This week the Riders got a little less lucky, the Bombers got a win in the new stadium, and the Eskimos just can't buy a break.
I stream-lined the chart a bit this week and it's presented in a slightly different format, as my stats are now in a database instead of a spreadsheet, so I can store more and do cooler things, like:
Big Win Percentage.
Big Win Percentage is a simple stat, created by Jim Glass. It's based on the premise that football by nature is a game that can be heavily influenced by luck. A bad call, a fumble recovery, a gust of win; these are all things which can turn a close game into a win or a loss. According to Brian Burke (the guru of NFL stats), the outcome of more than 40% of NFL games is determined by random chance. This makes judging a team by it's record a difficult proposition (especially in the NFL, where teams don't play every team in the league).
What Mr. Glass's formula does it try to account for that luck by giving teams credit for "Big Wins", defined as a game decided by 9 or more points. 9 points makes a good cut off because it is the border between 1 and 2 possession games.
The formula is simple - games won by 9+ points count as a "Big Win", games lost by 9+ points are considered a "Big Loss", and all the rest are considered ties. If you read the article linked above, you'll see that he's found that teams with a high number of "Big Wins" in a season tend to fare much better in the playoffs. We'll see if that holds true for the CFL (I'm compiling data back to 1990 for a post later this week), but in the mean time, I'm going to include it on the chart for this week.
|Py W = Pythagorean Wins, Projected = Py Wins over 18 games|
As noted above, I've been collecting data, back to 1990 so far. I plan to do a post to highlight some of the interesting points once I have a bit more information gathered.