Lies, Damn Lies, and Statistics: Part 154,256

This kid hit more net than Miettinen did last year.

Wild fans are excited. For the first time in recent memory, Wild fans are legitimately excited about the team they pay to watch. We may have mentioned the excitement a time or two. However, just like last year when Mikko Koivu signed his contract extension, there are a small number of people who just don't seem to get it. 

I have always enjoyed math. I am not particularly good at it, but the one thing I really like about math is that, in most cases, the answer is either right or wrong. There is little gray area in algebra. Statistics, however, is a branch of math that allows for answers in gray areas, to be debated, to have as much or as little meaning as the reader is willing to assign to them. 

What brings us to this conversation? The Dany Heatley trade, and the rounds being made on it around the web. More after the jump. (Warning: long post)

Editor's note: If you would like to save time, as @Knowsknothing on Twitter pointed out, just read the last 39 words. 

Before we go too far, I want to make something clear. I think advanced statistics are great. They show different details of any given game that basic stats cannot. Advanced stats changed the game of baseball, tracking every situation and every moment of the game. For many (including me) advanced stats ruined the game. It isn't about the players anymore, it is all about the numbers. They don't play baseball, they play moneyball, and to play fantasy baseball you need an engineering degree. Yawn.

Then, advanced stats came to hockey, a free flowing, fast moving game where the importance of a shot on goal is open to debate. Where who wins the battle in the corner and skates out with the puck has some modicum of importance, but that level of importance depends on the situation.

Example given - If Wes Walz holds the puck in the corner for thirty seconds and allows a line change behind him, but loses the puck... did he win or lose that battle? Is there a statistical category for that?

In a game with good icing and bad icing, good penalties and bad penalties, chance creation, ongoing psychological battles, hits that rattle not only bones but confidence, and where a fight can change the course of a game, you're telling me that there is any statistic that can tell me if a player is good or bad? That one can look at a stats book and tell me if a player is going to have a resurgence with new team mates or fall flat on his face?

Two days ago, we had a post comparing Heatley's stats, in what stat people call "counting stats" as though to make them seem like numbers a child would use, to those of the leader on last year's Wild team. It showed that even with Heatley's "down year" he still would have led the Wild in 12 statistical categories. Our friend Ryan over at Matchsticks and Gasoline, a Flames blog, doesn't see the value in such a comparison

 I'm not a big fan of this type of analysis as it fails to take context into account.  The Sharks were a much higher scoring team (2.96 g/gm) than the Wild (2.48 g/gm) and had a much better goal differential as well (+0.42/gm vs -0.30/gm). It also doesn't take into account the competition faced or the zone starts. It can represent an interesting first look, but ultimately can be deceiving.     

For the record, we are directed to another friend of the blog, Kent Wilson's post about the Decline of Dany Heatley, a post that gives us some advanced stats for the past three seasons in comparison to Martin Havlat. It tells us that Heatley has gone from a great all around player to a power play specialist. The stats say that may be true. The post uses relative Corsi, Corsi, and production over 60 minutes to prove the point. Point proven. I agree, the stats say exactly what Kent says they say. Keep in mind what I said above... lots of gray area in statistics. 

While I agree, 100%, that the stats I used do not tell the whole story, they do tell a story. Kent's stats also tell a story. The stats I used were not meant to deceive. They were meant to show that Dany Heatley had a "bad" year in San Jose, and still would have led the Wild in 12 statistical categories with those same numbers. To compare, straight up and down, cateogires that can be compared straight up and down. My question has always been... when did goals, assists, and points stop being solid measures of a player's ability? The question, for today, as brought up by Ryan, is context.

 Having watched 82 Wild games last year (plus preseason), plus the 82 the season before that, and the season before that, and so on, I feel I have a pretty good understanding of the team, and its issues. When those issues are then confirmed when the GM makes moves to correct them, and says directly that those are the problems he was attempting to address, I feel pretty confident I knew what I was talking about. That is, unless of course, someone who also watched the Wild play 82 games last year, and the nine previous, would like to offer a differing opinion, in which case, I am all ears. (Eyes?) 

Context is called for, absolutely. So, let's build in some context. The problem Kent suggests is that the Wild were terrible at even strength. I agree. You know who had 947 minutes of TOI at even strength? Antti Miettinen. I won't go into relative corsi, corsi, or any other advanced stat for Mittens. The stats that matter for him don't exist. Missed open nets, shots fired twelve feet over a gaping net, and flubbed Koivu passes directly on the tape with no defender within 15 feet.

You want advanced stats? There they are.

Ryan points out that the Wild were a lower scoring team than the Sharks. Without a doubt, that is the truth. Now I would like to know your reason why you think that is. I'm not trying to be a jerk here, just trying to understand how trading a passer for a shooter makes the Wild score fewer goals next season. Why trading a guy who had zero chemistry with the center he was brought in to play with is a bad idea. Why trading a guy, who by all accounts, was not fitting in with the team is a bad idea.

Because his relative corsi is higher than the player he was traded for? I'm just not buying it. 

You see, there are issues at play here that even advanced statistics don't take into account. I don't want to be the guy screaming "GET OFF MY LAWN" at the crazy kids and their funny math games. I really don't. I like the new stats, and I feel they have a real place in analyzing the game. What I don't feel they do is provide the context that those who use them claim they do. 

I'm not an extremist. I don't think advanced stats are useless. I'm not Joe Morgan, and I'm not a writer for the Toronto Sun. I like numbers. Business is one of my strong suits, with which comes a strong ability to use the numbers of a business to inform myself. Numbers are great, as they give us insight into what happened, and a marginal ability to predict what will happen.

They do not, cannot, and will not ever be able to predict with certainty, nor do they give us the context we so desperately need. Accounting ratios, stock ratios, and profit and loss analysis would tell you that investing in Microsoft in 1978 was a terrible idea, that Facebook is worthless, and that Enron & AIG were great buys just before they tanked and brought the US economy with them. Numbers are great, and they can be dangerous when relied on.

When it comes down to it, numbers in business, and numbers in hockey, give us only the basis of knowledge. We need to have the ability to take those numbers and apply them to new situations. Trading Martin Havlat for Dany Heatley changes the Wild in ways that numbers will not express. The only way to know if this is a good trade or not, is to see what effect it has on the Minnesota Wild. Even then, will it be this trade, or the loss of Brent Burns and addition of Devin Setoguchi? The addition of Darroll Powe? Mike Yeo behind the bench? The youth movement? 

Here is what we know. Dany Heatley has a past, and he has not had the same type of season he had with Ottawa since he left. We get that, and we accept that. We also know, at least here in Minnesota, that Mikko Koivu is one of the best passers in the game. Koivu also commands respect with his shot, meaning the goalies and d-men will not be able to simply cheat toward Heatley. Finally, we know, even without statistics, that Antti Miettinen had dozens of chances to score, and missed the net completely. Using the fact that no one says Heatley will do the same, that right there tells me this is a great trade.

While we can appreciate the numbers game, and that advanced stats tell a different part of the story (not a better part), the numbers don't tell the whole story. The Wild needed someone who will shoot the puck. They went out and got players who will shoot the puck. Production, corsi, relative corsi can all be damned.

Notice my complete lack of statistics to back my argument? Here's why:

Heatley is better than Antti Miettinen, and he is different than Marty Havlat. The Wild didn't need a player better than Havlat. They needed a player who was different than Havlat and better than Antti Miettinen. 

There's your context.

X
Log In Sign Up

forgot?
Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

Join Hockey Wilderness

You must be a member of Hockey Wilderness to participate.

We have our own Community Guidelines at Hockey Wilderness. You should read them.

Join Hockey Wilderness

You must be a member of Hockey Wilderness to participate.

We have our own Community Guidelines at Hockey Wilderness. You should read them.

Spinner

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9355_tracker