Is Boise State Screwed?

You say that like it’s a bad thing…

You’re hearing this here first – Boise State has virtually no chance to play for the BCS title.

And it doesn’t matter a bit what they do from here on out.

Let’s go with the facts we know so far:

Boise State is currently ranked 3^rd in the BCS standings. This is based on a 2^nd ranking in both the human polls (Coaches’ Poll and Harris Poll), and an average ranking of 7^th in the Computer Polls.

The math adds up as follows –

The BCS uses percentage of total points to calculate a score. Right now Boise gets .9504 for the Harris Poll, .9390 for the Coaches Poll, and .780 for the Computer polls. Average is .8898, putting Boise in 3^rd.

Here is your top 5 –

1 Oklahoma .9215

2 Oregon .8921

3 Boise State .8898

4 Auburn .8641

5 TCU .8573  

Oklahoma gets the top spot by being 4^th in the Harris and 3^rd in the Coaches’ but 1^st in the Computers. If the same scoring system were utilized without computers we would have this –

1          Oregon .98315

2          Boise State       .9447

3          Oklahoma        .8922

4          TCU    .886

5          Auburn .8462

Not only is Oklahoma 3^rd, it’s a distant 3^rd.  In other words, the computers alone pull up the Sooners to first place on average.

The computers also crush Boise, taking .0549 off their average.

Why are the Broncos so low in the computer rankings? After all, Boise State currently has the 3^rd best Scoring Offense, and the 3^rd Best Scoring Defense, in the NCAA. Their margin of victory is a close second to Oregon’s. So their prolific scoring isn’t helping them in the computers.

No, there is only one reason the Broncos are 7^th overall – their lousy Strength of Schedule

Not all of the computer polls reveal (or perhaps even utilize) SOS. But in three that do the Broncos presently rank as follows –

Sagarin – 55^th

Anderson & Hester (A&H) – 49^th

Colley – 73^rd

Notably, this is the Broncos rankings AFTER the toughest part of their schedule is behind them. There are no teams left like Virginia Tech or Oregon State on their schedule.

The reason we are sure SOS is the main weakness of the Broncos in the computers is fleshed out by a comparison to the other teams ranked above or near them by the computers. The following grid shows the BCS computers top 9 ranked teams, their NCAA SOS, and their SOS in both A&H and Colley –

Computer	Team	NCAA SOS	A&H	Colley
Ranking		Current	SOS	SOS
1	Oklahoma	20	29	13
2	LSU	15	24	28
3	Auburn	31	27	33
4	Michigan State	29	43	53
5	TCU	61	65	77
6	Missouri	26	48	44
7	Boise State	73	49	73
8	Oregon	111	82	113
9	Oklahoma State	63	76	89

Clearly the various team’s SOSs lines up very nicely with their present computer rankings. (In A&H, for instance, LSU is their number 1, Auburn their number 2).

So what can we expect for their future SOS?

The following is going to be necessarily based on speculation, as most computer polls don’t reveal their methods of calculation. Moreover, we have to guess as to what the schedule strength of the remaining teams on Boise State’s schedule will be.

However we can make some solid estimates. For one, Boise is only remaining opponents are in their conference, the WAC. Boise’s remaining schedule is –

Oct 26 vs Louisiana Tech

Nov 6 vs Hawaii

Nov 12 @Idaho

Nov 19 vs Fresno State

Nov 26 @Nevada

Dec 4 vs Utah State

So what future SOS can we expect from this schedule?

Obviously, not a very high one.  But looking at Boise’s future SOS in a vacuum isn’t as helpful as looking at the SOS’s of their main competitors for a BCS title game – the 8 other undefeated FBS teams (still in the order of their current BCS computer rankings).

Computer	Team	NCAA SOS
Ranking		Future
1	Oklahoma	14
2	LSU	5
3	Auburn	6
4	Michigan State	42
5	TCU	68
6	Missouri	24
7	Boise State	55
8	Oregon	34
9	Oklahoma State	4

Almost every team here has a future schedule per the NCAA that is better than the present, with the exceptions of Michigan State and TCU. And with the exception of TCU, all of them are better than Boise’s. Everything else remaining stable, Boise State’s SOS standing in the computers should stay about the same by early December as it is now.

Of course, everything WON’T remain stable. We know that Oklahoma plays both Missouri and Oklahoma State, and LSU and Auburn face off this weekend. Can we predict what the BCS might look like come early December?

We can try. Before getting hung up on what the human and computer rankings are, we need to focus on how the BCS is formulated. Rather than an average of rankings, it is an average of calculations. The calculations are all based on a number derived from vote placement. For example, a 1^st place vote in the polls is worth 25, 2^nd place 24, and so on. These are added together and made into a percentage of a perfect score. In the case of Oregon, the 1^st place team in both human polls right now, this percentage is .9819 in the Harris Poll, meaning they have 98.19% of a “perfect” (all number 1 votes possible) score.

So all 1^st place rankings aren’t created equal.

We have no way of knowing what the percentages will look like come December. But we can look at the average ratios for the rankings based on past polls. Here is the average percentages by ranking (top 10) in the human polls for the past 5 year’s final poll –

1	0.9922
2	0.9486
3	0.9129
4	0.8569
5	0.8269
6	0.7987
7	0.7495
8	0.7261
9	0.6741
10	0.6409

Oregon, the number one in the human polls, has an average of 0.9832 right now (human polls only). This is lower than average, but even in the chaos of 2007 Ohio State had the same 0.9832 human poll average.

So, recognizing Boise will likely never have a great computer poll percentage , what do they need by the humans to overcome this?

Right now the Bronco’s computer percentage is 0.780. If we give them the benefit of the doubt and give them a 1^st place ranking in December based on historical averages (a VERY generous scenario) and today’s computer ratio we get –

Average of (0.9922, 0.9922, 0.7800) = 0.9215

Where would a 0.9215 have landed a team in the last 5 BCS final (pre-Title game) polls?

2009 – 3^rd

2008 – 4^th

2007 – 3^rd

2006 – 4^th

2005 – 3^rd

In other words, a team with a very high ranking (and percentage) in both human polls but a 7^th or so ranking in the computers probably doesn’t play in the title game.

The 0.780 final computer percentage for Boise isn't necessarily far fetched. Historically, we have a comparison – and that’s the 2009 Boise team. In the first BCS standings of ’09, Boise State had a BCS computer percentage of 0.0840, or 5^th overall. Much like this year, Boise had played the “meat” of their schedule already, including a win over eventual 10-3 Oregon. By the end of the regular season Boise’s BCS computer percentage was 0.7800, or 6^th overall. This was after playing their weaker conference schedule. Notably, their record then (and presumably, this year) was undefeated at all times.  Looking at this, a final of 0.7800 this season may be too high.

Now, quite obviously, there is a lot of speculation involved in these calculations. A total chaos scenario that has all the BCS conference teams with one loss probably gets Boise high enough in the computers to slip in. That said, the calculations I used to get the 0.9215 final average were VERY generous to Boise on the human poll side. To get that high a human poll ranking/percentage, Oregon is going to have to lose a game. (The team I saw last night (Thursday) didn’t look like it will.) Moreover, with virtually all the other contenders standing to strengthen their computer standings, (not to mention at least one 1-loss team jumping Boise in the computer polls eventually, and probably more) Boise is in a serious predicament at this point in time. And if there are 2 undefeated BCS conference teams, they haven’t got a chance in Hades, regardless of human rankings.

So to the defenders of the current BCS and BCS conference teams – thank the machines. In this case they are going to keep the humans in check.