Previous post:

Next post:

Checkdowns: Worst. Monday Night Football Matchup. Ever. (By combined winning percentage)

by Neil Paine on October 21, 2013

in Checkdowns, Data Dump, History, Totally Useless, Trivia

(I originally posted this at the S-R Blog, but I thought it would be very appropriate here as well.)

Just a quick hit of a post to let you know that tonight’s MNF matchup between the 0-6 Giants and the 1-4 Vikings is, in fact, the worst ever this late in the season by combined winning percentage:

game_id
home
W
L
T
PF
PA
road
W
L
T
PF
PA
year_id
week_num
game_date
comb_wpct
comb_pt_diff
winner
201310210nygnyg060103209min1401251582013710/21/2013.091-12.6NULL
197512150sdgsdg1110148282nyj39022137819751312/15/1975.167-12.1sdg, 24-16
199411210otioti190147218nyg37017122019941211/21/1994.200-6.0nyg, 13-10
197211060nwenwe25092220clt160941451972811/6/1972.214-12.8clt, 24-17
197011020pitpit2407194cin150931561970711/2/1970.250-7.2pit, 21-10
198110190detdet240118126chi150891331981710/19/1981.250-4.3det, 48-17
199710200cltclt06088155buf3301221591997810/20/1997.250-8.7buf, 9-6
201211260phiphi370162252car28018424320121211/26/2012.250-7.5car, 30-22
200001030atlatl4110251351sfo41102664191999171/3/2000.267-8.4atl, 34-29
200512190ravrav490171253gnb310025525520051512/19/2005.269-3.2rav, 48-3
198310240crdcrd250137218nyg2501261561983810/24/1983.286-7.9tie (20-20)
201112120seasea570216246ram210014029620111412/12/2011.292-7.8sea, 30-13
201011290crdcrd370188292sfo37016021920101211/29/2010.300-8.2sfo, 27-6
200911160clecle17078209rav44020615420091011/16/2009.313-4.9rav, 16-0
198011240nornor0110181341ram74029422819801211/24/1980.318-4.3ram, 27-7
201112050jaxjax380138200sdg47024927520111312/5/2011.318-4.0sdg, 38-14
198312120tamtam2120212345gnb77039640719831512/12/1983.321-5.1gnb, 12-9
198711020daldal330135134nyg150991421987711/2/1987.333-3.5dal, 33-24
199910250pitpit33011793atl150741531999710/25/1999.333-4.6pit, 13-9
200211180ramram450194196chi27018223220021111/18/2002.333-2.9ram, 21-16
200511210gnbgnb270201184min45015422820051111/21/2005.333-3.2min, 20-17
200012040nwenwe390192253kan57028327420001412/4/2000.333-2.2nwe, 30-24
200412130otioti480231294kan48034132620041412/13/2004.333-2.0kan, 49-38
200712100atlatl390171272nor57026627920071412/10/2007.333-4.8nor, 34-14
199112090miamia760256275cin211021137419911512/9/1991.346-7.0mia, 37-13
197311260sfosfo370180232gnb35213819819731111/26/1973.350-5.6sfo, 20-6
197811130cincin190110184rai64019316419781111/13/1978.350-2.3rai, 34-21
200711260pitpit730269145mia010018327420071211/26/2007.350+1.7pit, 3-0
199311290cltclt370154233sdg46016419519931311/29/1993.350-5.5sdg, 31-0
198010270nyjnyj160114164mia430991441980810/27/1980.357-6.8nyj, 17-14
199710270miamia520143124chi0701011991997910/27/1997.357-5.6chi, 36-33
199811020phiphi16079162dal4301741151998911/2/1998.357-1.7dal, 34-0
201211050nornor250190216phi3401201552012911/5/2012.357-4.4nor, 28-13
198311070detdet450202188nyg26116621419831011/7/1983.361-1.9det, 15-9
199111250ramram380181256sfo56021815519911311/25/1991.364-0.5sfo, 33-10
199211300seasea110073218den74017520719921311/30/1992.364-8.0sea, 16-13
200812010htxhtx470252293jax47022424020081312/1/2008.364-2.6htx, 30-17
197111150sdgsdg350150179crd3501351491971911/15/1971.375-2.7sdg, 20-17
197910290atlatl350160181sea3501721811979910/29/1979.375-1.9sea, 31-28
200611130carcar440137163tam26010217320061011/13/2006.375-6.1car, 24-10
200711120seasea440167141sfo26010418620071011/12/2007.375-3.5sea, 24-0
201211120pitpit530191164kan17013324020121011/12/2012.375-5.0pit, 16-13
199010290pitpit340109128ram2401641731990810/29/1990.385-2.2pit, 41-10
201212170otioti490271386nyj67024530620121512/17/2012.385-6.8oti, 14-10
197012070otioti371177249cle56023623619701212/7/1970.386-3.3cle, 21-10
197711210waswas540126132gnb2708315219771011/21/1977.389-4.2was, 10-9
200311170sfosfo450202152pit36017621720031111/17/2003.389+0.5sfo, 30-14
201011220sdgsdg450239197den36020325220101111/22/2010.389-0.4sdg, 35-14
199410170denden140108146kan32090801994710/17/1994.400-2.8kan, 31-28
199911290sfosfo370163281gnb55019220919991211/29/1999.400-6.8gnb, 20-3

It is not, however, the worst by combined PPG margin. That honor belongs to this 1972 game between the 2-5 Patriots and the 1-6 Colts (Baltimore ended up winning 24-17):

game_id
home
W
L
T
PF
PA
road
W
L
T
PF
PA
year_id
week_num
game_date
comb_wpct
comb_pt_diff
winner
197211060nwenwe25092220clt160941451972811/6/1972.214-12.8clt, 24-17
201310210nygnyg060103209min1401251582013710/21/2013.091-12.6NULL
197512150sdgsdg1110148282nyj39022137819751312/15/1975.167-12.1sdg, 24-16
199710200cltclt06088155buf3301221591997810/20/1997.250-8.7buf, 9-6
200001030atlatl4110251351sfo41102664191999171/3/2000.267-8.4atl, 34-29
201011290crdcrd370188292sfo37016021920101211/29/2010.300-8.2sfo, 27-6
199211300seasea110073218den74017520719921311/30/1992.364-8.0sea, 16-13
198310240crdcrd250137218nyg2501261561983810/24/1983.286-7.9tie (20-20)
201112120seasea570216246ram210014029620111412/12/2011.292-7.8sea, 30-13
201211260phiphi370162252car28018424320121211/26/2012.250-7.5car, 30-22
197011020pitpit2407194cin150931561970711/2/1970.250-7.2pit, 21-10
199112090miamia760256275cin211021137419911512/9/1991.346-7.0mia, 37-13
198010270nyjnyj160114164mia430991441980810/27/1980.357-6.8nyj, 17-14
201212170otioti490271386nyj67024530620121512/17/2012.385-6.8oti, 14-10
199911290sfosfo370163281gnb55019220919991211/29/1999.400-6.8gnb, 20-3
197910150nyjnyj240128174min3301071421979710/15/1979.417-6.8nyj, 14-7
200611130carcar440137163tam26010217320061011/13/2006.375-6.1car, 24-10
199411210otioti190147218nyg37017122019941211/21/1994.200-6.0nyg, 13-10
200611060seasea430149177rai250921482006911/6/2006.429-6.0sea, 16-0
199710270miamia520143124chi0701011991997910/27/1997.357-5.6chi, 36-33
197311260sfosfo370180232gnb35213819819731111/26/1973.350-5.6sfo, 20-6
199311290cltclt370154233sdg46016419519931311/29/1993.350-5.5sdg, 31-0
198312120tamtam2120212345gnb77039640719831512/12/1983.321-5.1gnb, 12-9
201211120pitpit530191164kan17013324020121011/12/2012.375-5.0pit, 16-13
200911160clecle17078209rav44020615420091011/16/2009.313-4.9rav, 16-0
200712100atlatl390171272nor57026627920071412/10/2007.333-4.8nor, 34-14
199910250pitpit33011793atl150741531999710/25/1999.333-4.6pit, 13-9
201211050nornor250190216phi3401201552012911/5/2012.357-4.4nor, 28-13
198110190detdet240118126chi150891331981710/19/1981.250-4.3det, 48-17
198011240nornor0110181341ram74029422819801211/24/1980.318-4.3ram, 27-7
197711210waswas540126132gnb2708315219771011/21/1977.389-4.2was, 10-9
199510230nwenwe15069160buf510136951995810/23/1995.500-4.2nwe, 27-14
200611270seasea640203219gnb46018525220061211/27/2006.500-4.2sea, 34-24
201112050jaxjax380138200sdg47024927520111312/5/2011.318-4.0sdg, 38-14
200412060seasea650239223dal47019328920041312/6/2004.455-3.6dal, 43-39
198711020daldal330135134nyg150991421987711/2/1987.333-3.5dal, 33-24
200711120seasea440167141sfo26010418620071011/12/2007.375-3.5sea, 24-0
199010220clecle24098139cin4201541531990710/22/1990.500-3.3cin, 34-13
201110310kankan330105150sdg4201411362011810/31/2011.583-3.3kan, 23-20
197811060cltclt360120230was72018613519781011/6/1978.556-3.3clt, 21-17
200711190denden450153238oti63017815220071111/19/2007.556-3.3den, 34-20
197012070otioti371177249cle56023623619701212/7/1970.386-3.3cle, 21-10
200511210gnbgnb270201184min45015422820051111/21/2005.333-3.2min, 20-17
200512190ravrav490171253gnb310025525520051512/19/2005.269-3.2rav, 48-3
200512120atlatl750277237nor39018329520051412/12/2005.417-3.0atl, 36-17
200211180ramram450194196chi27018223220021111/18/2002.333-2.9ram, 21-16
199410170denden140108146kan32090801994710/17/1994.400-2.8kan, 31-28
197211130sdgsdg251152203cle5301411341972911/13/1972.469-2.8cle, 21-17
197710310crdcrd330124122nyg330911261977710/31/1977.500-2.8crd, 28-0
200512260nyjnyj3110189298nwe95032228920051612/26/2005.429-2.7nwe, 31-21

{ 9 comments… read them below or add one }

Chase Stuart October 21, 2013 at 2:19 pm

What I like about this is that Neil named the file bad_mnf.

[HEY LOOK AT ME, I HAVE BEHIND THE SCENES ACCESS.]

Reply

Neil Paine October 21, 2013 at 3:07 pm

It occurs to me that we can use this to measure what kind of a season MNF’s having in terms of watchability… For each matchup ever, I took the at-the-time record and did our little method of adding 11 games of .500 ball (this will capture the perception at the time of how good/bad a matchup is), then took the Harmonic Mean of the two numbers to give more weight to times that both teams were good or bad in a matchup. The averages of those numbers for each season were:

+---------+-------+-----------+
| year_id | games | harm_mean |
+---------+-------+-----------+
|    2013 |     8 | .494      |
|    2012 |    16 | .512      |
|    2011 |    17 | .503      |
|    2010 |    18 | .522      |
|    2009 |    17 | .520      |
|    2008 |    17 | .510      |
|    2007 |    17 | .497      |
|    2006 |    18 | .526      |
|    2005 |    17 | .506      |
|    2004 |    16 | .491      |
|    2003 |    16 | .504      |
|    2002 |    17 | .521      |
|    2001 |    15 | .505      |
|    2000 |    17 | .528      |
|    1999 |    17 | .496      |
|    1998 |    17 | .534      |
|    1997 |    18 | .534      |
|    1996 |    17 | .533      |
|    1995 |    17 | .530      |
|    1994 |    17 | .508      |
|    1993 |    18 | .523      |
|    1992 |    17 | .511      |
|    1991 |    17 | .528      |
|    1990 |    16 | .514      |
|    1989 |    16 | .552      |
|    1988 |    16 | .528      |
|    1987 |    16 | .514      |
|    1986 |    16 | .544      |
|    1985 |    16 | .552      |
|    1984 |    17 | .524      |
|    1983 |    16 | .498      |
|    1982 |     9 | .524      |
|    1981 |    16 | .515      |
|    1980 |    16 | .505      |
|    1979 |    16 | .531      |
|    1978 |    16 | .504      |
|    1977 |    13 | .528      |
|    1976 |    13 | .552      |
|    1975 |    13 | .517      |
|    1974 |    13 | .515      |
|    1973 |    13 | .533      |
|    1972 |    13 | .500      |
|    1971 |    13 | .531      |
|    1970 |    13 | .512      |
+---------+-------+-----------+

So MNF’s trendline is definitely going down over the years, and this is one of their worst years ever thus far. Here are the numbers thus far:

+------------+-------------+----------+-------------+----------+-----------+-------------+
|    date    |    home     | reg_wpct |    away     | reg_wpct | harm_mean |   winner    |
+------------+-------------+----------+-------------+----------+-----------+-------------+
| 2013-09-09 | sdg (0-0-0) | .500     | htx (0-0-0) | .500     | .500      | htx (31-28) |
| 2013-09-09 | was (0-0-0) | .500     | phi (0-0-0) | .500     | .500      | phi (33-27) |
| 2013-09-16 | cin (0-1-0) | .458     | pit (0-1-0) | .458     | .458      | cin (20-10) |
| 2013-09-23 | den (2-0-0) | .577     | rai (1-1-0) | .500     | .536      | den (37-21) |
| 2013-09-30 | nor (3-0-0) | .607     | mia (3-0-0) | .607     | .607      | nor (38-17) |
| 2013-10-07 | atl (1-3-0) | .433     | nyj (2-2-0) | .500     | .464      | nyj (30-28) |
| 2013-10-14 | sdg (2-3-0) | .469     | clt (4-1-0) | .594     | .524      | sdg (19-9)  |
| 2013-10-21 | nyg (0-6-0) | .324     | min (1-4-0) | .406     | .360      | NULL        |
+------------+-------------+----------+-------------+----------+-----------+-------------+

Compare that to 2004, the worst year ever:

+------------+--------------+----------+--------------+----------+-----------+-------------+
|    date    |     home     | reg_wpct |     away     | reg_wpct | harm_mean |   winner    |
+------------+--------------+----------+--------------+----------+-----------+-------------+
| 2004-09-13 | car (0-0-0)  | .500     | gnb (0-0-0)  | .500     | .500      | gnb (24-14) |
| 2004-09-20 | phi (1-0-0)  | .542     | min (1-0-0)  | .542     | .542      | phi (27-16) |
| 2004-09-27 | was (1-1-0)  | .500     | dal (1-1-0)  | .500     | .500      | dal (21-18) |
| 2004-10-04 | rav (2-1-0)  | .536     | kan (0-3-0)  | .393     | .453      | kan (27-24) |
| 2004-10-11 | gnb (1-3-0)  | .433     | oti (1-3-0)  | .433     | .433      | oti (48-27) |
| 2004-10-18 | ram (3-2-0)  | .531     | tam (1-4-0)  | .406     | .460      | ram (28-21) |
| 2004-10-25 | cin (1-4-0)  | .406     | den (5-1-0)  | .618     | .490      | cin (23-10) |
| 2004-11-01 | nyj (5-1-0)  | .618     | mia (1-6-0)  | .361     | .456      | nyj (41-14) |
| 2004-11-08 | clt (4-3-0)  | .528     | min (5-2-0)  | .583     | .554      | clt (31-28) |
| 2004-11-15 | dal (3-5-0)  | .447     | phi (7-1-0)  | .658     | .533      | phi (49-21) |
| 2004-11-22 | kan (3-6-0)  | .425     | nwe (8-1-0)  | .675     | .522      | nwe (27-19) |
| 2004-11-29 | gnb (6-4-0)  | .548     | ram (5-5-0)  | .500     | .523      | gnb (45-17) |
| 2004-12-06 | sea (6-5-0)  | .523     | dal (4-7-0)  | .432     | .473      | dal (43-39) |
| 2004-12-13 | oti (4-8-0)  | .413     | kan (4-8-0)  | .413     | .413      | kan (49-38) |
| 2004-12-20 | mia (2-11-0) | .313     | nwe (12-1-0) | .729     | .438      | mia (29-28) |
| 2004-12-27 | ram (6-8-0)  | .460     | phi (13-1-0) | .740     | .567      | ram (20-7)  |
+------------+--------------+----------+--------------+----------+-----------+-------------+

Reply

Richie October 21, 2013 at 3:19 pm

So MNF’s trendline is definitely going down over the years,

That makes sense, because when ESPN picked up the package in 2006, the system for picking the games changed. Sunday Night Football now gets the marquee matchups that used to be saved for MNF.

Reply

Neil Paine October 21, 2013 at 4:30 pm

Good point. I should do this for SNF as well, and see the quality of matchups rise meteorically after it moved from ESPN to NBC.

Reply

Neil Paine October 21, 2013 at 4:29 pm

One more way to look at this — for each MNF matchup ever, I took the average of the 2 teams’ final SRS (or current SRS for 2013 teams) ratings, then averaged those for each MNF season:

+---------+---------+
| year_id | average |
+---------+---------+
|    2013 | -1.1    |
|    2012 | +1.5    |
|    2011 | +0.3    |
|    2010 | +2.0    |
|    2009 | +3.2    |
|    2008 | +1.9    |
|    2007 | +0.4    |
|    2006 | +0.7    |
|    2005 | +1.6    |
|    2004 | +0.8    |
|    2003 | +1.5    |
|    2002 | +2.5    |
|    2001 | +1.5    |
|    2000 | +2.8    |
|    1999 | +0.6    |
|    1998 | +2.7    |
|    1997 | +1.9    |
|    1996 | +2.9    |
|    1995 | +3.4    |
|    1994 | +1.5    |
|    1993 | +1.7    |
|    1992 | +1.8    |
|    1991 | +3.2    |
|    1990 | +1.5    |
|    1989 | +4.5    |
|    1988 | +1.3    |
|    1987 | +1.9    |
|    1986 | +2.7    |
|    1985 | +3.1    |
|    1984 | +3.2    |
|    1983 | +1.9    |
|    1982 | +4.3    |
|    1981 | +0.9    |
|    1980 | +1.6    |
|    1979 | +2.1    |
|    1978 | +1.3    |
|    1977 | +1.8    |
|    1976 | +3.6    |
|    1975 | +1.7    |
|    1974 | +1.7    |
|    1973 | +3.5    |
|    1972 | +0.8    |
|    1971 | +2.5    |
|    1970 | +2.1    |
+---------+---------+

Granting that the range of SRS can be wider earlier in the season, and that we have less certainty about teams’ true ratings now than we will after 16 games, 2013 is the worst MNF season ever by this method (and it’s not even close).

Reply

Chase Stuart October 21, 2013 at 5:09 pm

Very cool stuff, Neil. I agree with Richie that SNF is now where the good games go. I also think TNF is taking some good matchups, although I don’t know how that arrangement works. Who gets first dibs?

Reply

George October 23, 2013 at 3:43 pm

Just as an out of interest – did anyone else see Monday night going differently? I’m aware that I didn’t see the Josh Freeman effect coming (was Cassell injured or did they just view Freeman as the better option??), and I probably didn’t pay enough attention to the Bayes Theorem post (and yes therefore I was underrating the Giants possibly), but I had (in pick terms only – I didn’t like it that much) the Vikings winning outright, covering and this one going over. Everything really pointed at the over on this one (even the weather was compliant), and I’m just struggling to see how the numbers were so far off (I know that they aren’t the be all and end all but they’ve just pointed in different ways at the moment).

Totally agree with the point above by the way, SNF seems to get the choice games at the moment (don’t know what the logic for this one is though?).

Reply

Neil Paine October 23, 2013 at 4:00 pm

Well, the Giants were favored by 3.5 in Vegas. Here’s what SRS would have said going into the games:

+-------------------+--------+-------+--------+
|       team        | pfr_id | games |  SRS   |
+-------------------+--------+-------+--------+
| Minnesota Vikings | min    |     5 |  -8.05 |
| New York Giants   | nyg    |     6 | -12.32 |
+-------------------+--------+-------+--------+
| hfa                                |   2.96 |
+------------------------------------+--------+

That would anticipate MIN being favored by 1.3, but we know the stdev of game-to-game margin about the expected MOV from point ratings is like 13.5, so a 16-point loss when favored by 1.3 is about a 1.28-standard deviation below the mean event, or it had roughly a 10% probability of happening by variance alone. Unlikely, but not phenomenally so… Chalk it up to the randomness of the game.

Reply

George October 23, 2013 at 4:21 pm

Hi Neil, thanks for the percentage (I would have probably got there by Friday). I had them by 0.83 with my ratings (and I know the numbers are not 100% yet), I just saw more likely to win yet getting 3.5 – that was at least a cover. The over was what really got me – extrapolating scores out I had 61 points being scored and a 97% chance of an over and it wasn’t even close.

There just feels like there is a lot of randomness and noise out there even after half a season almost and I just find it a bit disconcerting now I have numbers to give me a better feel for things (I know it wouldn’t be as much fun though if we all could predict what was going to happen before it did though, but I just really couldn’t handle the Jets losing to the Steelers but then beating the Patriots which just didn’t make sense)

Reply

Leave a Comment

You can use these HTML tags and attributes: <strong> <em> <pre> <code> <a href="" title="">