MythBusters: is an Aggressive Penalty Easier to Kill Off


Introduction:

My friends this is Esa Tikkanen doing what he did best, but I am sure all of you know this. I am on a personal level thankful that the Oilers don’t have a player like him: mean as cornered raccoon, spiteful, excellent penalty killer, and a decent scorer. He would put a major kink in my penalty project . . . as he would have taken a lot of lazy penalties when he was not being lazy but instead just a jerk, plus he was one of the Oilers’ best PK guys so that would have hurt the stats too. In all seriousness, Mr. Esa is exactly what the Oilers need these days: a real NHL player in multiple aspects of the game (shadow, PKer, offensive flair, veteran, and playoff performer).

When I started this project many months ago, it was a way to figure out if my sight was functioning right or was captain-two-minute-minor actually just taking a lot of “lazy” penalties, instead of his trademark aggressive play. Over time, posting on other fine blogs, and listening to gawd-awful colour commentators a trend has emerged: that people think that aggressive penalties are easier to kill then lazy penalties. This interesting (and unverified) idea was never part of the original project, instead the stats were invented to judge fighters and pests not rationalize people taking ‘good’ aggressive penalties and ‘bad’ lazy ones. I repeat, I do not think Lazy Penalties are ‘bad’ and Aggressive Penalties are ‘good’ (but I may do a blog on the idea in relation to Nietzsche’s good, evil, and bad concepts . . .).

Of course, no one ever listens to folks arguing on the internet (because it would get in the way of their point) so my project of Aggressive Penalties and Lazy Penalties has sunk in to the murk of this discussion. I have slowly become fine with this, so when Doogie2K and Derek pressed me on looking at the idea of Aggressive Penalties being easier to kill of the Lazy Penalties I reluctantly took this project. I never thought this is the case, and believe that penalty kill involves different factors—who takes the penalty, the time of the game, the PP & PK systems being used, and the intangible know as momentum—to determine of the penalty is easier or more difficult to kill off.

Methodology:

I counted all of the penalties by hand from the Oilers website’s "Boxscores" section. If the Oilers had a PPG against during the time the play was off for then it could as a penalty that was not killed off, henceforth know as power play goal causing penalty (I have no acronym). If there was a goal but not a PPG (so if the players had coincidental penalties, or a misconduct) then the goal was not attributed to the penalty. If the Oilers where two men down then I counted the first penalties based on the listing on the Oilers’ site to determine the penalty that caused the goal. I did not include ‘bench minors’ in penalties killed off, since they are neither Aggressive nor Lazy Penalties, but it does seem that the Oilers had a goal scored on them during one of these calls (based on the system of attributing goals that I used). Lastly I did not include the Penalty Shot for penalty killed off, but it was saved so it is an irrelevant point. Also please note that I counted more penalties then the computer did (by 27) for the 2009-2010 season; I know Bruce has tried to help find the missing penalties in the last post, but I still can’t figure out why. Because of this I am not going to use the totals of the penalties in the last post, but I am going to use the ratios and percents, OK?

Results:

Goal Causing Penalty

Penalty Successfully Killed off

Lazy Penalties

52

201

Aggressive Penalties

15

158

Conclusions, Ideas & Notes:

  • This means that the Oilers allowed goals on 18.66% of the penalties they took. This is of course different then the fact that the Oilers had a PK of 78% and 26th in the league (of course this means the Oilers let in goals on 22.00% of the short hand situations). On the one hand, this should be different based on misconducts, coincidental minors, and fighting penalties, that do not have a power play goal count against them. On the other hand, this test is about the fact that Aggressive Penalties are easier to kill off so the fact they produce less short handed situations they could be consider 'easier' to kill off.
  • Where goal causing penalty of a particular type was examined in relation to itself and penalties kill off of the same particular type, Aggressive Penalties only had 8.67% chance of being scored on, while Lazy Penalties had 20.55% chance of causing a power play goal. This is significantly higher and not what I expected but is the first piece of evidence linking the fact Aggressive Penalties maybe easier to kill off.
  • Also in relationship to the total number of penalties counted, if a penalty is take, it is only a 3.52% chance it is a goal causing Aggressive Penalty while it is 12.21% likely that it is a goal causing Lazy penalty. Since Aggressive penalties take up 39% of all penalties and Lazy Penalties take up 61%, you would expect that it is more likely that goal causing Lazy penalties would have a higher percent; but it is over three(ish) times more likely that a goal causing Lazy Penalty occurs then a goal causing Aggressive Penalty, it is only one and a half(ish) that a Lazy Penalty occurs instead of a Aggressive Penalty. This leads more support the myth that Lazy Penalties are harder to kill off then Aggressive Penalties.
  • I think that this difference is heavily relied on the fact that Aggressive Penalties are more likely to be offsetting (such as fighting) or do not result on a power play situation (such as misconducts). This of course is not worked into the stats above since they are attempting to examine the likely hood of a single penalty cause a power play goal against, not a subjective analysis of watching the goal and attributing the penalty responsible.
  • Also, there is a large difference on the type of player that takes each of the penalties. While I have not done the individual breakdowns of the Oilers’ players for 2009-2010, a hint can be gleamed at looking at the 2008-2009 numbers. Most players, 18 out of 25 players had negative differentials, and hence took more Lazy Penalties then Aggressive Penalties. The only players that had a positive differential were: Moreau, Strudwick, Staois, Smid, MacIntyre, Peckham, and Zorg. Other then Staois and Moreau none of these player spent a large amount of time on the 2009-2010 penalty kill, so a Lazy penalty would more likely be caused by one of the regular penalty killers (and I think this would cause it to be more likely that a goal was scored when a Lazy Penalty occurred.)

Conclusion:

I do not think this evidence leads us to support the myth of “the Lazy penalty to be harder to kill off then the Aggressive one”. But it does look like the fact that a power play goal is more likely to occur on a Lazy Penalty then an Aggressive Penalty. This is clearly not the same thing!

2 comments:

Doogie2K said...

Or, stated another way, we have a data point, but not a trend, and certainly nothing approaching statistical significance. Still, a bit of a surprising result: I was expecting something much closer to equivalence.

Maybe someone with pull can convince Vic to write a script to do this automagically for the rest of the League so we can look at this proper-like.

B.C.B. said...

I was quite surprised Doogie, as well.

My stats are not very good, as I slowly am using internet tutorials to learn how to do this. I did not use my 'fancy' new skills, I just counted and used the divide button on the computer.

Thanks for adding your thoughts and the proper words to the debate (aka 'data point').