Pages

JCI, Week 10

2/21/2006

The updated JCI rankings can be found here: http://home.comcast.net/~jcolton31/index.html

The list of teams that fall inside/outside the top 25 & 50 JCI vs RPI is growing. Check out the groups below:

Group A: Top 25 RPI, Non-Top 25 JCI
29. UCLA 20-6, RPI: 14 JCI SOS: 74 RPI SOS: 22
28. Oklahoma 18-6, RPI: 15 JCI SOS: 61 RPI SOS: 31
47. Arizona 16-10, RPI: 17 JCI SOS: 39 RPI SOS: 3
35. George Mason 21-5, RPI: 20 JCI SOS: 141 RPI SOS: 82
26. Northern Iowa 20-6, RPI: 23 JCI SOS: 65 RPI SOS: 55
30. Wichita St 20-7, RPI: 24 JCI SOS: 58 RPI SOS: 34
36. Missouri St 18-7, RPI: 25 JCI SOS: 72 RPI SOS: 42

Group B: Top 25 JCI, Non-Top 25 RPI
5. George Washington 22-1, RPI: 33 JCI SOS: 150 RPI SOS: 216
14. Georgetown 17-7, RPI: 29 JCI SOS: 5 RPI SOS: 51
17. West Virginia 17-8, RPI: 32 JCI SOS: 3 RPI SOS: 27
18. Boston College 20-5, RPI: 34 JCI SOS: 60 RPI SOS: 89
21. Bucknell 20-4, RPI: 39 JCI SOS: 108 RPI SOS: 155
24. Syracuse 19-8, RPI: 27 JCI SOS: 27 RPI SOS: 11
25. Cincinnati 17-9, RPI: 26 JCI SOS: 6 RPI SOS: 7

Group C: Top 50 RPI, Non-Top 50 JCI
56. UNC Wilmington 20-7, RPI: 38 JCI SOS: 123 RPI SOS: 100
60. Western Kentucky 19-5, RPI: 42 JCI SOS: 229 RPI SOS: 114
54. Air Force 19-5, RPI: 48 JCI SOS: 210 RPI SOS: 168
72. Saint Joseph's 11-12, RPI: 49 JCI SOS: 11 RPI SOS: 4
78. Utah St 17-6, RPI: 50 JCI SOS: 227 RPI SOS: 122

Group D: Top 50 JCI, Non-Top 50 RPI
38. Florida St 16-7, RPI: 60 JCI SOS: 52 RPI SOS: 98
42. Arkansas 17-8, RPI: 56 JCI SOS: 53 RPI SOS: 73
44. Louisville 16-9, RPI: 67 JCI SOS: 35 RPI SOS: 47
49. UAB 18-5, RPI: 53 JCI SOS: 155 RPI SOS: 157
50. Vanderbilt 14-9, RPI: 59 JCI SOS: 38 RPI SOS: 37

It's pretty astonishing that the Selection Committee would continue to have confidence in a tool that can be wrong 30 slots or more for any given team. Remember, the RPI being off has further ramifications for the other tools that the Committee looks at -- for example, each of the 10 teams that has beat Arizona this season gets credit for a top 25 win even though they only beat the 47th-ranked team.

Arizona is a perfect case study. Their RPI is 17th and they are 16-10!!! How can this be? Here's their wins and losses to date:

Wins
12/31 W at 33 Washington
11/21 W vs 32 Kansas
2/19 W at 100 Stanford
12/29 W at 115 Washington St
1/21 W 59 California
12/21 W 60 Western Kentucky
12/17 W at 138 Utah
11/27 W 64 Virginia
1/25 W at 153 Arizona St
1/7 W 82 Southern California
1/19 W 100 Stanford
12/19 W 121 Sam Houston
12/10 W 124 St. Mary's
2/9 W 133 Oregon
2/11 W 139 Oregon St
12/8 W 141 Northern Arizona

Losses
11/22 L 2 Connecticut
1/28 L at 19 North Carolina
2/4 L at 29 UCLA
11/23 L 16 Michigan St
2/16 L at 59 California
2/2 L at 82 Southern California
12/3 L at 89 Houston
1/5 L 29 UCLA
1/14 L at 133 Oregon
1/12 L at 139 Oregon St

Arizona is clearly benefitting from not having played any teams below 153rd in the JCI, and this is artificially boosting their RPI SOS and RPI. They have the 3rd strongest schedule according to the RPI, however their JCI SOS is 39th. [Note: The JCI estimates the expected winning percentage of the 30th ranked team against a given team's schedule accounting for the location of the game, so it provides a much more accurate SOS number. While the RPI accounts for home/road adjustments in the win-loss record (albeit inaccurately), the SOS number doesn't account for home/road.]

The RPI is clearly affected by the number of bottom feeders you play (or inArizona's case, don't play). Want to 'game' the RPI (say, for example, you know you're truly the 47th best team in the country, but you want to make your RPI look much better)? Instead of scheduling against the perennial bottom third of Division I, go slightly upmarket and schedule games against the teams in the 100-150 range (on the road would be even better, although the economics may not allow it). The probability of beating a 150th ranked team versus a 250th ranked team is not much different (see table below), but your RPI will receive a nice boost.

  Expected Win Percentage
47th Ranked Team vs.
Opp Rank Home Neutral Away
50 70.23% 53.37% 35.71%
100 85.11% 73.50% 57.38%
150 93.47% 87.42% 77.14%
200 96.90% 93.81% 88.04%
250 98.49% 96.94% 93.90%
300 99.43% 98.83% 97.61%

You can see that the probability of winning the game doesn't go down much when you move from 250th to 150th...a top-tier program should have no problem beating either team at home.

We should be able to show the impact to Arizona if we altered their schedule a little bit. Let's assume that for some of their lower-ranked non-conference wins, they played a team that was 100 slots below the team they actually played. I'm guessing their JCI won't move much but they're RPI will move dramatically.

12/8 141 Northern Arizona --> switch to 241 Charleston Southern
12/10 124 St. Mary's --> switch to 224 Bowling Green
12/17 at 138 Utah --> switch to 238 Delaware
12/19 121 Sam Houston --> switch to 221 Marshall

Switching just these 4 games in Arizona's schedule would cause their RPI to drop from 17th to 34th and their RPI SOS from 3rd to 25th. Meanwhile, their JCI would fall from 47th to 49th and JCI SOS from 39th to 48th. The best way to make sense of this information is to ask yourself: does beating first 4 teams as opposed to second 4th change my opinion of Arizona's season by 17 slots and their schedule by 22 slots? I would argue not. Sure, they should get the benefit of playing the slightly stronger schedule, and the JCI accurately reflects the difference.

To show this 'bottom feeder' effect a different way (without picking on Arizona), take two schedules. Schedule A: at #1 Duke, at #3 Villanova, at #333 Savannah State. Schedule B: home vs. #110 Pacific, home vs. #79 Northwestern St, home vs. #128 IUPUI. It's obvious that Schedule A is a far more difficult schedule -- the expected win percentage for the 47th-ranked team is 34.5% for Sch A and 86.3% for Sch B. However, the RPI SOS would be IDENTICAL for either schedule. Keep that in mind when you hear Jay Bilas talking about some team's strength of schedule.

Brett Petway

Check out this hilarious video compiled to a song written and performed by Michigan's very own Brett Petway. Among the classics moments are the line 'NIT 2004, other teams pursuing that' and the technique of rhyming 'Hunter' with 'Hunter' and 'Hunter' again. Pure lyrical genius.

This song will undoubtedly spawn an age-old debate: is Brett Petway a better rapper or a better basketball player.?

0 comments:

 
Wegoblogger #31 © 2011 | Designed by Bingo Cash, in collaboration with Modern Warfare 3, VPS Hosting and Compare Web Hosting