Ugh, baseball.
Let’s just draw a line in the middle of the thread. I’ll stay on my side, you stay on yours.
The diminishing returns would be very important. I’m not sure I am ready to see a college season in which teams have to ruthlessly crush every opponent by mathematical necessity.
Thanks, sir. I’m 1) a Bama fan, and 2) an SEC fan. That, as you well know, means that defensive strategy is paramount to me.
(Of course, apparently #2 above also makes me a malefactor and evil-doer, with questionable morals and shady personal habits. I can live with it.)
Roll Tide.
Perhaps what you meant to say was this (my personal opinion):
“A CLOSE loss to a hightly ranked team should be more valuable than a CLOSE victory against a bottom feeding team and Div I-AA team.”
I think that a team ranked #25 that loses to the #1 team in overtime should jump in the ratings. I might go so far as to say that if you beat the spread you should increase. I mean, if the #25 team is a 20 point underdog to the #1 team, and it’s a close game, then didn’t the #25 perform better than expected, better than their ranking indicates, and indicate that they are a better team than predicted when they were #25? Shouldn’t the opposite go for the #1 team?
That’s the beauty of the computer rankings. They don’t give a fuck about the preseason rankings, or even last weeks rankings. But again, their flaw is not taking margin of victory into consideration. I can’t remember the mathematic term (I’m thinking “logarithmic” or something similar. If you saw it graphed it would start as a 45 degree angle, then curve until it was approaching horizontal. What’s that term, math people?), but a win by 20 points should count for more than a win by 10… but not twice as much. Maybe there should be a cap, (somewhere between 30 and 50 points), to keep teams from running up the score, or maybe the formula could be such that trying to score a touchdown when you are up by 25 would be so minimal that it would not outweigh the negativity that would come with it. Sometimes a team winning by 20 is a closer game than another where a team wins by 20. That’s what the human polls are there for.
My biggest problem with the human rankings is how simplistic the logic is. For years I’ve been thinking about starting a thread where we dopers put together a computer program that ranks the teams. Writing a computer program that mimics the human polls would be much easier. The hardest part would be deciding how far a team that loses would drop. The rest would be simple. You start with the preseason poll, and teams that lose drop x number of spots. The rest fall into place like checker pieces in a Connect-4 grid.
I think it’s fairly safe to say that if Alabama or Texas or any of the currently undefeated teams had started the season as the preseason #1, they would still be there. (There has been some movement in the top 3, but if you look closely, you’ll see that they are so close that just a few shifts in first-place votes can make the difference between #1 and #3. This gives me some hope for the human polls.) However, for the most part, if you are the preseason #1, and don’t lose, you’ll be there at the end of the regular season.
ETA: “Roll Tide”. BTW, I’m not sure why I’ve paid attention to the polls and number of first-place votes. 'Bama could drop as far as 4th in the rankings, but as long as they win out, they’ll be in the NCG. I kinda feel sorry for teams like Cincinnati that can’t say that.
As I said in my post, the margin of victory would have diminishing returns, or logrithmic. A team gets a fixed amount for a win and then an incremental amount for margin of victory.
the points formula could be something like this:
Points = 10 + 3 * log(x) where x is the margin of victory, capped at 14 possible points, equivalent to a 21 point win.
A one point win would be worth 10 points.
A seven point win would be worth 12.5
A twenty point win would be worth 13.9
A forty point win would be worth 14.8 which would round down to 14 points
IMHO, this is a bit too steep.
IIUYC, you get 10 points for winning.
You would get an additional 1 point for a 3 point win.
You would get an additional 2 points for a 5 point win.
You would get an additional 3 points for a 10 point win.
You would get an additional 4 points for a 21 point win, and everything above that is worth nothing?
I don’t want there to be a reason to run up the score, but, as I indicated before, I’d set the cap at 30 at the lowest. JMHO
But isn’t that because recently, the SEC has had better showings in the BCS standings/bowl games? I honestly am guessing.
We get into real deep & murky waters very quickly here. Yes it’s a fact that 2 SEC teams have won the last 3 NCGs, and that the SEC did very well in last year’s bowl games. (The Pac-10 did better, btw).
But what’s less fact-based is how much last year’s results influence this year’s polls. Should Auburn, South Carolina and Ole Miss get the benefit of the doubt because Florida’s good? And conversely, do we look at Wisconsin and Iowa skeptically because Ohio State can’t seem to win the Big One?
Defining “Recently” as the last three years:
the SEC is 5-1.
Florida 2-0 winning two BCS Games (against OSU and OU)
LSU 2-0 winning one BCS game (against OSU) and the Sugar Bowl (against Notre Dame)
UGA is 1-0, winning the Sugar Bowl (against Hawaii)
Bama is 0-1, losing the Sugar Bowl (against Utah)
The Big 10 is 0-6.
OSU losing 3 games, Losing 2 BCS Games (UF, LSU), and the Fiesta Bowl (Texas)
**Michigan, Penn St, and Illinois **losing one game apiece, each loss to USC in the Rose Bowl.
Wiki has a summary of all the BCS games.
So many handsome, intelligent chaps in this thread.
FTR, I look at Iowa skeptically because they beat Northern Iowa by one point at home. A Division I-AA team. And they beat Arkansas St by three points at home.
I look at Wisconsin skeptically because they have not scheduled a BCS opponent in several years.
I look at Penn St skeptically because their OOC schedule was Syracuse, Temple, Akron, and Eastern Illinois.
Yes, U. Florida is SOOOOOOOOOOOOOO much more respectable for having an out of conference schedule of Charleston Southern, Troy, Florida Int’l., and FSU. :rolleyes:
And they were 4-0 before that (a Sugar and three Fiestas).
Again, these things are cyclical.
(1) Florida has won two BCS games in the last three years
(2) Florida is not a good football team right now, but they have won multiple National Championships, were basically the team of the 90’s and early 2000’s. Given that CFB schedules are made years ahead of time, there wsa no way to tell that FSU would be a mediocre football team this year.
on the other hand, what kind of football recent resume do Akron, E. Ill, Temple, and Syracuse that would say they would field a competitive team?
Akron: 1 bowl game in their history, the Motor City Bowl in 2005
E. Ill: , not a Div I team, however they did lose the 1948 Corn Bowl, against Illinois Wesleyan
Temple: Last Bowl Game 1979
Syracuse: A credible football history, but no bowl games since 2004, and only two bowl games this century.
Way to go out on a limb Penn St.
No doubt,
Again BobLibDem started this discussion by opining that a one loss SEC or Big XII team doesn’t rule them out of BCS discussion. Well, the fact is that OSU lost a late season game two years ago against Illinois and they were still in the BCS game.
IMO, given comparable records and comparable Strength of Schedule, the SEC deserves the benefit of a doubt over the Big 10 and any other conference, simply because of its performance in the big games.
This should say Florida State is not a good football team. I missed it when I proof-read. Sorry for the confusion.
Um, that’s because that was the only game OSU lost that year (besides the NCG, of course). The only other team in the top 25 BCS standings with only one loss was Kansas, and, in addition to being, you know, Kansas, they gained their loss even later than OSU did. (Hawaii was undefeated that year, and there was some outrage about how they got ‘shafted,’ but I’m ambivalent about that argument.) Every other potential NCG game participant had 2+ losses that year, so that’s why OSU made the NCG with a late-season loss.
Penn State didn’t win all their conference games. Florida has(until Bama rolls in the Dome, they’re my SEC team). When Penn State, Texas, and Florida all win out, that’s when non-conf. schedules matter.
And for those who complain that the BCS makes teams schedule pansies while those who shoot high and fall short get punished, that’s exactly why a 1-loss SEC, Big 12, or Big 10 team gets into the NC game over an undefeated Boise State.(Point taken about the difficulties or getting someone to play Boise). What the voters are doing is rewarding a tougher overall schedule over an unblemished record.
Except that that “tougher overall schedule” thing is a self-perpetuating myth, and has no bearing on whether or not Boise St. is or is not better than a one-loss SEC team (for example). The only way to know how Boise St. stacks up is to let Boise St. play in the Championship and see what happens. If Boise St., or TCU, or some other such non-BCS conference undefeated team makes it into the championship, and gets blown out, everyone could forever more point to that and say, “See, being undefeated from the MWC is meaningless.” But what the BCS conferences are really scared of is that, if they were to have Boise St. or TCU or some such come play in the big dance, and they WON, then they would have to admit that their whole gentleman’s club set-up, which is so thoroughly designed to be as exclusive as possible, in order to keep the riches of the system to themselves, needs to be abolished.