James Mirtle's post about the Colorado Avalanche's unexpectedly hot start got me wondering about historical precedents. Their numbers so far in regulation are impressive:
W | L | T | WPCT | GF | GA | WPCT | |
COL | 9 | 1 | 3 | .808 | 43 | 26 | .732 |
Since 1967-68, there have been 36 teams that have started the season with a winning percentage between .769 and .846, essentially one more win or one more loss than the Avs. The second WPCT is Colorado's pythagorean winning percentage, based on its goals for and goals against.
Here's how they finished the season - some numbers may not add up perfectly due to rounding:
W | L | T | WPCT | Pts | GF | GA | WPCT | |
Start | 9 | 2 | 2 | .793 | 22 | 54 | 32 | .743 |
Rest | 34 | 21 | 13 | .594 | 88 | 245 | 201 | .599 |
Total | 44 | 23 | 15 | .626 | 110 | 299 | 233 | .624 |
I've assumed that OT and SO performance is pure luck, and awarded 1.5 points per game that ends tied after regulation. Teams that have started hot have generally been pretty good - 110 points over a full season would be an amazing outcome for the Avalanche, a 41-point improvement over last season's disaster. Note that the GF and GA totals clearly reflect other eras in the history of hockey!
Now that's the average performance. What about the range of possible performances?
W | L | T | WPCT | Pts | |
Average | 44 | 23 | 15 | .595 | 110 |
-1 stdev | 36 | 29 | 17 | .543 | 98 |
Worst | 31 | 33 | 18 | .488 | 89 |
We can be completely sure that the 2009-10 Colorado Avalanche are not the 1976 Montreal Canadiens, who went 58-11-11 and whose 9-2-2 start was actually worse than their record the rest of the way. So if we compare the Avalanche only to the teams whose record over the rest of the season was in the bottom half of our sample, what does their playoff performance look like?
Miss | 1st Rd | 2nd Rd | 3rd Rd | Finals | Win Cup | Avg # Playoff GP |
2 | 10 | 2 | 2 | 0 | 2 | 8.3 |
So even among the bottom half of this group, we still have some very good teams - the cup winners were the 1998 Red Wings and the 1983 Islanders, both of whom swept the finals. But 2/3 of our comparative group either missed the playoffs or didn't make it out of the first round.
If Colorado finishes with 98 points (the -1 standard deviation record shown above), 6th in the West, and makes a 1st-round playoff exit, their fans should be overjoyed. There's really very little precedent for a team starting the season this hot and not actually being very good - except the 2008-09 Avalanche was, by six losses, by far the worst team ever to have this hot a start to the next season.
As best I can tell, NHL futures have the Avalanche in the middle of a dogpile at around 90 points, just outside of the last playoff spot. In other words, the people who have money riding on this are giving Colorado 50/50 odds to be the worst team ever to have this good a start to the season.