HOME
RATINGS & ARCHIVES
PREDICTIONS
TEAM PAGES
BLOG
ALGORITHM
CREDITS
FAQ & LINKS
Atomic Football


RANKINGS

Special thanks to Ralph Wallace at College Sporting News and the folks at I-AA.org for making our football rating part of every Gridiron Power Index since 2003.

AWARDS

The "best overall predictive system ... that is the most well rounded".
- Todd Beck, 2015

"The lowest [deviation from actual game scores] for any computer system was Ashburns's Atomic Football at 12.85."
- Todd Beck, 2008

BOOKS

Special thanks to Amy Langville and Carl Meyer for their reference to our paper in their book, "Who's #1? The Science of Rating and Ranking."
- 2012, ISBN: 9781400841677

PRESS

Here are some newspaper articles that have been written about AtomicFootball. See what others are saying about our rating system.

  • 22 October 2006 PDF
  • 8 July 2007 PDF
  • 9 July 2007 PDF

CONTACT

Please check out our site and let us know what you think. Comments and suggestions are welcome.

contact us


Frequently Asked Questions

#1 You rank team A above team B, even though A lost to B. It's simple head-to-head. Your rankings are broken.

PLEASE, no more 'simple head-to-head' emails. Here's the short answer. Up to about a month into the season, it is possible to rank all the winners ahead of all the losers. After that, it can't be done. Let's go all the way back to the 2007 season -- no undefeated teams in the FBS. So, who are you going to rank #1 -- ahead of a team or two that beat them? Maybe 'head-to-head' isn't so 'simple.'


#2 I've noticed on Prediction Tracker that your predictions have been the most accurate among individuals since the middle of the 2007 season. Would you say they are the best? Part 2: Is it possible to make money betting on football?

In answer to your first question, we suggest you look at the numbers and decide for yourself. In answer to your second question, my best answer is "probably not." To clarify, even if you could reasonably expect a positive return, it is highly unlikely that you could manage risk to the point of equalling the risk-reward record of the stock market. I will also say that I've never met a millionaire who made his fortune gambling.


#3 What is 'Lower-Higher?'

It's the numbers of a team's scheduled opponents that are ranked lower and higher, respectively, than that team. As a general rule, the win-loss record and the 'lower-higher' of a team should be pretty close. If they're not, then something is probably wrong with the rankings. We could try to make them match, but our ranking approach is a little more complicated (and better) than that.


#4 Your rankings and your predictions don't always seem consistent with each other. Why not?

There are a couple of reasons. First, home field advantage. Second, we rank based only on the current year's performance and give all games equal weight, but we predict using the current year plus the previous two years, with games having less weight as you go back in time. Plus, this helps us get lots of emails telling us how clueless we are.


#5 If I look on one of the team pages at your games predictions and compare them to the projected win/loss record, they don't match. What gives?

OK, this one is a little tricky, so I'll answer it with an example. Let's say I've predicted your team will win each of it's final four games (individually), and the probability of win is 75% in each game. In this case, I project their record assuming that COLLECTIVELY they will win only three of those games (0.75 * 4 = 3). They might not lose any, or one, or two, or three, or all four, but averaging across all the possible outcomes, expect one loss.


#6 Your early season rankings sometimes look squirrely. Especially for the smaller school divisions. What gives?

Since the smaller schools usually have smaller travel budgets, interconference games tend to be more limited and confined to relatively small regions. Teams such as those in the NAIA Frontier conference, which play very few non-conference games, are especially hard to rate, and the rankings of such teams tends to be rather sensitive to the outcome of those few non-conference games. We could better determine how good these teams are by using previous years' results, but then we would be rewarding or penalizing them this year for previous years' successes or failures.


#7 On several occasions, your score predictions predict scores of 2 and sometimes 1? I kinda understand the two, but what's with the one?

The score predictions represent the 'average expected result,' as if you could play the game hundreds of times and average the outcomes. So, if I expect a heavy underdog to notch a field goal two-thirds of the time and be shutout the rest, the expected result is 2. If there's only a one-in-three chance of the field goal, then their 'expected' score is 1. And before you complain that this is stupid, remember that the 'gambling professionals' often predict margins like '-3½'.


#8 This is a pretty weird hobby.

Assuming that is a question, the answer is 'yes.' It may be weird, but it's much cheaper than golf. Plus, if it ever makes money, we'll be able to deduct the cost of going to games as a business expense.


#9 Where did the name 'Atomic Football' come from?

We wanted to come up with something a little corny, actually. The word 'atomic' seems to have disappeared with the classic sci-fi movies of the '50s, and we wanted to bring it back.


#10 Doesn't the name 'Atomic Football' have another meaning?

Yes, there is a device used by the President of the United States to authorize the use of nuclear weapons. It is commonly called the Atomic Football or Nuclear Football.

'And so, because of the automated and irrevocable decision-making process which rules out human meddling, the Doomsday machine is terrifying and simple to understand... and completely credible and convincing.' --Dr. Strangelove


#11 What does the phrase 'Solutio Optimus' on your logo mean?

We're certainly not Latin scholars, but we think it means 'Optimal Solution' or 'the best answer.' It can also be translated as 'best loosening,' which I suppose means that our numbers can be an effective laxative. No doubt our critics would agree.


#12 Do you guys rank all college football teams?

Yes (almost), and finding some of the scores, even by Sunday afternoon, is not always easy. We rank all college football teams in both the NCAA and the NAIA with the exception of the members of the New England Small College Athletic Conference (NESCAC) in the NCAA Division III. More recently, we have had to similarly exclude the Great American Conference (D2), the Mid-America Intercollegiate Conference (D2), the Northern Sun Conference (D2), and the Heart of America Conference (NAIA).


#13 Do you have something against the NESCAC?

Since the New England Small College Athletic Conference (NESCAC) members decline to participate in the NAIA playoffs, we're cool with their isolationism. We're not so keen on the other conferences that have waived off interconference games as computer rankings become, for lack of a better word, impractical and the playoff selection committees are left to grade these teams with a level of subjectivity that exceeds that of Olympic synchronized swimming (I'm not knocking synchronized swimming, by the way).


#14 What factors do you consider in your predictions?

It might be easier to point out the things we DON'T consider. We do NOT consider playing surface, injuries, weather, or time of day. Many of these require years of archived data as well as volumes of detailed data on the teams (run/pass balance, etc., etc.). If someone would pay us to do this, of course, then we would gladly add them.


#15 Your year-end rankings don't jive with the FCS, D2, D3, and NAIA playoff results?

Our rankings give equal weight to all games, including playoff games. We are not trying to discredit the results of the playoffs. If a team can squeak into the playoffs and then go on a tear, I'll be the first to congratulate them on their championship. I realize this raises all kinds of "head-to-head" questions -- see FAQ #1 above and my blog for more on the topic.


Links

The following links and images are for reference only. No affiliation, support, sponsorship, or endorsement is to be inferred or implied.


BCS website

Football Bowl Association website

NCAA Football website

NAIA website

College Sporting News website

D2Football.com website

The Sports Network website

D3Football.com website
 

Valid XHTML 1.0 Transitional   Valid CSS!