March 23, 2009
Shots On Goal
Five For Fighting

by Timo Seppa

To get in the mood for writing this column, I went online to watch a vintage Flyers and Maple Leafs “linebrawl” from 1979. Does that make me a bad person? Though seeing Dave Schultz (NHL record 472 penalty minutes in a season) and Tiger Williams (NHL record 3,966 penalty minutes in a career) beat on each other did make me feel a tad nostalgic, that video was pretty mellow compared to the 5 fight marathon between the Flyers and Senators from March 5, 2004, featuring an NHL record 419 penalty minutes. Did you follow those links? We’re supposed to be more civilized as hockey fans nowadays, aren’t we?

While mainstream American media decries the barbarianism of a sport that –let’s admit it– they as a whole don’t understand or care for, the wonders of modern technology are fanning the flames of interest in hockey fisticuffs. Google “hockey fights” or “Dave Schultz” and in a moment you can be transported back to watching the golden era of hockey pugilism on YouTube. There are even websites dedicated to nothing but hockey fights, carrying fights from the minor leagues, juniors and foreign leagues in addition to the NHL.

Speaking of which, since when did they start fighting in the Finnish Elite League? That was news to me! While fighting still carries a match penalty and one game suspension in the SM-Liiga, their teams have either learned the necessity or the marketability of fighting since I used to watch Helsinki IFK games in person from 1989-1991.

Not only are all of these fights a fixture on the internet now –in addition to being guaranteed at one per game or your money back at your local AHL rink, but fighting majors and their common proxy, Penalty Minutes (PIM), have become a fixture in fantasy hockey scoring.

Pretend for a moment that you’re new to following hockey. Would you possibly consider penalties as being good? In fact, can you think of any sport where penalties are perceived as helping a team’s cause to win a game? While one could argue that fighting is hockey’s way of policing itself in the same fashion as intentionally hitting a batter in baseball, it would be hard to argue that a player that takes a high number of minor penalties, resulting in a high number of power plays against his team, is doing anything but hurting his team.

Now pretend for a moment that you’re new to fantasy hockey. Would you expect to get rewarded for PIM? Probably not. In comparison, fantasy soccer typically penalizes you for yellow cards and red cards received, while fantasy football ignores penalties. Yet when we think back to the popularity of fighting in hockey and the arguments for its necessity, it’s understandable that many fantasy hockey scoring systems reward PIM.

My two cents – What would make more sense is for scoring systems to reward fighting and misconduct penalties only, and perhaps for designated Enforcers only. As cool as it seems to be getting points for your pugilist and as simplistic as it is to use PIM as a proxy for fighting, it makes utterly no hockey sense to reward a real or virtual player for minor penalties that significantly increase the other team’s chances of scoring over the next two minutes.

Now that we have discussed the evils of rewarding PIM in your scoring system, we can deal with the reality of PIM likely being a part of your league’s scoring system. As with the last two Shots On Goal columns, we’ll look at what the best measures are for predicting future performance. For Goals, we learned that the last 8-10 weeks’ performance was the best indicator of future performance. For Assists, we learned that 6 weeks’ performance was the best indicator. In both cases, we found out that the previous 4 weeks’ performance gave a reasonable estimate.

Penalty Minutes, Goals and Assists: Past x weeks as a predictor for next week

Top 35        1     2     3     4    5    6    7   8    YTD

PIM, CV      87%   73%   67%   64%  62%  62%  61%  62%  64%

Goals, CV    77%   62%   58%   56%  55%  54%  53%  50%  57%

Assists, CV  62%   52%   48%   46%  45%  43%  43%  43%  47%

The table above shows data averaging the top players in Goals, the top players in Assists and the players highest in PIM. The Coefficient of Variation (CV) is the standard deviation divided by the average, a normalized statistic that allows equivalent comparisons between all players. Higher CV’s indicate greater variability in the data and hence a poorer predictor of future performance.

While PIM has a higher CV compared to Goals and Assists, indicating more variability in PIM performance, the overall trend is very similar to Goals and Assists. For instance, using last week’s PIM total as a predictor for next week’s PIM total is unlikely to be accurate (87% CV). Likewise, using the last 2 weeks (73% CV) or 3 weeks (67% CV) is not suggested. The last 6-8 weeks will give you the best results (61-62% CV), but are essentially equivalent to the last 4 weeks (64% CV) or year to date (64% CV), which are much more likely to be available as sortable statistics on your league’s website.

Keep in mind that these calculations assume 3 games/week, so adjust previous and future weeks if the number of games is not 3 games/week.

To apply what we’ve learned to fantasy hockey, let’s take a look at the NHL’s baddest boys to see what we can expect from them for next week:

Legend:

PIM: Penalties in Minutes (Year To Date)

PL: Predicted Low

PA: Predicted Average

PH: Predicted High

Player           Team  4 wks(CV) PIM    PL     PA     PH

Daniel Carcillo   PHI     61%    211    3.8    9.8    15.8 

Colton Orr        NYR     72%    162    1.3    4.8     8.3

Eric Boulton      ATL     73%    161    2.3    8.5    14.7

Riley Cote        PHI     48%    160    1.6    3.0     4.4

Eric Godard       PIT     77%    159    0.8    3.5     6.2

Jared Boll        CBJ     47%    156    3.7    7.0    10.3

David Backes      STL     64%    155    1.3    3.5     5.7

Zack Stortini     EDM     30%    154    5.1    7.3     9.5 

Cody McLeod       COL     57%    146    3.4    8.0    12.6

Arron Asham       PHI     62%    144    0.7    1.8     2.9

-After being traded from Phoenix, Daniel Carcillo took up the mantle of being the number one Broad Street Bully. Carcillo gave Philadelphia a league leading fourth player in the top 30 in PIM. The addition of “the Madman” has diminished Riley Cote’s and Arron Asham’s enforcer roles to a degree. While Carcillo is expected to tally 9.8 PIM next week, Cote and Asham are only expected to get 3.0 and 1.8 PIM respectively. Cote has had two majors over the past 12 games while Asham has had only one major penalty.

-The rough and tumble Atlantic Division sports nine of the top 30 bad boys in PIM. Colton Orr is one of two Rangers on the list, with Brandon Dubinsky falling just outside the list. Orr’s low recent PIM and high CV do not make him a good fantasy play right now. This trend should continue to hold, as Tortorella’s Rangers cannot afford bad penalties down the stretch.

-Eric Boulton of the Thrashers also has high variability (73% CV), but he’s been pounding out PIM at a steady 8.5 PIM/week clip recently, with 6 majors over the past 12 games.

-Eric Godard has the job of protecting Sidney Crosby and Evgeni Malkin during their Atlantic Division battles, but his recent track record of 3.5 PIM/week and his high variability suggest better Enforcer choices elsewhere.

-Coming in a distant second to Carcillo’s 324 PIM last year was Jared Boll of the Blue Jackets with 226 PIM. Boll is always a good fantasy play, with a low CV. Expect a reliable 7 PIM over the next 3 games.

-After the Atlantic Division, the hard hitting Central Division comes in second for the greatest number of top 30 players in PIM with seven. The St. Louis Blues have three of them. With 24 Goals and 21 Assists, David Backes is no run of the mill bully. In fact, if your league tallies points for majors instead of straight PIM, Backes is definitely not your man. He has not had a major since January 3rd, though he continues to get sent to the box for a steady stream of minor penalties. Even for PIM, the young Center is not a strong choice right now, averaging 3.5 PIM/week recently.

-Zack Stortini would be much higher on the PIM leaderboard if he had played more than 47 games this season. In addition to being a solid scorer, he’s put up 7.3 PIM/week at a low variability (30% CV). That said, Stortini was a scratch on Sunday, so keep your eye on his status. Incidentally, the Northwest Division comes in third with 6 players in the top 30 in PIM.

-Cody McLeod of the Northwest Division’s Avalanche has been a nice fantasy option of late, good for 8 PIM/week. He has no fighting majors to his name over the last 4 weeks though, with the PIM totals inflated by a pair of 10 minute Misconducts.

If you are looking for PIM from players outside of this top 10 list, remember to total the last 12 games’ PIM and to divide by 4 to get the expected number for next week. Then adjust for the number of games if their team is playing more or less than 3 games this week.

It’s been a good run for me in fantasy hockey again this season, pulling off three exceptional playoff weeks, capped by defeating a worthy opponent in the finals for the second year in a row – Fred’s Five for Fighting – With the name apropos to this week's column name.

I wish everyone the best of luck in your quests to bring home your fantasy league championship titles!

Timo Seppa is an author of Hockey Prospectus. You can contact Timo by clicking here or click here to see Timo's other articles.