Sunday, September 20, 2015

Weekly Playoff Probabilities - Week 3

Another great weekend leads to more upheaval in my playoff probabilities.  As always, the explanation behind my methodology can be found here.  First the rankings, then some thoughts.

Rank Team FPI Rank POFF Prob Change
1 Ohio State 6 58.140% -11.3%
2 Mississippi 1 46.488% 17.1%
3 Notre Dame 9 39.121% 8.2%
4 Baylor 5 36.183% 6.9%
5 Georgia 3 28.524% 4.9%
6 Texas Christian 8 26.169% -19.4%
7 Michigan State 16 23.105% 0.3%
8 Louisiana State 2 20.762% 6.8%
9 Oklahoma 11 16.807% -7.4%
10 Florida State 14 15.177% 0.6%
11 Brigham Young 38 9.209% 9.1%
12 UCLA 18 8.637% -3.3%
13 Clemson 19 8.594% 2.6%
14 Texas A&M 10 6.228% -4.6%
15 Wisconsin 28 5.464% 0.1%
16 Southern California 7 4.636% -17.5%
17 West Virginia 20 4.512% 1.5%
18 Northwestern 40 4.453% 3.5%
19 Iowa 48 3.613% 0.8%
20 Alabama 4 3.330% -3.2%
21 North Carolina State 35 3.263% 1.0%
22 Stanford 17 3.130% 2.7%
23 Oregon 15 2.894% -1.3%
24 California 22 2.316% 1.1%
25 Oklahoma State 26 2.265% 1.5%
26 Georgia Tech 12 2.210% -2.4%
27 Arizona 32 2.049% 1.2%
28 Texas Tech 27 1.839% 1.5%
29 Miami (FL) 30 1.676% -0.5%
30 Tennessee 13 1.494% -0.4%
31 Virginia Tech 23 1.421% 1.1%
32 North Carolina 24 1.262% 0.9%
33 Toledo 57 0.997% -0.1%
34 Utah 39 0.624% 0.3%
35 Florida 25 0.604% 0.2%
36 Kansas State 36 0.553% -0.1%
37 Michigan 37 0.436% -0.1%
38 Temple 53 0.381% 0.0%
39 Missouri 42 0.255% -0.3%
40 Ohio 71 0.184% 0.0%
41 Mississippi State 21 0.176% 0.1%
42 Penn State 45 0.165% 0.1%
43 Duke 50 0.106% -1.4%
44 Memphis 47 0.096% 0.0%
45 Nebraska 34 0.091% 0.0%
46 Houston 70 0.084% 0.0%
47 Kentucky 46 0.081% -0.4%
48 Illinois 52 0.048% -0.4%
49 Indiana 75 0.035% 0.0%
50 Auburn 31 0.032% -0.1%
51 Arizona State 43 0.020% 0.0%
52 Washington 49 0.014% 0.0%
53 Pittsburgh 54 0.012% 0.0%
54 Syracuse 72 0.012% 0.0%
55 Minnesota 59 0.012% 0.0%
56 Boston College 61 0.007% -0.1%
57 Navy 73 0.002% 0.0%
58 South Carolina 44 0.001% 0.0%
59 Arkansas 29 0.000% 0.0%
60 Texas 41 0.000% 0.0%
61 Purdue 79 0.000% 0.0%
62 Rutgers 85 0.000% 0.0%
63 Colorado 62 0.000% 0.0%
64 Virginia 58 0.000% 0.0%
65 Maryland 80 0.000% 0.0%
66 Oregon State 87 0.000% 0.0%
67 Iowa State 69 0.000% 0.0%
68 Wake Forest 89 0.000% 0.0%
69 Washington State 86 0.000% 0.0%
70 Kansas 114 0.000% 0.0%
71 Vanderbilt 64 0.000% 0.0%

1. Now that most small-conference teams have taken a loss, we're starting to see fewer teams tumble from the rankings each week.  This week, only eight teams received their walking papers.  Let's take a moment to say bye to Louisville, Northern Illinois, Western Kentucky, Air Force, Old Dominion, Tulsa, Connecticut, and Charlotte.  Louisville is our first major-conference victim, as their difficult opening slate eliminates them 15 days into the season.

2. BYU's one-point loss would have disqualified them from my rankings, but given that pollsters still love them enough to rank them, I decided to do what the ACC did, and declare BYU a power-five team.  This rockets them from a miniscule chance at the playoff to 11th place on the list.  Yeah, the old ranking was probably a bit low, and this one is probably a bit high.  The Cougars' weak October slate will probably at least partially ding them in the committee's eyes, so a 9% chance is pretty generous.  For now though, I will keep them as they are.  I may implement some new halfway classification for BYU next season, depending on how the rest of this season goes.

3. Western Kentucky and Air Force's losses eliminate their respective conferences from the race.  The MAC and American hold on with roughly the same long-shot odds they had last week (Toledo still has a better chance of making the playoff than all other small-conference teams combined).  Added clarity in the SEC race helps to elevate them to the top spot, while the general cooling off of pre-season favorites TCU and Ohio State dings their conferences' chances a bit:

Conference Exp Playoff Teams
SEC 1.080
Big 10 0.956
Big 12 0.883
Independent 0.483
ACC 0.337
Pac 12 0.243
MAC 0.012
American 0.006

4. Now that we are three weeks in, we have a fair number of two-loss major conference teams.  While it's very possible that a two-loss team makes the playoff, I feel confident in saying that none of these particular teams will.  That said, the best flyer from this group is by far Nebraska.  The Cornhuskers won out in 139 (of 10000) simulations, while all the other teams combined did so only once.

5. A major conference team has probably never had a more anonymous first quarter of the season than your 2015 NC State Wolfpack*.  I dare you to name me one of the three opponents they have defeated.  Don't try because you can't.  In spite of this, and in spite of minimal expectations for Dave Doeren's squad, they're sitting there virtually tied with Alabama of all teams.  College football is a wonderful thing.  Well, at the very least, it's a thing.

*Who?

6. I noted earlier that Alabama, Georgia Tech, and USC all looked like great teams entering the season, but could become undone by difficult non-conference schedules and/or strong cross-conference matchups.  Well, look what happened.  All three of these teams have the high-profile games to get themselves back into the race, but it's going to be an uphill climb.  As long as the committee places a primacy on win-loss record, difficult schedules are going to be more of a burden than a benefit.  We'll see if this gets any better this year, but for now, I'm going to keep my model as it is.


Week 4 Preview

Home Away Home Win Prob Playoff Teams Lost
Texas Tech TCU 32.02% 0.032
Michigan   BYU 58.34% 0.028
Arizona UCLA 41.73% 0.016
Arkansas Texas A&M 34.68% 0.007
Ohio State Western Michigan 96.42% 0.007
Oregon Utah 78.49% 0.005
Arizona State USC 18.86% 0.004
Florida Tennessee 43.78% 0.004
Syracuse LSU 5.80% 0.004
Ole Miss Vanderbilt 97.50% 0.004

Aside from the ugly Halloween slate I warned you about previously, this is probably the weakest remaining set of games (for reference, there are 22 games from weeks 5-13 that currently have more playoff impact that the Tech-TCU tilt, and 96 games with more impact than the #4 game from this week).  There is certainly good depth of semi-interesting games within the stronger conferences (for example, Duke-Georgia Tech and Auburn-Mississippi State didn't make the list), but the majority of top teams play a rather easy opponent.  From a playoff perspective, this weekend might not tell us much.  The main takeaway is likely to be a separation in the Pac 12, which might help to partially reverse their ever-weakening playoff case. 

No comments:

Post a Comment