It’s a fine Tuesday evening in December, and here we are with updated ECAC Hockey power ratings.

Yeah, I know I said I’d have these updated yesterday, but Harvard is playing Bentley as I type this, so I figured I just wait for that game to end and include it.

—

Last Week’s Games (via Collegehockeystats.net) – my picks in bold

Nov. 30

	Brown	1	at	Clarkson	5	EC	7:00 pm ET	Box / Text / Sheet
	Princeton	0	at	Quinnipiac	3	EC	7:00 pm ET	Box / Text / Sheet
	Yale	5	at	St. Lawrence	3	EC	7:00 pm ET	Box / Text / Sheet
	Rensselaer	1	at	Notre Dame	5	NC	7:05 pm ET	Box / Text / Sheet
	Cornell	2	at	Dartmouth	3	EC,IV	7:30 pm ET	Box / Text / Sheet
	Colgate	1	at	Harvard	6	EC	7:30 pm ET	Box / Text / Sheet

Dec. 1

	Yale	2	at	Clarkson	1	EC	7:00 pm ET	Box / Text / Sheet
	Cornell	2	at	Harvard	1	EC,IV	7:00 pm ET	Box / Text / Sheet
	Quinnipiac	6	at	Princeton	3	EC	7:00 pm ET	Box / Text / Sheet
	Brown	2	at	St. Lawrence	2 OT	EC	7:00 pm ET	Box / Text / Sheet
	Rensselaer	0	at	Notre Dame	3	NC	7:05 pm ET	Box / Text / Sheet
	Colgate	2	at	Dartmouth	2 OT	EC	7:30 pm ET	Box / Text / Sheet

Dec. 4

Harvard 2 at Bentley 1 NC 7:05 pm ET Box / Text / Sheet

—

And so, here are this week’s overall ratings:

Team	Overall Rating
Quinnipiac	2.866666667
Clarkson	2.153846154
Union	1.894417782
Yale	0.8727272727
Harvard	0.8022582083
Cornell	0.09090909091
Dartmouth	-0.2472697301
Princeton	-0.4947229551
Brown	-1.090909091
Colgate	-1.974683544
RPI	-2
SLU	-2.121890841

—

A few folks had asked for some clarification on how the ratings are computed, so I’ll oblige and explain in a little more detail.

First, let’s look at the stat table from which the overall rating is computed for one team — in this case, St. Lawrence:

Overall		Total Minutes	Pct of Total	GF	GA	Diff	GF/60Min	GA/60Min	Diff	PCT
5×5		646.9	0.7625839915	14	33	-19	1.298500541	3.060751275	-1.762250734	-1.343864199
5×4		75.9	0.08947306377	7	3	4	5.533596838	2.371541502	3.162055336	0.2829187787
5×3		0.7	0.0008251797713	1	0	1	85.71428571	0	85.71428571	0.07072969468
4×3		0	0	0	0	0	#DIV/0!	#DIV/0!	#DIV/0!	#DIV/0!
4×4		14	0.01650359543	0	1	-1	0	4.285714286	-4.285714286	-0.07072969468
3×3		0	0	0	0	0	#DIV/0!	#DIV/0!	#DIV/0!	#DIV/0!
3×4		0	0	0	0	0	#DIV/0!	#DIV/0!	#DIV/0!	#DIV/0!
3×5		3.9	0.004597430154	0	2	-2	0	30.76923077	-30.76923077	-0.1414593894
4×5		106.9	0.1260167394	1	14	-13	0.561272217	7.857811038	-7.296538821	-0.9194860309
		848.3	1							-2.121890841

The leftmost column (Column A) is each manpower situation. The total number of minutes played in each manpower situation (five-on-five, five-on-four, etc.) is calculated in Column B, and that total’s percentage of the team’s total minutes played is calculated in Column C. Columns D and E total goals for and goals against for each manpower situation and Column E is the situational goal differential. Column F calculates the team’s Goals For Average Per 60 Minutes for each manpower situation, and Column F is the differential between the two. Column G (PCT) is the differential in Column F multiplied by the manpower situation’s percentage of the team’s minutes played (Column C). The sum of Column G constitutes the team’s overall rating.

So far, I’ve broken this down for each team at home, on the road, neutral sites, conference and non-conference, and I’ll be posting it all here in the near future.

I chose St. Lawrence as the example table above both because I’m an alum of the Candle in the Wilderness and because using the team with the lowest rating is illustrative of where the Saints’ myriad weaknesses are and how this rating system works.

As I mentioned last week, the bulk of the rating is derived from the team’s goal differential in five-on-five play, because the bulk of the game is played five-on-five. St. Lawrence plays an average of 76.25 percent of a game at five-on-five, and they average 1.298500541 goals for per 60 minutes five-on-five. Unfortunately, the Saints also average 3.06 goals against per 60 minutes played five-on-five.

As you can see, the five-on-four power play is the only manpower situation where SLU has a positive goal differential (+4). I don’t think you need me to continue extrapolating this for each manpower situation, but if you’d like more explanation, feel free to ask.

—

 

You can support my work on this stuff by visiting my Patreon page at: https://www.patreon.com/leifskodnick

As always, if you have questions, email me or comment below.