Monday, 27 February 2012

Additional Statistics on the Impact of Fighting

Back in January I posted an article – More Facts and Less Emotion – that presented available facts on how fighting and fighters impacted the game of hockey.   I think that it is important to take emotion out of the argument and stop using perception as a valid reason to resist change.  With that in mind I have done some additional analysis that hopefully convinces NHL execs, coaches and fans to step back and focus on what is happening  versus what could happen.


Looking at Non-Fighting PIM and Fights Per Game

I have already presented this analysis but I was able to improve on the accuracy of the data through better online sources.    I started by researching Fights per Game since 1998, displayed in the chart below by the red line.  I then take the total PIM for each team, remove all penalties related to fighting (majors, misconducts and match penalties) to come up with Non-Fighting PIM per Game, shown as the blue line in the chart.



















I made this point in my previous blog - the chart shows a clear trend that as fighting decreases, all other penalties, those not related to fighting, also decrease.  This clearly is counter to the pro-fighting argument, that if fighting is eliminated then cheap shots will increase.   There is only 1 major anomaly, the season of 2005-2006 which was right after the lock-out.  My opinion is that the NHL implemented significant rule changes to speed up the game and eliminate hooking, tripping and interference.  Looking at game logs from that year it appears that it took some time for the players to adjust to the new rules and PIM were up dramatically while fighting was down.   However the downward trend returned immediately the next season.   Therefore the overall trend since 1998 is that if you decrease the violence attributed to fighting, overall violence also decreases.

Note:  I could not remove the impact of the 2-minute instigator from the analysis above as, unlike Majors, Game Misconducts and Match Penalties, they are not widely reported on stat sites.  However if they were removed then my analysis is even more accurate as the Non-Fighting PIM would be even lower.


Non-Fighting PIM for Teams That Fight

I went a step further and calculated the Non-Fighting PIM per Game for the Top 6 teams that fought (based on number of fights reported by one of those popular NHL fighting websites).  I also did the same calculation for the Middle 6 and the Bottom 6, the half-dozen teams that fought the least.   The chart below shows the results.
















The thing that quickly becomes clear is that the teams that fight the most (blue line) consistently rack up more non-fighting PIM than the Middle 6 or Bottom  6 (red line).   It could be the instigator penalty, which I noted earlier is not removed from my numbers.  But that penalty is not called very often – according to a recent Kerry Fraser column on TSN, in approximately 395 fights only 30 instigators were identified with the appropriate penalty so far in 2012.  Therefore it cannot explain the difference shown above.  

Two things come to mind when I look at this chart:
  1. The teams that fight the most supposedly have those enforcers to ensure that things don’t get out of hand by policing the game.  However the chart above shows the opposite; if you use violent players on your roster as a strategy, then you are more likely to generate more Non-Fighting PIM per Game.  Again this is the opposite of what we hear from proponents of fighting in the game.
  2. If you are going to have fighters then you should realize that you also should practice your penalty kill because you are going to give your opponent an extra man advantage at least every second game.  And get yourself a good goalie because you’re going to need it.

Average Points per Season for Teams That Fight

Lets take the same groups, the 6 top teams in fights, the middle 6 and the bottom half-dozen that fight the least.   If you average the season points earned by each team within their groups and plot them on a yearly basis you get the results below.

















As you can see, in most years the teams that fought the least scored more points on the season versus those that had the most fights.  The pro-enforcer crowd may point out that the Top 6 fighiting teams have shown greater improvement in earning league points over the past 12 years - and they would be correct.  But that's because fights per game are down over that period meaning if you fight less, you earn more points in the standings.

Despite the fact that these groups change from year to year, as teams either add enforcers or drop them, the pattern stays consistent.  Only in 2007-2008 and marginally in 2010-2011 did the top 6 in fighting manage to outpoint the 6 non-fighting teams.    It’s pretty clear that playing tough with enforcers is likely not going to result in success, based on the past 12 years of NHL history.


Tracking Success With The Fighting Groups
What if we took those same groups and looked at the success they had during the season; who finished out of the playoffs, who took 1st place in their division and how far they went in the playoffs. 
 







Over the past 12 seasons the teams that fought the least had more success, in terms of finishing first or going deep into the playoffs.   Knowing that fighting almost disappears in the playoffs (when winning is important, enforcers either play with a lot more discipline or sit in the press box) the key columns to look at would be Out of Playoffs and Finished 1st in Division.   Here the non-fighting teams have a pretty clear advantage.   Their strategy, of not having a fighter who takes away a roster spot from a more skilled player, appears to help them make the playoffs on a more consistent basis and they finish in 1st twice as often.



Summary

The pro-fighting crowd needs to dial back the emotion, put their perceptions aside for a moment and, with an open mind, study the real impact of fighting.  I believe that they would begin to see that fighting is not effective and a poor strategy for teams to employ.   As the data above shows, when fights per game drop then other PIM not related to fighting will also be reduced.   Leaving a position on your roster for someone whose primary role is violence means your team has a  player who is less disciplined and likely will put you on the penalty kill more often.   He won’t be enforcing but instead will be enforced by the referees.   And teams that fight are less likely to earn more points in the standings than those that don’t, and statistically will miss the playoffs more often.

Every team can and should play tough but without the fighting.   To be successful, you have to realize that it’s not part of the game.

5 comments:

  1. Hockey has been the same since it began, and the self-policing aspect has been what sets it apart from the other sports. The players understand the risks of fighting, and that is why they are paid alot of money to play.

    Please leave the game alone and watch a bevy of sports that will satisfy your need: figure skating, curling, golf, tennis, european hockey ... the list goes on.

    We don't want you as a fan. Get lost.

    ReplyDelete
    Replies
    1. Anonymous. Of course. Your tone implies a very tough man but your hiding in anonymity reveals you for the coward you are.

      Delete
  2. Actually hockey has not been the same since it began. A knowledgeable fan can tell you that it has evolved and changed quite a bit over the past 100 years. Although fighting has always been tolerated, it has never been part of the game as there are penalties in place to deal with it. The fighting today is not about policing it's about revenge and removing the few minutes of fights from a game will not destroy the game, it will make it better. There are lots of sports that would appeal to the fighting fan, like MMA and boxing. But I would not ask any hockey fan to leave the game just because they have an opinion different from mine.

    ReplyDelete
  3. Excellent analysis. You need to get together with Wired writer Brandon Keim (http://www.wired.com/playbook/2012/04/opinion_keim-hockey-fighting/).

    ReplyDelete
    Replies
    1. I have sent the link to Brandon already, via Twitter. He's responded with some interesting research that should be done, but we were not sure how we could ever collect the data. I did try to post a link to this article on the Wired site but they wouldn't allow it, likely the URL triggered some anti-spam block.

      Delete