New computer ranking with both Div A and Div B

[Hackalicious]

I made up a computer ranking system for the hell of it. This is based solely on points scored and does not take into account when the game was played, home advantage, or overtime.

Surprisingly, it spit out pretty decent rankings.

The numbers next to each team are arbitrary points. They give you an idea how close some of the teams are.

1. Michigan 12081
2. Colorado State 11137
3. Sonoma State 9140
4. Brigham Young 7890
5. Florida State 7749
6. UC Santa Barbara 7377
7. UC San Diego 7374
8. Colorado 6408
9. Oregon 5733
10. Chapman 5606
11. Arizona 5298
12. Cal Poly San Luis Obispo 5003
13. Lindenwood 4073
14. Utah 3797
15. Virginia Tech 3560
16. Georgia Tech 3150
17. Arizona State 3109
18. UCLA 2905
19. Oakland 2832
20. Northeastern 2529
21. Georgia 2477
22. Minnesota - Duluth 2285
23. Texas 2040
24. Boston College 1984
25. California 1952

Utah and Cal Poly seem underanked, while Chapman seems overranked. I think this is because it doesn't distinguish old games from new games. I tried adding in some weighting for time, but it just threw in more anomalies.

[benji]

Wow Hackalicious, not too far off from the poll, nice work.

Interesting that Texas is placed in #23, where Texas Tech was placed in the last poll ... is that a typo?

What were the conditions of your ranking system?

[Hackalicious]

Wow Hackalicious, not too far off from the poll, nice work.

Interesting that Texas is placed in #23, where Texas Tech was placed in the last poll ... is that a typo?

What were the conditions of your ranking system?

No typo. It came up with Texas.

The poll is based on a random walk on a graph.
http://en.wikipedia.org/wiki/Random_walk

Like I said, it only looks at raw scores and ignores homefield advantage, overtime games, or how late in the season the game was played.

I think there might be a bias for teams that play a lot of games. A team that loses a lot of close games might be ranked relatively high compared to a team that wins a couple of blowouts. (In fact, I edited out a few NCAA teams like Whittier and Notre Dame de Namur, who were under-ranked at around #12-13, probably for that exact reason)

[baskinginreflectedglory]

There is a reason Chapman is comming up with better rankings in the mathematical scores. It's highly possible the pollsters were looking at the wrong evidence to halo this team.

[Hackalicious]

I added in Division B games to see what would happen. San Diego's win over UCSD looks like it made a big impact.

Still can't fully explain why a team like Texas Tech is getting ranked lower than Texas, a team they beat head-to-head. The only difference I can see off hand is that Texas played more games.

If I average the rankings for number of games played, the gap closes, but Texas is still ranked higher. The way the program works, I might have to run the simulation much longer to get more accurate rankings outside the top 10.

Colorado still is under-ranked, too. I think the blowout against UCSB really hurt and might be over-valued.

1. Michigan 9236
2. Colorado State 8697
3. Sonoma State 6305
4. Brigham Young 5618
5. Florida State 5185
6. UC Santa Barbara 5078
7. UC San Diego 4816
8. Colorado 4558
9. Oregon 4469
10. Arizona 3823
11. Chapman 3754
12. Cal Poly San Luis Obispo 3580
13. San Diego 3235
14. Lindenwood 2923
15. Utah 2625
16. Claremont 2269
17. Virginia Tech 2114
18. Arizona State 2099
19. UCLA 1957
20. Georgia Tech 1919
21. Minnesota - Duluth 1752
22. Oakland 1692
23. Texas 1605
24. Northeastern 1568
25. Georgia 1422
26. Simon Fraser 1298
27. Boston College 1263
28. New Hampshire 1256
29. Florida 1176
30. Michigan State 1122
31. Illinois 1121
32. California 1088
33. Texas Tech 1053
34. Loyola Marymount 1048
35. St. John's - MN 1047