RS
Previous (2000) Regular Season Points
Previous season�s performance is the most significant determinant of the next year�s finish. Despite our dislike of timid prognostication, there really isn�t much mobility from year to year in the conference. We did an article about this about
ten years ago. It�s really time to update it.
Ordinarily, we take the points for each team over the balanced 22 game
schedule, but once again the UVM situation screwed that up. So, instead,
we apply a little bit of sophisticated math.
- First, we record the winning percentage (Pct), of each team last
year. In Vermont's case, we take their winning percentage in their
previous full season, 1998-99.
- Then, we normalize the winning percentages to have them come out to a mean
of .500. We take the mean of the Pct's (.489) and multiple each team's
percentage by the inverse function .500/.489 = 1.022. This gives the
adjusted percentage, AdjPct.
- Finally, we compute the number of points in a 22 game (44 point) schedule
that that pct would generate, RS_22. Because of rounding, we
need to decrement one value by 1, and we decrement the metric least above
the .5 value of the lower number, which turns out to be Brown's.
|
Team |
Pct |
AdjPct |
RS_22 |
|
Brown |
.238 |
.243 |
11 |
| Clarkson |
.525 |
.537 |
24 |
| Colgate |
.750 |
.767 |
34 |
| Cornell |
.525 |
.537 |
24 |
| Dartmouth |
.452 |
.462 |
20 |
| Harvard |
.476 |
.487 |
21 |
| Princeton |
.476 |
.487 |
21 |
| RPI |
.548 |
.560 |
25 |
| St. Lawrence |
.825 |
.843 |
37 |
| Union |
.310 |
.317 |
14 |
| Vermont |
(.364) |
.372 |
16 |
| Yale |
.381 |
.389 |
17 |