A lot has been made over the past few weeks (and rightfully so) about the glaring deficiencies in the game of the Calgary Flames. 

Shots, goals, ice time and matchups are always popular topics, and after Monday’s game it looks as though a couple of people even blamed goaltending somehow. But what about penalties? As they are such a big part of the game, it was weird to me that no one was talking about them. This piqued my curiosity because it looked to me like the Flames were sitting in the box more often this year as opposed to last year.

Now, of course we can’t actually fully measure the amount of penalties taken this year because the Flames haven’t played 82 games yet, but we can measure penalty kill and power play time as well as the actual numbers of calls from this year and last on a per-game basis.

If there happens to be something wrong with my math, point it out in the comments and I’ll correct it. There’s a reason I’m in the social sciences.



Obviously, one wants their team to draw more penalties then they take. The Flames have done a better job this year compared to last year in taking penalties, reducing them by about .2 per game. However, they’ve drawn about three quarters less penalties per game this year compared to last year. Currently, the Flames are on pace for a penalty differential 43 penalties worse then last year and a goal differential 10 worse this year, which is around 3 points in the standings. There’s still quite a bit of time to correct these things, and it could feasibly result in gaining maybe a point back by the end of the season.

  • PrairieStew

    It doesn’t appear you weighted for ice time. This makes your average not representative of the team total.

    Bouwmeester’s +.6 needs to be weighted ~2 times higher than Glencross’s -1 because he has played 20:09 ESTOI/G X 63 g for 1269.45 ESTOI total. Glencross is only at ~680 ESTOI (14:10 x 48 g).

    Also not sure why you would leave off the rest of the team to try and extrapolate out a team rate. That is like excluding 40% of the data. has Calgary in 10-11 Calgary was plus 36 in Power Play opportunities in favour over against. In 11-12, they are at -5 PP Opp differential over 63 games. If you extrapolate to 82 games, that is -6.51 PP Opps on the season.

    If you take an approximate league average PP of 18% (since we would have to work out which teams etc. they were taking them against) They got 6.48 goals last year on penalty differential. This year they will be -1.17 goals to the bad. That’s a -7.65 swing. So over one win ro so overall.

    Although I have to believe this is well swamped by variance at a team level season to season. Which is to say – blame the refs!

    • PrairieStew

      the stats are already weighted for ice time:

      Individual Penalties Taken per 60 Minutes. Excludes coincidental penalties.

      Individual Penalties Drawn per 60 Minutes. Excludes coincidental penalties.

      • SmellOfVictory

        Actually, Tach’s right. Since it’s normalized per 60 min of playing time, the aggregate statistic would be considering all players as having equal weight. You’d need to multiply the PT/60 and PD/60 by each player’s total ice time (or by their average TOIxGP) in order to get fully accurate data.

        Using raw pentalties drawn vs penalties taken would be an easier stat to work with, I should think. However, in response to Tach: given that the majority of players haven’t seen a substantial change in TOI between last season and this season, I don’t know that weighting the PT/PD ratio for ice time would result in a substantial change (it may, but it also may not).

        • PrairieStew

          The similarity in ice-time rationale is a fair comment, and probably why the team results don’t differ materially from the cumulative individual results.

          I still think that, unless you are going to critique particular players for taking more penalties, we are better off looking at the total team results than adding up player penalty rate results.

  • Bean-counting cowboy

    I wonder if the metric gets inflated by games where they are already out of it (i.e. Boston, Edmonton)- or when holding a signficant lead. At these times, taking a dumb penalty is not as bad an idea as a 1 goal close game.

    I didn’t go back and look at the prior year schedule and the goal differential of the games, but I wonder if you would need to take out penalties taken/drawn when the goal differential was more than 3 goals (in either direction).

    I won’t pretend to have the time or interest to do such a thing, but thought I would mention it as food for thought.

  • beloch

    Is an improvement in PD-PT always an improvement for a player? Stajan’s big improvement in PT could be due to reduced intensity in his play. His corresponding increase in PD could be blamed on that fact that he’s playing against fourth liners who might, on average, be less disciplined than the tougher competition he faced for large parts of last year (You’d need to check if PD-PT is correlated with QoC to see if this theory holds water).

    I’m not trying to slag Stajan, but it seems like the PD-PT stat could be interpreted in more than one way.