• Nothing provokes me like bad statistics

    by Tyler Dellow • September 6, 2011 • Uncategorized • 19 Comments

    I’m sure that this will come as a surprise to Laura Robinson, but not all hockey fans are troglodytes. I had something she wrote on the Canucks in the playoffs come to my attention after Neate Sager, retweeted it with the comment that “Agree or disagree, Laura Robinson usually has a provocative take…” When I responded that this is also true of Ernst Zundel, Sager asked if I was always this predictable and said “…I knew I’d get some knee-jerk angry-white-male response.”

    I don’t have Sager’s circa. 1993 worldview software installed but I do have a decent bullshit detector. And a quick glance at Robinson’s piece made it clear that she was most likely spouting bullshit statistics in support of her theory (her theory, which she’s been peddling since at least the early part of the century, is that sports make men violent. There’s a lot more but it’s never been well supported.) Here’s what caught my attention:

    Vancouver police and VANOC knew deeply disturbing behaviour was occurring during the Olympics, despite the report’s multiple glowing references to the contrary. Gangs of young males took over public space in Vancouver’s District 1 — where many sports bars are located. According to the police, District 1 saw an increase of 233.3 per cent in reported sexual offences during the Olympics. Women Against Violence Against Women saw a spike in the number of women they accompanied to hospital for the rape kit in the 24 hours after the gold-medal men’s game — all coming from hockey celebrations. Vancouver’s Battered Women’s Support Services reported an increase in domestic violence of 31 per cent.

    During the Stanley Cup playoffs, sexual offences rose 133 per cent in April and 72.2 per cent in May in District 1. By July they had decreased, 68.8 per cent lower than the year before.

    Last paragraph first. As a preliminary point, the Canucks were in the playoffs in 2010. Robinson, who appears more interested in having theories than in testing them, ought to explain why this year’s version of the Canucks caused so many more sexual offences in April than last year’s version. The same conditions – a Canucks playoff run – existed. The Canucks, of course, went out early in May in 2010 and played through May this year. So maybe she’s got a point on the 72.2% increase in May (from 11 to 19). After all, that’s an increase, albeit a relatively small one in absolute figures, which tends to lead to alarming swings in percentages.

    I assumed that she didn’t mention the June numbers – when the Canucks playoff run ended ignominiously because they were simply to horrible to mention. No Canucks playoff hockey in June 2010 and an awful ending to Canucks’ playoff hockey in June of 2011 must surely have unleashed the “…violent subculture of masculinity fed by hockey, UFC and other activities that define men through violence and (lead them to) start drinking hours before they commit public or private acts of violence.” I’m a brave man with a strong stomach so I looked. 13 reports of sexual offences in District 1 in June of 2010 and…11 this year.

    Huh. Guess she didn’t see that page. Oversight, I’m sure. She just went April, May, July. Whoops.

    As for the Olympics, the 233.3% increase in reported sexual offences in District 1 sounds horrifying. Until you realize that this meant that it spiked from 10 to 3 between February 12 and March 1 and remember that that was basically 18 Fridays or Saturdays in a row, about a 200% increase in the normal number of Fridays and Saturdays in an 18 day period. Curious how those numbers match up.

    That leaves two pieces of data: the increase in rape kit requests and the increase in domestic violence of 31%. On the information presented, we have no way of gauging how significant that might be. In another piece that Robinson wrote, she wrote: “Women Against Violence Against Women, reported they normally hear from five to six women a month who ask to be accompanied to the hospital for the gathering from their bodies evidence of rape. In the 24 hours after the men’s Olympic gold-medal hockey game, four contacted them, all from hockey celebrations.”

    Is four an unusual number for one 24 hour period? Sounds like it. Did WAVAW have a bigger presence during the Olympics? Do four calls in a 24 hour period happen on other occasions? We have no way of evaluating this? The same goes for the 31% increase in domestic violence. I’d trust Robinson to present the data fairly except…well, re-read everything up to this point if you don’t know why I don’t trust her.

    “Provocative” does not always mean “insightful.” And when someone is citing percentages that imply a “3″ in the denominator, you should be suspicious that they are citing data on relatively rare events in which large swings in percentage might not be uncommon. This is also true for when they skip a month in talking about the increases in sexual offences wrought by a playoff run.

    Does Laura Robinson have a point? I’ve got no idea. I’m pretty sure that the data doesn’t support her case though and if you want to make a provocative case, you’d best have facts to support it. She doesn’t.

    About Tyler Dellow

    19 Responses to Nothing provokes me like bad statistics

    1. September 6, 2011 at

      The alleged increase of 31% that Robinson so indisputably links to “hockey celebrations” makes no sense at all.

      Victims of domestic violence typically DON’T report their abuse for multiple reasons: they’re scared of their bf/husband, they’ve been manipulated into thinking they somehow deserved what happened, they’re afraid to ‘abandon’ him, they have little-to-no self esteem, they’re financially dependent, are afraid of what might happen if there are children involved… really, the list of factors contributing to the problematic of sexual and domestic violence goes on and on.

      With no source, no cred pertaining to abuse and assault, and no real breakdown of what she thinks she’s trying to say, there is no way I can simply accept Laura Robinson’s story and stats at face value, as they go in the exact opposite direction of what victimization actually involves.

      Because the topic of violence toward women is one that needs LESS stigma, LESS blame on the victims and LESS excuses for the assailants involved, her article was inflammatory and irresponsible at best. People like her are partly to blame for the culture preventing women from reporting violence against them. and THAT’s what’s dangerous. Not fucking hockey.

    2. Hemmertime
      September 6, 2011 at

      All four were coming from Hockey celebrations. It might be worth noting that every single place in that city was probably considered a hockey celebration. Every single street corner or sh*t hole bar, or house party – was a damn hockey celebration. I’d be more interested if they could find four people (unharmed even) that weren’t at “hockey celebrations” during after the gold medal game.

      • Doogie2K
        September 6, 2011 at

        Exactly. Not only is this small-sample stuff, but also easily explained: partying and chaos creating an atmosphere easily taken advantage of by those with ill intent. Spiked drinks, loud noises/crowds, dark places…

        But yeah, it’s those Goddamned sports and that Goddamned male testosterone. Should lop off everyone’s dicks then they turn 14.

    3. September 6, 2011 at

      Furthermore, can you imagine if she were right?

      “Sorry, ma’am, there’s nothing we can do about the violent sexual assault you just suffered. My suggestion is to lock yourself in a bomb shelter until the playoffs are over. Don’t talk to anyone, don’t know anyone [70% of victims know their assailant], and generally just don’t be born with a vagina. GO LOCAL SPORTS TEAM GO!”

    4. Peacecountry
      September 6, 2011 at

      Mr Dellow,
      I don’t often agree with your take on things but two articles in a row have been bang on. It looks like we may have found a common ground.
      Back on topic. The first day of my into stats course at UofL my prof showed exactly what your argue in this piece. He used murder going up by 50%. Sounds significant but not so much when the murder rate goes from 1 to 2. It is to bad that Ms. Robinson didn’t Dr. Connelly while she was guest lecturing there.
      Furthermore I find it unlikely that she would not have taken any statistics courses while persuing her post secondary education in sociology. If that is the case I would be shocked. The flipside of this is she is actually misrepresenting that facts knowingly.

      That would seem to be a horrible way to make a ‘provacative case’.

      • September 6, 2011 at

        Your stats teacher needs to be fired, valuable insight or no. The murder rate going from 1 to 2 is a 100% increase. A 50% increase is from 2 to 3.

        • DSF
          September 6, 2011 at

          Do they teach spelling at your school?

          PURSUE.

    5. Spartacus
      September 6, 2011 at

      Lawyered!

    6. Matt.N
      September 6, 2011 at

      The Edmonton Oilers have been keeping the streets safe for women for 5 years and counting.

      • SumOil
        September 8, 2011 at

        We need a rec button in here

    7. Edgar
      September 7, 2011 at

      Keep on going in that paragraph you’re dissecting…

      “During the Stanley Cup playoffs, sexual offences rose 133 per cent in April and 72.2 per cent in May in District 1. By July they had decreased, 68.8 per cent lower than the year before. Some say sex assaults increase with an increase of people, but Tourism BC stats show there are as many hotel stays in the average summer month as there were for the Olympics. There is no correlation between increased hotel stays and sexual offences, according to Vancouver police.”

      The logic makes no sense:
      1) Sex offences rose during the Stanley Cup
      2) Some argue that sex offences would naturally rise with an influx of people, but
      3) Hotel stays were not unusually high during the OLYMPICS (weren’t we talking about the Stanley Cup?) and
      4) Increased hotel stays do not correlate with increased sex assaults

      Point 3 makes point 4 irrelevant. It’s a safe assumption that the population of Vancouver increased during the Olympics, therefore since hotel stays didn’t increase during that time, there clearly isn’t a strict correlation between hotel stays and increased population.

      Also, notably absent are any statistics about hotel stays during the Stanley Cup. This paragraph just reeks of poor reasoning and manipulative use of statistics.

    8. Jeremy
      September 7, 2011 at

      Hey Tyler,

      I realize this is not the best place for this, but need to ask you couple questions about the history of Oiler fans posting pictures of game footage on line. Could you please email at your convenience. Thanks. Great stuff by the way about the crappy use of stats. Provocative should not be a license to lie.

    9. Ice Rink Eschatologists
      September 7, 2011 at

      I’m not going to try to defend this particular article by Robinson. To be honest, I didn’t read it and your analysis and critique seem perfectly reasonable to me. It’s clear she’s at least being selective about her evidence (and maybe worse).

      Still, I feel like someone should say something in defence of the idea she’s trying/failing to make a case for, lest we end up with a baby and bathwater scenario. I don’t have better evidence that violent sport viewing increases rates of sexual (or other) violence among men. To be honest, I think that phenomena may be so internalized as to appear invisible (or at least inconclusive) in social level statistics. After all, it’s early September and we’re still talking about and thinking about hockey. This isn’t a game that affects people only in its viewing, and from a sociological perspective, we can safely say that the social expectation for men to hold and express interest in sports has a normative impact on how masculinity is constructed and manifest.

      I’m a huge hockey fan – I watch a lot of games at many levels and I even play recreational hockey. I can’t honestly say how much of that is ‘my choice’ as an individual, and how much of it is me living up to social expectations as a masculine man. Regardless of the ratio of those forces, my joy for the sport is still true and meaningful to me.

      I’m tempted to share anecdotal evidence of other sports fans (because, of course, I think of myself as a model fan and certainly not a violent aggressor). I’m sure any rec hockey player has stories about the many, many players who take things Way too seriously, or become violent for their own apparent enjoyment. But they may be great boyfriends who always seek consent with their partners. And this is truly not a bad-apple scenario. Even as a self-identified pacificist, I engage in corner battles and push and shove in front of the net. ‘It’s part of hockey,’ but it’s also part of being a ‘man’ on the ice. The line between acceptable and not acceptable violence is obviously social and dynamic, and I believe should be critically considered whether Robinson can prove that line is in the wrong place or not.

      As I say, I have no definitive case to make on Robinson’s behalf. Yet I think it’s worth thinking about her idea in our own lives, and in how we raise our kids and even in how we react to our team mates and friends when they express violence. Do we as men (or do the men in our lives) sometimes treat women (or each other) with less respect because we can, because it’s expected of us, or because it makes us feel powerful? I think it’s important to honestly and critically consider this issue to help us sort out the violence we witness in sports and the violence we internalize as ‘normal.’

      Robinson is clearly being irresponsible, but that doesn’t mean she’s necessarily wrong.

    10. edgar
      September 7, 2011 at

      You, sir, should write Robinson’s articles for her. In both articles I’ve read by her, she makes no attempt at anything constructive, just using faulty logic and statistics to show how awful hockey is. I’m not even a huge hockey fan but her methods put me off.

    11. September 16, 2011 at

      This was Vancouver, right?

      Are we sure these women weren’t diving?

      P.S. “angry white male” – what does your being white have to do with anything (assuming you even are)?

    12. Colin
      September 27, 2011 at

      “her theory, which she’s been peddling since at least the early part of the century, is that sports make men violent.”

      I wouldn’t say that theory is hers alone – I’ve read similar elsewhere. The fact remains that violence against women still exists and identifying the causes is generally considered a good thing to do. I wouldn’t go as far as calling it theory – evolution by natural selection is a theory. It’s more of a hypothesis. As a hypothesis, perhaps it deserves to be rigorously tested (more so than the non peer reviewed effort of Robinson’s)? While the patterns in her statistics are weak, they also don’t contradict the idea that major sporting events lead to increased domestic and sexual violence toward women.

      Percentage changes based on small numbers do tend to show large swings. So the next logical question is whether those swings are even greater around major sporting events. Another obvious problem with Robinson’s stats is that they are pure correlation, and as we all know correlation does not mean causation. However, if two events are correlated it doesn’t mean there isn’t a causal link either. This is a question that deserves rigorous research across a large sample size where other factors (e.g. alcohol consumption, socio-economic, previous history) can be identified and if possible controlled for.

    13. A Man
      October 2, 2011 at

      If there’s a bigger man-hating d&ke than Laura Robinson, well…..

    14. Gerald
      September 11, 2012 at

      I think you buried the lede here in your last paragraph:

      “Does Laura Robinson have a point? I’ve got no idea.”

      It was thereafter followed by some excellent advice:

      “If you want to make a provocative case, you’d best have facts to support it.”

      Thanks for your time.

    Leave a Reply

    Your email address will not be published. Required fields are marked *