eCash is on the line SaturdayThe first batch of the BCS standings will be released on Sunday, October 19th, but that hasn't stopped curious minds from digging into the available data to get an idea of what to expect. We linked yesterday to a Dr. Saturday article highlighting the fact that LSU sits atop each of the five sets of rankings that have been released (Peter Wolfe's first set will be released the day the BCS rankings come out).
Of course, the computers can't take into account things like the ever-so important level of blue blood respect and coaching tree or conference affiliations you see in the Coaches Poll. Oh, and margin of victory:
Remember, though -- just to demonstrate how objective and definitive this process really is -- that the computers might spit out completely different results if they were actually allowed to consider all the factors that human voters are; the most glaring of the many, many fatal flaws in the BCS computer system is that the machines aren't allowed to account at all for margin of victory in their rankings. Beating Mississippi State by four points in the final minute, for example, is exactly the same to them as beating Mississippi State by 25 points in a game that was effectively over in the third quarter.
So basically, the Evil Tress/Hazell braintrust might be earning the Buckeyes some love from the humans, the computers don't care if you beat Illinois 30-0 or 3-01. This is reflected in the two human polls, which, on average, have the Buckeyes 6.4 spots higher than their silicon counterparts. In the five computer rankings, the Buckeyes are 8th, 9th, 18th, 18th and 19th.
Looking on the bright side, the three polls that have OSU outside of the top ten, Anderson & Hester, Colley, and Massey, have the Badgers 5th, 3rd and 3rd, respectively. And they like Iowa even more. There's a lot of football to be played, to be sure, but with that comes plenty of opportunity to make up some ground, both with the computers and with the human voters.
| IOWA | WISC | OSU | PSU | MICH | |
|---|---|---|---|---|---|
| Sagarin | 9 | 23 | 8 | 18 | 42 |
| Anderson & Hester | 3 | 5 | 19 | 24 | 29 |
| Billingsley | 8 | 30 | 9 | 16 | 39 |
| Colley | 2 | 3 | 18 | 23 | 27 |
| Massey | 2 | 3 | 18 | 26 | 20 |
| Computer Avg | 4.8 | 12.8 | 14.4 | 21.4 | 31.4 |
| Coaches Poll | 14 | 25 | 8 | 12 | NR |
| Harris Poll | 11 | 25 | 8 | 13 | NR |
The Big Ten as a whole is in pretty good shape with the computers. As I mentioned, they absolutely love the Hawkeyes, giving them an average rating of 4.8. Of the four that publish conference rankings, Massey has the league 2nd, Anderson & Hester 3rd, Colley's 4th and Sagarin is the set we ignore because he gives an edge to the Big East over the Big Ten (the UC love is getting a little out of hound).
Before I get into wild guesswork linking to other sites that know what they're talking about, it might be handy to review the BCS formula:
BCS = (CP/1575 + HP/2850 + (C1 + C2 + C3 + C4)/100) / 3
That would be a team's Coaches Poll points divided by the max plus the same formula for the Harris Poll, then the high and low computer rankings are thrown out, the remaining four are added up and divided by 100 and then finally, the whole thing is divided by three to reveal the magic tens of millions of dollars winning numbers.
The guys that actually study this stuff for a living predict Iowa would be 9th, followed by the Buckeyes 10th, Penn State 15th and Wisconsin 18th if the BCS standings were to be released today. Assuming the Buckeyes take care of business on the field the next two weekends (and the Badgers still make me more nervous than they probably should) and the spate of upsets marches on, they should find themselves in a slightly better position than they were in when the initial standings were released last season. Catch fire and finish strong against Penn State, Iowa and Michigan and who knows where this time might end up.
So Far, They're Biting as Pups
Marcus Hall saw the first action of his career Saturday night in Bloomington and this week, thanks to the flu bug hitting some teammates, he's repping with the ones. Tressel indicated Hall could play two or three series Saturday against the Badgers and may start to earn more time going forward. With Block O finally voltron'd up (and getting a chance to grow into one set and base of plays) and Jim Cordle expected to be out a couple of more weeks, a lot of young guys are getting quality snaps on the line. It's almost scary to think about this offense next season when everyone but Ballard returns and Pryor, Boom and Posey are juniors.
Improvements for iPhone Viewers
A couple of you have written in asking for access to polls, Good Shizzy and other portions of the site that weren't available when browsing with an iPhone and we finally got around to taking care of that. We upgraded to a new, more flexible iPhone theme that provides some additional features/options, not the least of which is the ability to view the site in it's "normal" state (or what you'd see in a non-mobile browser). Hopefully this works out for you guys and as always, feedback is welcome.
- 1 As Hinton points out, both Sagarin and Massey's ranking comes in two sets. The first is their actual rankings, while the second removes margin of victory to satisfy the BCS. When considering MoV, the Buckeyes jump from 18 to 12 in Massey's index and from 8 to 4 in Sagarin's set. Odder still, the Sooners jump from 70 to 19 for Massey and 16 to 3(!!) for Sagarin when factoring in win margin of victory.
Comments
Do these computer polls have any real criteria? I mean is there a documented reason why their computers spit out the data they do? This is what I think is Insane about the BCS. LSU is CLEARLY not the number 1 team in America, and they top 5 of the computer polls? I don't care if margin of victory plays into it or not where is this number coming from? Tebow or not, Florida will beat them Saturday. I would go on to say that Auburn, Bama, and Ole Miss will probably get LSU too. I don't understand all the love for that team this year.
Small sample sizes plus no margin of victory bias = what you see. They'll look more logical in about 9 weeks. Except Billingsley.
The computers always weigh a loss more than it should. That is why you have to ignore them until the end of the year, and also why there are such huge swings in the BCS numbers when somebody loses. Boise State will be at or near the top of the computers at the end of the year b/c they won't have a loss. It's crap, but it's the current system.
looks like in the long run if we win out, we will be doing good in the
polls. My only hope is the rest of the big team top tier teams can keep
winning which will vaildate are rankings...
Just curious i wonder where UC is ???
I swear, reporters looking to fill space could do a profile story on a "somewhat good player hailing for Ohio playing for a mid-tier Big 10 team that was not recruited by OSU but maintains that he dislikes OSU and that he would never have played for Buckeyes even if offered, ostensibly because School X is the best fit but more likely because he wouldn't have cracked the two-deep in four years of play" for every game of the year, except the big nonconference game, and maybe the Michigan game. If they get their recruiting back up to par.
http://www.jsonline.com/sports...
Isn't it fair to suggest there could be a major bias in the programming of the computer polls? I mean on a computer program you can manipulate the system to come out with something favorable to the person who made the whole program. Is this regulated at all? Does anyone have to approve the computer poll before it gets used. It's nuts that computers are used to come up with something like this.
The BCS lost credibility in my mind when they implemented the "USC rule" (strength of schedule is not considered) after the 2003 season.
Evidently, not much else is going on this week; "At UW, not at Ohio State."
It's not nuts that computers are used. It makes a *ton* of sense to use computers. Something that uses an ELO type system to calculate relative strength of schedule makes a lot of sense.
What *doesn't* make sense is that these algorithms are not published, and are essentially just made up by random guys with websites. Moreover, they aren't scientifically adjusted to match the actual results of bowl games, but rather to match the various polls as closely as possible.
Whether trying to judge which teams have the most accomplished resumes (ie, they have beaten the best teams) or who is most likely to win on a given day (these are two very different things) computers make a ton of sense. It is just ridiculous that such an important *business* decision (worth millions of dollars to various conferences) is not more open, consistent, or scientific.
I'm not sure that the system ever had credibility.
I dont' know, do you really think that Wisconsin is number 3? I dont' care if they are ealy results because of a small sample, they aren't number 3. I also think it's crazy that the coach's poll carries so much weight being as the coaches rarely vote themselves, they have major conflict of interest, and now their votes are revealed. How do you make sure that Ohio State goes up from number 3 at the end of the year vote yourself number 1 and Florida out of the top 10. I know it's only one vote, but we know SEC people carry a major bias towards other SEC teams, and I'm sure Big 10 coaches do as well. The whole system is ridiculous and there are way to many Millions being given out and made to be dependent on such a shotty system.
There was a link in the good Shizzy the other day, With a site trying to do a new kind of top 25 where you can vote any one have the link. I accidentally deleted it in my bookmarks.
Also not sure if any one has seen this but this is pretty good. The Dispatch's website now has a player data base. of all time players
http://www.dispatch.com/live/c...
I think that LSU being ranked the Number 1 team in the country makes a lot of sense, when you compare there schedule and wins to other undefeated teams. They have quality road wins over Washington (who btw beat USC) and Georgia (putting aside the fact the SEC rigged the results with the penalty call.) They win with sensational defense and solid mistake free offense. Remind you of any former national champions?
Who cares what I think, what have they done on the field? Anything other than that is preseason hype. Football isn't about style points.
The computers have been kind to the Buckeyes in years past and elevated us higher in the BCS standings than the human polls alone would have placed us. I think they've made some pretty reasonable adjustments to the system over the years and it will probably evolve some more...
My hope is that some day the BCS rankings end up doing more than just crowning a #1 vs. #2 for a MNC game... Would be a pretty decent methodology to seed an 8 team, single-elimination, playoff bracket, no? (Or, a 16 team bracket...)
What's the easiest way to make the Wisky game closer than it should be? Have the flu bug ravage your team. Boren and Ballard are sick, and I also read that Adams might have it as well...
As awesome as this is, EYEMA GUNNA GRYPE---- stats, please?
I hate practice, too.
No.
Some of them (Colley's or Massey) for example have pretty good descriptions of the algorithm available. Sagarin is secret I think but I've seen a back-engineered version that is pretty close. Billingsley's is weird. They are designed to match previous results using limited data (scores, win-loss) but since college football is both unpredictable and only plays a few games it is a hard problem. For example, the Buckeyes played UCLA twice in the 1975 season - the first game we won 41-20 in Los Angeles and the second game (Rose Bowl) we lost 23-10 - so what should a computer say? Not to mention all the times team A beat B and B beats C and C beats A. Most of these are (or started out as) hobby projects and are not terribly sophisticated (but free). I think any decent mathematician/programmer could probably code up Colley's method in a few days for example. Now, if someone threw 5 or 10 million $ at the problem and hired a team of data mining/AI/stats/math experts to work on it fulltime for a couple of years you might get somewhere. Or not. Maybe someone already has (like the big Vegas casinos).
I was at that UGA game.. LSU was far far from mistake free offensive football. Honestly, both teams looked awful. It was hard to call either defense stout because the offenses were so inept.
Boren and Ballard too? I totally missed that news. Not good. Where did you hear that?
I vote we quarantine Pryor and make him sleep in a sanitized bubble the rest of the reason. We lose him to the flu and we are in big trouble.
Too bad all of the Quants are on Wall Street or we'd have the perfect algorithm by now.
I do like your Vegas thought. I can't say I'd really object if the Coaches poll was tossed and the top lines makers in Vegas contributed a poll to replace it.
Being an 11W noob, I just have to share that via the new iPhone theme I have discovered the 11W terminology page and have been laughing continuously for about 15 minutes. There are some funny motherf***ers on this site!
They BArely beat washington.. whose 2-3
They beat Vanderbilt by 14... Whose 2-3
They Beat louisiana Lafeyette 31-3.. whose 2-2 (in the sun belt amd btw lost to nebraska 55-0)
They Barely beat Mississippi State... whose 2-3
And they barely beat georgia... who yes, has the only winning record of any team they played: 3-2, This team got waxed, by a team who got waxed, by a team who got waxed by utep
SO explain why you think LSU should be number 1?
This is me by the way ^
They need Gonzo's bubble! Stat!
Well, it looks like Adams will be a go. But it's looking more and more like we'll be without a combination of Miller, Boren. and Ballard. I found this out through Ken Gordon and The Ozone. I guess we'll learn more from Tress today.
I did explain. I think that they have the best quality wins of any of the undefeated teams in the country. I think that they have a stout defense, and even if the offense isn't mistake free it does enough to win. I do not really care about by how much they beat a team by. A win is a win. Case in point: how many points did Ohio State beat Cincinati the year they won the National Championship? And a beating a team who beat another team only has a one degree of separation before it loses legitimacy. (My personal rule. I do not expect to be yours.)
Alabama, Iowa, and boise state, all have atleast one quality win over a quality opponent... I'll give you that texas and Florida, really havent played anybody... but theres no reason to put LSU over the first 3 teams I mentioned, especially bama
One more quality win than the three mentioned teams is at least one reason.
By how many points did Iowa beat Iowa State? (Sorry just turning your logic against you.)
Boise State's quality win was at home and a season opener.
And I will grant that LSU/Alabama is a very close call to make. That is why there is still the rest of the season to play out. Heck I feel pretty certain that Florida will beat LSU. But if you asked me who should be the highest ranked undefeated team right now, I still pick LSU over 'Bama beacause of one more quality win.
You give me your reasons for picking the three teams you mentioned to be ranked higher than LSU.