If you read this blog, you know that I frequently use numbers. Not only do I enjoy statistics I actually teach an introductory class in this subject from time to time. When I teach the class I encourage my students to ask two basic questions whenever someone cites a statistic: "Compared to what?" and "Could this be caused by randomness?"
Joe Haggarty wrote a piece this week extolling the wonders of Bruin David Krejci. The Bruins are having a terrific year as a team and Krejci is solid player on a very strong club. Haggarty throws out this apparently "amazing" fact:
...the Bruins haven’t lost a game in regulation this season when young Krejci lights the magic red lamp — they’re an incredible 19-0-1 in 20 games this season Krejci has scored. The Black and Gold are a similarly dominant 27-2-5 in 34 games anytime the center has registered a point this season.
Sports casters LOVE to throw out these sorts of numbers "The Braves are undefeated when Chipper Jones hits a homer off of a left-handed reliever." These kind of statements are attractive because they are easy to understand--event __ ALWAYS happens when event ___ occurs. Simple. Easy. Clear. The problem is with apparently clear statements is that they are usually cited for the wrong reasons. There are no magical properties that result from a Krejci goal.
Let's consider the Krejci case more carefully. So the Bruins are 19-0-1 in regulation. Apparently the "Krejci magic" vanishes like Cinderella's pumpkin carriage when the clock strikes zero at the end of the third period.
Now consider randomness as a possible answer. There are 540 full time skater jobs (18 skaters X 30 teams) in the modern NHL. Now if have 540 guys playing out their season, what are the odds that at least one of those 540 guys just happens to always score when his team also happens to win? The odds are pretty good that such a "coincidence" will occur at least once every NHL season.
Consider this fact, the Bruins have won 47 games and will likely hit 50 wins--what are the odds that AT LEAST one Bruin player with 22 or fewer goals will just happen to score only in Bruins wins?
Or consider another basic fact-- EVERY NHL players who scores increases his team's chance of winning. At the most basic level of Player X scores that means your team is not going to be shut out. Every NHL team has a better record in non-shotout games than they do in games where they are shut out.
The logic employed by Haggerty behind Krejci's importance is explained this way:
In those instances Krejci — and linemates Michael Ryder and Milan Lucic or Blake Wheeler — are giving the Bruins a second legitimate scoring threat capable of challenging quality playoff-style opponents, and causing problems for defenses geared toward stopping Marc Savard on Boston’s top skating line.
Now I have no doubt that Haggerty is right that the Bruins are stronger when that particular line is playing well. But I would point out that when David Krejci gets an ASSIST rather than a GOAL, the Bruins are only 7-2-5 which is a much lower standing-points-per-game ratio than the19-0-1 when he scores a goal. But if Krejci picks up an assist isn't that also a sign that his line is playing well?
The truth is that the Bruins (and most teams for that matter) perform better as a team when their 2nd line is producing goals. Whether Krejci himself scores or assists is in some ways immaterial. It is probably just dumb luck that the Bruins have won most of the games in which Krejci has scored. It is NOT dumb luck that the Bruins win more often when their 2nd line is playing well. Show me a NHL team where that is not true.
So what's my point here--other than picking on a Bruins reporter? These sort of statements "Team X always wins when Player Y does Z" are just a lazy way of trying to illustrate someone's contributions. It is a lot harder and more time consuming to dig up things like a player's Corsi Number or his Zone Shift Number or discover what the teams' Even Strength Shooting and Save Percentage are--in fact those numbers tell you far more about a player's contributions to the team. I'll bet a could write a humdinger of post on Krejci's importance using such metrics.
Beware of sportscaster brandishing "always" statements. Even the Great One didn't win all the time--great players win more often. Anyone or anything that "always" happens is usually the product of both skill and luck. Even Gretzky and Orr had some bad luck nights.