Monday, June 9, 2008

Research at it’s worst

Just a couple of questions for John Haisken-DeNew and Matthias Vorell, the researchers that ginned up, “Blood Money: Incentives for Violence in NHL Hockey.”

1) Why only use wingers in your research? Enforcers can play defense as well.
2) Why use a very subjective source (wins/losses as determined by posters at as an important criterion in your research?
3) Why only assists? Enforcers score goals too you know…
4) Why propose a formula for eliminating fighting when fighting in hockey is supported by the majority of fans, players themselves, and has been positively correlated with television ratings?

From one researcher to another, tighten up!

1 comment:

dmg said...

(1) Not only wingers were used:
"On average in OLS, a
wing player who fights and wins the fights has an even larger wage penalty of $11,240. However, when controlling for fixed effects, a wing player who wins fights earns a wage premium of $18,135."
What they're saying is that the effect of a won or lost fight on salary is more significant if a player is a wing but all other factors are held equal. I think they then discuss the implications for wingers in more detail because this position has a much more significant effect than other positions. Results for all positions are given in the appendices.

(2) The subjective source isn't ideal, but it's probably the only place they can get that data.

(3) The reason assists are discussed that that the value for scoring an assist is less than that of having a fight, a point they wanted to make. What I dont understand is this passage: "he can either make an assist (1 point)
and earn an additional $10,930 or half a goal and make the same (1 goal =
2 points), or have a fight and earn $10,940, which is even $10 more". Are they saying that in their model they're assigning a value of one point to assists and two points to goals? I don't get that, and don't see any clarification on it.