2004 Predictions
NFC
North
Vikings: 12-4 predicted, 8-8 actual, -4 difference, finish 2nd
Packers; 9-7 predicted, 10-6 actual, +1 difference, finish 1st
Bears: 7-9 predicted, 5-11 actual, -2 difference, finish 4th
Lions: 6-10 predicted, 6-10 actual, HIT, finish 3rd
South
Panthers: 9-7 predicted, 7-9 actual, -2 difference, finish 3rd
Saints: 9-7 predicted, 8-8 actual, -1 difference, finish 2nd
Falcons: 7-9 predicted, 11-5 actual, +4 difference, finish 1st
Bucs: 6-10 predicted, 5-11 actual, -1 difference, finish 4th HIT
East (yes)
Eagles: 12-4 predicted, 13-3 actual, +1 difference, finish 1st HIT
Cowboys: 8-8 predicted, 6-10 actual, -2 difference, finish tie 2nd
Redskins: 7-9 predicted, 6-10 actual, -1 difference, finish tie 2nd
Giants: 5-11 predicted, 6-10 actual, +1 difference, finish tie 2nd
West (yes)
Seahawks: 12-4 predicted, 9-7 actual, -3 difference, finish 1st HIT
Rams: 6-10 predicted, 8-8 actual, +2 difference, finish 2nd HIT
Arizona: 4-12 predicted, 6-10 actual, +2 difference, finish 3rd HIT
49ers: 4-12 predicted, 2-14 actual, -2 difference, finish 4th HIT
AFC
North
Ravens: 10-6 predicted, 9-7 actual, -1 difference, finish 2nd
Bengals: 10-6 predicted, 8-8 actual, -2 difference, finish 3rd
Steelers: 7-9 predicted, 15-1 actual, +8 difference, finish 1st
Browns: 4-12 predicted, 4-12 actual, HIT, finish 4th HIT
South
Titans: 10-6 predicted, 5-11 actual, -5 difference, finish 4th
Colts: 10-6 predicted, 12-4 actual, +2 difference, finish 1st
Jaguars: 9-7 predicted, 9-7 actual, HIT, finish 2nd
Houston: 8-8 predicted, 7-9 actual, -1 difference, finish 3rd
East
Jets: 11-5 predicted, 10-6 actual, -1 difference, finish 2nd
Patriots: 10-6 predicted, 14-2 actual, +4 difference, finish 1st
Bills: 6-10 predicted, 9-7 actual, +3 difference, finish 3rd
Dolphins: 3-13 predicted, 4-12 actual, +1 difference, finish 4th HIT
West
Chiefs: 10-6 predicted, 7-9 actual, -3 difference, finish 3rd
Raiders: 8-8 predicted, 5-11 actual, -3 difference, finish 4th
Broncos: 7-9 predicted, 10-6 actual, +3 difference, finish 2nd
Chargers: 2-14 predicted, 12-4 actual, +10 difference, finish 1st
Findings:
- My average prediction was 2.97 games off, which is to say that my margin of error is roughly 19 percent across a 16-game schedule.
- Thirteen of my 32 record predictions (40 percent) were within one game of the actual results. Only three were hits.
- I tended to over-estimate performances. Of the 29 team records that were incorrectly predicted, I projecting extra wins for 16 of them. Despite this optimism, when I missed I missed big, actually registering an under-over ratio of +8 games.
- I was terrible at picking division winners, picking only two of eight (25 percent) correctly. I was much better at picking 4th-place teams, getting that right in five of the eight divisions.
- I was not good at predicting whether teams would finish with winning or losing records: If you count a hit as a match between a prediction of a winning or a losing season and the actual outcome, I got just 18 (56 percent) correct, just slightly ahead of random chance.
- I seriously underestimate the performance of eventual division winners, who are about 3.4 games better than I thought they'd be.
- My predicted division winners are predictably over estimated in terms of wins, requiring an average adjustment of -2.25 games.
Now let's compare that to how I did in 2005...
NFC
North
*(8-8) Vikings: 12-4 predicted, 9-7 actual, -3 difference, finish 2nd
*(5-11) Bears: 8-8 predicted, 11-5 actual, +3 difference, finish 1st
(6-10) Lions: 7-9 predicted, 5-11 actual, -2 difference, finish 3rd HIT
*(10-6) Packers: 6-10 predicted, 4-12 actual, -2 difference, finish 4th HIT
East
*(13-3) Eagles: 11-5 predicted, 6-10 actual, -5 difference, finish 4th
*(6-10) Cowboys: 9-7 predicted, 9-7 actual, HIT, finish 3rd
*(6-10) Giants: 7-9 predicted, 11-5 actual, +4 difference, finish 1st
(6-10) Redskins: 6-10 predicted, 10-6 actual, +4 difference, finish 2nd
South (yes/tie)
*(7-9) Panthers: 12-4 predicted, 11-5 actual, -1 difference, finish tie 1st HIT
*(11-5) Falcons: 10-6 predicted, 8-8 actual, -2 difference, finish 3rd
*(5-11) Bucs: 6-10 predicted, 11-5 actual, +5 difference, finish tie 1st
*(8-8) Saints: 3-13 predicted, 3-13 actual, HIT, finish 4th HIT
West
(8-8) Rams: 10-6 predicted, 6-10 actual, -4 difference, finish 2nd
(9-7) Seahawks: 9-7 predicted, 13-3 actual, +4 difference, finish 1st
(6-10) Cardinals: 7-9 predicted, 5-11 actual, -2 difference, finish 3rd
*(2-14) 49ers: 4-12 predicted, 4-12 actual, HIT, finish 4th HIT
AFC
North
*(8-8) Bengals: 10-6 predicted, 11-5 actual, -1 difference, finish tie 1st HIT
(9-7) Ravens: 10-6 predicted, 6-10 actual, -4 difference, finish tie 3rd
(15-1) Steelers: 9-7 predicted, 11-5 actual, +2 difference, finish tie 1st
(4-12) Browns: 2-12 predicted, 6-10 actual, +4 difference, finish tie 3rd HIT
East
*(14-2) Patriots: 12-4 predicted, 10-6 actual, -2 difference, finish 1st HIT
(10-6) Jets: 10-6 predicted, 4-12 actual, -6 difference, finish 4th
*(9-7) Bills: 8-8 predicted, 5-11 actual, -3 difference, finish 3rd HIT
(4-12) Dolphins: 4-12 predicted, 9-7 actual, +5 difference, finish 2nd
South (yes)
(12-4) Colts: 10-6 predicted, 14-2 actual, +4 difference, finish 1st HIT
*(9-7) Jaguars: 10-6 predicted, 12-4 actual, +2 difference, finish second HIT
*(7-9) Texans: 6-10 predicted, 2-14 actual, -4 difference, finish 4th
(5-11) Titans: 6-10 predicted, 4-12 actual, -2 difference, finish 3rd
West
*(12-4) Chargers: 11-5 predicted, 9-7 actual, -2 difference, finish 3rd
(10-6) Broncos: 10-6 predicted, 13-3 actual, +3 difference, finish 1st
*(7-9) Chiefs: 9-7 predicted, 10-6 actual, +1 difference, finish 2nd
(5-11) Raiders: 6-10 predicted, 4-12 actual, -2 difference, finish 4th HIT
Findings
- My average predictive error improved slightly, to 2.75 games per team, a margin of error of 17percent.
- I got much worse at predicting records. Only six of the records I predicted (18.7 percent) were within one game of actual results.Once again, only three were hits.
- This time I was slightly more likely to under-estimate teams' performance. Seventeen of my 29 incorrect predictions were under-estimates, and my over-under ratio for 2005 was -6.
- I continued picking division winners at a 25 percent rate, again no better than random chance. Strangely enough, I also continued picking cellar-dwellers at a 62.5 percent clip.
- I did much better at predicting teams as winners or losers in 2005, hitting that one correctly 21 times, or 65.6 percent of the time.
- I underestimated the performance of eventual division winners by 2.6 games per team.
- I overestimated the performance for my predicted division winners by 1.75 games per team and underestimated the performance of only one of them.
- I correctly predicted an upward or downward trend from 2004 for 18 of the 32 teams.
Thoughts about my predictive patterns:
My biggest problem is that I don't know how to project who will win divisions. While I'm just throwing darts when it comes to picking winners.
I'm slightly better than random when it comes to spotting winning and losing teams,
Comments