Welcome to the eighth annual NFL Prediction Tracker Awards. The purpose of these awards is to help raise the publicity for those systems that have been superior in various qualities of interest. Most awards will be based entirely on the numbers I've monitored with my prediction tracker web pages. In 2007 the number of NFL systems grew to 63, up six from last year. And extra big thanks to Fred Send, who contributed the bulk of the writeup to help me get this out quicker. BEST STRAIGHT UP WINNERS (Entire Season) Winner: Sonny Moore In a very tight race, Sonny Moore’s 184 wins against 83 losses (68.9%), straight up, puts him one win ahead of the updated Vegas Line and two wins ahead of the Computer Adjusted Line. The nearest independent computer models were Lee Burdorf, with 180 wins, and four systems – Beck Elo, CPA Retro Rankings, Pigskin Herald, and Pythagorean Rankings – with 179 wins. This year there were 13 systems that predicted 177 or more (two-thirds) of the winners. Winner 2007: Sonny Moore Winner 2006: Least Absolute Value Regression Winner 2005: Vegas Line (record 72.28%) Winner 2004: Least Absolute Value Regression Winner 2003: Beck Elo Winner 2002: Herman Matthews Winner 2001: JFM Power Ratings, ARGH Power Ratings Winner 2000: CPA Rankings Winner 1999: Kenneth Massey Ratings BEST AGAINST THE SPREAD (Entire Season) Winner: Sagarin Predictive Some consider the true measure of a ranking system to be its record against the Vegas line. Jeff Sagarin’s predictive rankings had the best season against the spread, with 151 wins against 107 losses. His winning percentage of 58.5% is just behind the record of 59.0% set last year. The Computer Adjusted Line deserves honorable mention; while it applied for less than a third of the games, it won 59.5% of those against the spread. Overall, this was the best year against the spread in recent memory, as about 80% of the systems won at least half the games against the spread. Winner 2007: Sagarin Predictiv e Winner 2006: Brian Gabrielle (record 59.0%) Winner 2005: Least squares with team HFA Winner 2004: Least Absolute Value Regression Winner 2003: CPA Retro Rankings Winner 2002: FreeSportsPlay.com Winner 2001: CPA Rankings Winner 2000: Yourlinx Winner 1999: PerformanZ Ratings SMALLEST MEAN ABSOLUTE ERROR (Entire Season) Winner: Computer Adjusted Line The Computer Adjusted Line becomes the second system in a row to beat the Vegas Line in this category over an entire season. The Computer Adjusted Line had an absolute error of 10.6966, just ahead of the updated Vegas Line and the System Average. Jeff Sagarin receives honorable mention for having the lowest absolute error of any independent computer model; both his base model and his predictive model had an absolute error of 10.7854. Winner 2007: Computer Adjusted Line Winner 2006: Brian Gabrielle Winner 2005: Vegas Line (updated) (record 9.9513) Winner 2004: Vegas Line (updated) Winner 2003: Vegas Line (updated) Winner 2002: FreeSportsPlay.com Winner 2001: Vegas Line (updated) Winner 2000: Vegas Line (updated) Winner 1999: Vegas Line (updated) SMALLEST BIAS (Entire Season) Winner: CPA Rankings Steve Wrathell’s CPA Rankings and CPA Retro Rankings had the smallest biases, -0.0246 and -0.0275 respectively. This season nine systems were within 0.2, including all three of Jeff Sagarin’s systems. Winner 2007: CPA Rankings Winner 2006: Tom Benson Winner 2005: Sagarin Elo (record 0.001) Winner 2004: System Average Winner 2003: CPA Rankings Winner 2002: Sagarin Pure Points Winner 2001: ARGH Power Ratings Winner 2000: Pigskin Index Winner 1999: Flyman Ratings MOST ACCURATE SYSTEM - SMALLEST MEAN SQUARE ERROR (Entire Season) Winner: Computer Adjusted Line The Computer Adjusted Line wins this category for the second consecutive year with a mean square error of 178.312. This also means that the perennial odds-on favorite Vegas Line gets shut of the full season awards for the second straight year. Among systems predicting all 267 games, nine finished with a mean square error of 185 or less. Kambour Rating was the lowest of the independent computer models with a mean square error of 183.402, just ahead of Sagarin Predictive and Jeff Sagarin’s base system. Winner 2007: Computer Adjusted Line Winner 2006: Computer Adjusted Line Winner 2005: Vegas Line (updated) Winner 2004: Vegas Line (updated) (record 166.808) Winner 2003: Vegas Line (opening) Winner 2002: FreeSportsPlay.com Winner 2001: Vegas Line Winner 2000: Pigskin Index SECOND HALF AWARDS Some systems don’t begin publishing rankings until every team has played a couple of games, while the season-long systems often rely on preset starting values in the early part of the season. For those reasons, I like to look at the results over the second half of the season when these systems have had time to become 'burned in' to the season's data. Second half data consists of all games from week 10 through the Super Bowl. BEST STRAIGHT UP WINNERS (Second Half) Winner: Computer Adjusted Line, Vegas Line (updated) This category reverts to its history of having multiple winners. This year the Computer Adjusted Line and the updated Vegas Line both had 96 wins against 41 losses (71.0%), straight up. They finished two wins ahead of the opening Vegas Line. CPA Retro Rankings, Eric Hollobaugh, and TheSportsCruncher receive an honorable mention for having the most wins, 93, among the independent computer models. Winner 2007: Computer Adjusted Line, Vegas Line (updated) Winner 2006: Least Absolute Value Regression Winner 2005: Vegas Line (updated), Nutshell Sports (record 100-37, 72.99%) Winner 2004: Pigskin Index Winner 2003: JFM Power Ratings, Kasulis Enhanced Spread, Beck Elo Winner 2002: Least Absolute Value Regression, JFM Power Ratings, CPA Rankings Winner 2001: Pythagorean Ratings, Elo Ratings, Ken Massey Ratings Winner 2000: PerformanZ Ratings, Scoring Effeciency Winner 1999: Ken Massey, Ed Kambor, Pythagorean Ratings BEST AGAINST THE SPREAD (Second Half) Winner: Least Squares with team home field advantages Least Squares with Team Specific Home Field Advantage had the best second half against the spread, 77 wins against 59 losses (56.6%), two wins ahead of JFM Power Ratings and three wins ahead of Roger Johnson. Honorable mention goes to Cover81 for its 58.3% record against the spread for the 125 games predicted. Most systems faded from the first half to the second half. Particularly notable this season was that the winners did about as well as the anti-winners (systems when you reverse the prediction) for a change. Winner 2007: Least Squares with Team Specific HFA Winner 2006: CPA Rankings Winner 2005: Tom Benson Winner 2004: ARGH Power Ratings Winner 2003: Logistic Regression, Viacheslav Ugolnikov Winner 2002: Least Absolute Value Regression (record 78-52, 60.0%) Winner 2001: Stat Fox Winner 2000: Pythagorean Ratings Winner 1999: Least squares using team specific home field advantages SMALLEST MEAN ABSOLUTE ERROR (Second Half) Winner: Vegas Line (updated) Vegas wins this category every other year. Since someone else won last year, it must be Vegas’ turn this year, with an absolute error of 10.8358. Kambour Rating had the lowest absolute error, 11.0061, among the independent computer models predicting 137 games. Honorable mention goes to Cover81, with an absolute error of 10.664 for the 125 games predicted. Winner 2007: Vegas Line (updated) Winner 2006: Jeff Self Winner 2005: Vegas Line (updated) (record 9.9489) Winner 2004: Least Squares Regression Winner 2003: Vegas Line (updated) Winner 2002: FreeSportsPlays.com Winner 2001: Vegas Line (updated) Winner 2000: Pythagorean Ratings SMALLEST BIAS (Second Half) Winner: Pigskin Index The lowest bias was Pigskin Index with a bias of 0.0077. This beat the previous record set in 2004. The bias went strongly negative over the second half of the season. Winner 2007: Pigskin Index (record 0.0077) Winner 2006: CPA Retro Winner 2005: Kasulis Enhanced Spread Winner 2004: Tom Benson Gridmarks (previous record 0.0216) Winner 2003: CPA Rankings Winner 2002: StatFox.com Winner 2001: Monte Carlo Markov Chain Winner 2000: Flyman Performance Ratings Winner 1999: Flyman Performance Ratings MOST ACCURATE SYSTEM - SMALLEST MEAN SQUARE ERROR (Second Half) Winner: Vegas Line (updated) This has been an interesting category because you would expect the Vegas line to win consistently but it didn’t for the first five years. Times change, however, and the updated Vegas Line now has the lowest mean square error for two of the last three years. The mean square error of 177.841 is quite high, but good enough to beat the Computer Adjusted Line. Game Time Decision finished third and receives an honorable mention for having the lowest mean square error among the independent computer models. The Pythagorean Ratings continue to hold the longest-standing PT record, set in 2000. Winner 2007: Vegas Line (updated) Winner 2006: CPA Rankings Winner 2005: Vegas Line (updated) Winner 2004: Computer Adjusted Line Winner 2003: Viacheslav Ugolnikov Winner 2002: FreeSportsPlay.com Winner 2001: Hank's Power Ratings Winner 2000: Pythagorean Ratings (record 165.954) BEST PREDICTIVE SYSTEM (Entire Season) Winner: Jeff Sagarin To come up with the best overall predictive system I give each system points for how well they do in all of the above categories. I then sum up the points and the system with the highest total is dubbed predictive system of the year. So this award goes to the system that is the most well rounded One flaw can totally take a system out of the running. The computer adjusted line actually accumated the most points but many of those points were based on against the spread numbers where it only makes a pick in roughly one third of the games. So I'll leave that as an honorable mention and give the official award to Jeff Sagarin and his main rating which was just a fraction better than his predictive rating. Eric Hollobaugh was also a very strong contender but didn't really have enough games to qualify for the full season award. Winner 2007: Jeff Sagarin Winner 2006: Brian Gabrielle Winner 2005: Vegas Line Winner 2004: Least Absolute Value Regression Winner 2003: Beck Elo Winner 2002: Matthews Grid Winner 2001: CPA Rankings Winner 2000: Ed Kambour BEST PREDICTIVE SYSTEM (Second Half) Winner: Winner: Cover 81 Cover 81 was a new system that joined the group during the second half of this season. He put up some really impressive numbers. I did not allow this system to be eligible for the individual awards because he decided not to post numbers for certain games towards the end of the season. The system skipped 12 games, which is a little less than a week's worth of games. But I will give Cover 81 the award for best overall predictive system in the second half because in the remainder of the games his total was almost twice as high as the second best system. Other strong systems during the second half where the Vegas line, and another new entry, The Sports Cruncher. Winner 2007: Cover 81 Winner 2006: CPA Retro Winner 2005: Hank Trexler Winner 2004: ARGH Power Ratings Winner 2003: CPA Rankings Winner 2002: Kenneth Massey Winner 2001: Monte Carlo Markov Chain Winner 2000: Pythagorean Ratings RETRO AWARDS The retrodictive categories are based on each systems final standings and measure things by applying those final ratings to the entire season in retrospect. As I did in the college awards I am going to break this down into big systems (estimate team specific home field advantage) and simple systems (do not estimate team specific home field advantage). MOST RETRODICTIVE WINS - Complex System Winner: The Sports Report SLOTS The Sports Report SLOTS system takes this category for the sixth consecutive year with 239 wins against 28 losses. This year’s winning percentage of 89.5% is just off the record of 89.9% set in 2004. CPA Retro Rankings and TSR Predictive were both 30 games behind. Winner 2007: The Sports Report SLOTS Winner 2006: The Sports Report SLOTS Winner 2005: The Sports Report SLOTS (record 89.9%) Winner 2004: The Sports Report SLOTS Winner 2003: The Sports Report SLOTS Winner 2002: The Sports Report SLOTS Winner 2001: The system average Winner 2000: CPA Rankings MOST RETRODICTIVE WINS - Simple System Winner: Least Absolute Value regression (L1) For the best retrodiction record among systems not fitting a team home field advantage, Least Absolute Value Regression (L1) becomes to fifth different winner in five years. Least Absolute Value Regression had 199 wins against 68 losses, a winning percentage of 76.4%. Winner 2007: Least Absolute Value Regression (L1) Winner 2006: Jeff Sagarin Winner 2005: Dunkel Index (record 78.3%) Winner 2004: Brian Gabrielle Winner 2003: Frank Alder SMALLEST RETRODICTIVE ABSOLUTE ERROR - Complex System Winner: The Sports Report Predictive The Sports Report’s home field advantage predictive model becomes the first system ever to beat Least Squares with Team Specific Home Field Advantage, ending a seven-year stranglehold on this category. This has been an intriguing matchup all season, as it appeared that both systems would smash the previous record. TSR Predictive set a record low absolute error of 8.0858, breaking the previous record of 8.4189 set in 2005. Least Squares with Team Specific Home Field Advanta ge must be given its due, however, as it also broke the previous record with a phenomenal average error of 8.2226. This marked the third time in four years that Least Squares has beaten its own record. Winner 2007: The Sports Report Predictive (record 8.0858) Winner 2006: Least Squares with Team Specific HFA Winner 2005: Least Squares with Team Specific HFA (previous record 8.4189) Winner 2004: Least Squares with Team Specific HFA Winner 2003: Least Squares with Team Specific HFA Winner 2002: Least Squares with Team Specific HFA Winner 2001: Least Squares with Team Specific HFA Winner 2000: Least Squares with Team Specific HFA SMALLEST RETRODICTIVE ABSOLUTE ERROR - Simple System Winner: Least Absolute Value regression (L1) The top five systems in this category use individual team home field advantages. The best system that doesn't is Least Absolute Value Regre ssion (L1), which set a new record absolute error of 8.9498 while winning this category for the third consecutive year. Winner 2007: Least Absolute Value Regression (record 8.9498) Winner 2006: Least Absolute Value Regression Winner 2005: Least Absolute Value Regression (previous record 8.9972) Winner 2004: Sagarin Points Winner 2003: Sagarin Points SMALLEST RETRODICTIVE BIAS Winner: The Sports Report SLOTS The Sports Report Slots system set a new record with a bias of -0.0006, finally beating the 2001 record of the legendary Herman Matthews. Well over half of the systems were biased by less than 0.2, and all but four had a bias between -0.77 and 0.21. Winner 2007: The Sports Report SLOTS (record -0.0006) Winner 2006: Least Absolute Value Regression Winner 2005: The Sports Report Elo Winner 2004: CPA Rankings Winner 2003: CPA Rankings Win ner 2002: PerformanZ Ratings Winner 2001: Herman Matthews (previous record -0.003) Winner 2000: Massey Ratings MOST ACCURATE SYSTEM - SMALLEST MEAN SQUARE ERROR - Complex system Winner: The Sports Report Predictive As with absolute error, we finally have a new winner in this category. The Sports Report’s home field advantage predictive model breaks the seven-year run by Least Squares with Team Specific Home Field Advantage. TSR Predictive’s mean square error of 113.175 is still off the 2004 record of 112.865. CPA Rankings continues to inch closer; CPA’s mean square error of 120.542 was just behind Least Square’s error of 120.255. Winner 2007: The Sports Report Predictive Winner 2006: Least Squares with Team Specific HFA Winner 2005: Least Squares with Team Specific HFA Winner 2004: Least Squares with Team Specific HFA (record 112.865) Winner 2003: Least Squares with Team Specific HFA Winner 2002: Least Squares with Team Specific HFA Winner 2001: Least Squares with Team Specific HFA Winner 2000: Least Squares with Team Specific HFA MOST ACCURATE SYSTEM - SMALLEST MEAN SQUARE ERROR - Simple system Winner: The Sports Report The most accurate system without a team specific home field advantage was The Sports Report’s non-home field predictive rating with a mean square error of 136.721, good for seventh place overall. This is TSR’s third win in a row. Winner 2007: The Sports Report Winner 2006: The Sports Report Winner 2005: The Sports Report Winner 2004: Least Square Regression (record 132.830) Winner 2003: Sagarin points BEST OVERALL RETRODICTIVE SYSTEM - Complex System Winner: The Sports Report SLOTS As with the other overall a wards, points are assigned to systems in each category then summed up. The system with the most points wins. TSR Slots wins for the third year in a row, and four of the last five years. The TSR Slots prototype sas designed to maximize retrodictive wins while minimizing absolute error, bias, and mean square error. Ironically, a new system from The Sports Report took away a lot of points that the SLOTS system would have otherwise earned in previous years, making the lead less than it has been before. In the end, the overwhelming lead in total wins makes the difference for a SLOTS win. The other contenders were The Sports Report Predictive, CPA Rankings, and Least Squares with team home field advantage. Winner 2007: The Sports Report SLOTS Winner 2006: The Sports Report SLOTS Winner 2005: The Sports Report SLOTS Winner 2004: Least Squares with team home field advantages Winner 2003: The Sports Report SLOTS Winner 2002: Least Squares with team home field advantages Winner 2001: CPA Rankings Winner 2000: CPA Rankings BEST OVERALL RETRODICTIVE SYSTEM - Small System Winner: The median of all predictions This is an interesting winner, the median of all predictions. I guess I can call that a simple system. It is good enough to come in 6th overall. I'll name an honorable mention, Least absolute value regression (L1) because it was in a virtual tie with the median. Then the system average was also very close behind the top two. Winner 2007: System Median Winner 2006: Jeff Sagarin Winner 2005: Sagarin Points Winner 2004: Sagarin Points Winner 2003: Frank Alder