On Any Given Night

In the last few years it has become common knowledge that the ECAC is significantly more competitive top to bottom than at any time in recent history. Unlike most common knowledge, this is actually true. We illustrate with the following statistical observations.

Data is Your Friend

Take a look at the best and worst winning percentages for each seed in the ECAC since the 12-team, full-schedule format was begun in the 1987 season. Records marked with asterisks were set or tied during the 1994-95 season. 

	 Pl	Best	Worst	96-97
	  1	.952	.705*	.773
	  2	.818	.636*	.681
	  3	.714	.568*	.659
	  4	.690	.568*	.614
	  5	.643	.545*	.568
	  6	.591	.523*	.523
	  7	.523	.409	.523
	  8	.500	.381	.454
	  9	.455*	.286	.364
	 10	.364*	.190	.341
	 11	.364*	.143	.273
	 12	.341*	.045	.227

What's with the Best and Worst Columns?

The '94-95 season accounted for an astounding 10 of 12 possible records skewed in the direction of a tighter conference (a better record for the bottom teams and a worse record for the top teams). Clearly, something interesting was happening in '95.

It is not surprising that more traditional measures of dispersion support the same hypothesis. For example, plots of the standard deviations of points and goal margins over the five seasons ending in '95 show a steady reduction of league dispersion. So, not only was the '95 conference race a log-jam relative to the typical ECAC season, but this tightening was the latest highlight of a five-year trend.

OK, But What's with the 96-97 Column?

Although the '97 results were not as dramatic as those of the '95 season, you can see that they were closer to the "muted" values (lower for better seeds, higher for worse seeds) than to the "extreme" values.

All of this data reinforces the hypothesis that we are in a period of strong competitiveness within the ECAC, in which the difference between good and poor teams has eroded. In such an environment, radical changes in team fortunes are possible, even likely. Keep this in mind as you review the TBRW? ECAC Predictions.

Note to readers: seems a little weird that we jump '96, doesn't it? Well, as of now your intrepid author is separated from his print records by approximately 3,500 miles and/or one helluva FedEx bill. So, more data as library resources and checkbook permit, and thanks for your patient indulgence.