The rankings seem fairly accurate, except can somebody explain how Rowan, who lost to TCNJ and Wagner and has mostly only beating low ranked teams is ranked 7th. Delaware, Maryland, and Rider should be ahead of them.
I don’t disagree, but they do have the second highest goal differential and SoS out of those 4 teams. Keep in mind SoS accounts for about 95% of a team’s rating.
The other thing to consider is that their “bad” losses (TCNJ, Wagner) were only by one goal. While their “good” losses have been close losses (within 2 or 3 goals) against very good teams (PSU 2x, Liberty 2x, WPU 1x, Rider 1x, Del 1x, UMD 2x). Simply being in the MACH probably gives enough SoS to be included, almost regardless of record. UMBC made regionals last year with a losing record.
SoS accounts for 95%??
Sorry for somewhat “off-topic” but look at the West’s #6 and #11. If SoS were 95% grade - it totally doesn’t make any sense over there
Sorry, I probably phrased that poorly. 95% was just my eyeball estimate. There is no hard and fast rule on that.
But a team’s overall rating is simply SoS + AGD + WP.
So looking at BSU (#6 west) their rating is 93.41, with a SoS of 89.71, which is 96.0%.
SDSU (#11 west) gets a rating of 92.42, with SoS at 89.47, which is 96.8%.
So I’d say the 95% guess was pretty close.
Even Aurora (#2 central) with a national best 5.11 AGD and a perfect win percentage gets 93.7% of its rating from its SoS.
Do you know if beating a higher ranked team and /or losing to a lower ranked team works into the equation. In other words does beating a higher ranked team provide more points than beating a lower ranked team or is it just considered a win. Vice versa does losing to a lower ranked team hurt you more or is just considered a loss.
I understand what you are asking, but unfortunately the answer just isn’t that simple.
Two reasons for this: 1. AGD, WP, and SoS all matter (even if SoS matters most) and 2. these criteria are dynamic and the values change over the course of the season.
Last year they showed the individual game breakdowns on MyHockeyRankings and it gave a lot of insight into how much or how little a particular result affected things. Unfortunately, they removed that feature this year, so it’s hard to isolate a particular performance. The ratings are designed to be cumulative, so it really matters how a team plays over the course of the season. A single win or loss doesn’t move the equation a lot when we getthis late in the season. It matters how you do against opponents, but it also matters how your opponents do against other common opponents.
In theory there is a way to figure out a line or expected outcome of a particular game and compare that with the actual result, but I’m not mathematically inclined enough to figure this out.
Basically, a really good team is expected to beat a really bad team by a lot of goals. If this happens the good team gets credit for the goal differential and a bump in win percentage. However, their strength of schedule goes down after playing a bad team with a low rating. You also need to consider that goal differential is capped at 7 goals. Since SoS is the most important factor (about 95% as discussed above), its actually possible for a good team to lose more in SoS rating than they earn in AGD and WP from a win over a bad team.
Obviously, if you win all your games and beat everyone by a ton of goals (PSU, Aurora) the rankings sort themselves out and you have nothing to worry about. However, if you are a bubble team I see two ways to play things. You can schedule a bunch of easy games and run the score up so you get maximum AGD and WP points and make up for the loss in SoS (see Cincinnati). Or you can schedule mostly tough games that you will likely lose but get big time SoS points. If you can also run the score up on a few really bad teams throughout the season you can cancel out the losses in AGD and WP that you would get with a losing record (see Rowan). Given the cap on goal differential, I would rather have the SoS points.
I hope that helps a little.
Thanks for the reply. Yes it helps and I did already consider what you explained so well; that being the AGD and SoS adjustments that occur in playing a weak or good team.