2006 Ragin' Cajuns Baseball RPI

[GoneGolfin]

Brian if the Cajuns make the regionals where do you see them going?

I have not started slotting my weekly selections into regionals yet. I will do that next weekend. But I can venture an educated guess.

Regionals in this area are likely to be ...

Rice
Texas
Oklahoma
Oklahoma State
Arkansas

I do not think Mississippi snags a fifth SEC regional unless they make the finals of or win the SEC Tournament. The Rebels are a #2 seed right now.

Texas, Oklahoma, Oklahoma State, and Nebraska are all #1 seeds. I think a lot of this comes down to what happens with Baylor and Houston. Baylor cannot go to a Big XII regional and Houston cannot play at a Rice regional. If Houston is a #1 seed (currently on the fence), they are likely traveling (maybe Arizona State). If the Cougars are a #2 seed, they could be sent to Austin. Baylor is currently sitting the fence w/respect to a #2/#3 seed. I can see Baylor in Houston as a #2 or #3 seed. Thus ...

With Houston as a #2 seed in Austin, you could see the Cajuns there as a #3 seed. Facing Brad Lincoln in the opener would not be fun. If Houston is a #1, you could see LSU as a #2 (on the fence w/respect to #2/#3) in Austin facing the Cajuns in the opener.

The Rice regional could see Baylor and LSU as #2/#3 seeds in any order. With Baylor as a #2 seed in Houston, you could see the Cajuns there as a #3 seed.

Thus, both the Texas and Rice regionals would be looking for a #3 seed that avoids conference matchups. This is why I like the Cajuns either in Austin or Houston. Don't be surprised to see LSU vs. Louisiana in the first round of the Austin regional. But if they are not sent to Austin, Houston (Rice) seems to be a logical choice as well.

Brian

[GoneGolfin]

Each year at this time of the season, I am inevitably asked various questions concerning RPI and at-large bids. For example, "What is the lowest RPI ever to receive an at-large bid?" or "What is the highest RPI ever to be excluded from the NCAA Tournament?". Other questions center around the chances that a given team will receive an at-large bid based on how they performed in conference, out-of-conference, and their RPI.

We all know how important conference finish is to the NCAA Selection Committee. Especially conference championships. We also know that Conference Tournament performances are an important consideration. The tournament games are also added in to the regular season records to calculate a given school's total performance in conference for that season. Many times, national seeds, #1 seeds, and host regionals are determined by this.

Now, in lieu of the Cajuns' current predicament, it is interesting to go back into what I call my "interesting data" archives. The below is data concerning the NCAA Tournament since the new 64 team tournament format commenced (1999). For purposes of the RPI, I use Boyd Nation's pRPI estimates. These estimates

1) Count neutral games as road games
2) Do not count ties
3) Apply bonuses for non-conference road wins against three tiers (Top 25, Top 50, and Top 75 of the Base RPI)
4) Apply penalties for non-conference hom losses against three tiers (Bottom 25, Bottom 50, and Bottom 75 of the Base RPI)

The RPI formula used by SEBaseball.com is very similar to that of Boyd Nation's, but slightly different with respect to #1 and #2 above.

Some answers to the above questions follow ...

What is the lowest RPI to ever receive an at-large bid?
1. Washington (82) (2002)
2. San Jose State (81) (2000) (Spartans made it to the CWS)
3. Louisiana-Monroe (78) (2000)
4. Evansville (74) (2000)
5. Washington (72) (2003)
Nevada (72) (2000)

What is the highest RPI ever to be excluded from the NCAA Tournament?
1. Virginia (22) (2003)
2. Georgia Tech (25) (1999)
3. Auburn (26) (2004)
4. Georgia (27) (2005)
5. Old Dominion (28) (1999)

What is the lowest RPI to ever receive an at-large bid and not win a conference regular season championship?
1. Washington (82) (2002 - Pac Ten Tied for 2nd place)
2. Evansville (74) (2000 - MVC 2nd place)
3. Washington (72) (2003 - Pac Ten 3rd place)
4. Creighton (64) (2000 - MVC 3rd place)
5. Fresno State (63) (2001 - WAC 2nd place)

Here is a year-by-year breakdown since 1999 of some fence sitting schools that did and did not receive at-large bids to the NCAA Tournament. Also included is conference regular season finish, as well as in some cases conference tournament performance.

2005
(Received at-large bids)
Louisiana (45) (Sun Belt champion, 2-2 Conference Tournament)
St. John's (59) (Big East champion, 0-2 Conference Tournament)
Creighton (48) (MVC champion, 4-2 Conference Tournament - 2nd place)
Michigan (50) (Big Ten 4th place, 1-2 Conference Tournament)

(Did not receive at-large bids)
Vanderbilt (30) (SEC 10th place)
Georgia (27) (SEC 11th place)


2004
(Received at-large bids)
George Mason (50) (Colonial Champion, 2-2 Conference Tournament)
Missouri (53) (Big XII 7th place, 3-1 Conference Tournament - 2nd)
Middle Tennessee (59) (Sun Belt co-champion and #1 seed, 3-2 Conference Tournament)
St. John's (49) (Big East T 2nd place, 1-2 Conference Tournament)
UCLA (45) (Pac 10 3rd place)
College of Charleston (48) (Southern Champion, 3-2 Conference Tournament)
Tulane (47) (CUSA T 2nd place, 0-2 Conference Tournament)

(Did not receive at-large bids)
Nebraska (31) (Big XII 8th place)
Auburn (26) (SEC 10th place)

2003
(Received at-large bids)
Rutgers (45) (Big East champion, 3-2 Conference Tournament)
Florida (32) (SEC 9th place, 13-16)
Southwest Missouri State (60) (MVC champion, 2-2 Conference Tournament)
East Carolina (52) (CUSA 5th place, 2-2 Conference Tournament)
Washington (72) (Pac 10 3rd place)
North Carolina-Wilmington (48) (Colonial 2nd place, 2-2 Conference Tournament)

(Did not receive at-large bids)
Virginia (22) (ACC 6th place)
West Virginia (38) (Big East 2nd place)

2002
(Received at-large bids)
Elon (45) (Big South 2nd place)
Washington (82) (Pac 10 T 2nd place)
Tulane (55) (CUSA 5th place)

(Did not receive at-large bids)
Mississippi (31) (SEC 9th place)
North Carolina State (39) (ACC 7th place)

2001
(Received at-large bids)
Ohio State (59) (Big Ten 1st place)
Fresno State (63) (WAC 2nd place)
Texas (51) (Big XII 3rd place)
Oklahoma State (44) (Big XII 5th place)
California (46) (Pac Ten T 3rd place)
Georgia Southern (62) (Southern 1st place)
Middle Tennessee (52) (Sun Belt T 1st place, #1 seed)

(Did not receive at-large bids)
Alabama (32) (SEC 9th place)
North Carolina (35) (ACC 6th place)

2000
(Received at-large bids)
Louisiana-Monroe (78) (Southland T 1st place)
Texas Tech (59) (Big XII 5th place)
San Jose State (81) (WAC T 1st place)
Evansville (74) (MVC 2nd place)
Florida International (46) (Sun Belt 3rd place)
Creighton (64) (MVC 3rd place)
Penn State (57) (Big Ten 2nd place)
Nevada (72) (Big West T 1st place)

(Did not receive at-large bids)
Kentucky (31) (SEC 8th place)
Georgia (29) (SEC 7th place)
Mississippi (37) (SEC 9th place)
North Carolina State (36) (ACC T 6th place)
Virginia Commonwealth (30) (Colonial 5th place)
Purdue (38) (Big Ten T 3rd place)
North Carolina-Wilmington (39) (Colonial 3rd place)

1999
(Received at-large bids)
Arizona (62) (Pac Ten 3rd place)
UCLA (54) (Pac Ten 4th place)
NE Louisiana (66) (Southland 1st place)
Nevada (45) (Big West 2nd place)
Southern California (46) (Pac Ten 2nd place)
Long Beach State (48) (Big West 3rd place)
Southwestern Louisiana (57) (Sun Belt 2nd place)

(Did not receive at-large bids)
Georgia Tech (25) (ACC 5th place)
Coastal Carolina (37) (Big South 1st place)
Old Dominion (28) (Colonial 5th place)

Brian

[GoneGolfin]

Some games to keep tabs on today that more directly affect the Cajuns ...

Louisiana is at #49 in the pRPI with a RPI rating of .557.

#60 (.550) Texas A&M at #42 (.567) Kansas
#44 (.563) Long Beach State at #63 (.548) California-Riverside
#46 (.561) Southern California at #45 (.562) Wichita State
#48 (.557) Texas Tech at #13 (.604) Oklahoma State
#251 (.426) Oakland at #53 (.555) Oral Roberts (Doubleheader)
#173 (.482) Iowa at #59 (.551) Michigan

Brian

As we can see below, the Texas Tech loss to Oklahoma State (OSU swept TT this weekend) has vaulted the Cajuns cleanly into the #48 spot entering the SBC Tournament. Southern California's loss to Wichita State (WSU swept USC this weekend) brings them closer to the Cajuns. Although, USC is now guaranteed to finish without a winning record and will not be playing in the postseason.

A threat to watch ... Missouri is moving up the ratings fast after sweeping Texas this weekend. Missouri finished 12-15 in the Big XII for 7th place, a game behind Baylor (#5) and Kansas (#6). Missouri opens up the Big XII Tournament against Oklahoma State. The Cajuns could use poor conference tournament performances from Kansas and Missouri. Kansas open the conference tournament against Oklahoma.

41 0.568 28 28 28 28 Florida
42 0.568 35 23 38 23 Kansas
43 0.566 33 23 33 23 Vanderbilt
44 0.565 29 24 29 24 Long Beach State
45 0.564 42 18 43 18 Wichita State
46 0.561 39 15 39 15 Old Dominion
47 0.559 25 29 25 29 Southern California
48 0.557 37 18 37 18 Louisiana-Lafayette
49 0.556 40 14 41 14 Notre Dame
50 0.555 26 30 26 30 Maryland
51 0.555 37 19 37 19 San Francisco
52 0.555 32 14 35 14 Oral Roberts
53 0.554 28 23 31 23 Texas Tech
54 0.553 32 23 32 23 San Diego
55 0.553 29 24 29 24 Missouri
56 0.553 38 16 38 16 Connecticut
57 0.553 32 24 32 24 East Carolina
58 0.552 33 17 35 17 Kent State
59 0.552 37 19 37 19 North Carolina-Wilmington
60 0.551 38 18 38 18 Michigan

Brian

[Turbine]

Brian I love the work you do and your in depth analysis, but I need you to get me over a hangup.

In my opinion RPI cannot tell you who the best team is.

What it can tell you is who has taken the most highly regarded tests.

It seems to me that if the worst team in the country played each of the top 25 RPI teams twice and lost every one they could still have a top 20 RPI

Conversely if the NY Yankees could get into the Southland and played SWAC and MEAC teams all year their RPI would say they were no good.

Please adjust my thinking. Thanks

[GoneGolfin]

Brian I love the work you do and your in depth analysis, but I need you to get me over a hangup.

In my opinion RPI cannot tell you who the best team is.

What it can tell you is who has taken the most highly regarded tests.

It seems to me that if the worst team in the country played each of the top 25 RPI teams twice and lost every one they could still have a top 20 RPI

Conversely if the NY Yankees could get into the Southland and played SWAC and MEAC teams all year their RPI would say they were no good.

Please adjust my thinking. Thanks

Just to clarify, I despise the RPI. It is a horrible rating system that should be junked. As a software professional, I am not convinced that you can design a reasonable mathematical rating system that is better than what the subjective human process can produce.

I use the RPI (pRPI) in my discussion only because it is so influential in the NCAA Selection process. If you want to project accurately, you need to understand the tools that the NCAA Selection Committee uses and how they use them.

I agree that the RPI cannot tell you who the best team is.

It does tell you on average how a given team would perform (W/L %) against an average schedule. But only if you have a proper distribution of scheduled games. Of course, if you have a complete distribution of scheduled games (balanced), you do not need the RPI because you could simply compare the W/L records (%).

If the worst team in the country played each of the Top 25 RPI teams twice and lost every game, they would not have a Top 25 RPI, much less a Top 20 RPI one. But they may (likely) very well have a higher rating than otherwise deserved. That same team could lose every game to the 26 -> 50 teams. But it would be better off in the ratings had they lost every game to the Top 25 RPI teams.

The biggest problem with the RPI in baseball is that even with the larger sample size (larger is better for RPI), play is still highly interregional and will get worse with the new shortened season and common start date looming. In other words, the least connective the schedules are, the more innaccurate the RPI results.

An example ...

Pool A
1. Rice (6-0)
2. Texas (4-2)
3. Clemson (2-4)
4. LSU (0-6)

Pool B
1. Bethune Cookman (6-0)
2. Delaware State (3-3)
3. Iona (2-4)
4. Alabama State (1-5)

The number of games above is purely arbitrary and be scaled to any desire. The important thing to note here is that there is no inter-pool play. Each team within a given pool plays each team in the same pool twice. What would you expect the RPI results to be?

In this example, Bethune Cookman and Rice would be tied for the #1 RPI position. Texas would be #3, Delware State #4, Clemson and Iona tied for #5, Alabama State #7, and LSU #8.

But if LSU were in Pool B, they could very well go undefeated.

In other words, with no connectivity and a completely distributed schedule, RPI would be the same as winning %.

Now this is obviously an exaggeration to illustrate a point. But you can see how in an imperfectly connected system, RPI begins to demonstrate its flaws depending on the severity of the unconnectivity.

Brian

[Square Head]

"Just to clarify, I despise the RPI. It is a horrible rating system that should be junked. As a software professional, I am not convinced that you can design a reasonable mathematical rating system that is better than what the subjective human process can produce."

That is why the NCAA uses the RPI system, to protect the big boys.
jmo

Don

[DeadRed]

"Just to clarify, I despise the RPI. It is a horrible rating system that should be junked. As a software professional, I am not convinced that you can design a reasonable mathematical rating system that is better than what the subjective human process can produce."

That is why the NCAA uses the RPI system, to protect the big boys.
jmo

Don

Exactly. The RPI is used to protect the bigger conferences and then seed the other conferences always against a number 1,2, or 3.

The RPI was used to finally, in essence, break the Lady Techsters program. They were ranked high but RPIed low. Time after time they have to beat a top 10-15 team just to advance. For the most part since Tech lost the National Champ. game to NC in 94 the seedings have not been in their favor year after year after year. Ranked high-seeded low.

Now, they are not what they used to be. (and the coach retired, but the downfall was prior to his departure)

Sounds alot what is happening to a softball program we know. Ranked high - seeded low.

[Turbine]

If the worst team in the country played each of the Top 25 RPI teams twice and lost every game, they would not have a Top 25 RPI, much less a Top 20 RPI one. But they may (likely) very well have a higher rating than otherwise deserved. That same team could lose every game to the 26 -> 50 teams. But it would be better off in the ratings had they lost every game to the Top 25 RPI teams.

Lets say every opponent of "worst team" went undefeated. Would not this alone give "worst team" an RPI of .500

[gorcajun]

It does tell you on average how a given team would perform (W/L %) against an average schedule. But only if you have a proper distribution of scheduled games. Of course, if you have a complete distribution of scheduled games (balanced), you do not need the RPI because you could simply compare the W/L records (%).


Seems like I have heard about successful win/loss percentages before?

[ulforlife]

As we can see below, the Texas Tech loss to Oklahoma State (OSU swept TT this weekend) has vaulted the Cajuns cleanly into the #48 spot entering the SBC Tournament.

Hey Brian...

what is the Cajuns RPI now after winning the first game to Arkansas State and how much to you think it would go up with a win against Middle Tennessee?

[GoneGolfin]

Hey Brian...

what is the Cajuns RPI now after winning the first game to Arkansas State and how much to you think it would go up with a win against Middle Tennessee?

Louisiana is sitting at #49, after having been surpassed by Notre Dame.

41 0.569 34 21 34 21 UC Irvine
42 0.568 42 15 42 15 Jacksonville
43 0.566 28 28 28 28 Florida
44 0.563 29 24 29 24 Long Beach State
45 0.562 42 19 43 19 Wichita State
46 0.559 39 16 39 16 Old Dominion
47 0.559 42 14 43 14 Notre Dame
48 0.558 25 30 25 30 Southern California
49 0.557 38 18 38 18 Louisiana-Lafayette
50 0.557 30 24 30 24 Missouri
51 0.556 33 24 33 24 East Carolina
52 0.556 38 19 38 19 North Carolina-Wilmington
53 0.555 37 19 37 19 San Francisco
54 0.555 26 30 26 30 Maryland
55 0.553 31 20 31 20 Missouri State
56 0.553 28 24 31 24 Texas Tech
57 0.553 34 17 36 17 Kent State
58 0.553 32 23 32 23 San Diego
59 0.552 32 14 35 14 Oral Roberts
60 0.552 39 17 39 17 Connecticut

Should the Cajuns defeat Middle Tennessee, I expect their rating to be at least .558 or .559 (currently .557). But where they will rank depends on how other teams do around them. See my post in another thread concerning how other school's in the vicinity of the Cajuns (w/ respect to at-large bids) are faring today (and yesterday).

Old Dominion losing today will help. Should the Cajuns win, I expect they will surpass Old Dominion. Notre Dame does not play today. I also expect the Cajuns to jump USC with a win today. Provided Wichita State holds on and wins today and East Carolina does not upset Houston tonight, I see the Cajuns likely being tied with Notre Dame (or just behind them) and positioned at about #47 (with Notre Dame at #46). But I am just guestimating and not running actual numbers.

[ulforlife]

thanks man, im trying to learn this whole RPI thing, you make it MUCH easier to follow and i think i can see how it plays out... best of luck to the cajuns, glass is going to take care of business