Sagarin this week

[URMite]

FWIW xlolx:D

1 Richmond
2 Appalachian State
3 Villanova
4 Massachusetts
5 Eastern Washington
6 Southern Illinois
7 Delaware
8 Georgia Southern
9 William & Mary
10 Montana
11 North Dakota State
12 James Madison
13 Northern Iowa
14 Sacramento State
15 New Hampshire
16 South Dakota State
17 Wofford
18 South Alabama
19 Pennsylvania
20 Montana State
21 Maine
22 Stephen F. Austin
23 Harvard
24 Elon
25 Northern Arizona
26 Jacksonville State
27 Weber State
28 Lehigh
29 Central Arkansas
30 Furman
31 Liberty
32 McNeese State
33 Cal Poly-SLO
34 Sam Houston State
35 Colgate
36 Chattanooga

Ok, the only thing I use the Sagarin for in football is to cut & paste names and records. xpeacex

[URMite]

http://www.usatoday.com/sports/sagarin/fbt11.htm

[Spider]

GO SPIDERS.....

[boonegoon]

GO SPIDERS.....

You go and beat 1 FBS team and jeez.

Seriously, kudos to UR

[1andDone]

What is funny about USA being up there?

[darell1976]

Great West:

Cal Poly 147
NORTH DAKOTA 151
UC Davis 152
Southern Utah 155
South Dakota 173

[Redbirdz]

Because USA hasn't beaten a real team yet.

[gasoutherneagle]

Can anyone explain how Va Tech can lay the SMACKDOWN on CrAppy State and they still come out #2?
xflaggedx

[alvinkayak6]

Do you guys know anything about Sagarin or even bother to read what he writes ABOVE the rankings?

He uses Elo Chess, which is a chess ranking measure. Of course, in order to rank teams/players/intelligence of comments/whatever, you need starting measures.

Until all the teams can be mapped onto one another (e.g. A has played B, which has played C, which has played D, which has played E, which has played A), there can be no fully effective Sagarin rating.

The start of the season is based almost totally on conjecture (his own and preseason expectations from last year). Villanova was his top rated FCS team coming into 2011 (don't ask me why).

I noticed that you also chose the numbers on the left because it's easiest. Foolish. Sagarin clearly indicates that PREDICTOR is the best choice (far right) since it includes margin of victory.

[URMite]

What is funny about USA being up there?

It is not a statement about the quality of the team. It is a statement about a start up that doesn't play a schedule with a majority of Div 1 teams until next week. You are the only start up or transistioning team on that list I believe.

[URMite]

Do you guys know anything about Sagarin or even bother to read what he writes ABOVE the rankings?

He uses Elo Chess, which is a chess ranking measure. Of course, in order to rank teams/players/intelligence of comments/whatever, you need starting measures.

Until all the teams can be mapped onto one another (e.g. A has played B, which has played C, which has played D, which has played E, which has played A), there can be no fully effective Sagarin rating.

The start of the season is based almost totally on conjecture (his own and preseason expectations from last year). Villanova was his top rated FCS team coming into 2011 (don't ask me why).

I noticed that you also chose the numbers on the left because it's easiest. Foolish. Sagarin clearly indicates that PREDICTOR is the best choice (far right) since it includes margin of victory.

Actually, I used the synthesised rating, since that is where USA today gets its numbers on the left. I do use predictor for individual games. I highlighted the two teams that seem the most indicative of the fact that there is very little mapping to date.

Once all teams are mapped, then Sagarin becomes a useful tool to find teams that you might otherwise not follow as closely from purely human analysis.

[Friend of Pounce]

I noticed that you also chose the numbers on the left because it's easiest. Foolish. Sagarin clearly indicates that PREDICTOR is the best choice (far right) since it includes margin of victory.

You raise good points on the need to understand the system(s) Sagarin uses. Just wanted to point out that we don't all agree that PREDICTOR is the best Sagarin rating. While it might be the best option for use in betting against the spread, it neglects to take into account that the point of football is to win, not to win big. And occasionally you get some weird results by using MOV, such as last year's Bama team being ranked two spots ahead of Auburn, despite losing three games (including one to that very Auburn squad).

[alvinkayak6]

You raise good points on the need to understand the system(s) Sagarin uses. Just wanted to point out that we don't all agree that PREDICTOR is the best Sagarin rating. While it might be the best option for use in betting against the spread, it neglects to take into account that the point of football is to win, not to win big. And occasionally you get some weird results by using MOV, such as last year's Bama team being ranked two spots ahead of Auburn, despite losing three games (including one to that very Auburn squad).

Sagarin says PREDICTOR is the best (or implies it), and I agree. Why throw away data? I'm not sure I agree either that the point of football is only to win (not win big). Obviously, outscoring an opponent by a lot in the 2nd quarter greatly increases the team's chances of victory. Counting a 3-2 victory the same as a 49-2 victory strikes me as totally stupid. Sure, there will be "odd" or counter-intuitive results in PREDICTOR, but your judgment between two choices can't be based on exceptional circumstances.

[Friend of Pounce]

Sagarin says PREDICTOR is the best (or implies it), and I agree. Why throw away data? I'm not sure I agree either that the point of football is only to win (not win big). Obviously, outscoring an opponent by a lot in the 2nd quarter greatly increases the team's chances of victory. Counting a 3-2 victory the same as a 49-2 victory strikes me as totally stupid. Sure, there will be "odd" or counter-intuitive results in PREDICTOR, but your judgment between two choices can't be based on exceptional circumstances.

You're right in that throwing away data handicaps your ability to fully evaluate the teams, and that exceptions can't drive consideration of the systems. In part, this comes down to the fact that I just don't value predictive rating/ranking systems as much as those that are acknowledged as purely retrodictive. It's quite possible that Bama would have beaten AU in a majority of games played last year. Maybe the smart money would have been on the Tide, even after losing the Iron Bowl. The thing is, they didn't win when they had the shot, and you'd be hard pressed to find a ranking system that doesn't include MOV that would place the Tide above the Tigers.

Also, it's probably a whole other discussion (and one in which the math would stretch beyond my capabilities), but I stand by the idea that MOV shouldn't be included in official rankings (predictive ratings that have no bearing on the postseason are fine). I do believe that winning (and not winning big) is the point. You're right - scoring a lot in the second quarter is important. But not because you want a huge MOV...it's so you can comfortably hold onto any MOV for the next two quarters. Maybe even give some extra shifts to your backups.

Yes, 49-2 and 3-2 should be considered differently when evaluating the strength of a team. I acknowledge that voters almost have no choice but to think about such things. But what about 56-45 vs. 10-7? Is it better to pass every play and play zero defense than it is to grind out a defensive win? I don't think so, but considering MOV necessitates evaluating these results differently. I suppose you could deal in percentages of opponents' scores to even things out, but I'm not sure when it just becomes too arbitrary.

I also like to think that "a win is a win" because it gives much more meaning to pulling out close games. Otherwise, you're stuck with, "Well, we won with a last second field goal, but we aren't completely winners because we didn't close 'em out with room to spare."

I suppose it's just personal preference as far as what we're looking for in a system.