Tuesday, October 07, 2008

Making a case for blocking shots

With the season approaching, the opportunity for me to write another lengthy post on a trivial subject is shrinking. So here's one more before things start getting interesting...

Something I've noticed about the Calgary Flames the last few years is they don't tend to block a lot of shots. I've heard it rumored they actually employ this as a strategy: the default setting in the defensive zone is "clear traffic so Kipper can see the shot."

Now, on one hand, this seems sensible. A failed shot block can end up in an unwitting screen or tip, increasing the chances of a goal against. Blocking a shot can also take a defender out of the play or cause an injury, depending on circumstances.

On the other hand, there is a 0% chance of a successfully blocked shot going in the net. That seems, to me, to be a worthwhile gamble assuming that most of your attempted blocks are successful and dont end up in screens/deflections.

I took a look at the numbers to see how the Flames did at blocking shots last season. Turns out, they were indeed at the low end of the scale (and were the only team without a player at or above the 100 plateau). First the info:


click to enlarge

What we have here is total blocked shots, average BlkS per player and per game. While collecting the data, I realized that BlkS numbers alone weren't going to be meaningful because some teams spend a lot more time in their end (due to taking more penalties, being worse than other teams, etc) and therefore tend to block more or less shots as a function of zone time/possession. For example, Detroit only blocked about 9.6 shots/game last season, but they only yielded about 23.5 shots against a night (league low) thanks to their dominance and tendency to keep the puck away from the other team.

As such, I looked at blocked shots relative to shots yielded to give me an idea of how well/often a club was getting in front of pucks in the proper context. Basically, I totaled both blocked shots and shots against and divided the resultant sum by blocked shots. I think this clarified things to a degree: Detroit, for example, only blocked 9.6 shots/game while about 33.16 shots were being directed at their net every night (blocked shots+shots against). Calgary, on the other hand, blocked slightly more shots per game (9.9), but mainly because they had more pucks flying around at the bad end of the ice (38.41). So while the Wings blocked less shots on average, we can probably say they were better at it than the Flames were (29% of total versus 26% of total).

The Flames poor BlkS rate was actually third worst in the league - above only Los Angeles (25%) and Anaheim (24%) - and trailed the league average by a full 3%. While that doesn't sound like much, we're talking about thousands of shots here. Hypothetically speaking (and assuming a stable rate of total shots directed at net) if the Flames could have blocked shots at the mean rate (29%), it would have resulted in about 1.23 less shots on net per game, or almost 101 less shots over the course of the year. At Kipper's .906 SV%, that would have saved about 9.5 goals against which is a fairly significant number. A GD of +11 would have been the 2nd best in the division behind Colorado's +12 and likely would have meant a higher finish in the conference (and thus, no date with the Sharks in the first round).

Of course, disallowing the opposition to shoot at your net at all is the best way to go. However, if the bad guys are going to be taking shots, you might as well get in front of them as frequently as possible. What I dont know here is what moderates blocked shots: player skill, coaching or luck. If it's the latter, then we can say the team was simply unfortunate last season and hope they rebound this year. However, if the coaching staff is promoting a "get out of the way" strategy and that is indeed the primary driver behind the Flames lower blocked shot rates, I would suggest that the strategy be ditched forthwith.