Big 12 Championship Game Projection

[Dale]

Hey, y'all. I've done this the past two years, and we're far enough into the season that I think it makes sense running the scenarios again.

Basically, this is 100,000 Monte Carlo simulations of the rest of the Big 12 football season in Python, using Ken Massey’s rating system for the odds for every remaining game.

Here are the chances for each team to make the Big 12 Championship Game. The parenthetical is their regular-season chances of finishing 1st and 2nd.

Oklahoma: 87.7% (61.6% 1st, 26.1% 2nd)
Texas: 72.8% (30.6% 1st, 42.2% 2nd)
Baylor: 17.1% (4.2% 1st, 12.9% 2nd)
Iowa St.: 6.9% (1.4% 1st, 5.6% 2nd)
Oklahoma St.: 5.1% (0.8% 1st, 4.3% 2nd)
Texas Tech: 4.1% (0.4% 1st, 3.8% 2nd)
TCU: 3.7% (0.6% 1st, 3.0% 2nd)
West Virginia: 2.1% (0.3% 1st, 1.8% 2nd)
Kansas St.: 0.4% (0.1% 1st, 0.3% 2nd)
Kansas: c'mon

And here are the Championship Game matchups that have at least a 1% chance of happening:

Oklahoma/Texas: 61.7%
Baylor/Oklahoma: 11.8%
Iowa St./Oklahoma: 4.8%
Baylor/Texas: 4.4%
Oklahoma/Texas Tech: 3.3%
Oklahoma St./Texas: 2.7%
Oklahoma/TCU: 2.6%
Oklahoma/Oklahoma St.: 2.1%
Iowa St./Texas: 1.6%
Oklahoma/West Virginia: 1.1%

(Slight boring caveat: The program only goes through the first two tiebreakers, since the third tiebreaker involves points and would be a pain in the butt to model. That said, the first two tiebreakers -- and situations with no tiebreaker necessary -- covered 99.6% of the simulations, so the odds wouldn’t change much.)

And finally, here’s Iowa State’s chances of ending up at any given record and their chances of making the championship game at that record:

ISU record chances:
3-9: 0.3%
4-8: 3.9%
5-7: 16.0%
6-6: 30.0%
7-5: 29.7%
8-4: 15.8%
9-3: 4.0%
10-2: 0.4%

ISU odds of championship game at record:
7-5: 0.4%
8-4: 19.9%
9-3: 82.8%
10-2: 100.0%

Anyone want the odds for different scenarios, just ask -- they're easy to do.

 

[Dale]

(One more caveat: For now Massey's a lot lower on Iowa State than e.g. SP+ or the Vegas oddsmakers, so Iowa State's true odds may be a fair deal higher than 6.9%.)

 

[Clonehomer]

Wouldn't have expected the 8-4 record still having that high of chance of making the championship game. 1:5 chance that we make it with 3 losses? Figured 2 losses would be the max.

 

[dirtyninety]

Spoiler Alert: 100% it is Texas v Oklahoma. The winner of that gets to lose to Alabama. Alabama is National Champion.

 

[cstrunk]

6-6 is the most likely outcome? Not buying it. 7-5 is probably the most likely at this point.

 

[Acylum]

Spoiler Alert: 100% it is Texas v Oklahoma. The winner of that gets to lose to Alabama. Alabama is National Champion.

I'm not sure the B12 gets a one loss team in this year, the way this season is going. Undefeated Oklahoma would be in for sure, Texas would really need LSU to go 12-1 to have a chance, IMO.

 

[Dale]

6-6 is the most likely outcome? Not buying it. 7-5 is probably the most likely at this point.

For fun, I ran it giving ISU a boost under the assumption they were around a top 15-20 team (where SP+ has them):

Oklahoma: 85.1% (60.6% 1st, 24.5% 2nd)
Texas: 64.2% (27.1% 1st, 37.2% 2nd)
Iowa St.: 18.2% (5.1% 1st, 13.1% 2nd)

ISU record chances:
3-9: 0.1%
4-8: 1.1%
5-7: 6.6%
6-6: 19.5%
7-5: 31.4%
8-4: 27.4%
9-3: 12.0%
10-2: 1.9%

So you very well might be right.

 

[cyfanatic13]

I don't like this. Can you crunch those numbers again?

 

[Dale]

Wouldn't have expected the 8-4 record still having that high of chance of making the championship game. 1:5 chance that we make it with 3 losses? Figured 2 losses would be the max.

Looking at situations where that happened: Most of them Texas fell apart and ISU won the two- or three- or four-team tiebreaker for second place.

 

[cyrocksmypants]

Baylor-Iowa State rematch. Book it.

 

[CloneFan4]

I LOVE this thread that pops up the past couple of years. The stat nerd in me goes bonkers when I see it pop up on the recent posts page.

@Dale thanks for all your work in this!!

 

[TykeClone]

I don't like this. Can you crunch those numbers again?

I reject your reality and would like to substitute it with a better one!

 

[cyfan15]

Just going off how good I think the teams are without crunching numbers, I think the one game we absolutely have to win is the Texas game.

I see our best case scenario being 7-2 with 6-3 being far more likely. We could still get to the championship game at 6-3 if one of those wins is over Texas. We basically need Baylor to lose to Oklahoma and Texas (both in Waco) plus at least one other game. We need Texas to lose to ISU, Oklahoma, and at least one other team that isn't Baylor. Then it comes down to tiebreakers and hopefully we come out on top.

This week we'll be rooting for Oklahoma over Texas and Texas Tech over Baylor in addition to ISU over West Virginia.

 

[cyrocksmypants]

Just going off how good I think the teams are without crunching numbers, I think the one game we absolutely have to win is the Texas game.

I see our best case scenario being 7-2 with 6-3 being far more likely. We could still get to the championship game at 6-3 if one of those wins is over Texas. We basically need Baylor to lose to Oklahoma and Texas (both in Waco) plus at least one other game. We need Texas to lose to ISU, Oklahoma, and at least one other team that isn't Baylor. Then it comes down to tiebreakers and hopefully we come out on top.

This week we'll be rooting for Oklahoma over Texas and Texas Tech over Baylor in addition to ISU over West Virginia.

So pretty much an identical situation to last year.

 

[TedKumsher]

I'm not sure the B12 gets a one loss team in this year, the way this season is going. Undefeated Oklahoma would be in for sure, Texas would really need LSU to go 12-1 to have a chance, IMO.

It generally seems like early season losses are de minimis in comparison to late season losses. I think Texas will be in easily if they win out, including the championship game, unless there are 4 teams with either 0 losses or 1 early season loss.

 

[HARMCYN]

What would this scenario look like on Massey?

Baylor 10-2 (loss @Okstate, @TCU)
ISU 9-3 (Ls @ BU, @OU)
Texas 9-3
(Ls @ISU, @BU)
Oklahoma 10-2 ( Ls to UT, @BU )

Baylor is League Champ on Head to Head.

3 way tie for 3rd goes past Head to Head and records vs. Each team in tie- 1-1, does the Big 12 use point differential as the Tiebreaker at this point?

 

[Dale]

Updated for this week.

Big 12 title game odds:

Oklahoma: 93.9% (82.5% 1st, 11.4% 2nd)
Texas: 51.4% (5.9% 1st, 45.5% 2nd)
Baylor: 26.2% (6.2% 1st, 20.0% 2nd)
Iowa St.: 17.8% (3.8% 1st, 14.0% 2nd)
Oklahoma St.: 4.4% (0.9% 1st, 3.5% 2nd)
TCU: 4.4% (0.7% 1st, 3.7% 2nd)
Texas Tech: 1.6% (0.0% 1st, 1.5% 2nd)
Kansas St.: 0.2% (0.0% 1st, 0.2% 2nd)
West Virginia: 0.2% (0.0% 1st, 0.2% 2nd)
Kansas: nope

Most likely title game matchups:

Oklahoma/Texas: 47.3%
Baylor/Oklahoma: 22.7%
Iowa St./Oklahoma: 15.3%
Oklahoma/TCU: 3.8%
Oklahoma/Oklahoma St.: 3.0%
Baylor/Texas: 1.9%
Oklahoma/Texas Tech: 1.5%
Baylor/Iowa St.: 1.1%
Iowa St./Texas: 1.1%

ISU record chances:

4-8: 0.2%
5-7: 3.3%
6-6: 15.8%
7-5: 32.7%
8-4: 32.1%
9-3: 14.0%
10-2: 1.9%

ISU odds of championship game at record:

7-5: 0.3%
8-4: 15.5%
9-3: 78.0%
10-2: 99.6%

 

[Dale]

What would this scenario look like on Massey?

Baylor 10-2 (loss @Okstate, @TCU)
ISU 9-3 (Ls @ BU, @OU)
Texas 9-3
(Ls @ISU, @BU)
Oklahoma 10-2 ( Ls to UT, @BU )

Baylor is League Champ on Head to Head.

3 way tie for 3rd goes past Head to Head and records vs. Each team in tie- 1-1, does the Big 12 use point differential as the Tiebreaker at this point?

Yep — which I don't model because of the difficulty. So, for example, in the run I did this evening, 0.61% of the simulations used point differential as a tiebreaker. In those situations, I just break the tie randomly. If it looks like it'll come into play in a particular season, I'll model it more carefully.

 

[Dale]

If TT had pulled it off over Baylor:

Oklahoma: 95.3% (85.7% 1st, 9.7% 2nd)
Texas: 55.1% (5.8% 1st, 49.2% 2nd)
Iowa St.: 22.5% (4.6% 1st, 18.0% 2nd)
Texas Tech: 9.4% (0.5% 1st, 9.0% 2nd)
Baylor: 9.0% (1.9% 1st, 7.1% 2nd)

 

[cyrocksmypants]

If TT had pulled it off over Baylor:

Oklahoma: 95.3% (85.7% 1st, 9.7% 2nd)
Texas: 55.1% (5.8% 1st, 49.2% 2nd)
Iowa St.: 22.5% (4.6% 1st, 18.0% 2nd)
Texas Tech: 9.4% (0.5% 1st, 9.0% 2nd)
Baylor: 9.0% (1.9% 1st, 7.1% 2nd)

****.