Week 1: Saturday Predictions

The first 10 or so games are under our belt, and now it's time to move on to the main course. Starting next week we'll be moving on to a full look at several key matchups, but as of right now there's not too much to say since all of our data is based on last year's standings. Most of these wouldn't be too competitive no matter which year they were played -- hello Kent State at Alabama -- but in a few other instances it'd be useful to have more info about this year's teams -- we're looking at you, Boise State versus Georgia.

Now, presented without further commentary, are the first batch of predictions for the 2011-2012 college football season.

Home			Visitors			Home			Visitors
3	Alabama	42	100	Kent St.	18	3	Alabama	45	98	Kent St.	3
10	Auburn	48	97	Utah St.	25	7	Auburn	49	108	Utah St.	17
105	Ball St.	32	90	Indiana	35	104	Ball St.	28	87	Indiana	31
2	Boise St.	38	23	Georgia	29	5	Boise St.	38	21	Georgia	20
58	Boston College	30	71	Northwestern	24	65	Boston College	31	72	Northwestern	20
69	Fresno St.	31	45	California	34	73	Fresno St.	27	49	California	31
34	Clemson	35	68	Troy	26	37	Clemson	41	81	Troy	20
74	East Carolina	30	20	South Carolina	41	77	East Carolina	24	19	South Carolina	38
17	Florida St.	44	109	LA-Monroe	23	16	Florida St.	41	102	LA-Monroe	7
11	Florida	43	113	FL-Atlantic	20	18	Florida	56	NA	FL-Atlantic	3
56	Hawaii	38	80	Colorado	30	62	Hawaii	35	67	Colorado	27
63	Houston	36	78	UCLA	29	68	Houston	34	66	UCLA	27
40	Illinois	39	88	Arkansas St.	27	35	Illinois	41	90	Arkansas St.	21
9	LSU	33	5	Oregon	34	11	LSU	29	4	Oregon	28
62	Michigan	35	67	Western Michigan	30	54	Michigan	45	85	Western Michigan	17
59	Mississippi	33	38	BYU	32	58	Mississippi	31	56	BYU	27
18	Missouri	37	89	Miami-OH	22	22	Missouri	38	93	Miami-OH	10
120	New Mexico St.	30	83	Ohio	36	117	New Mexico St.	21	82	Ohio	34
116	New Mexico	34	112	Colorado St.	32	118	New Mexico	24	107	Colorado St.	27
30	Northern Ill.	40	84	Army	27	55	Northern Ill.	38	79	Army	17
21	Notre Dame	32	49	South Florida	25	25	Notre Dame	31	44	South Florida	20
4	Ohio St.	47	119	Akron	16	6	Ohio St.	62	119	Akron	0
16	Oklahoma St.	48	110	LA-Lafayette	24	17	Oklahoma St.	52	113	LA-Lafayette	14
7	Oklahoma	42	47	Tulsa	29	8	Oklahoma	48	59	Tulsa	17
24	Pittsburgh	39	114	Buffalo	20	30	Pittsburgh	59	115	Buffalo	7
82	Purdue	32	96	Middle Tenn.	26	78	Purdue	34	NA	Middle Tenn.	20
25	USC	38	70	Minnesota	27	31	USC	38	83	Minnesota	20
60	Southern Miss.	39	79	LA Tech	31	57	Southern Miss.	38	86	LA Tech	24
6	Stanford	49	115	SJSU	20	1	Stanford	52	111	SJSU	6
33	Texas	44	98	Rice	26	32	Texas	45	101	Rice	17

Key
Close game							Certain victory

The 2011-2012 Tempo-Free Gridiron

Normally around this time I'd be writing a post about the changes I've made to my system over the last year and how those will affect the rankings and predictions for the coming year. I say "normally" because even though I've done that periodically over the last two years, generally the TFG predictor has done well enough to stay just ahead of Eddie's RBA system. With success comes a degree of complacency, but that complacency has allowed me to go through and automate a lot of the day-to-day running of the blog by generating the predictions, rankings, and projections, and uploading that to the blog for the final tweaks.

Besides, it's fun to watch Eddie fume and rage about trying to wipe the grin off my face.

Instead during the off-season I spent most of my efforts trying to support the massive amount of content we're planning to churn out this coming season. A basic overhaul of the blog, better weekly projections, more informative rankings, complete per-conference projections, integration with our Twitter feed, improved game selection for our not one but two pick'ems (both for TotalFark members and for the public at large), an upcoming countdown of the best BCS champions of the last decade, a regular look at the remaining undefeated teams (as opposed to the occasional glance last year), and more behind-the-scenes work on making it easier for us to get this information to you.

So while I've been busy contributing to the common good here, Eddie's been focused on himself and flailing around trying to catch up to me. I'm not worried just yet, and probably won't be until near the end of the year. After all, last year he started out strong, only to flag down the home stretch.

Maybe in the meantime I can start putting together predictions for that other football league.

Week 1: Mid-Week Predictions

So here we are with the first predictions of the new football season. Just a reminder that all the rankings here are the same as they were at the end of last year; they'll slowly start to adjust as our computers get data from this year. Given that, it's a little difficult to predict this first week or so, but we're still putting it out there. Also remember that we only examine games between two Division I (FBS) schools, so when Powerhouse U. plays Sister Mary's School of the Blind, that won't show up here, either. It's almost as if many of the big schools don't want to risk losing to a middle-tier out-of-conference opponent. Weird, right?

Home			Visitors			Home			Visitors
85	FIU	39	106	North Texas	30	90	FIU	34	104	North Texas	24
86	Idaho	38	101	Bowling Green	31	93	Idaho	34	108	Bowling Green	21
55	Kentucky	41	111	Western Kentucky	28	46	Kentucky	49	112	Western Kentucky	17
117	Memphis	23	32	Mississippi St.	37	111	Memphis	14	22	Mississippi St.	55
76	Syracuse	33	94	Wake Forest	27	64	Syracuse	28	65	Wake Forest	21
14	Wisconsin	48	107	UNLV	25	12	Wisconsin	55	103	UNLV	14
65	Baylor	26	1	TCU	39	56	Baylor	21	2	TCU	41

Key
Close game							Certain victory

The 2011-2012 Regression Based Analysis

I really want to beat Justin. Not just "sort of" beat Justin. I want to wipe that ACC smirk of his face and really give him the business.

The bad news is that RBA is up to 73.2% accuracy since 2000 while TFG is at 74.0%, implying that Justin gets approximately 5.3 games per year more than I do. However, I have been making significant strides on the difference between expected accuracy and actual accuracy. As of today, the difference between my predicted and actual performance is down to 0.2%, meaning that I miss 1.3 games per year more than I think I should. This manifests itself as an improvement in the confidence metric, implying that I may yet beat him in the TFD pick 'em pool.

The major update to RBA comes in the form of conference versus non-conference home field advantage. Teams that play each other every year are less phased by harsh environments, whereas teams that infrequently travel into opposing stadiums are more rattled. Based upon my data since 2000, home field advantage is worth approximately 1.9 PPH in conference games. In contrast, home field advantage is worth 3.6 PPH in non-conference games.

I experimented with a lot of different home field advantage metrics throughout the summer, including per-team advantage, linear regressions by strength, and linear regression by strength difference. These features generally caused drops by 1% accuracy or more. Justin suggests that attendance makes a difference, but I don't have a way to get these numbers into the algorithm.

One of the ideas that Justin and I have thrown around is changing predictors when we detect a lot of mispredictions. I took a shot at this by pruning history when encountering N consecutive mispredictions and had very poor results. In general, I don't tend to miss many consecutive picks. Furthermore, by time we detect the change (presumably due to injury, suspension, or whatever), the change has corrected itself, meaning that we miss even more games later.

In summary, there have been major experiments but only minor tweaks in the RBA algorithm during the summer of 2011. However, I'm relatively confident that RBA will keep it close during the 2nd Annual TFD College Football Pick 'Em.

Welcome to the 2011-2012 season

Welcome to the 2011-2012 college football season!

After a lengthy slumber the Tempo-Free Gridiron is back for the new year, and the theme of this year is "More Content." Our goal for this year is provide you with more interesting and more frequent updates to keep you coming back for more. In addition to a new look, we've got a slew of new regular features for our readers throughout the year. These include:

• Tuesday Trivia. Each Tuesday in September will explore a random trivia topic. We've done this before, but now we're making it official.
• Undefeated Countdown. We did this last year in a somewhat ad-hoc manner, but this year we're doing it every other Thursday starting in October. We'll examine the remaining undefeated teams, their possible road to a perfect season, and the main obstacles in their way.
• Conference Projections. Starting in October we'll have a weekly breakdown of the 12 conferences, spread out over three days. These breakdowns will include both the current standings and the projected final standings. The schedule will be:
• Tuesday: ACC, Big East, Conference USA, Independents
• Wednesday: Big Ten, Big XII, Mid-Atlantic, Mountain West
• Thursday: SEC, Pac-12, Sun Belt, WAC
• Games of the Week. Along with weekly predictions we'll focus on specific Saturday games that are interesting in some way or another. These may be the top-billed games on ESPN that week, or they may be a showdown of two middle-of-the-pack teams that pit an explosive offense against an impressive defense -- the Unstoppable Force/Immovable Object (UFIO) game -- or simply the game that's expected to turn into a shootout.

This is all on top of our weekly predictions, discussions, rankings, and complete bowl coverage that we've provided the previous years. Keep an eye on the blog or follow us on Twitter.

Revisiting Northwestern

Long-time readers may remember my claim that Northwestern was the luckiest team over the last decade. Well, now I get some validation from an independent source. The Iowa-centric blog Black Heart Gold Pants published an article today highlighting the luck of Big Ten teams since 2002. It's a great article, and I highly recommend reading the whole thing. However, one particular quote stood out to me:

In every year included in this study, Northwestern had a positive overall pythagorean margin, and in all but one the LOLcats had a positive margin in conference play. Their gigantic lead in the chart above isn't due to one season; it's consistently winning a game to a game and a half more than expected. Northwestern's nine seasons are all within the top 30 overall (eight are within the top 20), and all but 2002's 1-7 Big Ten mark are within the top 22 in conference play. Once may be a random occurrence, and twice a coincidence, but if three is a trend, nine is gospel. Northwestern wins close games.

Clearly, Northwestern continues to exceed expectations, despite its academic pedigree and talent gap.

The Big East: Great NCAA Tournament Disappointment, or Greatest Disappointment?

There's been a lot of noise in the sports media about the Big East and their record number of bids to the NCAA tournament. A week ago, the chatter was all about how dominant the Big East was this year and how they truly deserved the unprecedented eleven bids earned by teams from Division I-A's largest conference. There were some naysayers, of course, but they were generally at the margins of the discussion. Each of the teams have proven themselves during the grueling Big East schedule. It used to be that if you finished at or above .500 in the ACC, you had a guaranteed bid to the tournament; this year, if you broke even or better in the Big East, you were in.

On paper, the Big East looked dominant: nine of the top 24 seeds were Big East teams. Going by the chalk these teams would make it to the round of 32, with 9-seed Villanova likely to pull an upset over 8-seed George Mason. Even 11-seed Marquette -- pitted against 6-seed Xavier, and having just defeated eventual 5-seed West Virginia in the Big East tournament -- was an upset candidate. Big East supporters dreamed big. Eleven teams in the round of 32! An all-Big-East Final Four!

Sneak Peek: Game Predictions

Today's sneak peek gives you a preview of the new look and feel for our per-game predictions. There are a few changes to the table to better fit the look and feel of the new site and to convey the same info but with less information overload. Instead of listing the odds of victory next to each game, we're using a color-coding scheme where the likelihood of victory is related to the intensity of the background color.

Home			Visitors			Home			Visitors
13	Arizona	36	28	Oregon St.	30	8	Arizona	35	36	Oregon St.	21
18	Arkansas	38	44	Texas A&M	32	23	Arkansas	38	51	Texas A&M	24
98	Ball St.	31	101	Western Michigan	29	99	Ball St.	31	111	Western Michigan	20
56	Baylor	30	31	Texas Tech	34	38	Baylor	29	37	Texas Tech	28
3	Boise St.	46	96	Toledo	22	11	Boise St.	52	100	Toledo	7
50	BYU	36	86	SDSU	27	64	BYU	34	79	SDSU	21
58	Fresno St.	41	78	Hawaii	34	66	Fresno St.	38	83	Hawaii	27
21	California	33	53	UCLA	26	15	California	34	46	UCLA	20
34	Cincinnati	40	106	Miami-OH	24	50	Cincinnati	41	104	Miami-OH	13
104	FIU	41	116	Western Kentucky	32	101	FIU	34	113	Western Kentucky	24
6	Florida	34	11	LSU	27	6	Florida	27	12	LSU	20
40	Georgia Tech	34	73	Virginia	27	47	Georgia Tech	28	32	Virginia	24
37	Georgia	34	47	Tennessee	29	21	Georgia	31	48	Tennessee	20
49	Houston	34	43	Mississippi St.	32	61	Houston	24	20	Mississippi St.	31
65	Iowa St.	27	15	Utah	34	59	Iowa St.	24	18	Utah	31
111	Kent St.	33	118	Akron	28	105	Kent St.	34	119	Akron	21
54	Kentucky	31	19	Auburn	35	49	Kentucky	21	17	Auburn	31
87	LA Tech	33	84	Utah St.	30	85	LA Tech	34	91	Utah St.	24
76	Louisville	36	107	Memphis	27	77	Louisville	35	109	Memphis	17
17	Miami-FL	35	23	Florida St.	30	13	Miami-FL	31	16	Florida St.	24
33	Michigan	35	38	Michigan St.	32	34	Michigan	34	41	Michigan St.	24
29	Missouri	35	75	Colorado	27	40	Missouri	31	58	Colorado	24
27	North Carolina	30	22	Clemson	29	22	North Carolina	28	30	Clemson	24
32	Nevada	43	119	SJSU	24	28	Nevada	55	116	SJSU	13
120	New Mexico St.	30	117	New Mexico	32	120	New Mexico St.	25	118	New Mexico	24
46	North Carolina St.	32	66	Boston College	27	52	North Carolina St.	31	62	Boston College	21
114	North Texas	33	99	Arkansas St.	35	114	North Texas	24	98	Arkansas St.	27
115	LA-Monroe	32	105	FL-Atlantic	34	112	LA-Monroe	28	NA	FL-Atlantic	27
74	Northern Ill.	33	70	Temple	31	87	Northern Ill.	21	56	Temple	31
59	Northwestern	31	67	Purdue	28	65	Northwestern	31	80	Purdue	21
35	Notre Dame	32	30	Pittsburgh	31	27	Notre Dame	28	33	Pittsburgh	21
5	Ohio St.	44	80	Indiana	24	7	Ohio St.	42	72	Indiana	17
93	Ohio	33	85	Bowling Green	31	96	Ohio	28	94	Bowling Green	24
14	Penn State	31	62	Illinois	22	14	Penn State	31	73	Illinois	13
26	South Carolina	24	1	Alabama	34	26	South Carolina	17	1	Alabama	27
51	South Florida	32	90	Syracuse	24	60	South Florida	31	82	Syracuse	17
72	SMU	36	60	Tulsa	35	NA	SMU	35	69	Tulsa	31
61	Southern Miss.	34	63	East Carolina	31	57	Southern Miss.	34	67	East Carolina	24
16	Stanford	38	20	USC	31	2	Stanford	34	25	USC	21
2	TCU	40	102	Wyoming	17	5	TCU	45	97	Wyoming	6
94	UTEP	39	103	Rice	33	92	UTEP	35	103	Rice	24
110	Tulane	29	89	Army	32	106	Tulane	24	86	Army	28
24	Air Force	38	108	Colorado St.	21	31	Air Force	35	110	Colorado St.	10
88	Vanderbilt	39	113	Eastern Michigan	26	75	Vanderbilt	42	117	Eastern Michigan	14
8	Virginia Tech	40	68	Central Michigan	24	9	Virginia Tech	35	70	Central Michigan	10
81	Wake Forest	29	45	Navy	32	45	Wake Forest	27	55	Navy	24
112	Washington St.	22	7	Oregon	45	102	Washington St.	10	3	Oregon	56
55	Washington	32	42	Arizona St.	30	35	Washington	31	39	Arizona St.	27
36	West Virginia	37	91	UNLV	25	42	West Virginia	38	89	UNLV	17
41	Wisconsin	35	64	Minnesota	29	44	Wisconsin	34	78	Minnesota	21

Key
Close game							Certain victory

Sneak Peek: Conference Projections

A new feature for next year will be projections for all 11 FBS conferences -- plus all the independent teams -- starting around week five. We'll include current and projected conference and overall standings, plus a breakdown of AdjO, AdjD, and AdjPace for each team. We'll update projected standings for four conferences each day on Monday, Tuesday, and Wednesday, and provide some commentary and insight about the remaining weeks.

ACC
Atlantic
Team	Conference		Overall		Adjusted
	Now	Projected	Now	Projected	Off.	Def.	Pace
Florida St. 24	4 - 1 0.800	6 - 3 0.667	5 - 2 0.714	8 - 5 0.615	25.4	17.1	163.4
Maryland 77	3 - 1 0.750	4 - 4 0.500	5 - 2 0.714	7 - 5 0.583	16.2	19.8	166.0
North Carolina St. 35	3 - 1 0.750	5 - 3 0.625	5 - 2 0.714	8 - 4 0.667	23.0	18.4	174.3
Clemson 20	2 - 3 0.400	4 - 4 0.500	3 - 4 0.429	6 - 6 0.500	20.9	13.8	165.5
Wake Forest 78	1 - 4 0.200	2 - 6 0.250	1 - 6 0.143	2 - 9 0.182	17.5	21.4	170.4
Boston College 67	1 - 4 0.200	3 - 5 0.375	2 - 5 0.286	5 - 7 0.417	15.5	16.5	164.1
Coastal
Team	Conference		Overall		Adjusted
	Now	Projected	Now	Projected	Off.	Def.	Pace
Virginia Tech 6	4 - 0 1.000	8 - 1 0.889	6 - 1 0.857	10 - 3 0.769	29.2	13.9	158.5
Georgia Tech 36	3 - 2 0.600	4 - 4 0.500	4 - 3 0.571	6 - 6 0.500	24.8	19.9	159.3
Miami-FL 14	3 - 2 0.600	5 - 3 0.625	4 - 3 0.571	7 - 5 0.583	22.9	13.4	170.4
North Carolina 32	2 - 2 0.500	4 - 4 0.500	4 - 3 0.571	7 - 5 0.583	21.7	15.8	159.0
Virginia 81	1 - 3 0.250	3 - 5 0.375	2 - 4 0.333	4 - 6 0.400	15.0	18.7	166.8
Duke 97	0 - 4 0.000	1 - 7 0.125	1 - 6 0.143	2 - 9 0.182	14.5	23.1	172.6
Projected records may not sum to total number of actual games due to rounding errors.

Sneak Peek: Top 25

A preview of our look for next year. We've redone the rankings to look a bit cleaner, and to include rankings for SOS, AdjO, AdjD, and AdjPace directly in the table. We're also going to use color-coding to indicate whether the data is coming from Justin's TFG system (blue) or Eddie's RBA algorithm (red). Feedback and comments are welcome.

Rank	+/-	Team	WinPct	SoS		Adjusted
						Off.		Def.		Pace
1	--	TCU	0.934	0.452	78	28.0	5	9.1	1	80.7	99
2	--	Boise St.	0.933	0.429	82	28.6	4	9.3	2	82.8	63
3	+1	Alabama	0.925	0.646	8	27.8	6	9.5	3	77.7	118
4	-1	Ohio St.	0.910	0.527	60	25.4	11	9.5	4	80.5	101
5	+1	Oregon	0.869	0.569	38	26.0	10	11.8	7	92.3	1
6	+2	Stanford	0.862	0.599	29	29.6	1	13.7	20	80.9	93
7	--	Oklahoma	0.853	0.602	23	23.4	20	11.2	5	91.8	2
8	-3	Virginia Tech	0.848	0.586	34	27.7	7	13.3	14	78.7	114
9	--	LSU	0.837	0.641	9	24.4	15	12.2	9	78.3	117
10	--	Auburn	0.833	0.662	3	29.1	2	14.9	28	83.4	55
11	+2	Florida	0.807	0.658	4	23.9	17	13.2	13	81.7	79
12	+3	Arkansas	0.800	0.658	5	27.2	8	15.2	31	84.2	45
13	-2	Iowa	0.798	0.564	40	21.1	29	11.7	6	80.8	97
14	--	Wisconsin	0.784	0.547	46	28.8	3	16.7	44	78.9	111
15	-3	Nebraska	0.780	0.507	66	22.2	22	13.0	12	80.1	105
16	+2	Florida St.	0.767	0.652	6	25.0	13	15.2	32	81.4	85
17	+4	Oklahoma St.	0.767	0.534	57	24.4	14	14.8	27	89.6	7
18	-1	Missouri	0.759	0.545	48	20.2	38	12.3	10	85.2	31
19	--	South Carolina	0.755	0.684	1	24.1	16	15.0	29	79.2	108
20	-4	West Virginia	0.750	0.537	55	18.9	49	11.9	8	81.4	86
21	+3	Notre Dame	0.733	0.599	28	19.6	42	12.8	11	84.7	36
22	-2	Utah	0.730	0.488	71	21.9	25	14.5	25	81.0	88
23	--	Georgia	0.716	0.616	17	25.0	12	16.9	47	77.2	119
24	+1	USC	0.710	0.591	33	21.9	24	15.1	30	84.0	47
25	-3	Miami-FL	0.705	0.633	13	19.3	45	13.4	17	85.8	29

New entries: none.

Dropped out: none.