, he rebuts our argument, and passes on some love for our readers. (No, really.) Make the jump.
First off, let me thank Mr. Keating for reading Hockey Wilderness, and for the kind words he passes along for the blog and for the readers. While we in the hockey community have a strong disdain for ESPN, and its coverage of the NHL, we do respect the efforts of those that write about the game. In his post, Keating mentions a "genuine flame war" between us and him. That was never the intent, though it was the likely result.
To Mr. Keating: We use a healthy dose of snark in our posts, but none of it is meant with any disrespect. If anything written by Hockey Wilderness was construed as disrespectful, it was unintentional.
After reading the latest post from Mr. Keating (subscription required), it becomes clear that we are going to have to agree to disagree. Why? You'll see in a moment.
Back in July, Mr. Keating argued that just about everything in hockey that traditionalists called "intangible" can now be measured. I disagreed, citing my examples. To this, Keating responded (emphasis mine):
I wasn't thinking of statheads or fantasy league players (or Hockey Wilderness readers) who are familiar with faceoff wins. I was thinking about my dad, who has been a Rangers fan since the Truman administration but wouldn't recognize a Corsi number if one knocked on his front door. And my point was simply that statheads and fantasy players (and Hockey Wilderness readers) are way ahead of many fans, beat writers and even GMs in understanding just how much we can quantify about NHL performance. It was a compliment, not an insult!
Indeed, HW has some of the brighter readers I have come across, with some strong exceptions. We truly appreciate that Mr. Keating also picked up on that fact.
After this, we go back into seeming disagreement, so I will save the snark (well, most of it). Mr. Keating goes on:
And now on to the heart of the matter. Koivu ranked as the 19th-best forward in the NHL in 2009-10, according to either goals versus threshold or relative Corsi. So BReynolds and his readers wondered how I could possibly criticize Koivu's contract extension, because it ranks 17th in value among NHL forwards, according to this analysis by Ryan Popilchak of Sports Opinionated. Top 20 here, top 20 there ... But here my antagonists have made a fundamental mistake: There's no reason for value stats to line up according to magnitude.
Suppose I want to buy a car. A $40,000 model is a decent purchase if it's actually worth $40,000, but a $15,000 choice is just as good a value if it's really worth $15,000. A $20,000 car that is truly worth $30,000 would be an even better buy, even though it's more expensive, just as a $10,000 car that's actually worth $6,000 would be worse, even though it's cheaper. Similarly, hockey contracts can be terrific or terrible values whether they pay huge money or the league minimum. Value is always relative, and saying Koivu's contract ranks 17th among the top 25 NHL forwards means it's below-average value, wherever he ranks in on-ice performance.
To be clear, the argument he makes is that if you pay less than something is worth, it makes it a better deal. To that, I have no counter argument. If I can buy a $40,000 car for $20,000, it is a pretty good deal. My only counter argument would be to point out that in the real world, that does not happen very often. In a capitalist economy, the market tends to bear out. Workers are generally compensated at fair market value, and a $40,000 car usually sells for $39,999.
I am not going to include his charts here. Trust me when I say he has charts that show that Koivu comes in with a GVT/ $1million of 1.91. This is indeed, below the average number for the centers signed out of free agency this summer (including John Madden and Matt Cullen). Let me also point out that no contract on the list comes anywhere near what Koivu's is, and the players that rank above him are mostly role players for whom a tiny increase in production equals a huge increase in GVT / $1 million.
Koivu's Three year average GVT on the chart is 12.87. The next closest is a 9.67 from older brother Saku.
To sum it up... who do you want on your team? Koivu, or the leader of Mr. Keating's list, one Mr. Craig Conroy?
He goes on to somewhat agree with me:
There will be some flukes in any list: Craig Conroy rates highly here, for example, because he was excellent two seasons ago, but he's pushing 40 and might not even make the Flames' squad this year. Overall, though, NHL teams have invested $31.6 million in centers this offseason, buying an average of 2.6 GVT of performance for every $1 million of cap money they spend. Minnesota is getting 1.9 GVT for every $1 million it will blow on Koivu. That's not so hot.
Now, Koivu is a better player than anyone else on this list. Maybe teams pay a premium for All-Star performance? To find out, I also examined the highest-paid centers for every club other than the Wild:
Again, the chart is there, supporting his point. Pavel Datsyuk tops the chart, and Koivu once again is relegated to "below average" status. There are some big names on the second list, all with solid GVT / $1 million. But then, there's good ol' Vernon Fiddler, right there in 2nd place. The same Fiddler who has never put in more than 32 points.
But the list isn't about overall production, it is about value. That is what we are looking at here. The "value." The averages of goals vs threshold per $1 million of cap space used. We are not trying to find the best player, but the best deal.
To show that, Mr. Keating writes:
For these very good to great players in midcontract, teams are getting an average of 2.2 GVT for every $1 million of cap money that they spend on top centers. Koivu's contract would rank 17th on this list. Again, below-average. (To repeat, I'm ranking the value of deals here, not the abilities of players.)
To further prove his point, Mr. Keating has this:
Part of what's going on is that Koivu posted his best numbers in 2009-10, so he'll look worse in any analysis that relies on average GVT rather than last year's stats alone. But Koivu is 27, not 22, so there's no particular reason to expect him to extend his breakout; it's more likely that he'll perform in line with his career norms. And if you just use 2009-10 numbers, centers like Henrik Sedin and Nicklas Backstrom and Jonathan Toews would look even more superior in value than Koivu, anyway.
For the first time in his career, Koivu was used in a more offensive manner. That is why he posted his best numbers ever. A second year under Todd Richards, Bouchard likely moving to the top line, and a potent second line that has to pull some of the opposition's defensive pressure off of Koivu's line, and bottom six forwards who can handle the shut down role all add up to a continued upswing in Koivu's numbers.
The three players listed in this paragraph got their contracts for a reason. Sedin's deal would have been much higher outside of Vancouver, and much higher if he did not insist on he and his brother inking exact deals with the same team. Backstrom's contract is inflated due to his role as Ovechkin's center, and Toews took a deal to help the Hawks with a terrible cap situation and to be able to stay with Patrick Kane.
The contacts are better deals, sure, but they are the type of deal you get if your dad owns the dealership.
Finishing up, Mr. Keating states:
Look, I like Koivu. After studying his numbers and his style of play, I'm rooting for him. And even though I think Minnesota is spending too much on him, I can understand why Wild fans appreciate his contract extension in almost the same way Phillies fans are happy that their team is overpaying to keep Ryan Howard. I just wanted to point out that Koivu delivered more bang for the buck from 2006 through last season than he probably ever will again, and that when Minnesota had a player who was generating almost 3.5 GVT per $1 million of cap space, they should have done a better job capitalizing on him.
And now we also know that Vernon Fiddler is a steal. And that Detroit (Mike Modano) and the Flames (Olli Jokinen) got good deals on free-agent centers this summer, while Boston (Gregory Campbell) and Edmonton (Gilbert Brule) squandered serious coin.
Koivu likely provided more bang for the buck from 2006 to last season, yes. But that's what happens in the NHL. Good players sign contracts, are criticized for getting so much money, and by the end of the deal, they look like a steal. The next contract is signed and people gasp at how much money it is worth. By the end of the deal, they generally end up looking like a steal.
The Wild did not overpay. They paid market value because they don't have a reason not to. We will never know for certain, but my bet is that if Koivu makes the open market, Brian Burke, Glen Sather, and the like would have had the check books out, ready and willing exceed $6.75 million per year for Koivu.
For the record... Vernon Fiddler is not a steal. He is a role player, doing his job, who happened to have a bounce back season. The fact that GVT / $1million makes Fiddler look so good is exactly the reason why it scares me that we point to it as a viable stat.
Finally:
I did not set out for this post to be another volley in the debate with Mr. Keating, but it seems that is how it wound up. We appreciate that Mr. Keating is reading HW, and that he values our opinion enough to offer his counter argument to ours. We do have a great deal of respect for him, and never meant for it to be a flame war, just a hockey debate.
My argument continues to be that stats provide only a limited ability to judge talent in hockey. Without a way to judge talent, the best deal is subjective or incomplete. Stats are great for fantasy hockey and for fans who wish to rank players and debate who is the greatest goal scorer of all time. Stats, however, explain very little of the game.
There is a reason why there is no "Moneyball" for hockey. Hockey is a free flowing game, with constant changes and adaptations of the ice. What a player does away from the puck is just as important as what they do with it. Baseball is a series of moments. Every play in baseball can show up on a stat sheet. Hockey is different.
How do we quantify positioning for a defenseman? How do we quantify the ability to stand in front of the net and take a beating in order to deflect a shot? How do we quantify the battles in the corners? How do we quantify intimidation? How do we quantify good line changes vs bad? How do we quantify the ability to put the puck in the right place so only your player can get to it first? How do we quantify the ability to know when to ice the puck and when to skate it out?
I could go on for days with the moments in hockey that cannot be measured. My point is that hockey stats are great and giving us numbers to look at, but that it is the context in which we use them that makes them valuable or not. No one stat can fully measure a player's abilities. To be sure, using every stat we can find will still leave gaps in the puzzle. That is where the instinct, the "intangibles," come into play.
Mr. Keating is a stats writer, and my belief is that he cannot measure the game in the way he feels he can. We need to agree to disagree, and I'm OK with that.