Best First Half Early Season Mirage Teams

[BD Sullivan]

This one's a stretch, but the 1975 Bills came to mind. Started 4-0 with wins over playoff teams Colts and the Steelers. It probably looked like they were going to take the next step after the previous year's playoff appearance. Orenthal had 697 rushing yards and 5 TDs in those 4 games. The defense was horrendous that year though.

I thought about noting them, with that MNF loss against the 1-3 Giants killing their momentum. They were heavy favorites (one Buffalo writer said they'd win 63-7) and looked to be in control until late in the first half, when Morton drove the Giants 90 yards to reduce NY's deficit to 14-7. The Giants won at the gun, with the Bills presumably guilty of looking ahead to their game at home against Miami the following Sunday.

[CSKreager]

The 1986 Jets stick out to me. I thought they looked real good at 10-1 but they looked horrible losing their last 5 games.

Pretty incredible team. As you said, not only did they lose their last 5, they lost badly. 45-3, 45-24, 52-21 to teams that didn't even make the postseason.

I do like using the 1986 Jets as a 'case study' when the issue of "building momentum for the playoffs" or "which is the hot team going into the postseason" is mentioned. The Jets were possibly the 'coldest' playoff participant in NFL history, yet they still won their first game, and should have defeated Cleveland on the road in their divisional round game.

And it's not like the '86 Jets only started hot because of a soft schedule. In that 10-1 start, they beat DEN/SEA/NE (all of whom won 10+ games, the SEA/NE wins were on the road), ATL when they started off decently, an improved Saints team.

[Rupert Patrick]

I came up with a way to mathematically measure this, it is called Momentum. How it works is each week of the regular season is weighted, and as the season goes on, the weights get higher. The first week is weighted at 1.0, Week 2 is weighted at 1.2, Week 3 is weighted at 1.4, and so on until Week 16 is weighted at 4.0, with the bye week being skipped over as though it didn't happen. A win in a particular week gets 1.0 points, a tie gets 0.5 points, and a loss gets zero points. What happens for each week of the season is that the points the team acquires in that week (based on the game result) is multiplied by the weighting for that particular week, with the product of the two numbers known as Momentum Wins or MW.

It sounds confusing, but will quickly make sense when I give an example. Let's suppose a team wins their first game, they get 1.0 MW, which one for winning the game multiplied by the weight for the week, which is also one. They also win in Week 2, and get 1.2 MW, or 1.0 for winning multiplied by the Week 2 weight of 1.2, and they also win in Week three, and get 1.4 MW, or one times 1.4. They lose in Week 4, and get zero MW, or zero times 1.6, and in Week 5, they tie in the game, and get 0.9 MW, which is 0.5 for tying multiplied by the Week 5 weight of 1.8, and so on. At the end of the season, add up their MW for each week, and divide by the sum of the weights for every week, this is their MWP, or Momentum Winning Percentage. Subtract the actual winning percentage from the MWP, and you have Momentum or MOM. If MOM is positive, that means they improved as the season progressed, and if the MOM is negative, the team declined as the season progressed.

Here is the Momentum analysis for the 1978 Washington Redskins, who won their first six games along with Weeks 9 and 11. Their MW for the weeks they win would be, respectively, 1.0, 1.2, 1.4, 1.6, 1.8, 2.0, 2.6 and 3.0 The sum of these eight numbers is 14.6, which is their MW for the season. The sum of the numbers 1.0 + 1.2 + 1.4 +...+3.6 +3.8 +4.0 is 40.0, which is how many MW they would have gotten if they had won every single game. Dividing 14.6 into 40.0, the result is .365, which is their MWP. Subtract the actual winning percentage .5, from the MWP of .365, and the result is -.135, which is the 1978 Redskins Momentum of MOM. Their MOM of -.135 is the lowest of any team from 1933 to 2017.

Here is a list of the all time highest MOM's, with rank, team, season, Actual winning percentage, Momentum winning percentage, and Momentum:

RK TM YR APCT MWP MOM
1 S F 17N 0.3750 0.5200 0.1450
2 DET 89N 0.4375 0.5700 0.1325
3 OAK 64A 0.4167 0.5397 0.1230
4 CIN 84N 0.5000 0.6200 0.1200
5 S D 92N 0.6875 0.8050 0.1175
6 DET 95N 0.6250 0.7400 0.1150
7 NYJ 74N 0.5000 0.6149 0.1149
8 K C 15N 0.6875 0.8000 0.1125
T9 BUF 62A 0.5385 0.6503 0.1119
T9 PHI 71N 0.4615 0.5734 0.1119
11 CIN 70N 0.5714 0.6832 0.1118
12 N E 65A 0.3333 0.4444 0.1111
T13 PIT 58N 0.6364 0.7455 0.1091
T13 NYG 55N 0.5455 0.6545 0.1091
15 DET 15N 0.4375 0.5450 0.1075
16 DET 60N 0.5833 0.6905 0.1071
17 PIT 76N 0.7143 0.8199 0.1056
T18 STL 78N 0.3750 0.4800 0.1050
T18 WAS 98N 0.3750 0.4800 0.1050
T18 NYJ 99N 0.5000 0.6050 0.1050
T18 WAS 81N 0.5000 0.6050 0.1050
T18 HOU 93N 0.7500 0.8550 0.1050
23 CIN 08N 0.2813 0.3850 0.1038
25 CHI 59N 0.6667 0.7698 0.1032
T25 CLE 09N 0.3125 0.4150 0.1025
T25 N E 93N 0.3125 0.4150 0.1025
T25 ATL 97N 0.4375 0.5400 0.1025

The 2017 49ers have the highest Momentum of any team, as they lost their first nine games, won Week 10 and Weeks 12 thru 16. The maximum/minimum a team can get is +/- .2500, which they would get if they win (or lose) the first half of the games during the season and do the opposite over the second half of the schedule.


Here is a list of the all time lowest MOM's.

RK TM YR WPCT MWP MOM
1 WAS 78N 0.5000 0.3650 -0.1350
2 WAS 62N 0.4167 0.2937 -0.1230
3 SEA 06N 0.5625 0.4400 -0.1225
4 N E 74N 0.5000 0.3789 -0.1211
5 DEN 09N 0.5000 0.3800 -0.1200
6 S D 67A 0.6154 0.4965 -0.1189
7 DET 39N 0.5455 0.4273 -0.1182
8 S D 01N 0.3125 0.1950 -0.1175
T9 CHI 89N 0.3750 0.2600 -0.1150
T9 OAK 95N 0.5000 0.3850 -0.1150
T9 S D 02N 0.5000 0.3850 -0.1150
12 DEN 62A 0.5000 0.3851 -0.1149
13 STL 66N 0.6154 0.5035 -0.1119
T14 S D 87N 0.5333 0.4222 -0.1111
T14 WAS 43N 0.6667 0.5556 -0.1111
T16 PHI 94N 0.4375 0.3300 -0.1075
T16 K C 13N 0.6875 0.5800 -0.1075
18 NYJ 86N 0.6250 0.5200 -0.1050
T19 DET 34N 0.7692 0.6643 -0.1049
T19 DEN 60A 0.3077 0.2028 -0.1049
21 PIT 55N 0.3333 0.2302 -0.1032
T22 DAL 86N 0.4375 0.3350 -0.1025
T22 SLR 95N 0.4375 0.3350 -0.1025
T22 MIA 93N 0.5625 0.4600 -0.1025
T22 NYG 68N 0.5000 0.3975 -0.1025
T22 L A 69N 0.7857 0.6832 -0.1025

[74_75_78_79_]

Pretty interesting/intriguing, Rupert! Another cool formula of yours! So in order for a team to score high in this they’d have to have started the season bad, meaning ’48 Browns, ’72 Dolphins, nor ’07 Pats could make this list. Same with ’76 Bucs nor ’08 Lions at the other end.

I guess ’86 Colts is a case of “too little, too late”. On the opposite end, I’m surprised ’98 Steelers aren’t on that second list. If ’93 Dolphins (also an 0-5 finish) are on it...

[rhickok1109]

Interesting, but I'm not at all sure what it's showing. The way momentum is usually spoken of, it seems to me that if a team wins, say, its first four games and then loses, they've completely their momentum...in other words, momentum should be down to 0.

But, if I understand your formula correctly, a loss wouldn't stop momentum, it just wouldn't add to it.

[Rupert Patrick]

Interesting, but I'm not at all sure what it's showing. The way momentum is usually spoken of, it seems to me that if a team wins, say, its first four games and then loses, they've completely their momentum...in other words, momentum should be down to 0.

But, if I understand your formula correctly, a loss wouldn't stop momentum, it just wouldn't add to it.

Yes, what I have come up with is actually a weighted winning percentage, with each week of the season having a different weight and the weight progressively increasing each week to where the results of the Week 16 game is four times that of Week 1:

Week Weight
1 = 1.0
2 = 1.2
3 = 1.4
4 = 1.6
5 = 1.8
6 = 2.0
7 = 2.2
8 = 2.4
9 = 2.6
10 = 2.8
11 = 3.0
12 = 3.2
13 = 3.4
14 = 3.6
15 = 3.8
16 = 4.0

If you win in a particular week, you get whatever the weight is for that particular week; if you tie, you get have that weight, and if you lose, you get zero. (The bye week is ignored.) To get Momentum Winning Percentage, it is the sum of the weights in the weeks you win divided by the sum of the weights of every week, or 1 + 1.2 + 1.4 ... + 4.0. Because the later weeks carry much more weight than the earlier weeks, whether a team does well or not in this is based on how well they do later in the season, whether a team picks up or loses momentum from early in the season.
Here is a sample team, the 1981 Detroit Lions (WK is week number, RES is result win loss or tie, WPCT is winning percentage for the season up to that week, WT is weight of the week, MW is Momentum Wins or Result (1 for win, .5 for tie, zero for loss) times WT for that week, SMW is sum of MW for the season up to that week, SWT is Sum of Weights for the season up to that week, MWP is Momentum Winning Percentage or MW divided by SMW, and MOM is Momentum, or MWP minus WPCT for the season up to that week:



If you graph the data, WPCT vs MWP (y-axis) vs. Week number (x-axis), you would get the following graph:



One way to visualize momentum is that it is the area between the two graphs. The Lions won their first game, then, lost their next two. After three games, their MWP (the thin dashed line) was below the WPCT (the thicker line), this indicates negative momentum as the MWP dropped below the WPCT. After losses in weeks six and seven, the Lions found themselves in negative momentum once again. Two wins in a row were followed by two losses in a row, and then, they won three in a row in weeks 11 thru 13, which built up a lot of positive momentum, so much that even though they lost two of their final three games, their momentum was on the positive side at the end of the season.

I played with different ways to try to capture Momentum over the years, but in the end, it came down to weighing each week of the season and turning it into a weighted winning percentage that differs slightly from actual winning percentage.

If you've studied higher mathematics or basic physics of motion, I liken Winning Percentage to be akin to the first derivative in Calculus (rate of winning or velocity) and Momentum Winning Percentage to be akin to the second derivative in Calculus (rate of rate of winning or acceleration/deceleration). Momentum is actually acceleration (positive momentum) or deceleration (negative momentum)

As for the Momentum of teams like the 2007 Patriots or 1972 Dolphins, or for that matter, the 1976 Bucs or 2017 Browns, this approach finds their momentum every week to be zero as there is no difference between their WPCT and MWP (weighted winning percenage) at any point in the season. The two graphs for WPCT and MWP would sit directly on top of each other. These teams are not picking up or losing steam as the season increases, they win (or lose) every single week and for that reason, they are consistent every week.