Wednesday, December 07, 2011

Lies, damn lies, and .500

Allow me to take a break from my irregular soccer-themed postings to provide you with a rant from a secondary (actually tertiary, behind CFL gridiron football) passion.

Did anybody catch the big Jets game yesterday? The one where they knocked off the defending Stanley Cup champs from Boston, a team that was unbeaten since October (actually, they weren't)? You know, the game that put the Jets over .500 (they're not)?

If you've spent any time reading this blog, you'll know that I have a thing for spreadsheets. Mostly, they're an elegant way of organizing data, and of course the spreadsheet was the first killer app for the PC (remember those?) But sometimes it's about the numbers contained therein and the story that they tell.

Moreso than a lot of anecdotal tripe and breathless speculation that constitutes sports writing these days, numbers are able, concisely and precisely, to narrate happenings on the pitch, ice, or court. Statistics is a powerful tool to analyze teams and players, and to suggest means of improving both.

Which is why it is so endlessly infuriating when the language of numbers and statistics are so carelessly misused in the popular media. And more than anything, it is the use of the term ".500" when referring to NHL teams that has me blowing my lid.

WHAT IS .500?

.500 is a decimal number. Mathematical speaking, the trailing zeros are redundant. .5 would be more correct, but gives some people other ideas.

In sport, .500 is used to refer to teams:
  • that have lost exactly as many games as they have won
  • whose records are statistically average

In North American sport, this stat is rarely problematic. In baseball and basketball, where rules have legislated ties out of the game, a team with a winning percentage of .500 (also infuriating, as 0.500 is a decimal number, 50% is a percentage) will have won exactly as many games as they have lost. In gridiron, where ties are rare, a 7-7-1 team is rightly called .500 and will have amassed exactly half of the available points.

Only the NHL and other lower divisions in hockey, with the recent abomination of points awarded for overtime and shootout losses, do scribes and fans alike play fast and loose with the .500 truth.


This is the common use of the term, at least around these parts:

The 12-11-4 Jets — first time this season over .500 — skated vigourously for an hour at the MTS IcePlex today, now looking ahead to their Friday night home game against the Carolina Hurricanes.

Forget for now that the dashed and bolded clause is grammatically incomplete. Instead, look to the record: 12-11-4. If this were 1985, you could say the Jets were over .500. But it's not, and you can't. First off, Dale Hawerchuk is like 100 now while my Jets lovechild Alex Burmistrov is barely 14. Of more importance, the meaning of that third column has entirely changed.

Let's break it down for the uninitiated. The first number, 12, represents wins. Wins of any kind, whether recorded in regulation, overtime, or the shootout (please, Gary, kill it). The second column represents losses in regulation. Period. These two are not created equal.

The final column represents losses in overtime and the shootout, for which a point is awarded. So my beloved Jets (long may they remain) have 12 wins and 15 (11 regulation + 4 extra time) losses. By any of the definitions of .500 given above, they are above it.

The only case one can make is to state that the Jets have earned more than 50% of the points available to them. That's great, but so have all but 7 teams in the 30-team NHL.

This is a term that has outlived its usefulness, at least in referring to hockey. My own preference would be for records to be listed thusly: total wins (OT/SO wins) - total losses (OT/SO losses). Reporters would quickly remember what decade we are in and could easily compute which teams are over .500 by doing the old math exercise with the < symbol (the little snake eats the bigger animal!)

Maybe sports style guides have already proscribed the correct use of .500(hey, it's only been over a decade) and it's only a problem here in Winnipeg where things take a while to catch on (we've only recently completed the Zubaz pants craze). Of course the easy solution would be to do away the unnecessary loser's point and even less necessary shootout.

But that's me. Rant over.

No comments: