Thursday, March 25, 2010

March Madness Contest Prize Teaser

Posted by Danny Tarlow
Doug has been brainstorming logo ideas for the eventual official Smell the Data sticker that will go to the winner of the March Madness Predictive Analytics Challenge. I really like some of the directions he's going in. Here's one of my favorites:

Sunday, March 21, 2010

Sweet Sixteen starts Thursday

Posted by Lee
What a crazy week of basketball. Who would have expected Kansas to be knocked out so early? Turns out none of the prediction algorithms had Kansas leaving this early, but a few of them don't have Kansas going to the Final Four -- and that's good news for those competitors!

If your bracket is busted or if you didn't participate yet, it's not too late! You can still submit a Sweet Sixteen bracket to our second chance bracket by the beginning of play Thursday for a chance at a $25 Amazon.com Gift Certificate.

Full contest info: http://blog.smellthedata.com/2010/03/official-march-madness-predictive.html

Saturday, March 20, 2010

Current Standings and Introductions

Posted by Lee
We have eight entries in our inaugural March Madness Predictive Analytics Challenge. The standings after the first two rounds of play look like this:
  1. My Robots Wicked Smaht
  2. ebv
  3. Danny's Dangerous Picks
  4. Hugues
  5. The Pain Machine
  6. FTW
  7. Simple PageRank
  8. BrentsBracket
With the first week of basketball in the tournament over, let's introduce our competitors!

Entry Name: My Robots Wicked Smaht
Team Members: Rolf and Andrew
In their own words: Our backgrounds are more on the human learning side of things, so we took a fairly simple approach to creating a bracket picking robot. Our robot uses a simple regression to identify key variables in order to enhance the RPI rankings.

Rolf has blogged about their entry.

Entry Name: ebv
Team Members: Eric Venner (venner at bcm dot edu)
In his own words: I'm using a very simple model based on PageRank. A loss is treated like a link from the losing team to the winning team, and weighted based on the point in the season at which it was played - later games are weighted higher.

Entry Name: Danny's Dangerous Picks
Team Members: Danny Tarlow (you're reading his blog)
In his own words: I generated the predicted score using my probabilistic matrix factorization model's offensive and defensive rankings to determine each game's winner, like described here: http://blog.smellthedata.com/2010/03/march-madness-2010-offense-and-defense.html

Entry Name: Hugues
Team Members: Hugues Salamin
In his own words: I am currently doing a PhD in Glasgow, Scotland. For the prediction, I use a CRF with one variable (winner) and the features are the some of the states of the team members. I got an accuracy of 0.75 when training on 2006 to 2009 and testing on 2010. I was planning to extend the model (predict overtime and score delta) but did not have enough time. Maybe for the sweet 16 part. The code is in Python and training uses the SciPy LBFGS implementation for the gradient descent.

Name: The Pain Machine
Team Members: Dr. Scott Turner (srt19170 at gmail)
In his own words: My Ph.D. is in Artificial Intelligence from UCLA, where I wrote a program (MINSTREL) to tell stories about King Arthur and his knights as a way to look at creativity and storytelling. For the past twenty years I've worked for the Aerospace Corporation as a software architect and ground system expert for satellite programs.

My approach wasn't particularly sophisticated; it processed the first part of the season to develop a ranking for all the teams, and then did a simple genetic algorithm to evolve an equation to predict outcomes based upon the ranking, RPI, and a few other stats. It was able to correctly predict the outcomes of my test set of games at about 80% (not particularly good, IMO).

Entry Name: FTW
Team Members: Matt Curry (matt at pseudocoder dot com); @mcurry - http://pseudocoder.com
In his own words: A bunch of years ago I wasted hours per day writing programs to predict the outcomes of sporting events, mostly for pretend gambling purposes. My best was a program that could pick select NBA games with a tremendous success rate (focused on home teams that were huge underdogs). I didn't have the testicular fortitude to trust it with more then a few small bets. My program for this contest is awful. I fully expect to get destroyed.

Entry Name: Simple Page Rank Bracket
Team Members: Daniel Mack (dmack at isis dot vanderbilt dot edu) and @manieldack
In his own words: As a first go around, I decided to step back and look at the problem as a network. With teams as nodes, and wins being represented as edges from the team that was beaten to the team that won. This structure actually has some interesting properties, but one of the most fascinating, is that it resembles in some fashion a web infrastructure. Good teams are linked frequently from other teams that are also linked frequently. Using the barest of page rank algorithms, I calculated the teams' ranks and propagated the winner through in the brackets, this meant that when I predict a team to win, the page rank is calculated to take that win into account, and thus teams in the Final Four have been impacted.

Entry Name: BrentsBracket
Team Members: Brent Castle
In his own words: Matrix Completion for Power Rating Differences

Thursday, March 18, 2010

Team "My Robots Wiked Smaht"

Posted by Danny Tarlow
This is a guest post by Rolf, the driving force behind our early co-leader, "My Robots Wiked Smaht". After three games, they are one of only two teams to have a perfect 3 of 3 bracket.

Entry Name: My Robots Wicked Smaht
Team Member Names: Rolf and Andrew
Description: Our backgrounds are more on the human learning side of things, so we took a fairly simple approach to creating a bracket picking robot. Our robot uses a simple regression to identify key variables in order to enhance our variant on RPI rankings.


As an early leader in the March Madness Predictive Analytics Challenge, Danny suggested that I write a guest blog post detailing some of the finer points of my prediction algorithm - "My Robots Wicked Smaht." First off, I must mention that although I am a regular reader of "This Number Crunching Life," my formal mathematics and computer science background is pretty limited. I'm currently doing a PhD in Education, so I've been more focused on human learning than machine learning. But enough chit chat.

My first task in creating this robot was to learn how to use excel. That was the trickiest part, actually. Once I had organized the 2009 data into some statistics that I thought might be relevant, I consulted with my colleague Andrew. He then ran a regression and determined that wins in January and February were more highly correlated with wins in march (our goal) than those from November or December. Taking this information, we created a kind of modified RPI ranking ( had to look this one up as well), with extra weight placed on late season wins in addition to the standard factors of winning percentage, opponent's winning percentage, and opponent's opponent's winning percentage.

I was hoping to also factor in the average team height, but that seemed like a whole lot to do. Next Year!

The Tournament Is Underway!

Posted by Lee
The tournament is in full swing. A few games are done and others have just tipped off! I've made the Yahoo bracket public so that anyone can view it.

Also, if you're still interested in participating, it's not too late! We have the Sweet Sixteen bracket coming up in a week and we will be accepting entries for that bracket through the morning of 3/25.

I'll be posting soon about all our wonderful contestants. And, get ready, because we'll also begin featuring guest posts from the contestants themselves!

Wednesday, March 17, 2010

Tournament Bracket Closing Soon! Deadline Extended

Posted by Lee
I hope you have those predictions ready! We already have 8, count 'em, EIGHT teams participating in the "Tournament" bracket of our inaugural March Madness Predictive Analytics Challenge. I'm really excited to see how everyone does!

Since it's hard to check that the 1am deadline is enforced via the Yahoo bracket, I've extended the deadline to noon ET on March 18th, which is when the Yahoo brackets should close (right before the tipoff of the first game for Florida @ BYU).

I'll be making a post once the competition is under way about who our competitors are and a little bit of information about each competitor

If you have a last-minute submission, please make sure to request your password from me tonight so that I have time to respond in order for you to join the competition.

Tuesday, March 16, 2010

More Data Updates

Posted by Lee
Sorry the player data has been a bit of a mess. Don't worry, all the data you have is technically correct. However, the schema for the 2010 data is a bit different. I have updated the previous blog posts to reflect what the schema actually is.

You can continue to use the existing data you have, assuming you use the schema changes mentioned. Alternatively, I've posted the most up-to-date version of the data here: http://cs.stanford.edu/~lzen/All_Player_Data.zip

This merged data set of the 2006-2010 seasons uses the original schema:
  • ID (GUID)
  • Name
  • Height
  • Position
  • Team
  • Year
  • Class (Freshman, Sophomore, Junior, Senior)
  • Games - the number of games the player participated in
  • Field goals (shots) made, excluding three point shots
  • Field goal attempts, exlcuding three point shots
  • Three point shots made
  • Three point shots attempted
  • Free throws made
  • Free throw attempts
  • Assists
  • Blocks
  • Rebounds
  • Steals

Starter Code

Posted by Danny Tarlow
There's been some demand for this. I leave it to you to decipher the mystery (which means I don't have time to explain it right now):
http://www.cs.toronto.edu/~dtarlow/march_madness_public.tgz

If you improve it in any great ways, though, please let me know.

PS. It's Python code for my march madness model.

Monday, March 15, 2010

Updated Player Data

Posted by Lee
In the original announcement post, I did not have 2010 player data included. I have now included that data in the player data.

Though the data has the same format as the other player data and I tried as hard as possible to match players to players, it's possible that some of the players are matched inaccurately. This might be due to two players with the same name on a team or players who I have assumed transfer schools but maybe actually be another player with the same name appearing on a different team.

In the process of doing some checks on the 2010 data, I realized that I had made a mistake in the hashing of the original data set. Players with the same name, instead of the same hash, were being mapped to each other. I've fixed that and re-uploaded all the player data. Sorry about this and I hope this doesn't severely inconvenience anyone.

The most up-to-date version of the data are available at

The 2010 player data has a slightly different schema (sorry!) It includes three sets of field goal figures -- field goals made and attempted without 3 pointers, field goals made and attempted including 3 pointers, and 3 pointers made and attempted. Also note that the last four columns are in slightly different order.
  • ID (GUID)
  • Name
  • Height
  • Position
  • Team
  • Year
  • Class (Freshman, Sophomore, Junior, Senior)
  • Games - the number of games the player participated in
  • Field goals (shots) made, excluding three point shots
  • Field goal attempts, exlcuding three point shots
  • Field goals (shots) made, including three point shots
  • Field goal attempts, including three point shots
  • Three point shots made
  • Three point shots attempted
  • Free throws made
  • Free throw attempts
  • Rebounds
  • Assists
  • Steals
  • Blocks

March Madness 2010: Offense and Defense Ratings

Posted by Danny Tarlow
I quickly ran my algorithm from last year on this year's score data. See the old post for the details. I ran the exact same 1D model with this year's scores data, which gives offensive and defensive ratings for each team.

How it works:
If you want the details, read the old post. If you just want the quick story, the output here can be viewed a rating of each team's offense (higher is better) and defense (lower is better). It's actually more than that, though. The numbers are calibrated so that you can predict the outcome of a game between team A and team B as follows:
Team A's predicted score: Team A offense rating * Team B defense rating
Team B's predicted score: Team B offense rating * Team A defense rating

For example, if Kansas played Lehigh, the predicted score will be:
Kansas: 10.27 * 9.17 = 94
Lehigh: 8.70 * 7.04 = 61

A real blowout!

I've also included a combined measure (higher is better) at the bottom, which is just the offensive rating divided by the defensive rating for each team.

So have fun with it. Look up a few games from the season and see how well these predictions match the true outcomes. For my real predictions, though, I will most likely be using a higher dimensional version of the model, so don't pin me down to these predictions.

Update: I lied. I actually did just use the 1D version, though there's no good justification for it over other settings of parameters. I didn't get a chance to do proper validation to set latent dimension or regularization strength.

Offenses
1 Villanova (10.6096071747)
2 Providence (10.4961023481)
3 BYU (10.3145090472)
4 Syracuse (10.2897143393)
5 Texas (10.2737615755)
6 Kansas (10.2705742443)
7 Maryland (10.2565209896)
8 Washington (10.2331268368)
9 California (10.2154362322)
10 Duke (10.2054254119)
11 Kansas St. (10.1591288426)
12 Xavier (10.121168382)
13 Virginia Military (10.0635213108)
14 Seton Hall (10.02517276)
15 Kentucky (10.0157560113)
16 Houston (9.98686919786)
17 Vanderbilt (9.90921563852)
18 Mississippi (9.79158837692)
19 Missouri (9.78065560781)
20 Louisville (9.70612426736)
21 Notre Dame (9.68648004228)
22 St. Mary's (9.68403578665)
23 Texas Tech (9.68259249492)
24 Ohio St. (9.66670247978)
25 New Mexico (9.65916226491)
26 Nevada (9.60499808476)
27 North Carolina (9.58002470987)
28 Georgetown (9.54401370366)
29 VCU (9.5113590692)
30 Baylor (9.50523410129)
31 Marshall (9.50445111215)
32 Minnesota (9.50421932536)
33 Michigan St. (9.50198567669)
34 Gonzaga (9.49881517781)
35 New Mexico St. (9.48362123051)
36 Colorado (9.47117793387)
37 Rhode Island (9.4689437856)
38 Arkansas (9.426477063)
39 Wake Forest (9.41800936615)
40 Seattle (9.40062565121)
41 Tennessee (9.39149332818)
42 Auburn (9.39083686601)
43 West Virginia (9.38665180217)
44 Clemson (9.38072138583)
45 Sam Houston St. (9.37071466958)
46 Siena (9.3263895148)
47 Georgia Tech (9.32282841693)
48 Arizona (9.31339708463)
49 Illinois (9.28819011497)
50 Oklahoma St. (9.26270937933)
51 UTEP (9.23343702695)
52 Memphis (9.22475741799)
53 Purdue (9.21034444121)
54 Marquette (9.20800477412)
55 Connecticut (9.18296576043)
56 Lipscomb (9.18236722172)
57 Loyola Marymount (9.16892800575)
58 Iowa St. (9.1571824487)
59 UNLV (9.14354681667)
60 Washington St. (9.13168384574)
61 South Carolina (9.12252944278)
62 Portland St. (9.09746808172)
63 Mississippi St. (9.08765594201)
64 Weber St. (9.0831999882)
65 Oklahoma (9.05438021562)
66 Miami (FL) (9.04895695302)
67 Texas A&M (9.04386975172)
68 Cornell (9.03644880214)
69 Texas St. (9.03095128356)
70 Ohio (9.02215074799)
71 Charlotte (8.99381210537)
72 Virginia Tech (8.99308457727)
73 Charleston (8.956239944)
74 Florida (8.95097964854)
75 Valparaiso (8.92759560363)
76 Oakland (8.9186263927)
77 Oregon (8.91203196461)
78 Duquesne (8.90419433875)
79 Boston Coll. (8.90010038244)
80 Morgan St. (8.8952970081)
81 Northwestern (8.88963818909)
82 Harvard (8.8895080359)
83 Mercer (8.88181308695)
84 San Diego St. (8.85067344293)
85 Stanford (8.84532617832)
86 Louisiana Tech (8.84197368222)
87 San Jose St. (8.82173364911)
88 Tulsa (8.82170452554)
89 Missouri St. (8.81315886007)
90 Cincinnati (8.79493855176)
91 Niagara (8.76603702345)
92 Florida St. (8.76475396565)
93 Boise St. (8.76355302968)
94 Hofstra (8.75730714315)
95 Buffalo (8.74189303958)
96 Murray St. (8.73611793116)
97 Southern Ill. (8.73280668662)
98 Utah St. (8.72402191384)
99 Georgia (8.72398768514)
100 CSU Fullerton (8.72154974619)
101 Wisconsin (8.70891624191)
102 Lehigh (8.70264803831)
103 Creighton (8.6997766943)
104 Appalachian St. (8.68598851192)
105 Butler (8.67851699281)
106 South Florida (8.65946082027)
107 Kent St. (8.65787713043)
108 La Salle (8.65580569834)
109 Indiana (8.65247255758)
110 IUPUI (8.65188290013)
111 Northwestern St. (8.650235705)
112 Alabama (8.64614610947)
113 Tennessee Tech (8.64465023399)
114 Portland (8.63953820513)
115 South Dakota (8.6393257918)
116 Dayton (8.62156605483)
117 Massachusetts (8.62118474604)
118 Pittsburgh (8.61917708428)
119 Florida Atlantic (8.61684915412)
120 Rutgers (8.60939294212)
121 Illinois St. (8.60220828937)
122 South Dakota St. (8.58818128717)
123 St. Joseph's (8.58635760433)
124 Belmont (8.57947885872)
125 St. Bonaventure (8.57627623419)
126 UCLA (8.56735122638)
127 N.C. State (8.56484924061)
128 Arizona St. (8.56228116852)
129 Western Carolina (8.55781508298)
130 Northern Colorado (8.55566403218)
131 Akron (8.55509217116)
132 Boston U. (8.54478891175)
133 Wichita St. (8.54473529915)
134 Geo. Washington (8.52543739356)
135 Long Beach St. (8.52103754504)
136 St. John's (8.51878683872)
137 West. Kentucky (8.50587746016)
138 Penn St. (8.50373836683)
139 UAB (8.4925456171)
140 North Texas (8.49212894777)
141 Bradley (8.49115606127)
142 Wyoming (8.46965791013)
143 CSU Northridge (8.46053573273)
144 Wis. Milwaukee (8.46041313033)
145 Austin Peay (8.44759396845)
146 Towson (8.43043210642)
147 Fairfield (8.42743029488)
148 Davidson (8.42734534402)
149 Rider (8.42711072139)
150 Troy (8.42688530197)
151 Nebraska (8.42014334254)
152 Longwood (8.40286077115)
153 Richmond (8.39797612089)
154 Quinnipiac (8.39744540759)
155 W. Michigan (8.38682961886)
156 Virginia (8.37602397935)
157 TCU (8.37093507936)
158 Old Dominion (8.36246957369)
159 East Carolina (8.35925965558)
160 Idaho (8.35268923895)
161 Michigan (8.3365433049)
162 Sacred Heart (8.32808333199)
163 Lafayette (8.32625599744)
164 Detroit (8.32542281924)
165 Drake (8.31709561953)
166 Indiana St. (8.31341329676)
167 Jacksonville (8.30868848135)
168 Utah (8.30390692076)
169 Colorado St. (8.29295481735)
170 Norfolk St. (8.29219191754)
171 Cal Poly (8.28455631283)
172 Vermont (8.27729930135)
173 Georgia Southern (8.27678648167)
174 Northern Illinois (8.27057671426)
175 San Francisco (8.25424694999)
176 William & Mary (8.25208642378)
177 Cleveland St. (8.23586167401)
178 Wis. Green Bay (8.23082932485)
179 N.C. Asheville (8.22769377309)
180 Wright St. (8.19144625554)
181 UCF (8.18654496997)
182 SE Louisiana (8.18149519102)
183 Navy (8.16042522414)
184 Oral Roberts (8.15086234093)
185 Houston Baptist (8.14243450471)
186 N.C. Wilmington (8.1400577915)
187 Montana (8.13988183176)
188 TX Arlington (8.12712309124)
189 Morehead St. (8.11426184637)
190 Long Island (8.10772355358)
191 Eastern Kentucky (8.106981612)
192 E. Tennessee St. (8.09684175744)
193 James Madison (8.08667648927)
194 UCSB (8.07427827314)
195 UC Davis (8.07190032751)
196 SMU (8.06687909503)
197 North Dakota St. (8.06186227025)
198 Canisius (8.06106458325)
199 Wofford (8.05958276878)
200 Hawaii (8.05942849129)
201 Northern Arizona (8.05878912917)
202 Centenary (8.04635244693)
203 Brown (8.03501954313)
204 Temple (8.03261201059)
205 Florida Intl. (8.02774849995)
206 Fresno St. (8.02272203563)
207 Holy Cross (8.01169361973)
208 Arkansas St. (8.00963777856)
209 Chattanooga (8.00745524164)
210 Pepperdine (8.00657585665)
211 George Mason (8.00084105845)
212 Tennessee St. (7.98946385038)
213 Furman (7.98087495336)
214 Northeastern (7.97733855598)
215 Radford (7.96513211378)
216 AR Little Rock (7.96289967102)
217 East. Michigan (7.95995718537)
218 LA Lafayette (7.95789934659)
219 Iona (7.95702027986)
220 Jacksonville St. (7.95636634842)
221 S. Carolina St. (7.95398725818)
222 UTSA (7.95272384668)
223 Robert Morris (7.94917766814)
224 Youngstown St. (7.94838383395)
225 Stony Brook (7.9417013915)
226 Idaho St. (7.93729628843)
227 Delaware (7.91199423416)
228 Iowa (7.9088096721)
229 Middle Tenn. St. (7.8810085041)
230 Fordham (7.87892005172)
231 Loyola (MD) (7.87272847122)
232 Texas A&M C.C. (7.86758877119)
233 Manhattan (7.86516248955)
234 Lamar (7.86410573215)
235 Evansville (7.86311218223)
236 Coastal Carolina (7.85161633959)
237 Colgate (7.84802475514)
238 Northern Iowa (7.84218230398)
239 Montana St. (7.83560428312)
240 LA Monroe (7.81963610353)
241 Loyola Chicago (7.81714457956)
242 Eastern Wash. (7.81449468244)
243 CSU Bakersfield (7.80991967038)
244 Santa Clara (7.79518235909)
245 Southern Miss (7.78346772869)
246 Elon (7.77743909896)
247 Eastern Illinois (7.7760680638)
248 McNeese St. (7.77366837427)
249 IPFW (7.76983760159)
250 Campbell (7.76633283401)
251 UC Irvine (7.764649057)
252 Drexel (7.75562001336)
253 N.C. Greensboro (7.74896701072)
254 Tulane (7.74292641597)
255 Ill. Chicago (7.73972600341)
256 Saint Louis (7.73412222778)
257 Yale (7.72699256894)
258 N.C. A&T (7.72138799967)
259 DePaul (7.71328141092)
260 Rice (7.70807536381)
261 South Alabama (7.70764782786)
262 LSU (7.69950003223)
263 Cent. Michigan (7.69586780381)
264 Pacific (7.68724054707)
265 Mo. Kansas City (7.67736402677)
266 High Point (7.66522423916)
267 Bucknell (7.64268499918)
268 Miami (OH) (7.63081561921)
269 Mount St. Mary's (7.63048414746)
270 Denver (7.62180749134)
271 Bowling Green (7.61975926885)
272 Fla Gulf Coast (7.6131476021)
273 Kennesaw St. (7.59963686767)
274 Gardner-Webb (7.58850969258)
275 San Diego (7.56885211032)
276 USC (7.54594795722)
277 Texas Southern (7.52570746444)
278 Georgia St. (7.51597035452)
279 Hampton (7.514234785)
280 Pennsylvania (7.50925883307)
281 Tenn-Martin (7.50182722172)
282 UMBC (7.4873574332)
283 Liberty (7.47148187962)
284 Florida A&M (7.4349682354)
285 Sacramento St. (7.42424021272)
286 SE Missouri St. (7.42136521258)
287 Nicholls St. (7.41215783519)
288 Albany (7.40098997539)
289 AR Pine Bluff (7.39173225453)
290 Oregon St. (7.38679861661)
291 Monmouth (7.37489634877)
292 Fair. Dickinson (7.3356201816)
293 Binghamton (7.33260132598)
294 St. Peter's (7.31955349901)
295 UC Riverside (7.29725090775)
296 St. Francis (PA) (7.28129141613)
297 Southern Utah (7.27886711148)
298 Maine (7.26674029118)
299 Ball St. (7.25577667672)
300 Charleston S. (7.24025306611)
301 Hartford (7.23842840017)
302 South Carolina Upstate (7.23537470181)
303 Stephen F. Austin (7.22833231784)
304 Miss. Valley St. (7.22022084486)
305 Alabama St. (7.21965990029)
306 American (7.1989899691)
307 Citadel (7.19776887055)
308 Alabama A&M (7.15863976235)
309 Central Arkansas (7.14821031528)
310 SIU Edwardsville (7.14017516397)
311 MD Eastern Shore (7.13988606988)
312 Texas Pan Amer. (7.1395551766)
313 Grambling St. (7.10377729696)
314 New Hampshire (7.10016962468)
315 North Dakota (7.08799906924)
316 Central Conn. St. (7.08791824285)
317 Columbia (7.06121962841)
318 Winthrop (7.05229738658)
319 Stetson (7.05058324169)
320 Jackson St. (7.03502114953)
321 Alcorn St. (7.03412739453)
322 North Florida (7.02615483067)
323 Wagner (7.01714656847)
324 Coppin St. (6.982664216)
325 St. Francis (NY) (6.93553118848)
326 New Orleans (6.90034577172)
327 Marist (6.88547344299)
328 Presbyterian (6.88275297017)
329 Princeton (6.87238530992)
330 Chicago St. (6.86467540963)
331 Air Force (6.86156157659)
332 Prairie View A&M (6.85885459159)
333 Howard (6.84415147973)
334 Samford (6.83629259329)
335 Winston-Salem (6.83109766563)
336 N.C. Central (6.80707521442)
337 Delaware St. (6.76011692275)
338 Army (6.75601767785)
339 Southern (6.75205622186)
340 Utah Valley (6.74731423289)
341 Bethune-Cookman (6.74090336909)
342 Savannah St. (6.70615731815)
343 Western Ill. (6.67671544068)
344 N.J.I.T. (6.64872787082)
345 Toledo (6.53500233042)
346 Dartmouth (6.29772838065)
347 Bryant (5.95480032938)


Defenses
1 Wisconsin (6.32057065963)
2 Temple (6.44605516553)
3 USC (6.44692532578)
4 Northern Iowa (6.45024371697)
5 Princeton (6.71832788491)
6 Arizona St. (6.75973258441)
7 Old Dominion (6.81807850841)
8 Florida St. (6.8242040874)
9 Pittsburgh (6.87076188071)
10 Duke (6.89518504414)
11 Oregon St. (6.91918358722)
12 Purdue (6.92284973952)
13 Utah St. (6.9599625176)
14 West Virginia (7.00124414689)
15 Michigan (7.02679704679)
16 Saint Louis (7.03835353947)
17 Kansas (7.04030322554)
18 Southern Miss (7.07228715031)
19 Ohio St. (7.07884259933)
20 Butler (7.08964174039)
21 Northeastern (7.09629072959)
22 Texas A&M (7.12530348174)
23 Marquette (7.1480705149)
24 Richmond (7.15339157418)
25 Dayton (7.15407151931)
26 Virginia (7.18353766063)
27 Pacific (7.19891611498)
28 Georgetown (7.21675750724)
29 Clemson (7.23483082807)
30 UAB (7.23922253822)
31 Tennessee (7.27184878438)
32 Stephen F. Austin (7.28604960942)
33 San Diego St. (7.29372227606)
34 Michigan St. (7.30339210826)
35 Alabama (7.30903904967)
36 Air Force (7.31649872138)
37 Cincinnati (7.31906298288)
38 N.C. State (7.32095956951)
39 Baylor (7.32165324762)
40 Wright St. (7.33416688762)
41 Minnesota (7.3365113623)
42 Kentucky (7.34945752954)
43 St. John's (7.36255627257)
44 Connecticut (7.36459502867)
45 Mississippi St. (7.39821704997)
46 Florida (7.40017780545)
47 Montana (7.40285526754)
48 UNLV (7.40595866694)
49 St. Peter's (7.41269436385)
50 Georgia Tech (7.41589079417)
51 Syracuse (7.44868324078)
52 South Florida (7.45089659866)
53 Wichita St. (7.46348451676)
54 Missouri (7.4729284621)
55 LSU (7.47573255956)
56 Miami (OH) (7.47900971338)
57 Wofford (7.48742664695)
58 Iona (7.48867594252)
59 Penn St. (7.49814888772)
60 Citadel (7.49830311569)
61 Nebraska (7.49904269456)
62 Miami (FL) (7.5006777214)
63 UTEP (7.50283671305)
64 DePaul (7.5052997362)
65 Memphis (7.51440983833)
66 Virginia Tech (7.51613488033)
67 Iowa (7.56420047514)
68 Kansas St. (7.56509818048)
69 Murray St. (7.58529581953)
70 Delaware St. (7.58895799063)
71 Boston Coll. (7.59333620215)
72 SMU (7.59480253644)
73 Oklahoma St. (7.60612280358)
74 Samford (7.62059057461)
75 Drexel (7.62293362159)
76 Tulsa (7.62459657347)
77 Illinois (7.62532578585)
78 Portland (7.63482056947)
79 Morehead St. (7.66066378862)
80 Northwestern (7.67233635766)
81 Utah (7.69836822656)
82 Coastal Carolina (7.7033515457)
83 San Diego (7.70396346086)
84 William & Mary (7.72423302282)
85 Winthrop (7.72487329491)
86 UCLA (7.72648162538)
87 Army (7.72722115146)
88 Louisville (7.73382821831)
89 Georgia (7.74376644246)
90 Western Ill. (7.74385661906)
91 Illinois St. (7.76426797661)
92 Maryland (7.77528026697)
93 Wake Forest (7.79036526764)
94 Denver (7.79401961232)
95 Ball St. (7.80074237613)
96 Notre Dame (7.80089366125)
97 Georgia St. (7.80197909369)
98 Gonzaga (7.80662609984)
99 BYU (7.81416990538)
100 California (7.82606819299)
101 Vanderbilt (7.83014433876)
102 Iowa St. (7.83028613886)
103 Texas (7.8335621828)
104 Fresno St. (7.84611580197)
105 St. Mary's (7.84645378187)
106 New Mexico (7.8479444928)
107 Middle Tenn. St. (7.84942637318)
108 UC Riverside (7.85475895144)
109 Bethune-Cookman (7.87396971794)
110 Kent St. (7.87629754562)
111 George Mason (7.90025735892)
112 Campbell (7.91598773648)
113 Colorado St. (7.93210203831)
114 Missouri St. (7.94094032558)
115 Tulane (7.95101601543)
116 Mississippi (7.95367851061)
117 Washington (7.95432647712)
118 Stanford (7.95720907604)
119 South Carolina (7.96277587671)
120 UCSB (7.966640255)
121 VCU (7.96746197122)
122 Maine (7.98920434411)
123 Bradley (7.98920860735)
124 North Carolina (7.99006864652)
125 Oral Roberts (7.99080464416)
126 UCF (8.00605473831)
127 Mount St. Mary's (8.01404551663)
128 Xavier (8.02351076537)
129 Texas A&M C.C. (8.02446615457)
130 Savannah St. (8.04569467886)
131 Cent. Michigan (8.05402586387)
132 West. Kentucky (8.05887288397)
133 Geo. Washington (8.06038065862)
134 Indiana St. (8.07246235979)
135 New Hampshire (8.07774962112)
136 Akron (8.09083851123)
137 Oklahoma (8.09402823024)
138 North Florida (8.10379974475)
139 Cleveland St. (8.10389606057)
140 E. Tennessee St. (8.10471837755)
141 Siena (8.10595282816)
142 Detroit (8.11149093745)
143 Arizona (8.11247555657)
144 Cornell (8.12109349834)
145 Oregon (8.12499261218)
146 Bowling Green (8.13087582324)
147 Montana St. (8.14627271973)
148 Fairfield (8.15012446845)
149 Villanova (8.17587846299)
150 Idaho (8.18000737199)
151 Creighton (8.18360558613)
152 Wis. Green Bay (8.19445627815)
153 Southern Ill. (8.1950580689)
154 Hampton (8.20326126486)
155 Toledo (8.21302925358)
156 IUPUI (8.21576028368)
157 Manhattan (8.22130057055)
158 Rice (8.22253013708)
159 Washington St. (8.22288790194)
160 UTSA (8.22574472187)
161 Eastern Kentucky (8.23109725059)
162 Drake (8.23649402409)
163 Rhode Island (8.23982869893)
164 TCU (8.24110941024)
165 Indiana (8.24621895551)
166 Marshall (8.24996953102)
167 Rutgers (8.2538701799)
168 Eastern Illinois (8.25416720774)
169 Northern Colorado (8.2630252777)
170 Loyola (MD) (8.27909029889)
171 New Orleans (8.2793863026)
172 Louisiana Tech (8.28055379758)
173 Wis. Milwaukee (8.28255600529)
174 Hofstra (8.28449899266)
175 Stony Brook (8.28483270871)
176 Charlotte (8.3041660791)
177 Loyola Chicago (8.31062184407)
178 East. Michigan (8.31182296875)
179 St. Bonaventure (8.31310274207)
180 AR Pine Bluff (8.32110536489)
181 Columbia (8.3272917188)
182 Vermont (8.33870032527)
183 Evansville (8.34641283057)
184 LA Lafayette (8.36296635221)
185 Robert Morris (8.36585424925)
186 Santa Clara (8.37058557082)
187 Colorado (8.38106259999)
188 Arkansas St. (8.38127263458)
189 Ohio (8.39718772894)
190 South Carolina Upstate (8.4007082262)
191 Dartmouth (8.40335946266)
192 Utah Valley (8.40337517228)
193 W. Michigan (8.40440404052)
194 Jacksonville (8.40782902276)
195 Hawaii (8.41135487474)
196 American (8.42065268017)
197 San Francisco (8.42601745261)
198 Harvard (8.42749126568)
199 St. Francis (NY) (8.44054055968)
200 Seton Hall (8.44771710686)
201 Texas Tech (8.46818959437)
202 Appalachian St. (8.47215601275)
203 Long Beach St. (8.47595388304)
204 Boston U. (8.48855576781)
205 Central Arkansas (8.48921146639)
206 South Alabama (8.49126224194)
207 Belmont (8.49141453145)
208 Arkansas (8.49621673644)
209 Binghamton (8.50035868758)
210 UC Irvine (8.5006142522)
211 Canisius (8.50775907628)
212 Prairie View A&M (8.51091532265)
213 IPFW (8.51524783412)
214 N.C. Wilmington (8.5152581359)
215 James Madison (8.52000467171)
216 Auburn (8.52461452915)
217 Jackson St. (8.52910457893)
218 Weber St. (8.53362984696)
219 Niagara (8.53659694399)
220 Radford (8.54006564549)
221 La Salle (8.54693640652)
222 Nicholls St. (8.54913856799)
223 SE Louisiana (8.56077726702)
224 Duquesne (8.59773053405)
225 Delaware (8.60280698837)
226 Ill. Chicago (8.60808215306)
227 Troy (8.60982491205)
228 Bucknell (8.61386897796)
229 Monmouth (8.62278504593)
230 Charleston S. (8.62680629702)
231 North Texas (8.63825319561)
232 Boise St. (8.64434713111)
233 Western Carolina (8.64809265887)
234 Marist (8.65338106147)
235 Nevada (8.65608723751)
236 Oakland (8.65642781706)
237 Kennesaw St. (8.67869116985)
238 N.C. Greensboro (8.67981044751)
239 Massachusetts (8.67988477401)
240 Albany (8.69120190998)
241 Rider (8.69757385785)
242 Youngstown St. (8.70061407371)
243 Central Conn. St. (8.70393489342)
244 Quinnipiac (8.70612999176)
245 Texas Southern (8.72030476475)
246 Davidson (8.73101338061)
247 Wyoming (8.7360652008)
248 St. Joseph's (8.74362472809)
249 Coppin St. (8.75493806185)
250 Furman (8.75980614742)
251 Sacramento St. (8.77326310285)
252 Bryant (8.78048912511)
253 Austin Peay (8.78487613032)
254 Alabama St. (8.79203844695)
255 Liberty (8.79750178724)
256 UC Davis (8.79994976191)
257 Pepperdine (8.80675284012)
258 Yale (8.81463822777)
259 Northern Arizona (8.8327407083)
260 North Dakota St. (8.83329331039)
261 Houston (8.86681547748)
262 N.J.I.T. (8.89190042364)
263 Elon (8.89508768935)
264 San Jose St. (8.90228645909)
265 Holy Cross (8.9102852514)
266 Buffalo (8.91496425265)
267 Jacksonville St. (8.91622612699)
268 East Carolina (8.92111643715)
269 Sam Houston St. (8.92669268469)
270 SE Missouri St. (8.92849481542)
271 Winston-Salem (8.93934219245)
272 Hartford (8.94077611862)
273 Presbyterian (8.95632457717)
274 Florida Atlantic (8.96338451873)
275 LA Monroe (8.97012356545)
276 Chattanooga (8.98154423833)
277 McNeese St. (8.98263439309)
278 Eastern Wash. (8.99090298468)
279 Charleston (8.99428844871)
280 Stetson (8.99529851335)
281 Mo. Kansas City (9.0006530765)
282 Morgan St. (9.00163300792)
283 AR Little Rock (9.00213672152)
284 TX Arlington (9.00256935542)
285 Tennessee St. (9.01003374672)
286 Lamar (9.01530528868)
287 Pennsylvania (9.01734872767)
288 Loyola Marymount (9.01919725003)
289 Fla Gulf Coast (9.02014638581)
290 CSU Bakersfield (9.02103555006)
291 MD Eastern Shore (9.04143668998)
292 New Mexico St. (9.05158842458)
293 N.C. Central (9.05196304571)
294 Howard (9.05294825394)
295 S. Carolina St. (9.05605147407)
296 Texas Pan Amer. (9.05996521795)
297 CSU Northridge (9.06285169277)
298 Wagner (9.06589569738)
299 St. Francis (PA) (9.06639894959)
300 Colgate (9.07605299997)
301 Northern Illinois (9.08285569398)
302 South Dakota St. (9.09169089921)
303 Fair. Dickinson (9.09364072219)
304 Idaho St. (9.10398788438)
305 Cal Poly (9.11776657342)
306 CSU Fullerton (9.12936899069)
307 Southern Utah (9.1310335987)
308 N.C. A&T (9.15651666505)
309 Lehigh (9.17181198377)
310 Valparaiso (9.19006331902)
311 Chicago St. (9.22816567117)
312 High Point (9.2314108556)
313 Lafayette (9.25065942627)
314 North Dakota (9.27399334025)
315 Providence (9.27454860001)
316 Brown (9.2806385682)
317 SIU Edwardsville (9.33791078455)
318 Towson (9.34944644617)
319 Fordham (9.35448135911)
320 Lipscomb (9.38079698141)
321 Mercer (9.41301764102)
322 Portland St. (9.43760160698)
323 Long Island (9.4473457865)
324 UMBC (9.45742098012)
325 Alabama A&M (9.48291716302)
326 Tenn-Martin (9.48881486106)
327 Florida Intl. (9.48960691055)
328 Navy (9.52218904341)
329 South Dakota (9.52469625509)
330 Miss. Valley St. (9.5409902926)
331 Southern (9.57825041251)
332 Florida A&M (9.59913215592)
333 Sacred Heart (9.62362981881)
334 Georgia Southern (9.69349029777)
335 Tennessee Tech (9.70248177492)
336 Norfolk St. (9.7145731196)
337 Grambling St. (9.7645185558)
338 N.C. Asheville (9.78288772817)
339 Seattle (9.78365887659)
340 Gardner-Webb (9.80860538609)
341 Texas St. (9.85033995182)
342 Centenary (9.85333899769)
343 Longwood (10.0855988691)
344 Northwestern St. (10.1669651218)
345 Houston Baptist (10.2985052311)
346 Alcorn St. (10.6680173801)
347 Virginia Military (12.5710938304)

Combined (Offensive Rating divided by Defensive Rating)
1 Duke (1.48007999011)
2 Kansas (1.45882555272)
3 Syracuse (1.38141386962)
4 Wisconsin (1.37786866264)
5 Ohio St. (1.3655766948)
6 Kentucky (1.36278847398)
7 Kansas St. (1.34289451376)
8 West Virginia (1.34071196565)
9 Purdue (1.33042674444)
10 Georgetown (1.32247947836)
11 BYU (1.31997501617)
12 Maryland (1.31911913621)
13 Texas (1.3115057155)
14 Missouri (1.30881162016)
15 California (1.30530887034)
16 Michigan St. (1.30103731743)
17 Baylor (1.29823603766)
18 Villanova (1.29767183095)
19 Clemson (1.29660549206)
20 Minnesota (1.29546849395)
21 Tennessee (1.29148633403)
22 Marquette (1.28818046142)
23 Washington (1.28648564604)
24 Florida St. (1.28436281409)
25 Texas A&M (1.26926099006)
26 Arizona St. (1.2666597475)
27 Vanderbilt (1.26552145271)
28 Xavier (1.26143887358)
29 Georgia Tech (1.25714208524)
30 Louisville (1.25502196239)
31 Pittsburgh (1.25447180879)
32 Utah St. (1.25345817478)
33 Connecticut (1.24690709057)
34 Temple (1.24612833808)
35 Notre Dame (1.24171415006)
36 UNLV (1.23462028724)
37 St. Mary's (1.23419267555)
38 Mississippi (1.23107671046)
39 New Mexico (1.23078881021)
40 UTEP (1.23065946656)
41 Mississippi St. (1.22835757327)
42 Memphis (1.22760903603)
43 Old Dominion (1.22651412174)
44 Butler (1.22411220631)
45 Illinois (1.21807125044)
46 Oklahoma St. (1.21779645406)
47 Gonzaga (1.21676317737)
48 Northern Iowa (1.21579627811)
49 San Diego St. (1.21346455321)
50 Florida (1.20956278131)
51 Wake Forest (1.20893039576)
52 Miami (FL) (1.20641857831)
53 Dayton (1.20512718269)
54 Cincinnati (1.20164815801)
55 North Carolina (1.19899153983)
56 Virginia Tech (1.19650388404)
57 VCU (1.19377527041)
58 Seton Hall (1.1867315907)
59 Michigan (1.18639306776)
60 Alabama (1.18293883104)
61 Richmond (1.17398523956)
62 UAB (1.17312951388)
63 Boston Coll. (1.17209354959)
64 USC (1.17047236875)
65 N.C. State (1.16990800991)
66 Iowa St. (1.16945693763)
67 Virginia (1.16600265427)
68 South Florida (1.16220386441)
69 Northwestern (1.15866116587)
70 St. John's (1.15704200054)
71 Tulsa (1.15700607115)
72 Marshall (1.15205893506)
73 Murray St. (1.1517174991)
74 Siena (1.15056054637)
75 Rhode Island (1.1491675533)
76 Arizona (1.14803391637)
77 South Carolina (1.145646893)
78 Wichita St. (1.14487211435)
79 Texas Tech (1.14340761824)
80 Penn St. (1.13411169799)
81 Providence (1.13171031829)
82 Portland (1.13159675811)
83 Colorado (1.13006886906)
84 Georgia (1.12658197403)
85 Houston (1.12631972812)
86 Northeastern (1.12415610633)
87 Nebraska (1.12282909773)
88 Oklahoma (1.11864944847)
89 Wright St. (1.11688844569)
90 Cornell (1.11271330689)
91 Stanford (1.1116116334)
92 Washington St. (1.11052028857)
93 Missouri St. (1.1098381928)
94 Nevada (1.10962353096)
95 Arkansas (1.10949112475)
96 UCLA (1.10882956069)
97 Illinois St. (1.10792264194)
98 Auburn (1.10161425293)
99 Southern Miss (1.10055878152)
100 Montana (1.09955976952)
101 Kent St. (1.09923185104)
102 Saint Louis (1.09885389877)
103 Oregon (1.09686646992)
104 Charlotte (1.0830481977)
105 Utah (1.07865805796)
106 Wofford (1.0764155896)
107 Ohio (1.07442527656)
108 William & Mary (1.06833732222)
109 Pacific (1.06783304935)
110 Louisiana Tech (1.06779979919)
111 Oregon St. (1.0675823995)
112 Southern Ill. (1.06561864641)
113 Weber St. (1.06440051316)
114 Creighton (1.06307380075)
115 Bradley (1.06282818219)
116 Iona (1.0625403397)
117 SMU (1.06215784496)
118 Morehead St. (1.05921132558)
119 Geo. Washington (1.05769662186)
120 Akron (1.05738016638)
121 Hofstra (1.0570714235)
122 West. Kentucky (1.05546738144)
123 Harvard (1.05482257479)
124 IUPUI (1.05308365889)
125 Sam Houston St. (1.04974092876)
126 Indiana (1.04926543962)
127 New Mexico St. (1.04773005418)
128 Iowa (1.0455579143)
129 Colorado St. (1.04549270512)
130 Rutgers (1.04307346184)
131 Duquesne (1.03564473247)
132 Northern Colorado (1.0354154495)
133 Fairfield (1.03402473514)
134 St. Bonaventure (1.03165767347)
135 Oakland (1.03028946595)
136 LSU (1.02993251442)
137 Indiana St. (1.02984850548)
138 DePaul (1.02771130828)
139 Niagara (1.02687723)
140 Detroit (1.02637392847)
141 Appalachian St. (1.02523944305)
142 Princeton (1.02293091788)
143 UCF (1.02254421654)
144 Fresno St. (1.02250874676)
145 Wis. Milwaukee (1.02147370026)
146 Idaho (1.02111023366)
147 Miami (OH) (1.02029759442)
148 Oral Roberts (1.02003023524)
149 Coastal Carolina (1.01924679057)
150 Drexel (1.01740621109)
151 Loyola Marymount (1.01660133952)
152 Cleveland St. (1.01628421841)
153 TCU (1.01575342137)
154 Boise St. (1.01379004068)
155 UCSB (1.01351109309)
156 La Salle (1.01273781466)
157 George Mason (1.0127316991)
158 Belmont (1.01037098436)
159 Drake (1.00978591075)
160 Boston U. (1.0066245832)
161 Long Beach St. (1.00531900747)
162 Wis. Green Bay (1.00443873827)
163 Middle Tenn. St. (1.0040234954)
164 E. Tennessee St. (0.999028143885)
165 W. Michigan (0.997908903288)
166 Charleston (0.995769703748)
167 Massachusetts (0.993237234191)
168 Vermont (0.992636619434)
169 Stephen F. Austin (0.992078383393)
170 San Jose St. (0.990951447097)
171 Western Carolina (0.989560984201)
172 Jacksonville (0.988208544543)
173 Morgan St. (0.988187032316)
174 St. Peter's (0.987434951414)
175 Eastern Kentucky (0.984921130827)
176 North Texas (0.98308405131)
177 San Diego (0.982462098732)
178 St. Joseph's (0.982013509425)
179 Campbell (0.981094601526)
180 Buffalo (0.980586437795)
181 Texas A&M C.C. (0.980450115889)
182 San Francisco (0.979614271679)
183 Lipscomb (0.978847238664)
184 Troy (0.978752226445)
185 Denver (0.977904582032)
186 Tulane (0.973828552343)
187 Valparaiso (0.971440053644)
188 Wyoming (0.969504887549)
189 Rider (0.968903611411)
190 UTSA (0.966808977859)
191 Davidson (0.965219611591)
192 Quinnipiac (0.9645439955)
193 Portland St. (0.963959749582)
194 Georgia St. (0.963341514284)
195 Montana St. (0.961863732373)
196 Austin Peay (0.961606497705)
197 Florida Atlantic (0.961338781808)
198 Seattle (0.960849695374)
199 Citadel (0.959919699097)
200 Stony Brook (0.958583193014)
201 Hawaii (0.958160559305)
202 East. Michigan (0.957666833774)
203 Manhattan (0.956681053327)
204 N.C. Wilmington (0.955937877818)
205 SE Louisiana (0.955695369221)
206 Arkansas St. (0.955658899045)
207 Cent. Michigan (0.955530555015)
208 CSU Fullerton (0.955328868302)
209 Mount St. Mary's (0.952138858162)
210 LA Lafayette (0.951564195219)
211 Loyola (MD) (0.950917091975)
212 Robert Morris (0.950193181867)
213 James Madison (0.949139912578)
214 Lehigh (0.948847191123)
215 Canisius (0.947495634393)
216 South Dakota St. (0.944618705407)
217 Mercer (0.943567028733)
218 Evansville (0.942094806697)
219 Eastern Illinois (0.942077846026)
220 Loyola Chicago (0.940620897717)
221 Air Force (0.937820375277)
222 Rice (0.937433519282)
223 Bowling Green (0.937138806999)
224 East Carolina (0.937019454289)
225 CSU Northridge (0.933540128377)
226 Radford (0.93267808989)
227 Santa Clara (0.93125890574)
228 Ball St. (0.930139251736)
229 UC Riverside (0.929022895911)
230 Delaware (0.919699145274)
231 UC Davis (0.917266637412)
232 Texas St. (0.916816204084)
233 Hampton (0.916005786284)
234 Youngstown St. (0.913542856471)
235 UC Irvine (0.913422115937)
236 Winthrop (0.912933729441)
237 North Dakota St. (0.912667788442)
238 IPFW (0.912461710211)
239 Northern Arizona (0.912376961501)
240 Furman (0.91107894616)
241 Northern Illinois (0.910570088628)
242 Maine (0.909569961938)
243 Pepperdine (0.909140520009)
244 Cal Poly (0.908616846694)
245 South Alabama (0.907715202787)
246 South Dakota (0.907044756119)
247 TX Arlington (0.902755954482)
248 Towson (0.901703876797)
249 Lafayette (0.900071618008)
250 Holy Cross (0.899151193669)
251 Ill. Chicago (0.899123157259)
252 Samford (0.89708173223)
253 N.C. Greensboro (0.892757630778)
254 Jacksonville St. (0.89234685562)
255 Chattanooga (0.891545488077)
256 Tennessee Tech (0.890973096836)
257 Delaware St. (0.890783284227)
258 AR Pine Bluff (0.888311339707)
259 Bucknell (0.887253453558)
260 Tennessee St. (0.886729625546)
261 AR Little Rock (0.884556624427)
262 New Hampshire (0.878978670758)
263 S. Carolina St. (0.878306321575)
264 Yale (0.876609155053)
265 Kennesaw St. (0.875666240328)
266 Elon (0.874352155997)
267 Army (0.874314005699)
268 Lamar (0.872306092843)
269 Idaho St. (0.871848292114)
270 LA Monroe (0.871742295017)
271 Eastern Wash. (0.869155711697)
272 North Florida (0.867019799597)
273 Nicholls St. (0.867006397925)
274 Brown (0.865783047587)
275 CSU Bakersfield (0.865745360057)
276 McNeese St. (0.865410750797)
277 Sacred Heart (0.865378603374)
278 Colgate (0.864695782975)
279 Texas Southern (0.863009684577)
280 Binghamton (0.862622578114)
281 Western Ill. (0.86219512694)
282 South Carolina Upstate (0.861281514247)
283 Long Island (0.858201206647)
284 Navy (0.856990465842)
285 Bethune-Cookman (0.856099732481)
286 Monmouth (0.855280087523)
287 American (0.854920662629)
288 Georgia Southern (0.853849978432)
289 Norfolk St. (0.853582737548)
290 Mo. Kansas City (0.852978551836)
291 Albany (0.851549653552)
292 Northwestern St. (0.850817879413)
293 Liberty (0.849273130066)
294 Columbia (0.847961121918)
295 Sacramento St. (0.846234761877)
296 Florida Intl. (0.845951636946)
297 Fla Gulf Coast (0.844015970082)
298 N.C. A&T (0.843266962986)
299 Fordham (0.842261558846)
300 Central Arkansas (0.842034662887)
301 N.C. Asheville (0.841029152302)
302 Charleston S. (0.839273865302)
303 Savannah St. (0.83350880015)
304 New Orleans (0.833436865913)
305 Longwood (0.833154369927)
306 Pennsylvania (0.832756840159)
307 SE Missouri St. (0.831200036065)
308 High Point (0.830341576067)
309 Jackson St. (0.824825289035)
310 St. Francis (NY) (0.82169277423)
311 Alabama St. (0.821158818157)
312 Centenary (0.816611754535)
313 Central Conn. St. (0.814334933526)
314 Hartford (0.809597321769)
315 Fair. Dickinson (0.806675830473)
316 Prairie View A&M (0.805889182488)
317 St. Francis (PA) (0.80310732592)
318 Utah Valley (0.802929072494)
319 Virginia Military (0.800528692777)
320 Coppin St. (0.797568659729)
321 Southern Utah (0.79715697383)
322 Marist (0.79569747294)
323 Toledo (0.795687209756)
324 UMBC (0.791691249542)
325 Houston Baptist (0.790642362361)
326 Tenn-Martin (0.790596858677)
327 MD Eastern Shore (0.78968490459)
328 Texas Pan Amer. (0.788033398015)
329 Stetson (0.783807589179)
330 Florida A&M (0.774545877131)
331 Wagner (0.774015806347)
332 Gardner-Webb (0.773658373834)
333 Presbyterian (0.768479626979)
334 SIU Edwardsville (0.764643754766)
335 North Dakota (0.76428770317)
336 Winston-Salem (0.764161111474)
337 Miss. Valley St. (0.756758011846)
338 Howard (0.756013542522)
339 Alabama A&M (0.754898481056)
340 N.C. Central (0.751999889974)
341 Dartmouth (0.749429845128)
342 N.J.I.T. (0.747728556783)
343 Chicago St. (0.743882983276)
344 Grambling St. (0.727509221921)
345 Southern (0.704936280747)
346 Bryant (0.678185491097)
347 Alcorn St. (0.65936594813)

Sunday, March 14, 2010

March Madness Bracket Predictions

Posted by Lee
With the bracket released and data available from past seasons, it's time to get those algorithms running!

I've created a "pace" bracket within our tournament bracket hosted on Yahoo. The pace bracket will simply pick the higher seeds over the lower seeds. The #1 seeds are actually ranked, too, so the pace bracket will pick based upon the overall ranking. Hopefully, the best algorithm in the contest will do better than the pace bracket. Submissions should be made to the Yahoo site; please see the contest details for instructions.


I'm looking forward to seeing everyone's algorithms at work and the resulting bracket submissions. Remember, there are prizes! If you need any help, have questions, etc. please use the comments, and I'll try to reply.

Selection Sunday Complete! 2010 NCAA College Men's Basketball Bracket Released

Posted by Lee
March Madness begins! The bracket is out!


(click on the image for full size)

Image courtesy ESPN.com

Related Posts:

Saturday, March 13, 2010

Official March Madness Predictive Analytics Challenge Announcement

Posted by Lee
NEW UPDATE Please see the 2011 March Madness Algorithm Prediction Contest.

OLD UPDATE Player data has been updated to fix a bug. 2010 player data is also now included. Please refer to the more recent posts for the latest data.


Welcome to the inaugural March Madness Predictive Analytics Challenge! I'm very excited about this event and I hope you are, too! We already have some prizes lined up (keep reading for details) and we're hoping to get some more prizes set up.

As Danny said in his previous post, I'll be acting as the commissioner for this contest. In this post, I'll be explaining the format for the challenge, rules, available data, and prizes.

Background

Most readers of this blog are probably familiar with the general idea of what this contest is about. In case you aren't a frequent reader or a fan of college basketball, this section will serve as a brief introduction. Tomorrow is "Selection Sunday" where the teams for the NCAA College Basketball tournament will be selected. In total, there will be 65 teams with 2 teams playing a "play-in" game to determine the field of 64. These 64 teams are then pit against each other in a bracket with one national champion emerging at the winner. Every year, millions of people fill in their predictions of who will be the winners and losers of the games. People participate in leagues or pools with other people to see who has the best bracket. We would like YOU to participate in our algorithm-only pool. That is, your bracket must be completed by a computer algorithm based upon historical data without the use of human judgment. With that said, let's take a quick look at the format.

Contest Format

The format is fairly simple. We will have two pools: a Tournament pool and a Sweet Sixteen pool. Entries in both pools will be evaluated on the typical exponential point scoring system. Correct picks get 1, 2, 4, 8, 16, and 32 points depending on the depth in the bracket (1 point in the first round, 2 points in the second round, etc). The entry only needs to pick the winning team. Thus, if the other team is no longer in the tournament, but the winning team is picked, points are still awarded. Each person is limited to one entry per pool. Each pool will have a winner determined by the submission scoring the most points.

Deadlines

TOURNAMENT pool entries must be submitted no later than March 18, 2010 at 1am.
SWEET SIXTEEN pool entries must be submitted no later than March 25, 2010 at 1am.
Entries past the deadline will not be accepted.

Rules

  • Your bracket must be chosen completely by a computer algorithm.
  • The computer algorithm must base the decision upon historical data.
  • You may not hard code selections into your algorithm (e.g., "Always pick Stanford over Cal")
  • Your algorithm may only use the data published on this blog. This includes the data described in this post as well as the other data Danny has published beforehand.
  • The above rule is fairly restricting, but I believe this provides a more even playing field. The contest should be about your algorithm's predictive capabilities and not a data advantage one person has over another.
  • You must be able to provide code that shows how your entry is chosen. In other words, your bracket and the selection of winning teams in your bracket must be reproducible by me on a machine.
  • In the event of a tie, the entry with the EARLIER submission time wins.

Submissions

EDIT Thanks to Matt's suggestion in the comments, we'll be using Yahoo's bracket system for the contest submissions. Please send an e-mail to leezen+MarchMadness at gmail for the group password to join. UPDATE Sweet Sixteen prediction bracket is open to contestants.

Prizes

Tournament Bracket: First Place - a custom vinyl sticker, or laser etching featuring the yet to be revealed, super secret "Smell the Data" logo, courtesy of Doug Tarlow
Sweet Sixteen Bracket: First Place - $25 Amazon.com Gift Certificate

Data

As described above, data previously posted on this blog is acceptable for use in this contest (description of the previous data). In addition, we've provided CSV dumps of the 2006-2009 seasons. These are available here: I apologize that the format is not exactly the same as Danny's as it includes some additional attributes. The GameDataCsv.zip files contain game result data while PlayerDataUpdated.zip contains two files: one for 2006-2009 (Players2.csv) and one for 2010. Please see http://blog.smellthedata.com/2010/03/updated-player-data.html for why there is updated player data. The player data columns are:
  • ID (GUID)
  • Name
  • Height
  • Position
  • Team
  • Year
  • Class (Freshman, Sophomore, Junior, Senior)
  • Games - the number of games the player participated in
  • Field goals (shots) made, excluding three point shots
  • Field goal attempts, exlcuding three point shots
  • Three point shots made
  • Three point shots attempted
  • Free throws made
  • Free throw attempts
  • Assists
  • Blocks
  • Rebounds
  • Steals

The 2010 player data has a slightly different schema (sorry!) It includes three sets of field goal figures -- field goals made and attempted without 3 pointers, field goals made and attempted including 3 pointers, and 3 pointers made and attempted. Also note that the last four columns are in slightly different order.
  • ID (GUID)
  • Name
  • Height
  • Position
  • Team
  • Year
  • Class (Freshman, Sophomore, Junior, Senior)
  • Games - the number of games the player participated in
  • Field goals (shots) made, excluding three point shots
  • Field goal attempts, exlcuding three point shots
  • Field goals (shots) made, including three point shots
  • Field goal attempts, including three point shots
  • Three point shots made
  • Three point shots attempted
  • Free throws made
  • Free throw attempts
  • Rebounds
  • Assists
  • Steals
  • Blocks

The game result data columns are:
  • Game Date
  • Days Count
  • Home Points
  • Away Points
  • Overtime
  • Home Team
  • Away Team
  • Home Team Name
  • Away Team Name
  • Game Type

I tried to make this file as backward compatible with Danny's file as possible. I've kept most of the columns, including Days Count. Note that some are negative from 2009-11-08. "Overtime" is either "True" or "False" and True indicates that the game went into Overtime. The Game Type will be either Other, Regular, NCAA Tournament, or Conference Tournament; with the latter two being postseason. Regular refers to the regular season while Other is usually because the game is played at an invitational or other tournament. One thing to note here is that not all games are home or away, there are often neutral courts (the NCAA tournament for example is considered neutral). To preserve the formatting however, I had to pick a home or away team for neutral court games. In these games, I decided to pick the winning team as the home team. If you'd prefer this broken out or altered, please let me know and I can change it sooner than later. One should note that all NCAA tournament games are considered as neutral courts.

Additional Information

If you have other questions, concerns, etc. please comment on this post and I'll do my best to answer.

March Madness Algorithm Contest: Your Commissioner

Posted by Danny Tarlow
I am in a bit of scramble mode trying to get a paper submission out to ECCV by Wednesday, so I haven't had as much time as I'd like to devote to the March Madness algorithms contest. However, this is turning out to actually be a good thing: the great Lee-Ming Zen has volunteered to step in be the contest commissioner.

Lee and I were undergrads at Stanford together ('06), and we were in many of the same introductory computer science courses. I don't remember officially meeting until the summer after sophomore year, though, when we both took undergraduate research positions with the Stanford Logic Group. I was working on automatic grading of logic-based homework problems, and I believe Lee was working on logical spreadsheets. After that, we worked on several projects together, including a General Game Playing program and more recently the Open Red Balloon project. He was an announcer for Women's Basketball and Men's Baseball at Stanford, and he is also the force behind the in-progress NewSlant site. Also, in his spare time (ha), he holds down a job at Microsoft as an Applied Researcher Lead.

So it is with great pleasure that I tell you that Lee's capable hands will be taking the lead on organizing the contest. He has been working on the official rules, and he is also pulling more data for you to work with. Look for posts from him in the near future with the specifics.

And remember, get those algorithms running! We've already got one prize, but we're looking into getting more.

Friday, March 12, 2010

Machine Learning Contests: Life and Death

Posted by Danny Tarlow
There is a lot happening with data analysis and machine learning focused contests.

Yahoo has launched a learning to rank challenge. Given a set of query-dependent features of documents (e.g., how well the query matches words in the document) and relevance labels denoting how well a document matches a particular query, the goal is to learn a ranking function that performs best on a held out set of queries where you have the features but not the labels. This looks to be the largest publicly available data set of the sort, so it's very exciting. I have some ideas and may take a stab at it:
http://learningtorankchallenge.yahoo.com/

On the death side, the Netflix contest sequel is cancelled due to privacy concerns and lawsuits: http://blog.netflix.com/2010/03/this-is-neil-hunt-chief-product-officer.html
(via @hmason)

Finally, I have found a commissioner to run this year's march madness prediction challenge (for which there is a prize!). Get your algorithms tuned up and stay tuned--more details to come.

Tuesday, March 9, 2010

2009 and 2010 March Madness Data

Posted by Danny Tarlow
As promised, here is data from the 2009 and 2010 NCAA Division 1 college basketball seasons. Available for each game is the date, who played (in id, short code, and full name format), who the home team was, and the resulting score. The database view that the data is coming from looks like this:
 
+-------------+-------------+---------+---------+------------+------------+-----------+-----------+---------------+---------------+
| date_played | day_counter | home_id | away_id | home_score | away_score | home_code | away_code | home_name     | away_name     |
+-------------+-------------+---------+---------+------------+------------+-----------+-----------+---------------+---------------+
| 2009-03-18  |         128 |      61 |      28 |         73 |         71 | cbt       | bap       | Creighton     | Bowling Green | 
| 2009-03-18  |         128 |      88 |     131 |         84 |         62 | fak       | jab       | Florida       | Jacksonville  | 
| 2009-03-18  |         128 |     135 |     122 |         83 |         79 | kab       | iag       | Kansas St.    | Illinois St.  | 
| 2009-03-18  |         128 |     309 |     209 |         68 |         59 | tba       | nbd       | Tulsa         | Northwestern  | 
| 2009-03-18  |         128 |      15 |     294 |         87 |         82 | abb       | tae       | Auburn        | Tenn-Martin   | 
| 2009-03-18  |         128 |      22 |     101 |         74 |         72 | bae       | gae       | Baylor        | Georgetown    | 
| 2009-03-18  |         128 |     236 |     166 |         66 |         78 | pas       | mav       | Providence    | Miami (FL)    | 
| 2009-03-18  |         128 |     323 |      77 |        116 |        108 | vah       | dav       | Virginia Tech | Duquesne      | 
| 2009-03-19  |         129 |     211 |     195 |         70 |         68 | nbf       | nam       | Notre Dame    | New Mexico    | 
| 2009-03-19  |         129 |      33 |     300 |         66 |         79 | baw       | tan       | BYU           | Texas A&M     | 
+-------------+-------------+---------+---------+------------+------------+-----------+-----------+---------------+---------------+
And the data itself is just a CSV export of this. For example:
"2008-11-10",0,76,234,80,49,"dau","paq","Duke","Presbyterian"
"2008-11-10",0,113,98,63,65,"ham","gab","Houston","Georgia Southern"
"2008-11-11",1,76,98,97,54,"dau","gab","Duke","Georgia Southern"
"2008-11-11",1,124,204,60,73,"iak","nay","IUPUI","Northeastern"
"2008-11-11",1,113,234,76,57,"ham","paq","Houston","Presbyterian"

In the linked file, there is one CSV for the 2008-2009 season, and one for the 2009-2010 season. Two notes
  • There are still a few more games to be played this season, so I will update the data when the season officially ends and the brackets are announced.
  • I do not have all the tournament data from last year. If some kind soul would like to transcribe last year's final bracket into text form, I'd appreciate it (I don't have the time or inclination to do it right now).

So with that all out of the way, and without further ado, here is the data (Update: this is version 3 of the data, which has entries for four teams that were missing in the first version and has removed some game duplicates that were in version 2).

I encourage you all to try out your favorite machine learning algorithms on it. If you need a starter idea, I still like my strategy from last year, which you can read about here, but there are likely plenty of other good approaches. I'm curious to see what other people can come up with.

I'll try to get my model's predictions posted by March 16, when the first game begins. If you have a bracket and a model that you'd like to share, please send me the predictions along with a brief description of your approach. It'd be fun to see the different ideas then see how the different approaches stack up. I can't promise a prize for the winner (other than a showcase on this blog), but perhaps there is some generous sponsor out there who would like to donate a better prize to the pool? =P

Update:
We have a prize (!!) generously offered up by Doug in the comments: "a custom vinyl sticker, or laser etching featuring the yet to be revealed, super secret "Smell the data" logo."



*For those curious, here is the SQL query that generated the data. If you'd like to give me an alternate SQL script to run, I might be willing to give you a custom export. The raw table descriptions are here.
SELECT  gr.date_played, TO_DAYS(gr.date_played) - TO_DAYS(date('2008-11-10')) as day_counter,
  tc1.team_id as home_id, tc2.team_id as away_id, 
  home_score, away_score,
  gr.home_code, gr.away_code, 
  tc1.team_name as home_name, tc2.team_name as away_name
INTO OUTFILE '/tmp/2009_season.csv'
FIELDS TERMINATED BY ',' OPTIONALLY ENCLOSED BY '"'
LINES TERMINATED BY '\n'
FROM 
 game_result gr, team_code tc1, team_code tc2 
WHERE 
 gr.date_played < '2009-06-01' and
 gr.home_code = tc1.team_code and gr.away_code = tc2.team_code
ORDER by gr.date_played;

Monday, March 8, 2010

Data-driven March Madness 2010

Posted by Danny Tarlow
Some of you may remember from last year that I don't really follow college basketball, but I still like to enter into the March Madness bracket pools with my friends. How do I keep from being woefully uninformed, you might ask? Simple: Data-driven march madness predictions.

I just quickly checked my scripts from last year, and the scraping and parsing seems to work. So consider the data gathering officially started. Anybody else who is interested in applying some machine learning and making data-driven predictions, I suggest you start thinking about your strategy now. I will provide the data shortly.

Update:
The data is being put in a MySQL database with two tables. First is the game result:
mysql> describe game_result;
+-------------+------------+------+-----+---------+----------------+
| Field       | Type       | Null | Key | Default | Extra          |
+-------------+------------+------+-----+---------+----------------+
| game_id     | int(11)    | NO   | PRI | NULL    | auto_increment | 
| date_played | date       | YES  |     | NULL    |                | 
| home_code   | varchar(3) | YES  | MUL | NULL    |                | 
| home_score  | int(11)    | YES  |     | NULL    |                | 
| away_code   | varchar(3) | YES  |     | NULL    |                | 
| away_score  | int(11)    | YES  |     | NULL    |                | 
+-------------+------------+------+-----+---------+----------------+
And second is the team codes, as used by rivals.yahoo.com:
mysql> describe team_code;
+-----------+-------------+------+-----+---------+----------------+
| Field     | Type        | Null | Key | Default | Extra          |
+-----------+-------------+------+-----+---------+----------------+
| team_id   | int(11)     | NO   | PRI | NULL    | auto_increment | 
| team_code | varchar(3)  | YES  | UNI | NULL    |                | 
| team_name | varchar(64) | YES  |     | NULL    |                | 
+-----------+-------------+------+-----+---------+----------------+
I haven't set the exact format of the output, but it will be some simple export of this.