100k Simulations of All Texas Private Six-Man Brackets

It was just time to get down to it. I had been delaying the inevitable, running 100,000 simulations of each and every private school six-man state bracket. For details on how I did this, please read the earlier posts I have written about the public school brackets and other Monte Carlo simulations I have written. This was very similar….

First build the start bracket using this week’s ratings from my website (www.sixmanfootball.com). Then calculate the probability of each first round game and simulate the result. After each round I update the ratings (not 100% like my formula, but a close enough estimation) and continue…. do this 100,000 times and see what happened.

Well, here’s what happened.

TAPPS D1          
Boerne Geneva 0 16679 33496 12466 37359
Midland Trinity 0 13728 45784 12329 28159
Baytown Christian 0 38588 25904 24432 11076
Watauga Harvest 25725 17419 26487 20469 9900
Rockwall Heritage 35535 37235 13252 10377 3601
Sugar Land Logos Prep 24633 61253 8646 2274 3194
Houston Emery-Weiner 74275 9537 10169 4741 1278
Pasadena First Baptist 64465 24177 6168 3917 1273
Abilene Christian 45374 38952 10399 4169 1106
Round Rock Christian 49480 43205 5492 864 959
Katy Faith West 50520 43067 4867 711 835
Austin Hill Country 54626 34092 7621 2879 782
Waco Vanguard 75367 22068 1715 372 478
TAPPS D2          
Waco Live Oak 0 9711 16354 17471 56464
Austin Veritas 0 20885 34691 29735 14689
Orange Community Christian 0 46910 27849 18231 7010
Dallas Tyler Street 36085 17714 34388 5019 6794
SA Castle Hills 22194 38296 19682 14086 5742
Cedar Park Summit 25660 66166 3416 2406 2352
Denton Calvary 0 69007 26428 2318 2247
Kerrville Our Lady of the Hills 63915 13279 18663 1980 2163
Bulverde Bracken Christian 36088 49169 9136 4452 1155
Dallas Lakehill 77806 14794 4581 2204 615
Conroe Covenant Christian 63912 29946 4061 1649 432
Lubbock Christ The King 74340 24123 751 449 337
TAPPS D3          
Longview Trinity 0 25590 27022 16279 31109
Fredericksburg Heritage 0 29268 21774 25772 23186
WF Notre Dame 0 36128 35968 11996 15908
Fort Worth Covenant Classical 11814 54054 22060 5912 6160
Granbury North Central Texas Academy 43834 15878 24039 10298 5951
Seguin Lifegate 15010 58039 11590 10003 5358
Richardson Canyon Creek Christian 41133 42480 8406 3951 4030
San Marcos Hill Country Christian 56166 14491 19126 6837 3380
Alvin Living Stones 0 69631 22245 5739 2385
WF Wichita Christian 58867 31930 5010 2152 2041
Brenham Christian 84990 12693 1226 805 286
Selma River City Believers 88186 9818 1534 256 206
TCAF D1          
Fort Worth Nazarene 25645 32238 42117
Wylie Preparatory 26385 35665 37950
Dallas Inspired Vision 74355 15287 10358
Waco Methodist Childrens Home 73615 16810 9575
TCAF D2          
Azle Christian 20861 16452 62687
Granbury Cornerstone 32698 49507 17795
Weatherford Christian 79139 7600 13261
Arlington St. Paul Prep 67302 26441 6257
TCAL D1          
Bryan Allen Academy 1340 2173 2839 93648
SA The Atonement 47104 23154 28492 1250
Tyler King’s Academy 48830 49877 289 1004
Greenville Phoenix 44302 28918 25840 940
EP Faith 52896 22695 23532 877
Bryan Christian Homeschool (BVCHEA) 51170 47717 255 858
Houston Mount Carmel 98660 233 266 841
Clear Lake Christian 55698 25233 18487 582
TCAL D2          
Stephenville Faith 2150 4906 24891 68053
Sugar Land HCYA Fort Bend 4188 21572 49549 24691
Corpus Christi Annapolis 12883 64578 18001 4538
Killeen Memorial 37562 58609 2464 1365
SA Sunnybrook 62438 35860 1165 537
Corpus Christi Abundant Life 97850 625 1092 433
Corpus Christi WINGS 87117 11393 1290 200
Lockhart Lighthouse Christian 95812 2457 1548 183
TAIAO D1          
Tyler HEAT 9493 28388 28029 34090
SA FEAST Homeschool 23317 29535 20494 26654
Capital City Christian Home School 34148 35468 15424 14960
Temple Centex Homeschool 39095 38257 12965 9683
Fort Worth THESA 65852 22050 7102 4996
Crosby Victory and Praise 60905 27484 7228 4383
Bryan Aggieland Home School (BCAL) 76683 12947 6135 4235
Plano CHANT 90507 5871 2623 999
TAIAO D2          
Austin NYOS 0 18228 29436 52336
Bastrop Tribe Consolidated 0 29389 40156 30455
Waco Parkview 19490 54627 17568 8315
San Marcos Homeschool 21741 62559 8584 7116
Weatherford Home School 78259 19213 1626 902
Victoria Home School 80510 15984 2630 876

Obviously for TCAF, I am moving straight into this week since the first round was played last weekend.

Another thing to notice is that teams like Austin NYOS do not lose in the first round. Why? They got a bye.

The biggest shocker at first glance – the fact that Bryan Allen Academy is such a huge favorite. I expected it to be high, but 93.6% to win it all is a little obscene.

So I hope everyone enjoys this… and remember, no wagering.

East and Throckmorton likely to rule UIL D2 Six-Man Playoffs

After 100,000 simulations, the Throckmorton Greyhounds appear to have a 29.8% chance to win the UIL D2 Six-Man State Championship. The biggest challenge it appears will be the dominance of the East bracket, which won a dominating 80.1% of the time in the simulation.

Yesterday I wrote about how the Crowell Wildcats are a somewhat dominant 33.1% to repeat as the D1 UIL State Six-Man Champions. If you would like to read more details on the methods, I have several posted below.

Basic note: The table represents how many times each team LOST in that round or became the champion (final column).

Throckmorton 2277 17879 26568 18537 4942 29797
Guthrie 7473 13276 45288 15376 3977 14610
Calvert 7171 26565 28201 20608 3668 13787
Richland Springs 16212 9781 33384 23044 3859 13720
Groom 14392 22605 26298 11473 18726 6506
Follett 20907 9637 30748 13218 19457 6033
Jonesboro 21822 51329 15095 7807 1096 2851
Motley County 36812 49576 7304 3538 821 1949
Buena Vista 24382 31007 18346 15296 9149 1820
Balmorhea 35900 17918 21475 14798 8314 1595
Blanket 26142 37455 17133 12373 5848 1049
Southland 16387 56498 15533 5156 5424 1002
Chillicothe 14573 70298 12063 1898 356 812
Oglesby 83788 4289 8126 2777 309 711
Lueders-Avoca 63188 31122 3323 1401 289 677
Mt. Calm 26113 62182 8582 2343 252 528
Sands 64100 13218 12862 6674 2706 440
McLean 79093 4955 10141 2791 2601 419
Blackwell 30030 45928 14678 6701 2320 343
Mullin 78178 17599 2828 1014 112 269
Sierra Blanca 75618 14193 5708 3167 1158 156
Whitharral 41417 49108 6784 1462 1082 147
Jayton 92527 2983 3809 455 90 136
Lefors 85608 7191 4864 1239 973 125
Trinidad 92829 4507 1933 566 61 104
Loraine 73858 17345 5047 2663 984 103
High Island 73887 23748 1851 406 26 82
Rising Star 69970 22936 4751 1763 507 73
Kress 58583 36300 3781 770 511 55
Lazbuddie 83613 13706 1851 456 333 41
Harrold 97723 1423 667 125 28 34
Forestburg 85427 13443 978 105 21 26

It is interesting to note that while Richland Springs and Calvert have higher ratings at the current time, Guthrie actually has the second-highest chance to win the tournament (14610 to 13720 and 13787, for RS and Calvert, respectively). This is due to the fact that Guthrie has it easier in the first two rounds.

Out West, Groom and Follett (6506 and 6033 wins) have a combined probability that’s less than any of the top-4 from the East. On the bright side, they reach the finals more than each of these, mostly due to the fact that Throckmorton is not in their half of the draw.

It certainly looks like the West is more competitive in the sense that the teams are more even and quite a few more have solid opportunities to reach the semis and finals.

Coming Next: All of the private school draws.


Crowell Favorite to Win Six-Man Title with 33.1% Win Probability

I have created several Monte Carlo simulations over the past year to try and determine probabilities for various sporting events. This week I decided to tackle the Texas Six-Man state tournament. (I will publish more bracket evaluations as the week goes on)

For the past 21 seasons, I have been producing rankings for six-man football. For those of you who do not know the history, I would fax my rankings to newspapers across the state and several would actually publish them. I eventually put together a newsletter, The Huntress Report, where I would add scores, game stories, stats and schedules to the rankings and mail (or fax) to subscribers. Eventually I moved to a website, where I would update the information a week behind, so that my subscribers would be getting the freshest information first. That all was scrapped in 1999 when I decided to go 100% to the website (www.sixmanfootball.com).


You can read some of my earlier posts (see below here at sixmanguru.com) where I discuss Monte Carlo simulations if you are interested. In this case I played the UIL Division I tournament 100,000 times using probabilities calculated from the ratings on my website. To account for upsets and a more Bayesian methodology, I modified the teams ratings to also simulate my rating systems (generally) after each round. I also recorded each round a team lost and below are the results.

Crowell, the defending DI state champions, wins the title again a whopping 33.1% of the time and reached the finals over 41% of the time.

Crowell 8686 12684 25264 11843 8426 33097
Ira 16846 9048 41178 9920 6573 16435
May 1911 8450 19585 28915 26882 14257
Blum 9396 2817 28089 27327 22565 9806
Borden County 16195 24064 21597 25036 4836 8272
Happy 9925 21253 34686 24447 4075 5614
Abbott 26692 4293 40184 16474 9330 3027
Water Valley 22717 63399 8220 2512 1390 1762
Valley 21460 50424 14276 10825 1381 1634
Gordon 30614 18265 35123 9414 5227 1357
Knox City 83154 4192 9627 1484 664 879
Grady 42027 41247 11329 4300 508 589
Highland 91314 3270 3476 935 478 527
Aquilla 73308 2813 17360 4297 1795 427
Sterling City 44824 47109 6529 782 373 383
Zephyr 17303 55678 21972 3477 1296 274
Anton 83805 8484 4666 2544 254 247
Newcastle 69386 11555 15094 2678 1046 241
Garden City 55176 39651 4328 411 224 210
Ropes 57973 32379 6938 2288 218 204
Marfa 77283 20647 1378 374 151 167
Nazareth 78540 17028 2724 1409 143 156
Milford 90604 972 5262 2335 697 130
Santa Anna 48258 46737 2667 1705 539 94
Rochelle 51742 44020 2323 1416 429 70
Spur 90075 5121 3784 890 64 66
Leverett’s Chapel 35599 59216 4515 531 121 18
Eden 82697 14502 2461 249 77 14
Chester 44796 52912 1717 462 99 14
Tioga 98089 793 775 280 50 13
Campbell 55204 43299 1154 281 51 11
Savoy 64401 33678 1719 159 38 5

The good news is every teams has a chance to win it all — even Savoy. The bad news — it appears they only an approximate 5 in 100,000 chance. I did run this a few times and they did get as high as 12 in one of the iterations. Tioga, a team that loses 98.1% of the time in the first round actually has a better chance than Savoy with 12 wins.

Another thing that stands out would be the fact that Ira, despite winning the title a theoretical 16.4% also seems to lose in the first round (16.8%) much more often than teams like Crowell (8.7%) or May (an amazing 1.9%). This goes to show that despite the 45-point expected spread on the Ira-Knox City game, it is still a much more difficult match-up for the Bulldogs than Highland or Tioga will be for Crowell and May, respectively.

Also interesting to note is that the East wins a dominant 70.2% of the time.

The most common final is a rematch of last year’s, May v. Crowell, with Blum v Crowell coming in next. The good news for May is they reach the final 41.1% of the time, which is a very good season. Blum is expected to reach the final about 32.4% of the time.

Wednesday I will release my UIL DII simulation results (they are already done, but it is my anniversary and we are going out for dinner). I will release the private school results either late Wednesday or early Thursday.

Creating Maps on the Fly For UIL Realignment

This morning the high school football season officially started with the release of the much anticipated 2014-2016 UIL Football Alignments. This usually started with the UIL servers crashing due to the high volume of traffic (it did briefly, prior to release). This year the UIL was prepared and had a back-up plan to divert traffic off their site.

So at exactly 9:00 am, the Twitterverse was alive with the ramblings of everyone who cares about Texas high school football.

I downloaded the files and immediately started sorting the teams into an Excel spreadsheet I had prepared for the occasion. Once done, I placed the data into my main database online.

I had been modifying my map code I created and showed in a previous post to handle the different divisions, sort them by division and district and color code them. By recycling this code, I easily created three maps.

Here’s the one where I sort them by division (blue for d1, red for d2)

Here’s a look at it

2014 UIL Realignment

2014 UIL Realignment

Then here are the ones where I separate Division 1 and 2 and then split up the districts.

Here’s an example shot of what they look like

UIL Division 1 snapshot

UIL Division 1 snapshot

Using the new google maps API, it took less than an hour to get everything up and running. All that was left was a little formatting and fine-tuning. The conversion of the data to an XML file makes all of the debugging so much easier.

Super Bowl Least Squares Predictions — take the points

For the past few months, I have been applying Least Squares Optimization principals to the NFL in making predictions. The method is fairly well established, simply to do and so far — very effective. (here’s a link to the first article explaining it all)

Up-To-Date Results
In games where the (expected line-actual line)/actual >100%, the line went 15-5 -1 since I started this in week 13 and 2-2 during the playoffs

In games where the absolute value of the (expected-actual)>2.5, it has gone 25-14-1 since week 13 and 3-2 during the playoffs.

In all games the method has gone 53-30-2 since week 13 and 6-4 during the playoffs.

Visitor V-Rat Home H-Rat Line Expected Diff (raw) Diff/line
Denver Broncos 31.21 Seattle Seahawks 32.67 -2.5 1.45 3.95 -158%

As you can see, the system expected Denver to be getting points and they are getting them, so all signs point to taking the Seahawks and the points.

Least Squares Predictions 3-0-1 During NFL Wild Card Round

With the first weekend of the NFL Playoffs completed, it seemed like a good time to catch up on how well the Least Squares Optimization predictions did the past two weeks.

If this is your first time reading about this, please refer to my initial article here.

First, let’s recap the final week of the regular season. Using only the games where the percentage difference between the expected line and the actual lines (from by sportsbook.com when published) was greater than 100%, the predictor went 3-1. In games where the raw absolute value of the expected and actual lines was greater than 2.5, the predictor went 5-3. Overall the predictor went 12-3 (the Bears-Packers game did not have a line sure to the unsure status of Aaron Rodgers.

Regular Season Totals
For the final five weeks of the season, using least squares predictions vs. actual vegas lines, the predictor went 15-4 in games where the percentage difference was greater than 100%, 22-13 in games where the raw absolute value was greater than 2.5 and 47-27-1 overall.

Wild Card Games
That bring us to the playoffs, where only four wild card games were on the slate this past weekend.

Visitor Home Vegas Exp Raw Diff % Diff
Kansas City Chiefs Indianapolis Colts -1.5 1.1 2.6 -173%
New Orleans Saints Philadelphia Eagles 3 -3.64 -6.64 -221%
San Diego Chargers Cincinnati Bengals 7 5.83 -1.17 -17%
San Francisco 49ers Green Bay Packers -3 -10.11 -7.11 237%

Reading the tables, the predictions were Indianapolis (+1.5), New Orleans (+3), San Diego (+7) and San Francisco (-3). Of those selections, the games went 3-0-1 overall, 2-0-1 for those where the percentage difference was greater than 100% and 2-0-1 also where the raw difference was greater than 2.5 points.

Of course about half way into that first game, when Kansas City was 28 points ahead of Indianapolis.

Up-to-date Totals
Percentage difference Vegas v Expected: 17-4-1
Raw Difference Greater than 2.5: 24-13-1
Overall: 50-27-2

Why isn’t Purdue in the Sugar Bowl? A study in graph theory

Why isn’t Purdue in the Sugar Bowl?

Yes, 1-11 Purdue, with their big time win over Indiana State. It sounds absurd, doesn’t it? But like 118 other teams in the NCAA Division I BCS, they have an indirect win over Alabama (and Auburn for that matter).

This is one of the reasons I love college football.

You hear all of the talk about how on any given day, TEAM A can beat TEAM B. But we don’t believe it, until some Saturday in the fall, Georgia Southern beats Florida or Appalachian State beats Michigan.

This is a story, not so much that a team like Purdue has a win over the mighty SEC teams like Alabama, but the amazing journey of how we get there.

A little background: I have been working on a project that connects College Football and graph theory – actually graph databases. I am going to skip through this, but if you are interested at all, just go here (http://en.wikipedia.org/wiki/Graph_database or download the free book located here, http://www.neo4j.org/learn). I use the Neo4j software.

The central concept is to connect teams together as nodes in a huge network. The data presented here was not compiled using this exact software, but by other means, as I attempted to verify parts of the database I am building. I will write more on that at a later date.

Back to Purdue, a horrible team (sorry Jerry and Drew), that went 1-11 and only beat NCAA Division I FBS Indiana State. How can they have an indirect win over Alabama? With a little luck and some help from the NCAA and NAIA playoffs in other divisions is the short answer.

It is a long and winding trip, which starts deep in the heart of the Midwest, travels through the backwoods of collegiate football not known to the casual reader and ends at Jordan-Hare Stadium… and it takes an incredible 40 games to connect the dots. Yes, 40 games.

The 40 games is by far the longest path this season, by a long shot, which starts and ends at a NCAA I BCS school. The next longest paths are Kentucky, Tulsa and UTEP’s 15-game paths to Alabama.

That is quite a jump and it is a long journey, so let’s get started.

September 7, 2013 (Lafayette, IN)
Purdue 20 Indiana State 14
It began in September when the Boilermakers, full of hope and promise, took it to the Sycamores. Unfortunately, they never won again.

September 14 (Terre Haute, IN)
Indiana State 70 Quincy 7
Not many realize this, but Indiana State went 1-11 as well, but their lone victory came the next weekend in a 70-7 rout of NCAA Division II, Quincy University.

Quincy is a small, Franciscan school of about 1300 students located on the banks of the Mississippi River in West Central Illinois. It was the college of Father Augustine Tolton, the first African-American Catholic priest.

September 28 (Quincy, IL)
Quincy 36 Lindenwood University-Belleville 7
The Quincy Hawks of Quincy, Illinois did not soar this season, going a paltry 2-9.

Lindenwood-Belleville, was opened in 2003 as an extension of Lindenwood University, but is now a stand-alone college with almost 2000 student. The football team is known for having the barbershop-striped field that alternates grey and red stripes every five yards.

what the?

what the?

LU-Belleville is an NAIA Independent and the path towards Alabama stayed within the NAIA for 15 games.

September 5 (Fayette, MO)
Lindenwood-Belleville 42, Central Methodist 16
Central Methodist, located in Fayette, Missouri, was founded in 1854 (or 1853 or 1855, depending on where you look). Former Missouri Governor Roger Wilson, who served the remainder of Governor Mel Carnahan’s term when he was killed in an airplane crash in October of 2000, is an alum of Central Methodist.

September 14 (North Newton, KS)
Central Methodist 38, Bethel KS 31
The Bethel College struggled through a 2-9 season. The school is the oldest Mennonite college in North America, founded in 1887 by Russian Mennonites, who has flocked to the area in the 1870’s.

November 16 (North Newton, KS)
Bethel KS 26 Bethany KS 14
Bethany College is located in Lindsborg, Kansas and is home to a mere 600 students. The college was founded in 1881 by Swedish Lutherans.
According to Wikipedia, “since 1903, when the ‘Terrible Swedes’ were feared and respected by all opponents, students and alumni have rallied Bethany athletic competition with the ‘Rockar! Stockar!’ cheer.” This apparently means “Rocks! Sticks!”.

November 2 (Hillsboro, KS)
Bethany KS 24 Tabor 17
Tabor College is another Mennonite-based college. Founded in 1908, it has fewer than 700 students. NFL Pro Bowler, Rolland “Bay” Lawrence, who played eight seasons at cornerback for the Atlanta Falcons during the 1970’s attended Tabor.

September 7 (Lincoln, NE)
Tabor 10 Nebraska Wesleyan 9
Founded by Methodists in 1887, Nebraska Wesleyan is known as the Prairie Wolves, which was only adopted in 2000. They were previously known as the Sunflowers, Coyotes and Plainsmen.

Glenn and Grace Hefner, parents of Hugh Hefner, are listed among their notable alums.

September 28 (Orange City, IA)
Nebraska Wesleyan 22 Northwestern IA 8
Northwestern College was founded in 1882 as an academy and became a four-year institution in 1961. It is affiliated with the Reformed Church of America.

The Red Raiders reached the NAIA playoffs this season and went 8-3 before falling to Missouri Valley.

Former NAIA Women’s Player of the Year (2006 and 2008) and record holder of most in-game consecutive free throws (133 straight), Deb Remmerde attended Northwestern.

November 9 (Orange City, IA)
Northwestern IA 38 Morningside 28
Morningside, located in Sioux City, Iowa, is affiliated with the Methodist Church. Pro Football Hall of Famer George Allen coached the football team from 1948-1950.

The Morningside Mustangs rebounded from this loss and actually reached the NAIA national semifinals, finishing with an 11-2 record.

November 23 (Sioux City, IA)
Morningside 40 Rocky Mountain 21
Arlo Guthrie attended Rocky Mountain College, which is located in Billings, Montana, but never graduated. Former Kansas City Chief, Chris Horn, also played football for the Battlin’ Bears.

October 19 (Billings, MT)
Rocky Mountain 45 Eastern Oregon 13
Eastern Oregon University is part of the Oregon Universities System and located between Portland, Oregon and Boise, Idaho. In 2011, 99-year old Leo Plass received his diploma from the university. He had dropped out less than one semester away from graduation during the Great Depression in 1932 to get a job as a teacher.

September 28 (La Grande, OR)
Eastern Oregon 35 Carroll MT 31
Carroll has a long football history, their 1931 team went undefeated and the more current version won six NAIA national championships between 2002 and 2010. Carroll is also where Hall of Fame coach John Gagliardi graduated and began his coaching career. Interestingly enough, Bobby Petrino also graduated from Carroll.

November 23 (Helena, MT)
Carroll MT 38 Georgetown KY 28
Founded in 1829 when the Kentucky General Assembly chartered the Kentucky Baptist Education Society to form a Baptist college within the state. Just recently the college of about 1300 students has severed ties with the Kentucky Baptist Convention and will operate as an independent university. The school has also considered becoming an NCAA DII affiliate, but their application was denied in 2012.

November 16 (Georgetown, KY)
Georgetown KY 20 Lindsey Wilson 10
Located in Columbia, Kentucky, Lindsey Wilson was only grades one through 12 from 1903-1922, mostly to train students to become teachers, many of whom continued their schooling at Vanderbilt. They also have a mascot named Blue Raider Bob. Seriously.

September 28 (Columbia, KY)
Lindsey Wilson 37 Faulkner 30
Faulkner University was founded in 1942 as Montgomery Bible School. It was eventually renamed to Alabama Christian College then Faulkner University. The Eagles did not even begin football until the 2007 season, when Jim Nichols, who had been a graduate assistant for Tommy Tuberville at Auburn, became their head coach.

September 7 (Ave Maria, FL)
Faulkner 47 Ave Maria 7
Ave Maria University was founded in 2003 as the dream of Thomas Monaghan, the founder of Domino’s Pizza, in his mission to found a Catholic university. He is the Chancellor of the school. The name Gyrenes refers to the Marines, of which Monaghan was a member. BTW, the Gyrenes went 7-2 this season, despite having fewer than 900 students.

Kind of looks like a pizza

Kind of looks like a pizza

October 12 (Ave Maria, FL)
Ave Maria 45 Florida Tech 41
Ave Maria defeated NCAA Division II, Florida Tech, who started football this season. Yes children, you can start football this season and still have an indirect win over Alabama. Good work, Panthers!

The school has been around for awhile and is the alma mater of at three astronauts and the original Survivor winner, Richard Hatch.

October 19 (Melbourne, FL)
Florida Tech 28 Shorter University 24
Shorter University of Rome, Georgia, was originally founded in 1873 as Cherokee Baptist Female College. It unsuccessfully attempted to break away from the Georgia Baptist Convention in 2005 and created a public relations storm when it required all faculty and staff to sign a public lifestyle statement in 2011. (http://jezebel.com/5854929/university-forces-employees-to-sign-anti+gay-pledge)

November 2 (Atlanta, GA)
Shorter 58 Clark Atlanta 14
Clark Atlanta is a historically black university founded in 1988 with the consolidation of Clark College (founded in 1869) and Atlanta University (1865). The school boasts many prominent alumni, including Henry Flipper, who after his freshman year at Atlanta University during the Reconstruction, was given a West Point appointment and later became the first African American to graduate from the United States Military Academy.

October 5 (Atlanta, GA)
Clark Atlanta 21 Morehouse 17
All-male Morehouse College is the alma mater of Dr. Martin Luther King, Jr., Spike Lee, Edwin Moses, Samuel L. Jackson and Herman Cain, to only mention a few. It was founded in 1867 as the Augusta Institute. Along with Wabash and Hampton-Sydney, it is one of only three remaining traditional liberal arts male colleges in the United States.

September 21 (Soldier Field, Chicago, IL)
Morehouse 42 Central St OH 20
The Central State Marauders were NCAA Division II national runner-up in 1983 and won NAIA titles in 1990, 1992 and 1995. Then, due to financial difficulties, the school dropped football in 1997, only reinstating the sport in 2005.

The school counts quite a few dignitaries as alum, including former President of Malawi, Hastings Kamuzu Banda. Pro Bowl lineman Erik Williams, baseball player Eddie Milner and ‘actress’ Omorosa (from The Apprentice) also attended Central State.

October 12 (Wilberforce, OH)
Central St OH 25 Miles 21
Located in Fairfield, Alabama, Miles is another historically black university founded in 1898. The football team was part of history this year when their game against Lane was officiated by a crew that included four women, marking the first time in history a predominantly female crew had officiated at any NCAA level.

The school is also the alma mater of Autherine Lucy, who graduated in 1952 then applied to graduate school at Alabama, eventually becoming the first African American student in the school’s history. She was expelled three days into school, as the university felt it could not provide a safe environment for her. The University of Alabama overturned her expulsion in 1980 and she earned a Master’s degree in 1982.

ADDITIONAL FUN FACT – Four of those teams ONLY won two games each. Shorter, Clark Atlanta, Morehouse and Central St OH won a combined eight games.

November 9 (Tuskegee, AL)
Miles 41 Tuskegee 36
Of course everyone knows that Tuskegee University is world-renowned as one of the first historically black universities, the home of the Tuskegee Airmen and was founded by George Washington Carver and Booker T Washington. But it also happens to be the shortest distance geographically to Jordan-Hare Stadium, where this journey will end. It is a mere 20.5 miles from Abbott Stadium.

September 7 (Huntsville, AL)
Tuskegee 23 Alabama A&M 7
With Alabama A&M, we are back in the NCAA Division I FBS, where the SWAC will exchange this chain for a few games. Famous Alabama A&M alumns are American Idol Season 2 winner Ruben Studdard and Pro Football Hall of Famer John Stallworth.

November 2 (Lorman, MS)
Alabama A&M 19 Alcorn State 18
The first black land-grant institution in the United States, Alcorn State was founded in 1871. Again, a school chock full of famous alumni including Medgar Evans and Alex Haley. A few football Braves include Donald Drive and the late Steve McNair.

November 7 (Lorman, MS)
Alcorn State 50 Prairie View A&M 35
Prairie View is part of the Texas A&M University System and was the organizing body of interscholastic sports and academic contests for black high schools in Texas prior to integration.

September 28 (Nacogdoches, TX)
Prairie View A&M 56 Stephen F. Austin 48
SFA is one of four public universities in Texas that is not part of one of the six university systems. The Lumberjacks of the Southland Conference, had their only bowl appearance in the 1973 Poultry Bowl, where they defeated Gardner-Webb.

September 21 (Nacogdoches,TX)
Stephen F. Austin 52 Montana State 38
Montana State earned a share of the 1956 NAIA title when they played St. Joseph’s of Indiana to a 0-0 tie in the Aluminum Bowl. They won the 1976 NCAA Division II title and the 1984 NCAA Division I-A title. They are the only team to win national titles in three different divisions. It also happens to be the alma mater of former ESPN anchor Craig Kilborn NFL kicker Jan Steneroud.

October 5 (Bozeman, MT)
Montana State 36 Northern Arizona 7
Northern Arizona’s initial graduating class consisted of four women who received teaching credentials for the then Arizona Territory. It is a far cry from the 26,000+ who now attend.

September 21 (Flagstaff, AZ)
Northern Arizona 22 South Dakota 16
Located in Vermillion, South Dakota, school was founded in 1862 as the University of Vermillion and is the oldest postsecondary institution in the Dakotas. The Yotes (formally Coyotes) now play in the Summit League.

October 19 (Cedar falls, IA)
South Dakota 38 Northern Iowa 31
UNI may be best known for upsetting top-seeded Kansas in the 2010 NCAA Basketball Tournament when Ali Farokhmanesh hit a crucial three. NFL Pro Bowler Bryce Paup attended UNI.

August 31 (Ames, IA)
Northern Iowa 28 Iowa State 20
After 32 games outside of Division I BCS, the trail returns.

November 30 (Morgantown, WV)
Iowa State 52 West Virginia 44
These two teams were by far the strangest in the Big 12 this season. West Virginia lost to both Kansas and Iowa State, but also defeated Oklahoma State to extend our chain.

September 28 (Morgantown, WV)
West Virginia 30 Oklahoma State 21
I still don’t know how this happened, other than this was OSU’s usual letdown loss.

August 31 (Reliant Stadium, Houston, TX)
Oklahoma State 21 Mississippi State 3
The rest fall in the hands of the mighty SEC.

November 28 (Starkville, MS)
Mississippi St 17 Mississippi 10
Jamerson Love’s fumble recovery in the end zone of what looked like it was going to be an easy Ole Miss touchdown shockingly ended the Egg Bowl.

October 19 (Oxford, MS)
Mississippi 27 LSU 24
Ole Miss gets the biggest win of Hugh Freeze’s tenure.

September 21 (Baton Rouge, LA)
LSU 35 Auburn 21
When LSU defeated Auburn, the Tigers were in the middle of the SEC and national title hunts. Little did we know how the fates of the two schools would reverse several months later.

November 30 (Auburn, AL)
Auburn 34 Alabama 28
Finally, in the Iron Bowl, Auburn shocks Alabama on the ridiculous field goal return by Chris Davis on the final play of the game. The chain is complete.

So there it was: 40 games. Purdue -> Alabama.

Texas Football Fan Sentiment Analysis During Valero Alamo Bowl

With Monday night’s Alamo Bowl being Coach Mack Brown’s final game as coach of the Texas Longhorns, it seemed like a good opportunity to test fan sentiment on the occasion via Twitter. I captured tweets containing certain words in an attempt to follow sentiment towards Mack Brown and Texas over time, leading up to the game, during the game and afterwards for a brief period.

I began collecting data around 2:25 PM CST and stopped just after 10:00 PM. The search terms I used were: Mack Brown, mackbrown, Texas Football, Texas Longhorn, hookem and hook em. During that time period, over 51,000 tweets were collected using these search terms. Please not that these terms could be used as regular words, a part of words as well as hashtags.

Alamo Bowl Pre-Game
Mack Brown is a good man and I have had the opportunity to meet him several times and have always found him amazing. I won’t go into the details, but he’s the type of guy that makes you feel important, despite the fact he’s probably the most important guy in whatever room he is in.

But things didn’t really end well on the 40 Acres, so I thought it might be interesting to see what transpired over the day.

To find sentiment, I utilized the R package, qdap (Quantitative Discourse Analysis Package), created by Tyler Rinker. The polarity function is based on Jeffrey Breen’s work.

Tweets are evaluated based on the words within them, utilizing the polarity function, which assigns this based on the sentiment dictionary (Hu and Liu, 2004). Approximately 50% of the tweets are neutral, earning a score of 0.

As I stated above, there were over 51,000 tweets, so plotting them over time would basically give you a huge cluster centered around the y-axis.

Instead, I decided to collect the tweets in 5-minute intervals and find average polarity. (You can click on the chart for a larger view)

Texas Fan Sentiment

Texas Fan Sentiment

As you can see, the average polarity was always positive during the pregame. Reading through the early tweets, you can see a bunch of congratulatory and well-wishing tweets directed at Coach Brown. As we near kickoff, the ‘positive’ flavor of the tweets drops dramatically, but the volume of tweets begins to take off.

The First Half
There’s a quick peak as kickoff approaches (almost 2k tweets in the five minute period around kickoff), then it tapers off a bit (300-500 tweets every five minutes).

Tweet Volume

Tweet Volume

The first big change comes when Case McCoy through the pick six on the first possession of the game. Not only does sentiment fall, but the volume of tweets begins to rise following the interception.

Sentiment fluctuates back and forth throughout the first half. I am guessing these are directly related to the Texas defense (positive) and offense (not-so positive) being on the field.

Again, when Oregon makes the big drive and scores at the end of the half, Texas fans react with expected disdain.

Second Half
Sentiment stabilizes during halftime and doesn’t really dip too badly until the final pick six by McCoy at the end of the game.

The most interesting discovery are the peaks in volume and what I call ‘Plus-Minus’. Plus-Minus is the number of positive tweets minus the number of negative tweets during the five minute interval.



The first major peak comes in the second half when Tyrone Swoopes enters the game. Swoopes is a very popular freshman quarterback many fans believe should have been given more playing time when Case McCoy struggled during the season.

Interestingly enough, the language was not as strongly positive as other times during the game, despite a preponderance of positive tweets.

The final big negative dip of the night came on Case McCoy’s second interception that went for a touchdown. The sentiment was so strong, it was the only time all evening where the average dipped below neutral.

Post Game
Texas fans appear to rebound fairly quickly after the game. It appears that fans were more interested in wishing Coach Brown well, than bashing the teams’ performance. Positive polarity reached its second-highest peak of the night, while tweet volume and Plus-Minus reach night-high peaks.

Final Thoughts
Obviously, when you limit yourself to search terms specifically tied to a person, team and school, you get skewed results. I did not include the term ‘Alamo Bowl’ specifically for this reason. I did not want Oregon fan or ‘bowl watcher’ sentiment included.

Since this past summer when I posted a twitter cloud of terms used in tweets at the 2013 NCAA Division I Tennis Championships, I have wanted to use R to do some ‘sentiment analysis’ on tweets during an event. I was initially stymied by Twitter’s move to OAuth, then by my fall class load.

Luckily, my cohort and friend, Taylor Smith was thinking the along the same lines and created an awesome pathway to try this out. Taylor details his methods here.

I had to make some modifications to his code and also altered my methods a bit, but the main collection, cleaning and initial evaluation of the data was the same. I will detail my specific changes in a later post.

Week 17 NFL Lines and Least Squares Predictions

If you have not read any of my previous Least Squared posting, please refer to the initial post here.

Week 16 in review: Only one game was in the range where we have been fairly confident on the selections. New England was a favorite in the system, but getting 2.5 points from Vegas. New England destroyed the Ravens, so the >100% difference between the expected and Vegas lines, bring that rule to 12-3 over the past four weeks.

Overall, the LS method went 8-7 (the Green Bay game again did not have a line, due to the unexpected status of Aaron Rodgers). This brings the Least Squares method to 35-24-1 over the last four weeks.

In games where the absolute raw difference was greater than 2.5 went 4-2, bringing its totals to 17-10 for the past four weeks.

Week 17 schedule: Five games fulfill the greater than 100% difference between expected and Vegas lines. Miami (-6) is a heavy favorite over the New York Jets. Detroit (+3) is considered a favorite over Minnesota, despite getting points. Washington (+4) is in the same position, as they are getting points at the New York Giants, despite being the favorite. The Kansas City Chiefs (+9.5) are getting huge points, despite being straight up favorites against the San Diego Chargers. Dallas (+6.5) is getting points as well, although the system is considering them a favorite.

Why these lines are probably so different
Being the final weekend of the season, there are obvious reasons for the dramatic differences in the lines. One is that some games have almost zero playoff implications to one team, but mean everything to the other (Kansas City at San Diego, NY Jets at Miami). The other main reason is probably injuries and general chaos among one of the teams (Washington at NY Giants, Philly at Dallas).

Both Washington and Dallas are disasters. In Washington, Shanahan has benched RGIII and virtually ticked off everyone, including his son and OC, Kyle. Talk out of Dallas is that Tony Romo and maybe Dez Bryant are out or probable. With a playoff spot in the line, who knows who will actually play for the Cowboys? I suspect they will make a game of it, but then again, they may just be ready to implode. I am also surprised that Washington game is so out of whack, since the Giants have not been too great, despite how bad Washington may tank this game. Again, gamble at your own risk.

I have listed the data below.

Visitor Home Line Expected Diff (raw) Diff/line
Carolina Panthers Atlanta Falcons -6.5 -10.28 -3.78 58%
Baltimore Ravens Cinncinati Bengals 6.5 8.30 1.80 28%
Houston Texans Tennessee Titans 7 9.88 2.88 41%
Jacksonville Jaguars Indianapolis Colts 11.5 16.83 5.33 46%
New York Jets Miami Dolphins 6 12.93 6.93 116%
Detroit Lions Minnesota Vikings 3 -1.90 -4.90 -163%
Washington Redksins New York Giants 4 -2.46 -6.46 -161%
Cleveland Browns Pittsburgh Steelers 7 7.01 0.01 0%
Denver Broncos Oakland Raiders -13 -15.39 -2.39 18%
Buffalo Bills New England Patriots 9.5 11.19 1.69 18%
Tampa Bay Buccaneers New Orleans Saints 13 13.42 0.42 3%
San Francisco 49ers Arizona Cardinals -1 -1.99 -0.99 99%
Kansas City Chiefs San Diego Chargers 9.5 -2.11 -11.61 -122%
St. Louis Rams Seattle Seahawks 10.5 14.53 4.03 38%
Philadelphia Eagles Dallas Cowboys -6.5 3.73 10.23 -157%

College Bowl Season Predictions Based on Least Squared Non-Linear Programming Model

There’s no shortage of data out there when it comes to college football, so I decided to take the time to create a least squares model, based on the same principals I have been using for the NFL, and outlined here.

The idea was to take all 752 schools that played college football, cross that with all 4138 games that were played (up to last weekend) and see how they predict the games (especially versus the Vegas lines).

When you try to create a model with 753 variables (I also included home field advantage for the regular season games as a variable), you quickly begin to test the limitations of Excel and Solver. After a little digging, I discovered that Frontline Solver, the company that developed Solver for Excel, limits the version included with the Microsoft product to 200 variables.

Not to be stopped, I found they have some advanced pro models and engines that will handle significantly more and downloaded a 16-day trial of these.

You set it to pick the best engine and it runs. For this particular problem, it happened to run the Gurobi Optimization engine.

From there, it is fairly plain and simple. You run the solver and it calculates a ranking for each team. I just plugged in the bowl games and here’s what we get.

Team A Team B Exp Vegas Diff % Diff
Colorado State Washington St 7.68 5.5 2.18 40%
Southern Cal Fresno State -8.22 -5 -3.22 64%
Buffalo San Diego St -6.86 -1.5 -5.36 357%
Louisiana-Lafayette Tulane -0.67 1.5 -2.17 -145%
Ohio U. East Carolina 16.48 14 2.48 18%
Oregon State Boise State -0.74 -3.5 2.76 -79%
Pittsburgh Bowling Green 5.69 5 0.69 14%
Utah State Northern Illinois -5.46 2 -7.46 -373%
Marshall Maryland -5.11 -2 -3.11 156%
Syracuse Minnesota 6.41 5 1.41 28%
Washington Brigham Young -8.75 -3 -5.75 192%
Rutgers Notre Dame 20.75 14 6.75 48%
Cincinnati North Carolina 5.93 3 2.93 98%
Miami FL Louisville 3.60 3.5 0.10 3%
Michigan Kansas State 3.23 3.5 -0.27 -8%
Middle Tennessee St. Navy 9.60 6 3.60 60%
Mississippi Georgia Tech 3.71 -3.5 7.21 -206%
Texas Oregon 18.33 14 4.33 31%
Texas Tech Arizona St 20.19 14.5 5.69 39%
Boston College Arizona 12.46 7.5 4.96 66%
Virginia Tech UCLA 11.22 7.5 3.72 50%
Mississippi State Rice -9.02 -7 -2.02 29%
Duke Texas A&M 9.69 13 -3.31 -25%
Nebraska Georgia 10.82 9 1.82 20%
Nevada-Las Vegas North Texas 7.91 6.5 1.41 22%
Wisconsin South Carolina -2.33 0 -2.33 -233%
Iowa LSU 7.21 8 -0.79 -10%
Michigan State Stanford 9.24 6 3.24 54%
Central Florida Baylor 24.64 17 7.64 45%
Oklahoma Alabama 14.31 16 -1.69 -11%
Oklahoma State Missouri -1.74 1.5 -3.24 -216%
Clemson Ohio State 2.49 2.5 -0.01 0%
Houston Vanderbilt -5.73 3 -8.73 -291%
Arkansas St Ball St 10.52 9.5 1.02 11%
Auburn Florida State 19.81 7.5 12.31 164%

Here’s what the table means. Vegas has Auburn as a (+7.5) underdog to Florida State, however, the model shows FSU as an almost 20-point favorite, an almost 12-point difference.

One of the most interesting lines is Clemson as a 2.5-point underdog to Ohio State. The LS model has that game as a 2.49-point spread, almost exactly spot on.

Several others are extremely close like that: Miami-Louisville (3%), Michigan-Kansas State (8%), Iowa-LSU (10%), Oklahoma-Alabama (11%) and Arkansas State-Ball State (11%).

Here’s a quick rundown of which side the computer is on for each game (including the Vegas lines): Washington State (-5.5), USC (-5), Buffalo (-1.5), Louisiana-Lafayette (+1.5), East Carolina (-14), Boise State (+3.5), Bowling Green (-5), Utah State (+2), Marshall (-2), Minnesota (-5), Washington (-3), Notre Dame (-14), North Carolina (-3), Louisville (-3.5), Michigan (+3.5), Navy (-6), Georgia Tech (+3.5), Oregon (-14), Arizona State (-14.5), Arizona (-7.5), UCLA (-7.5), Mississippi State (-7), Duke (+13), Georgia (-9), North Texas (-6.5), Wisconsin (0), Iowa (+8), Stanford (-6), Baylor (-17), Oklahoma (+16), Oklahoma State (+1.5), Clemson (+2.5), Houston (+3), Ball State (-9.5) and Florida State (-7.5).

Of course bowl games are strange for a variety of reasons, (hitting the buffet circuit, poor practices, lack of focus…) so there always seems to be a lack of consistency.

We will check back in a few weeks on how well this did.