HighSchoolHockeyForums.com

WestMetro wrote:I’ve been trying to schedule to make sure see each of top 20 in person at least once against quality opponent

But Karl keeps moving teams in and out of top 20 ! No fair!

Karl , any chance you can take my schedule into account as a legitimate ranking factor ? Or speak to Mrs West Metro so I can go to more games ?

Looks like I’ll be taking a long road trip up Hansen Blvd to Andover Arena for first time since Maddie Rooney era

karl(east) wrote:I won't deny my rankings tend to be fairly conservative in that I don't move teams around much unless they give me good reasons to do so, and winning a lot with few quality wins doesn't really do that for me. Historically, I've found this is safer and saner than launching a team upward on the strength of some good wins only to drop them all the way back down again when they finally face good competition and get killed. The downside is that I can be slow to pick up on the ones that are legit. Speaking to Rosemount specifically, they did get a pretty good win over Eagan earlier this week, so should be on the way up, and will be even further up if they beat North tomorrow. The difference between Rosemount and some of the other non-traditional powers that have already cracked the top 15 (Andover, Brainerd) is that those two have some legitimate good wins.

The other thing that can take a while to flush out are teams that I had ranked highly in the preseason that do log a few quality results. (As Blaine has, not only vs. East but also the Hill and EP ties...I realize they're both down, but that has also taken some time to become obvious, and both have had flashes too.)

I also think one can read plenty into results, even when a team is playing a poor schedule. If Rosemount were blowing out Eastview, Lakeville South, and Eagan, I'd really be paying attention. As it is, they lost 5-1 to Eastview (though that was early) and have one-goal wins over the other two, and their bodies of work do not put them in the top 20. On the flip side, a team can have a bad record, but if they're losing narrowly to good teams, they stay high because we know they don't need that many things to go right to beat a top-flight team.

All ranking attempts are an attempt to make up for a lack of information; no one can watch every game. (Well, I guess that's at least plausible with video now, but come on; and even if we could, people don't always interpret games the same ways.) Computerized algorithms solve this by cutting out a lot of stuff and focusing just on wins/losses and goals for/against. It has the benefit of cleanliness, but sacrifices helpful explanations like injuries or suspensions, etc., and there are a lot of human decisions on what to value or not value that go into them. The LPH poll uses many different sets of eyes and tries to mash that all together into something coherent. I just try to rely on pre-existing knowledge and adjust, both as I watch and collect others' accounts on games and follow things like PS2.

tl;dr I have a system that usually works and I'm sticking with it, much to goldy's delight.

Stang5280 wrote:

karl(east) wrote:I won't deny my rankings tend to be fairly conservative in that I don't move teams around much unless they give me good reasons to do so, and winning a lot with few quality wins doesn't really do that for me. Historically, I've found this is safer and saner than launching a team upward on the strength of some good wins only to drop them all the way back down again when they finally face good competition and get killed. The downside is that I can be slow to pick up on the ones that are legit. Speaking to Rosemount specifically, they did get a pretty good win over Eagan earlier this week, so should be on the way up, and will be even further up if they beat North tomorrow. The difference between Rosemount and some of the other non-traditional powers that have already cracked the top 15 (Andover, Brainerd) is that those two have some legitimate good wins.

The other thing that can take a while to flush out are teams that I had ranked highly in the preseason that do log a few quality results. (As Blaine has, not only vs. East but also the Hill and EP ties...I realize they're both down, but that has also taken some time to become obvious, and both have had flashes too.)

I also think one can read plenty into results, even when a team is playing a poor schedule. If Rosemount were blowing out Eastview, Lakeville South, and Eagan, I'd really be paying attention. As it is, they lost 5-1 to Eastview (though that was early) and have one-goal wins over the other two, and their bodies of work do not put them in the top 20. On the flip side, a team can have a bad record, but if they're losing narrowly to good teams, they stay high because we know they don't need that many things to go right to beat a top-flight team.

All ranking attempts are an attempt to make up for a lack of information; no one can watch every game. (Well, I guess that's at least plausible with video now, but come on; and even if we could, people don't always interpret games the same ways.) Computerized algorithms solve this by cutting out a lot of stuff and focusing just on wins/losses and goals for/against. It has the benefit of cleanliness, but sacrifices helpful explanations like injuries or suspensions, etc., and there are a lot of human decisions on what to value or not value that go into them. The LPH poll uses many different sets of eyes and tries to mash that all together into something coherent. I just try to rely on pre-existing knowledge and adjust, both as I watch and collect others' accounts on games and follow things like PS2.

tl;dr I have a system that usually works and I'm sticking with it, much to goldy's delight.

Thanks for the extremely thorough and thoughtful response, Carl. I certainly appreciate all of the work that goes into compiling these rankings and hope it didn’t feel like I was completely attempting to undermine your methodology and credibility. And I certainly wouldn’t want to go through this exercise every week, particularly in a year where there is so much parity beyond the top handful of teams (in both classes). My main issue boils down to trying to spread some love two teams that are having surprisingly good seasons and stripping away some of our earlier expectations, which you acknowledged eloquently and is admittedly not easy to do.

FYI, I want to go on the record as not stumping for Rochester Century, who received a 9-0 spanking at the hands of Northfield today. That’s all goldy

goldy313 wrote:9-0 is the girls score, Century boys won 3-1.
Tartan wins in ot with Hastings.....trap game just like WBL had last week with Century. Losing in ot to Mahtomedi but pulls out the victory....unlike WBL.
Rosemount shuts out Lakeville North 2-0.....playing and losing to better teams seems to have not paid off in this one.

1AA seedings might be a bit wonky, Century may win 20 games with only 1 real quality win (out of 2 possible) but that could probably be the best win of anyone in 1AA. North wil struggle to be a .500 team so they probably need to sweep South to have any hope of a 1 or 2 seed. Farmington plays Hastings late in the year.

The Lakeville teams do not play the Rochester schools in just about every sport in the regular season, even though they are in the same section in just about every sport.

I didn't know this until today but the teams I think deserve a mention all have top goalies....at least in save percentage....Tartan #1, Irondale #4, Rosemount #8, and Century #11.

goldy313 wrote:I must have stirred the 5 AA hornets nest.....Champlin Park beats top 10 Duluth Marshall, Centennial beats Elk River.......maybe Blaine beats Edina tonight?