You Are Here > Pro-Football-Reference.com > PFR Blog > NFL and College Football Analysis

The Best 2-Game Stretches of the Decade (2000-08)

Posted by Neil Paine on Tuesday, November 3, 2009

5340416 SBXXXV RavensA few weeks ago, I threw out this crazy idea about how to isolate a team's peak performance by SRS:

"Obviously, we can't run the SRS on single-game samples, because it requires multiple opponents to 'work'. But what if we broke each team-season down into 15 to 19 'mini-teams' based on 2-game stretches of the season? Like, the Patriots' win over the Jets in Week 1 of the 2007 season would be part of the 'New England Games 1-2' team, and so would their win against the Chargers the next week. And that Chargers team would be part of the 'San Diego Games 2-3' team, who played 'Green Bay Games 3-4' the next week, who played 'Minnesota Games 4-5' the week after that... and so on and so forth. Now, every 'team' connects to every other team, just like in the regular SRS, but we've also isolated team performance down to the most specific time period possible using the SRS method."

As a follow-up, I calculated the best and worst 2-game stretches by teams this decade (2009 isn't included because not all teams have played the same # of games yet). Remember, the SRS is focused on measuring a team's point differential vs. the point differential you'd expect an average team to have based on the game's location and the strength of the opponent; this method takes it even further and is only concerned with the strength of the opponent at the time of the game, meaning wins against teams with mediocre records can still be positive for a team's SOS if they play them either before or after a strong performance. I think there's definitely some logic to this, because (as Chase pointed out in the post that inspired this series) every game features a different version of the same team; sometimes the differences are so small as to be imperceptible, but sometimes they're huge (think the '07 Giants early in the season vs. late), so it really does matter when you catch a particular opponent.

Anyway, here are the 100 best 2-game stretches of the Oughts:

Year_Tm_G#_G# MOV SOS SRS
2000_rav_19_20 21.52 29.13 50.64
2000_rav_18_19 16.53 27.67 44.20
2002_tam_18_19 23.52 16.20 39.71
2000_rav_17_18 16.00 17.66 33.66
2003_buf_1_2 25.34 7.71 33.06
2005_sea_12_13 40.00 -7.69 32.31
2000_rai_17_18 -2.70 34.21 31.51
2003_gnb_15_16 31.00 0.28 31.28
2001_ram_3_4 33.50 -2.39 31.11
2002_tam_17_18 21.00 9.13 30.13
2006_buf_13_14 19.50 10.15 29.65
2001_ram_2_3 18.00 10.77 28.77
2000_tam_1_2 27.99 0.63 28.62
2005_pit_19_20 15.52 13.05 28.57
2008_nyg_9_10 12.50 15.98 28.48
2007_nwe_9_10 28.03 0.38 28.41
2006_nwe_16_17 19.00 9.08 28.08
2001_ram_17_18 13.47 14.57 28.03
2000_rai_16_17 31.97 -4.00 27.96
2004_clt_10_11 34.53 -6.73 27.80
2007_nwe_7_8 33.00 -5.31 27.69
2004_buf_14_15 28.03 -0.38 27.65
2003_gnb_14_15 28.53 -1.05 27.49
2001_nwe_18_19 6.52 20.84 27.36
2007_nwe_1_2 22.99 4.26 27.25
2004_nwe_17_18 15.50 11.70 27.20
2007_nwe_2_3 24.47 2.46 26.92
2000_nyg_18_19 5.48 21.26 26.75
2000_tam_2_3 31.00 -4.40 26.60
2007_nwe_6_7 24.03 2.53 26.57
2007_nwe_8_9 24.50 1.99 26.49
2004_gnb_6_7 24.50 1.91 26.41
2001_ram_18_19 -0.52 26.76 26.24
2001_phi_18_19 4.37 21.79 26.16
2004_nwe_18_19 10.02 16.09 26.11
2007_gnb_1_2 16.68 9.37 26.05
2006_chi_4_5 28.97 -2.93 26.03
2006_jax_12_13 20.50 5.42 25.92
2002_phi_2_3 30.50 -4.71 25.79
2001_phi_17_18 18.00 7.61 25.61
2005_pit_14_15 31.03 -5.47 25.56
2006_nyg_5_6 16.53 8.54 25.07
2004_pit_7_8 15.97 9.09 25.06
2007_nwe_5_6 19.00 5.82 24.82
2000_nyg_17_18 22.47 2.29 24.75
2005_sdg_3_4 23.00 1.67 24.67
2007_nyg_3_4 10.00 14.53 24.53
2002_rai_17_18 15.47 8.97 24.44
2007_clt_6_7 26.03 -1.71 24.33
2000_oti_15_16 27.50 -3.25 24.25
2003_gnb_16_17 17.63 6.57 24.20
2004_pit_8_9 19.00 5.08 24.08
2002_rai_16_17 18.97 4.96 23.93
2000_oti_16_17 -2.03 25.87 23.83
2004_clt_9_10 33.00 -9.17 23.83
2008_crd_18_19 13.50 10.30 23.80
2008_rav_11_12 30.00 -6.26 23.74
2004_clt_11_12 29.50 -5.80 23.70
2007_nwe_3_4 26.00 -2.36 23.64
2003_oti_6_7 20.50 3.14 23.64
2005_nyg_7_8 27.00 -3.40 23.60
2003_rav_12_13 24.97 -1.37 23.60
2005_nyg_6_7 15.47 7.76 23.23
2003_oti_7_8 19.53 3.54 23.07
2002_kan_12_13 40.97 -17.95 23.02
2006_rav_15_16 21.01 2.00 23.01
2003_kan_2_3 24.50 -1.56 22.94
2005_nwe_13_14 28.00 -5.09 22.91
2007_clt_5_6 20.50 2.29 22.79
2008_nyg_10_11 14.00 8.59 22.59
2007_min_11_12 28.00 -5.53 22.47
2006_clt_17_18 12.00 10.36 22.36
2005_sea_11_12 22.50 -0.18 22.32
2007_det_7_8 23.00 -0.71 22.29
2006_clt_18_19 6.50 15.76 22.26
2006_nwe_15_16 10.70 11.55 22.25
2001_ram_1_2 6.70 15.49 22.19
2004_kan_6_7 24.97 -2.98 21.99
2007_nwe_4_5 19.00 2.94 21.94
2004_phi_2_3 14.00 7.92 21.92
2004_nwe_11_12 24.00 -2.08 21.92
2003_sea_1_2 32.01 -10.11 21.90
2000_oti_14_15 28.32 -6.43 21.89
2006_buf_12_13 7.50 14.35 21.85
2006_chi_1_2 25.66 -3.86 21.79
2004_buf_13_14 23.00 -1.35 21.65
2007_gnb_2_3 14.50 7.10 21.60
2005_den_7_8 13.50 8.09 21.59
2003_rav_15_16 25.34 -3.80 21.55
2007_dal_8_9 19.03 2.50 21.54
2008_nyg_8_9 13.00 8.47 21.47
2004_phi_18_19 5.48 15.99 21.47
2002_atl_10_11 24.00 -2.59 21.41
2007_sdg_5_6 26.00 -4.68 21.32
2001_ram_16_17 19.97 1.29 21.26
2003_kan_8_9 23.97 -2.72 21.25
2006_nwe_6_7 26.03 -4.81 21.22
2005_pit_15_16 33.01 -11.85 21.16
2007_clt_7_8 10.00 11.10 21.10
2006_sdg_12_13 15.50 5.56 21.06

And the bottom 100 performances of the decade:

Year_Tm_G#_G# MOV SOS SRS
2001_nor_15_16 -37.03 3.31 -33.73
2003_nyg_13_14 -25.50 -6.43 -31.93
2008_ram_2_3 -26.00 -4.76 -30.76
2008_ram_10_11 -21.50 -9.19 -30.69
2008_ram_11_12 -17.03 -13.28 -30.31
2002_crd_9_10 -22.50 -6.48 -28.98
2008_ram_3_4 -20.50 -8.36 -28.86
2006_car_13_14 -27.03 -1.35 -28.39
2000_crd_6_7 -30.00 2.05 -27.95
2008_ram_12_13 -14.00 -13.79 -27.79
2003_crd_1_2 -32.34 4.92 -27.43
2000_cin_1_2 -15.00 -12.04 -27.04
2007_atl_13_14 -27.00 0.13 -26.87
2001_atl_12_13 -16.00 -10.67 -26.67
2008_ram_9_10 -28.47 2.21 -26.26
2003_crd_13_14 -19.50 -6.47 -25.97
2001_was_2_3 -34.50 8.57 -25.93
2004_chi_15_16 -11.50 -14.32 -25.82
2002_crd_8_9 -20.03 -5.64 -25.67
2000_cle_13_14 -39.47 13.98 -25.49
2007_ram_15_16 -22.01 -3.32 -25.33
2008_ram_1_2 -31.34 6.13 -25.22
2005_nor_13_14 -18.00 -7.20 -25.20
2002_crd_11_12 -35.00 9.89 -25.11
2000_cin_2_3 -21.97 -3.14 -25.11
2004_rai_6_7 -19.53 -5.52 -25.06
2001_nor_14_15 -28.50 3.52 -24.98
2004_rai_5_6 -24.50 -0.31 -24.81
2000_crd_12_13 -17.50 -7.18 -24.68
2007_mia_12_13 -24.00 -0.62 -24.62
2006_sfo_6_7 -30.00 5.39 -24.61
2008_ram_8_9 -32.50 7.94 -24.56
2008_det_2_3 -20.50 -3.99 -24.49
2003_crd_12_13 -27.47 3.07 -24.40
2005_det_14_15 -13.02 -11.21 -24.23
2008_det_15_16 -27.68 3.45 -24.23
2003_nyg_12_13 -18.03 -6.18 -24.21
2003_nyg_14_15 -23.97 -0.03 -24.00
2008_det_11_12 -30.53 6.56 -23.98
2003_chi_15_16 -12.50 -11.44 -23.94
2001_car_15_16 -30.53 6.67 -23.87
2003_atl_6_7 -32.00 8.17 -23.83
2001_atl_11_12 -26.53 2.78 -23.75
2007_kan_13_14 -21.50 -2.15 -23.65
2005_det_13_14 -13.50 -10.11 -23.61
2007_ram_6_7 -19.97 -3.61 -23.58
2005_sfo_5_6 -30.00 6.59 -23.41
2005_gnb_13_14 -21.00 -2.29 -23.29
2008_ram_13_14 -13.50 -9.65 -23.15
2003_crd_9_10 -22.47 -0.68 -23.14
2003_was_14_15 -15.00 -7.98 -22.98
2005_nor_5_6 -24.48 1.63 -22.85
2008_det_9_10 -16.50 -6.23 -22.73
2000_crd_7_8 -26.00 3.32 -22.68
2008_cin_7_8 -28.50 6.10 -22.40
2008_det_1_2 -20.68 -1.48 -22.16
2007_atl_12_13 -16.00 -6.10 -22.10
2005_nor_14_15 -10.52 -11.57 -22.09
2006_atl_14_15 -11.53 -10.42 -21.95
2007_kan_12_13 -24.00 2.05 -21.95
2008_det_3_4 -22.50 0.56 -21.94
2004_sfo_3_4 -22.00 0.14 -21.86
2006_tam_8_9 -15.50 -6.30 -21.80
2006_was_9_10 -10.47 -11.26 -21.72
2007_sfo_3_4 -20.50 -1.02 -21.52
2005_htx_10_11 -20.03 -1.45 -21.48
2002_crd_12_13 -23.00 1.57 -21.43
2002_crd_10_11 -22.50 1.10 -21.40
2007_rav_8_9 -22.50 1.15 -21.35
2003_det_5_6 -19.00 -2.34 -21.34
2003_was_13_14 -7.00 -14.25 -21.25
2003_det_1_2 -9.66 -11.53 -21.18
2000_cle_6_7 -17.97 -3.10 -21.07
2000_crd_11_12 -24.50 3.50 -21.00
2003_det_13_14 -17.50 -3.46 -20.96
2000_buf_13_14 -25.50 4.57 -20.93
2000_sdg_13_14 -24.50 3.57 -20.93
2003_was_6_7 -19.50 -1.41 -20.91
2008_den_7_8 -21.50 0.63 -20.87
2002_cin_2_3 -16.97 -3.88 -20.84
2005_nor_4_5 -16.98 -3.81 -20.80
2005_htx_4_5 -23.00 2.24 -20.76
2002_nyj_2_3 -32.00 11.26 -20.74
2006_gnb_10_11 -22.50 1.88 -20.62
2002_cin_13_14 -17.50 -3.10 -20.60
2004_sfo_10_11 -19.50 -1.07 -20.57
2000_crd_13_14 -19.47 -1.10 -20.56
2008_crd_14_15 -30.50 9.98 -20.52
2008_mia_11_12 -8.00 -12.50 -20.50
2008_det_8_9 -14.00 -6.47 -20.47
2005_gnb_14_15 -26.00 5.65 -20.35
2001_was_1_2 -30.63 10.39 -20.24
2003_cin_15_16 -12.99 -7.17 -20.16
2002_cin_3_4 -27.50 7.44 -20.06
2004_sfo_2_3 -15.47 -4.56 -20.03
2003_det_14_15 -13.97 -6.05 -20.01
2005_buf_4_5 -4.52 -15.47 -19.99
2006_tam_9_10 -5.50 -14.49 -19.99
2006_nyg_14_15 -21.53 1.55 -19.98
2005_buf_3_4 -11.52 -8.40 -19.92

This entry was posted on Tuesday, November 3rd, 2009 at 2:44 PM and filed under Simple Rating System, Statgeekery. Follow comments here with the RSS 2.0 feed. Skip to the end and leave a response. Trackbacks are closed.

8 Responses to “The Best 2-Game Stretches of the Decade (2000-08)”

  1. Roby said:

    As a Miami fan, I looked to see if the Dolphins showed up on the list. Interestingly, they made the list twice. Once in an 1-15 season (no surprise), and again in an 11-5 season. the 2 games that made it in the 11-5 season were a win against Oakland, and a loss against New England (who also finished that season 11-5). I found it somewhat surprising that they went 1-1 in those games.

  2. Roby said:

    I wonder how many wins show up on the bad list. Do you know without having to manually check all of them?

  3. MattieShoes said:

    As a retrodictive tool for fun, it's a very cool idea, but in terms of actually measuring team quality... Well, the Ravens were VERY good, but 50.64? I think 2 game stretches allows too much "creative accounting". I'm thinking a 4 or 5 week stretch would balance that out more. I'm curious as to what time period would give the best predictive values. I'll have to think more about it.

  4. MattieShoes said:

    Just to see what'd happen, I did the same thing with 4 week timeframes instead of 2 week timeframes. I like it a bit better because the timespan is still short but there's 4x as much interconnectivity.

    The top and bottom ends of the list:
    NWE 2007 07_10 (30.58)
    RAV 2000 17_20 (30.19)
    NWE 2007 06_09 (28.10)
    NWE 2007 05_08 (27.78)
    NWE 2007 08_11 (26.74)
    NWE 2006 14_17 (26.03)
    TAM 2002 16_19 (25.40)
    RAM 2001 01_04 (24.85)
    NWE 2007 03_06 (24.05)
    CLT 2004 09_12 (23.58)

    ...

    NOR 2005 05_08 (-22.54)
    DET 2008 01_04 (-22.73)
    NYG 2003 13_16 (-23.86)
    NYG 2003 12_15 (-24.02)
    RAM 2008 11_14 (-24.61)
    RAM 2008 01_04 (-25.33)
    CRD 2002 09_12 (-25.58)
    RAM 2008 10_13 (-25.78)
    RAM 2008 09_12 (-25.96)
    RAM 2008 08_11 (-27.47)

  5. Neil Paine said:

    It's tough to balance the issue of needing more data for significance vs. wanting to zero in on how well a team played at as specific a moment in time as possible. I went with 2 games to lean as far toward the latter as possible, but it's essentially a personal choice, especially since it can only be retrodictive at best.

    However, I'm not sure the 2000 Ravens looking so good is that weird -- in the Super Bowl, they trounced a Giants team 34-7 that had rolled up a 41-0 win vs. a Minnesota team that thoroughly beat New Orleans 34-16 a week after the Saints beat the Rams 31-28, on the heels of St. Louis beating the Saints in New Orleans by 5... etc. You can trace back their final 11 games of the 2000 season and it was as dominant a stretch as you'll find: they allowed >10 pts only 3 times, and other than a 1-point win at Tennessee and a 6-point win at Arizona, they won all of the games in that stretch by double-figures, including 8 by 2 TDs or more. The fact that the final 4 games of the run came in the playoffs, ending with the Super Bowl, only makes it more impressive.

  6. MattieShoes said:

    Well here's the thing. Do you think they'd beat the 2007 Pats at their peak by more than 3 touchdowns? Because that's what the numbers suggest, yes? But it'd have been fun to watch...
    .
    But with 2 games per mini-team in the playoffs, you've got very few connections and the connections are all to other teams at their extreme ends (and are thus poorly connected too). So the Ravens have an absurdly high score solely because the Giants won by 41 in their previous game.

    You'll get the same answer any time the superbowl is a blowout and/or the loser won by a significant margin in the previous game. 1992, Dallas has +8 or so for the season on average but +49 in their final one because they blew out the Bills who won their penultimate game handily.

    Using the 2000 ravens as an example with their last rating of the year:

    G | Tms | ConPT | RAV
    2 | 487 | 01.9 | +51.01
    3 | 456 | 04.2 | +37.77
    4 | 425 | 07.1 | +30.28
    5 | 394 | 10.7 | +26.39
    6 | 363 | 14.7 | +17.23

  7. Karl said:

    The 2000 Ravens, while impressive, are a bit overrated, particularly their defensive performance.

    They played in a division with a terrible expansion Cleveland Browns team, a Bengals team trying to develop Akili Smith, a Steelers team with the dreadful Kordell Stewart, and a generous slate of non-divisional opponents with weak offenses.

    That they won a convincing Super Bowl is again due in large part to a great draw, as they played possibly the worst team to ever appear in a Super Bowl.

    A good team, but not a great one, in my view. (For reference, I should add that I think the '85 Bears are overrated, too, although the hyperbole around that team is so extraordinary that they could hardly be anything else).

    Anyway, interesting post!

  8. MattieShoes said:

    I was thinking one solution to the connectivity issue would be to tack a year on either end of the one in question then simply number the games sequentially. So for the 2000 Ravens, instead of games 1-20, you have games 1-54, of which the "real" season is 17-36. Now there's a "crossover" team containing week 17 of 1999 and week 1 of 2000, and another containing the superbowl and week 1 of 2001. While the playoff mini-teams are still poorly connected to the league in general, it's less so than before, and the extreme ends where whipsaws occur is far away. It also has the small bonus of giving you a number at week 1 too. I also did the same with 4 weeks just for kicks. The regular season numbers were pretty similar in all 4 iterations, but the postseason numbers changed dramatically.
    .
    http://farm3.static.flickr.com/2801/4078394203_7541d77c14_o.jpg

Leave a Reply

*Required
*Required (Not published)