Computer Rating System Prediction Results for College Football (NCAA IA)
2008 Season Totals
Through
| Rank |
System |
Pct. Correct |
Against Spread |
Absolute Error |
Bias |
Mean Square Error |
games |
suw |
sul |
atsw |
atsl |
| 1 | Line (updated) | 0.73816 | | 12.5912 | 0.1024 | 255.867 | 718 | 530 | 188 | | |
| 2 | Computer Adjusted Line | 0.73816 | 0.47893 | 12.6072 | 0.0738 | 256.741 | 718 | 530 | 188 | 125 | 136 |
| 3 | Line (opening) | 0.72905 | 0.47458 | 12.7884 | 0.0216 | 264.500 | 716 | 522 | 194 | 280 | 310 |
| 4 | Atomic Football | 0.75209 | 0.49215 | 12.8479 | 0.0911 | 266.434 | 718 | 540 | 178 | 345 | 356 |
| 5 | Harmon Forcast | 0.73780 | 0.49178 | 12.9010 | -0.1646 | 270.614 | 717 | 529 | 188 | 329 | 340 |
| 6 | System Median | 0.75487 | 0.48058 | 12.8794 | -0.2401 | 270.866 | 718 | 542 | 176 | 334 | 361 |
| 7 | System Average | 0.75487 | 0.48080 | 12.9140 | -0.2240 | 270.949 | 718 | 542 | 176 | 338 | 365 |
| 8 | Dunkel Index | 0.73603 | 0.50427 | 12.9018 | 0.6649 | 273.168 | 716 | 527 | 189 | 354 | 348 |
| 9 | Stat Fox | 0.74791 | 0.53881 | 12.8705 | 0.9515 | 273.741 | 718 | 537 | 181 | 361 | 309 |
| 10 | Bihl System | 0.73708 | 0.50114 | 13.0810 | -0.3526 | 274.910 | 445 | 328 | 117 | 219 | 218 |
| 11 | Pigskin Index | 0.73625 | 0.45692 | 13.0734 | -0.2268 | 275.667 | 709 | 522 | 187 | 297 | 353 |
| 12 | Hank Trexler | 0.72981 | 0.47939 | 13.0362 | -0.3343 | 275.688 | 718 | 524 | 194 | 314 | 341 |
| 13 | Covers.com | 0.73538 | 0.50857 | 13.0702 | -0.5363 | 278.854 | 718 | 528 | 190 | 356 | 344 |
| 14 | Ashby AccuRatings | 0.73816 | 0.46875 | 13.1393 | 0.4153 | 279.280 | 718 | 530 | 188 | 315 | 357 |
| 15 | Catherwood Ratings | 0.70455 | 0.51701 | 13.0714 | -0.5325 | 279.343 | 308 | 217 | 91 | 152 | 142 |
| 16 | Born Power Index | 0.72981 | 0.50710 | 13.2065 | -0.1040 | 280.310 | 718 | 524 | 194 | 357 | 347 |
| 17 | Sagarin Predictive | 0.73677 | 0.49716 | 13.2170 | -0.4181 | 280.841 | 718 | 529 | 189 | 350 | 354 |
| 18 | Dokter Entropy | 0.74652 | 0.47585 | 13.1575 | 0.5936 | 281.082 | 718 | 536 | 182 | 335 | 369 |
| 19 | Moore Power Ratings | 0.74373 | 0.49148 | 13.1593 | -0.4516 | 281.757 | 718 | 534 | 184 | 346 | 358 |
| 20 | Lee Burdorf | 0.73677 | 0.49785 | 13.2455 | -0.3589 | 281.820 | 718 | 529 | 189 | 348 | 351 |
| 21 | Laz Index | 0.74234 | 0.50285 | 13.1206 | -0.6036 | 282.273 | 718 | 533 | 185 | 353 | 349 |
| 22 | NutShell Sports | 0.74373 | 0.50852 | 13.1522 | 0.0385 | 283.134 | 718 | 534 | 184 | 358 | 346 |
| 23 | Dave Congrove | 0.72423 | 0.50356 | 13.2581 | -0.8393 | 285.592 | 718 | 520 | 198 | 354 | 349 |
| 24 | Edward Kambour | 0.74591 | 0.49469 | 13.3416 | 0.4994 | 287.046 | 673 | 502 | 171 | 326 | 333 |
| 25 | ARGH Power Ratings | 0.73259 | 0.47858 | 13.2695 | 0.1560 | 288.412 | 718 | 526 | 192 | 324 | 353 |
| 26 | Linear Regression | 0.74561 | 0.50148 | 13.3796 | -0.8116 | 288.757 | 342 | 255 | 87 | 169 | 168 |
| 27 | Keeper | 0.73120 | 0.48295 | 13.4905 | 0.4086 | 288.850 | 718 | 525 | 193 | 340 | 364 |
| 28 | CPA Rankings | 0.74930 | 0.52489 | 13.3203 | -0.2382 | 289.026 | 718 | 538 | 180 | 369 | 334 |
| 29 | Bassett Model | 0.69948 | 0.47803 | 13.3794 | -0.1183 | 290.850 | 579 | 405 | 174 | 272 | 297 |
| 30 | Sagarin | 0.73677 | 0.48506 | 13.3644 | -0.4172 | 291.403 | 718 | 529 | 189 | 341 | 362 |
| 31 | Howell | 0.72563 | 0.49102 | 13.3831 | -0.0236 | 291.982 | 718 | 521 | 197 | 328 | 340 |
| 32 | Beck Elo | 0.72284 | 0.47863 | 13.4931 | -0.4882 | 294.166 | 718 | 519 | 199 | 336 | 366 |
| 33 | DP Dwiggins | 0.71429 | 0.50903 | 13.5109 | -3.0640 | 296.301 | 567 | 405 | 162 | 282 | 272 |
| 34 | Marsee | 0.72065 | 0.49606 | 13.5691 | 1.3284 | 298.360 | 673 | 485 | 188 | 315 | 320 |
| 35 | NutShell Retro | 0.71727 | 0.47937 | 13.5750 | 0.0755 | 299.352 | 718 | 515 | 203 | 337 | 366 |
| 36 | Massey Consensus | 0.74095 | 0.49148 | 13.6242 | 0.1080 | 300.364 | 718 | 532 | 186 | 346 | 358 |
| 37 | Nutshell Girl | 0.70760 | 0.50742 | 13.5932 | -1.5051 | 302.535 | 342 | 242 | 100 | 171 | 166 |
| 38 | Imes Comprank | 0.72331 | 0.48308 | 13.7986 | -0.3628 | 304.556 | 665 | 481 | 184 | 314 | 336 |
| 39 | Stephen Kerns | 0.72385 | 0.53912 | 13.6591 | -2.9885 | 305.183 | 717 | 519 | 198 | 379 | 324 |
| 40 | PerformanZ Ratings | 0.73955 | 0.47511 | 13.9043 | -0.4733 | 308.905 | 718 | 531 | 187 | 334 | 369 |
| 41 | Massey BCS * | 0.71637 | 0.49258 | 13.5732 | -0.8457 | 313.232 | 342 | 245 | 97 | 166 | 171 |
| 42 | Warren Claassen | 0.72563 | 0.49075 | 14.0176 | -1.0525 | 316.050 | 718 | 521 | 197 | 345 | 358 |
| 43 | Anderson/Hester * | 0.70710 | 0.47748 | 13.8421 | -1.3609 | 320.219 | 338 | 239 | 99 | 159 | 174 |
| 44 | Colley Rankings * | 0.70175 | 0.47917 | 13.9185 | -0.9162 | 320.895 | 342 | 240 | 102 | 161 | 175 |
| 45 | Sagarin Elo | 0.70613 | 0.48011 | 14.1104 | -0.3101 | 326.758 | 718 | 507 | 211 | 338 | 366 |
| 46 | CPA Retro | 0.71588 | 0.48153 | 14.1334 | -0.1177 | 326.773 | 718 | 514 | 204 | 339 | 365 |
| 47 | Martien Maas | 0.70221 | 0.49487 | 14.0884 | -0.1875 | 328.068 | 497 | 349 | 148 | 241 | 246 |
| 48 | Wolfe * | 0.70175 | 0.51039 | 14.0233 | -0.3427 | 329.651 | 342 | 240 | 102 | 172 | 165 |
| 49 | Billingsly | 0.68663 | 0.45739 | 14.4297 | -0.5499 | 335.311 | 718 | 493 | 225 | 322 | 382 |
| 50 | Tom Benson | 0.69923 | 0.48148 | 14.6618 | -1.0881 | 347.855 | 522 | 365 | 157 | 247 | 266 |
| 51 | Sportrends | 0.69350 | 0.49564 | 14.8414 | -1.3575 | 367.346 | 708 | 491 | 217 | 341 | 347 |
| 52 | Super List | 0.73398 | 0.48222 | 15.3791 | 0.4898 | 369.664 | 718 | 527 | 191 | 339 | 364 |
| 53 | Least Squares w/ HFA | 0.65789 | 0.49258 | 16.1102 | -1.1122 | 412.168 | 342 | 225 | 117 | 166 | 171 |
| 54 | Logistic Regression | 0.67251 | 0.49852 | 19.9079 | -4.7427 | 967.244 | 342 | 230 | 112 | 168 | 169 |
* This system does not make predictions. I make predictions for this
system by translating it to a new scale that allows for making predictions.
Retrodictive records are found by taking the ratings from the current week
and applying them to the entire season to date.
The ideal system would be one that has the highest correct game decisions,
has the smallest mean error(deviation from the actual game result), and has
a bias of zero.
Mean Error = average[abs(prediction-actual)]
Bias = agerage(prediction - actual)
Std. = Standard Deviation of individual game biases