2008 NFL Graphs/Rankings – Inter/Intra

Here’s a quick runthrough of the entire 2008 season if we were to exclude all inter/intra matches. I’m also posting picks for each week, and the season performance compared to the current system’s season performance.

Week 1 had no games to exclude, so it’s identical to the Week 1 already posted – picks record 8-8.

Week 2:
Excluded Games: NO@WAS(24@29), BUF@JAC(20@16)
Picks: 11-4 (19-12) (even)
Rankings: CAR NO NE DAL DEN ARI NYG TEN GB PIT CHI BUF TB BAL PHI NYJ OAK SF JAC HOU IND WAS ATL SD CLE CIN SEA KC MIA STL MIN DET
Graph:
2008-2-Nfl-Ic-Clean

Week 3:
Excluded Games: DAL@GB(27@16), DET@SF(13:31)
Picks: 9-7 (28-19) (-1)
Rankings: DEN NYG DAL NO WAS TEN PHI CAR ARI BUF GB TB BAL NE PIT JAC OAK SF ATL SD MIN IND CHI NYJ CIN MIA HOU CLE SEA DET KC STL
Graph:
2008-3-Nfl-Ic-Clean

(more below the fold)

Week 4:
Excluded Games: SF@NO(17:31), PHI@CHI(20:24)
Picks: 5-8 (33-27) (-2)
Rankings: NYG DEN WAS CAR NO DAL SD TEN PHI TB BUF NYJ NE JAC PIT GB ARI ATL IND BAL CHI OAK SF MIN CLE DET KC MIA HOU CIN SEA STL
Graph:
2008-4-Nfl-Ic-Clean

Week 5:
Excluded Games: (none)
Picks: 9-5 (42-32) (-2)
Rankings: NYG WAS DAL DEN TEN NYJ PHI NE CAR TB ARI PIT GB ATL NO SD BUF CHI MIN SF JAC BAL MIA DET KC IND CLE SEA OAK HOU CIN STL
Graph:
2008-5-Nfl-Ic-Clean

Week 6:
Excluded Games: GB@SEA(27@17), JAC@DEN(24@17), CIN@NYJ(14@26), MIA@HOU(28@29)
Picks: 6-8 (48-40) (-2)
Rankings: WAS NYJ DEN ARI TB DAL TEN PHI CAR NYG SD PIT GB ATL NE NO BUF MIN CHI JAC MIA IND KC BAL DET SF CLE OAK HOU SEA CIN STL
Graph:
2008-6-Nfl-Ic-Clean

Week 7:
Excluded Games: SEA@TB(10@20), TEN@KC(34@10), BAL@MIA(27@13)
Picks: 7-7 (55-47) (-5)
Rankings: DEN PHI CAR TB TEN PIT ATL WAS ARI JAC GB CHI NYG NE BUF NYJ DAL SD MIA IND SF KC SEA MIN STL HOU BAL NO DET CLE OAK CIN
Graph:
2008-7-Nfl-Ic-Clean

Week 8:
Excluded Games: ARI@CAR(23@27), TB@DAL(9@13), ATL@PHI(14@27), OAK@BAL(10@29), WAS@DET(25@17)
Picks: 8-6 (63-53) (-4) (First week of better results)
Rankings: DEN WAS CAR TB PHI ATL ARI MIA TEN PIT GB CHI DAL NYG BUF IND NE NYJ JAC BAL CLE SEA MIN SF KC STL NO HOU SD DET OAK CIN
Graph:
2008-8-Nfl-Ic-Clean

Week 9:
Excluded Games: NE@IND(15@18)
Picks: 9-5 (72-58) (-3)
Rankings: NYJ CAR TB ARI ATL TEN MIA PIT CHI NYG PHI GB NE WAS DEN IND BAL BUF DAL JAC CLE MIN SEA SF KC STL HOU CIN NO DET SD OAK
Graph:
2008-9-Nfl-Ic-Clean

Week 10:
Excluded Games: DEN@CLE(34@30)
Picks: 11-3 (83-61) (-1)
Rankings: CAR NYJ TB ATL TEN ARI CHI GB NYG MIA WAS IND PIT NE PHI DEN BUF DAL MIN BAL KC SEA SF CLE STL JAC HOU NO SD DET OAK CIN
Graph:
2008-10-Nfl-Ic-Clean

Week 11:
Excluded Games: SD@PIT(10@11), CLE@BUF(29@27)
Picks: 10-5-1 (93-66-1) (-3)
Rankings: CAR NYJ TB ATL TEN ARI PIT GB NYG MIA CHI PHI NO IND WAS NE BUF DAL SD DEN BAL OAK MIN KC SEA SF CLE STL JAC HOU DET CIN
Graph:
2008-11-Nfl-Ic-Clean

Week 12:
Excluded Games: CHI@STL(27@3), IND@SD(23@20), NYJ@TEN(34@13)
Picks: 9-7 (102-73-1) (-4)
Rankings: NYG NO NYJ ATL TEN TB ARI CAR PIT GB NE MIA CHI PHI MIN WAS SD JAC IND BUF DAL BAL OAK DEN SF SEA STL CLE HOU KC DET CIN
Graph:
2008-12-Nfl-Ic-Clean

Week 13:
Excluded Games: PIT@NE(33:10)
Picks: 9-7 (111-80-1) (-6)
Rankings: TB TEN NYG MIN ATL CAR NYJ NO PIT NE ARI PHI DAL GB MIA CHI DEN WAS IND SF SD BAL BUF OAK HOU SEA STL CLE KC CIN JAC DET
Graph:
2008-13-Nfl-Ic-Clean

Week 14:
Excluded Games: (None)
Picks: 10-6 (121-86-1) (-5)
Rankings: DEN TB TEN MIN CAR NYG ATL NO ARI NE PHI GB MIA CHI SF IND BUF PIT NYJ DAL SD BAL SEA OAK STL JAC HOU WAS CLE DET KC CIN
Graph:
2008-14-Nfl-Ic-Clean

Week 15:
Excluded Games: MIN@ARI (35@14)
Picks: 8-8 (129-94-1) (-6)
Rankings: DEN TEN NYG TB CAR ATL ARI NE PHI CHI MIN IND PIT NO MIA SF NYJ BUF BAL DAL JAC GB SD SEA HOU OAK STL WAS CLE DET CIN KC
Graph:
2008-15-Nfl-Ic-Clean

Week 16:
Excluded Games: CAR@NYG(28@34), HOU@OAK(16@27)
Picks: 9-7 (138-101-1) (-3)
Rankings: TEN IND PIT BAL MIN NYG NE DAL HOU DEN ATL CAR ARI CIN WAS PHI NO SD TB NYJ CHI MIA OAK CLE SF BUF SEA JAC GB STL DET KC
Graph:
2008-16-Nfl-Ic-Clean

Week 17:
Excluded Games: STL@ATL(27@31), KC@CIN(6@16), NYG@MIN(19@20)
Picks: 10-6 (148-107-1) (-2)
Rankings: TEN IND NE NYG ARI PIT BAL MIA ATL CAR PHI MIN TB NYJ DAL HOU DEN CHI SF NO SD CIN BUF SEA WAS CLE OAK STL JAC KC GB DET
Graph:
2008-17-Nfl-Ic-Clean

So, the system is two games worse in the picks for this year.

2 Responses to 2008 NFL Graphs/Rankings – Inter/Intra

  1. Tom says:

    I’d wager that two games difference in picks is not a statistically significant deviation from the standard method. I wonder how much information we’d have to remove to get a really significant degradation in the pick accuracy.

    It’s interesting that you can more clearly see the divisions grouped together, and get a better picture of the relative divisional strength from these graphs than from the normal graphs.

  2. mm says:

    Interesting stuff…thanks for putting it up.

    It pretty much does what I thought. It removes most of the hard to remove ‘flukes’ that some felt was screwing up the graph (NYJ over Tennessee, and now Minn over NYG), while leaving us with a graph much less vertical (because the connections between divisions become weaker). Two games worse in picks over the year should be within a margin of error.

    Some observations:
    1)The Jets are still on top of a section, but its a very small section
    2)Dallas is almost completely disconnected from the rest of the NFC.
    3)The Bengals don’t look quite as good as they do in the main graph, but again they look a bit better than they are perceived. Here, the Browns also rise some.
    4)Minnesota looks good even without the Giants victory.
    5)The NFC South suffers from the fact that only 1 of the 6 head-to-head series was not split (Carolina beat New Orleans twice). This leaves Carolina, Atlanta, and Tampa Bay each at the top of a small chain, rather than one leading a bigger chain.
    6)The NFC East also has only 1 direct connection (NYG over Washington) between its teams, but Dallas is firmly trapped under an AFC chain.

    “I’d wager that two games difference in picks is not a statistically significant deviation from the standard method. I wonder how much information we’d have to remove to get a really significant degradation in the pick accuracy.”

    Well, 2 games out of 16 for each team is 12.5% of the teams, so its a pretty large amount. Of course this wasn’t removing a random 12.5% of games, but the specific 12.5% of games that had the least amount of independent ‘checks’ from the other games.

Leave a Reply

Your email address will not be published. Required fields are marked *