Breaking down the AP Top 25 by using the RPI

[SelectionSunday]

Breaking down the AP Top 25 using the RPI

Here's how the AP rankings could look (using the same 25 teams) based strictly on the number of RPI top 50 wins each team has earned through games played Monday, Jan. 5. An * represents a road or neutral win.

1. Butler 4: *Xavier (4), Evansville (25), UAB (34), Northwestern (39)
2. Pitt 4: Miami-Ohio (9), *Georgetown (13), *Florida State (17), Siena (42)
3. Michigan State 4: *Texas (21), *Oklahoma State (32), *Minnesota (38), *Northwestern (39)
4. Duke 3: *Xavier (4), *Michigan (40), Rhode Island (41)
5. Georgetown 3: *UConn (8), Memphis (26), *Maryland (50)
6. West Virginia 3: Miami-Ohio (9), *Ohio State (12), *Iowa (48)
7. Xavier 3: Miami-Ohio (9), *Memphis (26), Mizzou (49)
8. Boston College 3: *North Carolina (11), *UAB (34), Iowa (48)
9. Tennessee 3: *Georgetown (13), *Marquette (28), *Siena (42)
10. Texas 3: *Wisconsin (14), *Villanova (23), UCLA (47)
11. Oklahoma 3: Davidson (22), Utah (29), *UAB (34)
12. North Carolina 2: *Michigan State (6), Evansville (25)
13. Minnesota 2: Ohio State (12), *Louisville (44)
14. Notre Dame 2: Georgetown (13), *Texas (21)
15. Marquette 2: Wisconsin (14), Villanova (23)
16. UConn 2: *Wisconsin (14), *Gonzaga (35)
17. Clemson 2: *Illinois (19), *Temple (24)
18. Purdue 2: *Davidson (22), *Boston College (30)
19. Villanova 2: Temple (24), *Rhode Island (41)
20. Syracuse 2: *Memphis (26), *Florida (46)
21. Arizona State 2: *Stanford (36), *BYU (37)
22. UCLA 1: Miami-Ohio (9)
23. Baylor 1: *Arizona State (18)
24. Louisville 1: UAB (34)
25. Wake Forest 1: *BYU (37)

 

[rrjackIII]

Here's how the AP rankings could look (using the same 25 teams) based strictly on the number of RPI top 50 wins each team has earned through games played Monday, Jan. 5. An * represents a road or neutral win.

1. Butler 4: *Xavier (4), Evansville (25), UAB (34), Northwestern (39)
2. Pitt 4: Miami-Ohio (9), *Georgetown (13), *Florida State (17), Siena (42)
3. Michigan State 4: *Texas (21), *Oklahoma State (32), *Minnesota (38), *Northwestern (39)
4. Duke 3: *Xavier (4), *Michigan (40), Rhode Island (41)
5. Georgetown 3: *UConn (8), Memphis (26), *Maryland (50)
6. West Virginia 3: Miami-Ohio (9), *Ohio State (12), *Iowa (48)
7. Xavier 3: Miami-Ohio (9), *Memphis (26), Mizzou (49)
8. Boston College 3: *North Carolina (11), *UAB (34), Iowa (48)
9. Tennessee 3: *Georgetown (13), *Marquette (28), *Siena (42)
10. Texas 3: *Wisconsin (14), *Villanova (23), UCLA (47)
11. Oklahoma 3: Davidson (22), Utah (29), *UAB (34)
12. North Carolina 2: *Michigan State (6), Evansville (25)
13. Minnesota 2: Ohio State (12), *Louisville (44)
14. Notre Dame 2: Georgetown (13), *Texas (21)
15. Marquette 2: Wisconsin (14), Villanova (23)
16. UConn 2: *Wisconsin (14), *Gonzaga (35)
17. Clemson 2: *Illinois (19), *Temple (24)
18. Purdue 2: *Davidson (22), *Boston College (30)
19. Villanova 2: Temple (24), *Rhode Island (41)
20. Syracuse 2: *Memphis (26), *Florida (46)
21. Arizona State 2: *Stanford (36), *BYU (37)
22. UCLA 1: Miami-Ohio (9)
23. Baylor 1: *Arizona State (18)
24. Louisville 1: UAB (34)
25. Wake Forest 1: *BYU (37)

Great AP / RPI breakdown SS. This likely partly explains why Butler jumped ahead of the Gophers from last week in the AP poll. Though I doubt very many of the actual AP Poll voters put together the info that you provide for us in the GopherHole.

Thanks.

 

[SelectionSunday]

Thanks. Next week I'll do the same thing, except keep the teams in order of their AP ranking. That might be easier to translate who's overrated & who's underrated among the top 25.

 

[GoGoGophers]

SS...

Do you think the RPI is still the best indicator out there? I've heard/read alot of rumblings that the committee is beginning to weight this less and less. Are there other metrics out there that are being used?

I know that realtimerpi also has "power ratings" that have always appeared to make more sense to me.

 

[SelectionSunday]

The RPI

RPI is simply a measuring tool, one of many, many things the committee looks at. Certainly it isn't, nor should it be, the be-all and end-all when determining at-large bids. It has flaws like every other measuring tool.

But Ope1169, who used to post here a lot, I think summed up best where the RPI is most effective. A team's individual RPI shouldn't necessarily be all that important (though certainly you don't want be lower than the 60's if you're on the bubble), but where the RPI is most helpful is in determining how teams have performed against the top 25, top 50, top 100, etc., in the RPI.

People like to criticize computers because they lack the human element in evaulating teams, but that's exactly what I like about the computer element. What computers can do is completely take "name and perception" out of the equation. Computers judge teams by performance and quality of competition, not by their name. To me, the single most important element in evaluating bubble teams is how they've done against the top 50 and top 100, as well as road/neutral record & nonconference strength of schedule. A team's individual RPI, per se, doesn't really carry a lot of weight with me, but nevertheless it is helpful.

 

[GoGoGophers]

To me, the single most important element in evaluating bubble teams is how they've done against the top 50 and top 100, as well as road/neutral record & nonconference strength of schedule.

Agreed. The talking heads (Lunardi, et al) have started to evaluate a team's record vs. Top 25/50/100 independently from the RPI for their in/out predictions and power rankings. This makes more sense to me, as simply looking at the RPI for Teams A, B and C seems too simplistic at times.

Teams like Miami, Oh have great RPI rankings due to SOS, but have are getting rewarded even though they've lost all those tough games they scheduled. I understand that this will correct itself by the end of the season, but it shows that the RPI rankings may not be a real good tool for "snapshot" looks during the season.

Anyway...thanks for the info and for posting them every week. Keep 'em coming, Hodger.

 

[SelectionSunday]

Perfect example

You hit the nail on the head with Miami-Ohio. Temple is another one. Both have really good RPI's right now, but neither really has beaten anyone, save for Temple's win over Tennesse. At some point a team has to win a few of their tough games, no matter how challenging their schedule.