A couple of weeks ago I stumbled across a great study dealing with the effects of uniform colour conducted by a University of Florida professor and his colleagues and published in the May edition of Social Psychological and Personality Science. Building on past studies on the association between black uniforms and perceived aggression, the authors studied a 25 year period in the NHL from the 1984-1985 season to the 2009-2010 season – a sample size of 52,098 games and nearly one million penalty minutes – to determine whether uniform colour has an effect on aggression and penalties called in the NHL.
After the jump, I'll get into the results and potential real-world applications.
First off, the reasoning:
"Color not only represents psychological associations but can also influence them. Relationships between color and attitudinal valence are pervasive in nearly every culture (Adams & Osgood, 1973), particularly associations between ‘‘black’’ and ‘‘bad’’—and between ‘‘white’’ and ‘‘good’’ (Williams, Moreland, & Underwood, 1970).
Evidence of such color-valence associations comes from experimental social cognition. Individuals implicitly associate the color black with negativity (Meier, Robinson, & Clore, 2004), immorality (Sherman & Clore, 2009), and criminality (Vrij, 1997). This association exists early in development; young children believe dark boxes contain negative things (vs. white boxes; Stabler & Johnson, 1972). Research on racial prejudice shows a strong connection between skin color and attitudinal valence (Dasgupta, McGhee, Greenwald, & Banaji, 2000; Williams, 1966; Wittenbrink, Judd, & Park, 2001). Even within particular racial or ethnic groups, lighter-skinned individuals have more status and face less prejudice than darker skinned individuals (see Maddox, 2004, for a review). This effect also holds in 48 of 51 cultures (Iwawaki, Sonoo, Williams, & Best, 1978); preferences for lighter skin are nearly universal among cultures that value skin color as a sign of beauty (Russell, Wilson, & Hall, 1992). Furthermore, implicit attitudinal evaluations of the colors white and black are significantly related to implicit racial prejudice (Smith-McLallen, Johnson, Dovidio, & Pearson, 2006)." – Webster et al
Essentially, we associate darker colours – and particularly black – with bad. Similarly, we associate lighter colours – white, in particular – with good. The purpose of the study was to determine whether or not our colour-valence associations transfer to uniform colours in the NHL and perceptions of aggression, or "good" vs. "bad".
One interesting – and very important – finding of the study was the lack of correlation between bench minor infraction (BMI) penalties and non-BMI penalties. Why is this important?
"Distinguishing between BMI and non-BMI penalties is important for two reasons. First, non-BMI penalties are typically aggressive acts (fighting, roughing, elbowing, kicking, kneeing, tripping), whereas BMI penalties are typically nonaggressive mental lapses (being on the ice when one should not). Second, a quasi-experiment can benefit from the addition of a nonequivalent dependent variable because it helps rule out threats to internal validity and helps bolster the construct validity of the effect (Shadish et al., 2002)."
In other words, if the uniform colour effect were to apply to all penalties across the board (both "aggressive" and "nonaggressive" ones), then there should be a correlation between BMI penalties and non-BMI penalties. If this were the case, teams wearing dark uniforms should receive more BMI penalties than teams wearing coloured uniforms. However, the study found that "BMI and non-BMI penalties were not significantly correlated". In other words, the uniform colour effect was not observed for BMI penalties – typically "nonaggressive" penalties – but was observed for non-BMI ("aggressive") penalties. This provides evidence that the uniform colour effect applies to the NHL and supports the idea that the uniform colour effect is "generalized to aggressive penalties but not nonagressive ones".
Moving on to the crux of the results, the study showed that teams were assessed significantly more minutes in penalties when wearing darker uniforms – and in particular, black jerseys – than when they were wearing lighter uniforms. In fact, the average team wearing black jerseys was assessed 1,528 minutes in penalties compared to 1,386 minutes in penalties for jerseys of another colour. This is a difference of an additional 142 penalty minutes over the course of a season, or about 104 seconds per game. Furthermore, for the 10 teams in the study that switched to or from black uniforms at some point during the study (i.e. they only wore black uniforms for part of the duration of the study), the effects were even larger.
If you're wondering how this relates to the real world, consider that the average NHL team this season spent 445.4 minutes on the penalty kill, giving up 46.6 goals against. That's one goal every 9 minutes and 33 seconds. We should also consider that because fighting penalties are pretty clear-cut (and thus unlikely to increase based on uniform colour) and misconducts are incredibly rare (just 2.2% of all penalties called last season were misconducts), the increase in penalties will mostly come from additional minor penalties being called and from some minor penalties being called major penalties. In other words, a large chunk of that 142 additional penalty minutes will result in having to kill a penalty. At the average rate of a goal every 9 minutes and 33 seconds, the additional 142 penalty minutes (on average) given to teams wearing black as opposed to another colour would result in approximately 15 more goals against if all that time was spent killing penalties.
The Winnipeg Jets, however, are not your average hockey team. The Jets gave up 58 power play goals last season, third most in the NHL, and spent 468 minutes killing penalties. That works out to about one goal against for every 8 minutes and 4 seconds spent killing penalties. At the Jets’ rate, 142 additional minutes killing penalties would result in about 18 more goals against.
Now granted, not all of those 142 additional penalty minutes racked up by teams wearing black will result in having to kill a penalty; some of those penalty minutes will result in offsetting minors, some penalties will end early due to a power-play goal being scored or another penalty being taken, and there might even be a misconduct in there. And the Jets' uniforms aren't truly black, but rather a very, very dark navy. Still though, even taking those factors into account, we could easily still be talking about a difference of a couple of wins per season.
If you're a skeptical fellow like myself, you might have done a quick eyeball test to see if this matches up with our perceptions. And if you did, you've have found, like I did, that the five NHL teams with the fewest penalty minutes this season were, in order:
- Nashville Predators (yellow)
- Detroit Red Wings (red)
- Phoenix Coyotes (maroon)
- New York Islanders (blue)
- Carolina Hurricanes (red).
Notice anything about those five teams? I sure did. All five of those teams wear coloured uniforms at home and none are particularly dark. Last year's top 5 was Florida, Nashville, Chicago, Detroit, New Jersey, with Carolina and Phoenix at 6th and 8th. Again, each of these teams' "dark" uniforms are light coloured uniforms. In 2009-2010 the top 5 was Nashville, Detroit, New Jersey, Buffalo and Phoenix, with Chicago finishing just 3 minutes out of the top 5.
Perhaps it's just a happy coincidence. Or perhaps the Predators are on to something with those mustard yellow uniforms of theirs.
In addition to the chief study, the researchers also compared penalty minutes at home between the same teams wearing different coloured uniforms. With the benefit of the NHL switching from white home uniforms to dark home uniforms after the 2002-2003 season, the researchers had the benefit of what they called a "quasi-experiment" between the same teams wearing white uniforms vs. dark uniforms at home.
Sure enough, they found that NHL teams were assessed significantly more penalty minutes in home games after they switched to wearing dark uniforms at home at the start of the 2003-2004 NHL season. For example, in 2003-2004, the average NHL team was assessed 615 penalty minutes when wearing their darker home jerseys, an increase of 54 minutes from an average of 561 minutes in the previous season when teams wore white at home. That amounts to a 9.6% increase, or an additional 1.32 minutes in penalties per home game. In addition, similar to with BMI vs. non-BMI penalties above, there was no significant difference between home goals between the two seasons, supporting that the white vs. dark/coloured uniform effect is "generalized to aggression (penalties) but not performance (goals)".
Although the results show that teams wearing dark are assessed more penalty minutes, as the study mentions:
one issue these studies cannot empirically resolve is whether (a) black jerseys are causing more aggression, (b) white jerseys are causing less aggression, (c) both are affecting aggression in opposite directions, or (d) the supposed causal link is spurious (a ‘‘third-variable’’ problem).
In addition, the study does not address whether uniform colours affect aggression or simply affect our perceptions of aggression. Nevertheless, there is a clear association between colour and penalty minutes; teams wearing dark colours clearly get penalized more, and teams wearing lighter colours clearly get penalized less.
With that in mind, is this something that the NHL should take into account in the future? Furthermore, is this something that the Jets should take into account? That is, should the Winnipeg Jets consider switching to a lighter coloured home uniform? And should the Jets' third jersey, when it inevitably arrives in a season or two, also be lighter in colour? The evidence would suggest that they should and I tend to side with evidence. At the very least, I figure it can't hurt.
I wish I could take credit for personally going over 50,000+ NHL games worth of data, but unfortunately I cannot. Alas, I've just interpreted some of the data that I found particularly interesting or noteworthy. The source of the data is a study conducted by Gregory D. Webster, Geoffrey R. Urland and Joshua Correll and published as "Can Uniform Color Color Aggression? Quasi-Experimental Evidence From Professional Ice Hockey" in the May edition of Social Psychological and Personality Science. If you'd like to read the full article, you can purchase it here. Alternatively, for those readers with access to journal databases (for example, university students, or certain professionals, or those readers in possession of a valid library card), I'd suggest searching through your university's/employer's/library's databases to see if you can obtain the article for free.
