Toggle sidebar

Sagarin rankings

N

NavyTider

Scout Team
Jan 4, 2007
147
0
0
????!!!!!?!?!?!?!!!??!!!

The Sagarin rankings have USC #2 with the 3rd-hardest strength of schedule in the nation?

???!!!!?!?!???!!?!?!!!?!

I know they're supposed to use unbiased formulae to come up with the final results, but come on -- you cannot seriously tell me that USC (USC?!?!!?!) has played the 3RD HARDEST SCHEDULE of any team in the nation to date. This sounds suspiciously like when Florida State used to be good, and every year when the BCS results came out they would show FSU as having played one of the top 5 hardest schedules in the nation, even though your average football fan couldn't even name any of the other teams in the ACC because they sucked so badly.

Further, 6 of the top 10 hardest schedules to date have been played by PAC-10 teams, according to Sagarin. I know this is supposed to become more accurate as the season progresses, but seriously -- come on!

Oh, and we're #4, behind Texas, USC, and Penn State. Our SOS? 38th.
 
Last edited:
there is something extremely flawed in that system. EXTREME BIAS.....again.........PLAYOFF WOULD ELIMINATE....computer error AND personal opinion.

Personal "OPINION" + Computer nonsense=National championship contenders.

ACTUAL results + ACTUAL Schedules=meaningless.
 
NavyTider said:
????!!!!!?!?!?!?!!!??!!!

The Sagarin rankings have USC #2 with the 3rd-hardest strength of schedule in the nation?

???!!!!?!?!???!!?!?!!!?!

I know they're supposed to use unbiased formulae to come up with the final results, but come on -- you cannot seriously tell me that USC (USC?!?!!?!) has played the 3RD HARDEST SCHEDULE of any team in the nation to date. This sounds suspiciously like when Florida State used to be good, and every year when the BCS results came out they would show FSU as having played one of the top 5 hardest schedules in the nation, even though your average football fan couldn't even name any of the other teams in the ACC because they sucked so badly.

Further, 6 of the top 10 hardest schedules to date have been played by PAC-10 teams, according to Sagarin. I know this is supposed to become more accurate as the season progresses, but seriously -- come on!

Oh, and we're #4, behind Texas, USC, and Penn State. Our SOS? 38th.
Click to expand...


I've wondered the same thing.
 
I believe most of the strength of schedule is based on W/L record only, something that will change between now and the end of the year...

USC - played Arizona State.
Cal - played AZ State and Washington St.
Oregon State's only played one team with a losing record...
and so on.

So, not only have they rarely played teams with losing records, the teams they have played have rarely played teams with losing records.

Compared to:
Alabama - played Western KY, Arkansas, Tulane.
LSU - played North Texas, MS State, and Appalachian State
Georgia - Georgia Southern, Arizona State, Tennessee

It just appears that the PAC-10 has, so far, played teams with better records. That doesn't mean that much right now. It will level out. Beating a team can hurt your strength of schedule - even moreso right now than at the end of the year...

I may have some of this wrong, but I believe that is why the Pac-10 has better SoS right now.
 
It's neither bias nor computer error. Come on people. The strength of schedule is determined, predominantly, by opponents wins and losses and opponent rankings.

Since the WAC and MWC have been beating up on the PAC 10, their losses aren't nearly as high. ;)

You have to remember that those wins and losses, early in the season, can result in some funky rankings. A 5-1 WAC team who beat Baylor, lost to Iowa State, and played no other BCS teams would be ranked higher than a 4-2 SEC team who lost to Georgia and LSU - if they're opponents win/loss record was roughly the same.

The longer you get in the season, the more the wins and losses of and against non-BCS schools get leveled out.

If you're trying to see how far and accurate you usually hit a golf ball, you need a decent amount of data. If you take 3 swings and take your results from that, you could get wildly inaccurate results. If you're a 10 and you hit 1 bad shot out of three, you look like a 20. If you're a 30 and hit three luckily good shots, you look like a 5. Hit 15 or 20 balls and you'll get a much more accurate reading.

The computer rankings are the same way. The closer we get to the end of the season, the more accurate they get, especially Sagarin.

If a #15 team is playing a #10 team in a bowl game then they should win, assuming the ranking is accurate. If the AP has team A at 15 and team B at 10, the AP poll is saying that team B should win. If Sagarin has team A at 12 and team B at 16, his ranking is saying that team B should win. I'd take Team A almost every time.
 
I couldn't begin to tell you how the SOS ratings work, but I can point out that USC's opponents to date currently have a significantly higher won-loss record than Alabama's opponents and I'm sure that factors into it. Going by current records, that will change dramatically in the second half of the year when USC starts to play teams will worse records, so their SOS will fall accordingly.
 
Wow. That just doesn't look/sound right, does it. I guess we just have to trust that it'll be "right" by the end of the season.
 
The explanations above are all correct, but I wanted to elaborate. The PAC 10 teams play a 9-game conference schedule (all other PAC 10 schools). SEC schools play an 8 game conference schedule (5 divisional games + 3 cross divisional games). Because they have to get one more game, they start conf. play earlier. Prior to Sept. 13, the PAC 10 had played 4 conf. games, and most of its members played conf. games on that date.

Comparatively, the SEC had played only one conf. game prior to Sept. 13, and only about half the teams played in conf. that weekend. While I know some of you will argue that our non conference cupcake games to start the season are as hard as a PAC 10 schedule, the fact is that most of their teams were squaring off against PAC 10 schools while the SEC was still in warmup mode against FCS opponents and WKU.

As the season progresses, and we play more in-conference games, our strength of schedule will generally rise, faster than theirs. It also does not help that UcheaT got beat by UCLA; Ohio St. has maintained its high ranking after getting drubbed by USC; and Clemson has been in utter chaos.

The real equalizer always occurs in the SECCG when we usually square off 2 top flight opponents, with the winner getting a big boost in the polls/rankings. PAC 10 = no champ. game.
 
The problem with SOS rankings is that it considers the 151st team to be 50 teams worse than the 101st, while also considering the 51st team to be 50 worse than 1. So you can play a much harder schedule (I consider # of very good teams to be the key factor) against teams, but if you play any really low ranked teams it's totally voided in terms of math. I think USC always has a high SOS because they don't play teams lower than the 50-75 range, even though they don't play a lot of really tough teams.
 
Alabama is 4th in the Sagarin composite. That is the average of the ELO-Chess score and the Predictor. The BCS uses the ELO-Chess number as part of the formula.

The Predictor takes margin of victory into consideration.

ELO-Chess uses the same system used to create chess ratings [ame]http://en.wikipedia.org/wiki/Elo_rating_system[/ame]

The strength of schedule Sagarin uses is based off of pre-season ranking,but only until all the teams are connected in the matrix (Team A played Team B, who played team C, etc.) Once all teams are connected the SOS is based off opponents' W-L records.


TL;DR: The only column you need to worry about in the Sagarin rating is the ELO-Chess column, where we are #1
 
Interesting, I didn't realize that the ELO-Chess rankings were the ones used by the BCS. Makes me feel a little bit better. Of course, I know in the long run we control our own destiny, but sometimes the stuff that people come up with in regards to football rankings just makes my head hurt.

In reference to the earlier poster's call for a playoff system, I'd be for it, as long as it involves no more than 8 teams, and only includes conference champions. I am not interested in a format that makes it possible for a team that didn't win their conference to claim a national championship because they happened to go on a run at the right time.
 
Is it just me or is those rankings a little flawed? It says on the site where his rankings are listed that alabama is 1-0 against top 10 teams this year but we've played and beat two top 10 teams? How does that work?

HOME ADVANTAGE= 2.78 RATING W L SCHEDL(RANK) VS top 10 | VS top 30 | ELO_CHESS | PREDICTOR
1 Texas A = 95.21 6 0 70.47( 43) 1 0 | 1 0 | 93.95 2 | 94.82 4
2 Southern California A = 94.56 4 1 78.57( 3) 1 0 | 2 1 | 85.79 13 | 106.07 1
3 Penn State A = 93.02 7 0 67.44( 74) 0 0 | 1 0 | 87.56 6 | 96.50 3
4 Alabama A = 92.85 6 0 70.76( 38) 1 0 | 1 0 | 94.95 1 | 90.57 6

Also if those are schedule rankings then however the man gets
those rankings are really flawed.. USC definitely does not have
the 3rd hardest schedule in the country, maybe the 3rd easiest.
 
UA_Fan87 said:
Is it just me or is those rankings a little flawed? It says on the site where his rankings are listed that alabama is 1-0 against top 10 teams this year but we've played and beat two top 10 teams? How does that work?
Click to expand...
It's "current" top 10, which is the way it should be. Who cares at the end of the year if a team had a preseason ranking of #5 if they finish the year 0-12. They obviously weren't a top 10 team.
Also if those are schedule rankings then however the man gets
those rankings are really flawed.. USC definitely does not have
the 3rd hardest schedule in the country, maybe the 3rd easiest.
Click to expand...
Quite a few people have explained why, specifically, USC's schedule (to date) has indeed been tougher than most.

Just compare USC's schedule to Bama's, for example.

Bama's opponents to date have current ratings of: 55, 126, 117, 68, 7, and 32. (Average = 67.5)
USC's opponents have ratings of: 53, 10, 17, 24, and 58. (Average = 32.4)

That's what we get for playing Tulane, Western Kentucky and Arkansas State, while USC's OOC slate is Virginia, Ohio State and Notre Dame. I know Virginia and ND aren't very good right now, but they're way better than W. Kentucky and Arkansas State. In a "normal" year, we could expect our tough in-conference slate to balance out our poor out of conference slate, but with Tennessee and Auburn being terrible this year, and Clemson starting to look terrible too... it is what it is.

As others have said, as the season progresses and more of USC's opponents start to play more difficult games, and USC itself starts to play teams that already have bad records, this will begin to even out. But right now, the AVERAGE rating of Bama's opponents is lower than ANY of the teams USC has played to date.

FWIW, here are the current ratings for the teams left on the schedules (assuming a SECCG against Florida).

Bama: 49, 59, 96, 23, 74, 43, 6 (Average = 50) Overall season average = 58
USC: 134, 37, 107, 15, 47, 39, 73 (Average = 64.6) Overall season average = 51.2

Would you care to add any evidence to support your opinion that it's been easier than most?
 
Last edited:
You must log in or register to reply here.

New Posts

Advertisement

Trending content

Advertisement

Latest threads

Top Bottom
Back