The Book: 2002 Baseball

With respect to the RPI
Turbine said:
What exactly flaws it?
--
The RPI is flawed because of the regional nature of college baseball. College baseball lacks sufficient inter-regional play to make the tool worthwhile in my opinion. Remember that your RPI score is based on your W-L record, your opponents' W-L record, and your opponents' opponents' W-L record.

If you have a bunch of teams on the West Coast that are very good, but all they do is beat up on one another, you are going to have low RPIs. On the other hand, northeastern teams play predominantly against each other. W-L records are inflated because of the weak regional competition. Since these schools do not appropriately mix games with other parts of the country, you do not get a true reading as to the strength of a given team. This is why you see a number of northeastern schools with inflated RPIs. You also see this in smaller conferences on the east coast. On the other hand, western teams tend to get brutalized. The ISRs seem to do a decent job of offsetting these shortcomings.

If you had a round-robin schedule where each team played each other team the same number of times, you would get an accurate reading (Ex. MLB, or very close).

Did I explain it well enough? I can provide an example if you like.

Brian

Quote:

---

Originally posted by gonegolfin
Did I explain it well enough? I can provide an example if you like.

---

Very clear and concise, although an example would be nice.

Also I see the NE and FW scenario, so where does mid south and the Sunbelt come in as far as rpi accuracy is concerned?

but why not have the following regionals out west?

Palo Alto:
1. Stanford
2. Long Beach State
3. San Jose State
4. BYU

Mesa:
1. Houston
2. Arizona State
3. California State-Fullerton
4. New Mexico State

Los Angeles:
1. Southern California
2. California State-Northridge
3. San Diego
4. Maine

I did not introduce any new regional sites to the picture nor did I move any of the current western regional teams to regionals outside of the west. I simply moved some teams around in the three western regionals. This more evenly distributes the #4 seeds while requiring the same travel overhead.

BYU has about the same flight to Palo Alto as it does to Los Angeles. It is about the same trip from Fullerton to Palo Alto as it is from Fullerton to Mesa. So why not send CSF to Mesa as the #3 seed? San Diego would then have a 200 mile shorter trip to Los Angeles (as opposed to Mesa). Maine stays in Los Angeles, but now as the #4 seed.

Stanford, a national seed, now plays a real #4 in the first round and the Palo Alto regional is appropriately weakened. Los Angeles is slightly tougher, but still easy. Mesa is more difficult, but not nearly as difficult as the current regional in Palo Alto. Besides, Houston is not one of your national seeds. Although I think they should have been. I have to wonder if the NCAA made Stanford a national seed just to ensure a western super-regional. National seeds and #1 seeds are supposed to be awarded on merit.

Brian

How much of an effect will all of this have on crowning the true national champion? Will tougher regionals wear some teams down?

Quote:

---

Originally posted by Turbine

Very clear and concise, although an example would be nice.

Also I see the NE and FW scenario, so where does mid south and the Sunbelt come in as far as rpi accuracy is concerned?

---

--
This is an extreme example, but it better illustrates my point.

Let's say we have two pools of teams, one from the northeast and one from the west coast. Let's also assume that the teams from the west are very strong and that the teams from the northeast are very weak. Finally, teams can only play games within the pools (consider these the regions) and play a twelve game schedule (4 games against each opponent alternating home/road) with the W-L records below.

Pool A:
Central Connecticut State (10-2)
Harvard (9-3)
Northeastern (5-7)
Stony Brook (0-12)

Pool B:
Stanford (8-4)
Southern California (6-6)
California State-Fullerton (5-7)
Arizona State (5-7)

Your RPI rankings are simply going to be rankings by W-L record. That is,

1. Central Connecticut State
2. Harvard
3. Stanford
4. Southern California
5. Northeastern
California State-Fullerton
Arizona State
8. Stony Brook

The system is flawed because there is no inter-regional play. We know that all of the west coast schools would beat the pants off of the northeastern schools in this example. But according to the RPI, the two strongest teams are Central Connecticut State and Harvard.

Now, a little inter-regional play will bring us closer to the truth. But to achieve truly accurate results with the RPI, you must have perfect inter-regional play. This is why the RPI is even less effective in college baseball than in college basketball. Because of costs, inter-regional college baseball is minimal when you take into account the entire schedule played.

As far as the south and sun belt is concerned, I think they are slightly inflated because of some of the southern tours some northern schools make. But it is cetainly not signficantly inflated. The accuracy of an individual schools's RPI is really going to be dependent on the schedule that school plays.

Another problem with the RPI is that you can boost your RPI in many cases simply by playing (win or lose) to high RPI teams. This does not measure a damn thing. But this is the reason why I have always voiced the opinion that UL should play the toughest schedule they can afford budget-wise. Road games are especially good since bonuses are worked into the RPI formula for road wins against highly rated RPI schools.

I do think that FIU's RPI is somewhat inflated. They schedule smart with respect to the RPI. They play many home games and some games against northern schools early in the season. They also occassionally hit the road against a highly rated RPI team that can potentially get them bonus points. The massive number of home games is not a penalty in the RPI system.

As for UL this season, I think their pRPI is slightly inflated. I think the ISR is more accurate. I would put the Cajuns somewhere between 22 and 27.

Brian

Quote:

---

Originally posted by Turbine
How much of an effect will all of this have on crowning the true national champion? Will tougher regionals wear some teams down?

---

I do not think that it will wear down teams for the super-regional round the following week as there is enough recovery time between the regional and super-regional (unless rain delays ensue) . But I do think that it tests the pitching depth of all of the teams in that regional in a way that is not tested in other regionals (which does not promote a consistent theme throughout the regionals). Hence, it would give an unfair advantage to a team with superior pitching depth.

Whether this really has an effect on the eventual national champion is arguable. But I still think there is much to be gained by simply winning a regional championship ($$$ and reputation).

As an example, I think that this is the type of regional in which LSU probably does not advance if you replace Stanford with the Tigers. Not that LSU does not have the best team. But they really only have a couple of solid starters and one reliever (3 pitchers). I do not consider Petit to be a regional caliber starter that could beat a CSF.

Brian

Quote:

---

Originally posted by Turbine
[B]Tallahassee, Fla.
1. Florida State*
2. <b>Central Florida</b> (reg season opponent 3-0)
3. South Florida
4. Stetson
]

---

The Cajuns also played and beat South Florida 6-2 in the Rice Invitational early in the year. The Cajuns are 18-12 versus teams in the 2002 NCAA Tournament.

Brian

Wow 30 games against field of 64 teams.

So if you take away the 5-1 record in the BR regional.
Louisiana is still 13-11 against the field of 64 opponents.

The following is a breakdown of the Baton Rouge regional schools w/respect to their schedule against schools making the field of 64. I call it, battle tested.

#1 Louisiana State
National Seeds: (9 games) 5-4
#1 Seeds: (15 games) 8-7
#2 Seeds: (12 games) 8-4
#3 Seeds: (5 games) 4-1
#4 Seeds: (1 game) 1-0
--
Totals (33 games) 21-12

#2 Louisiana
National Seeds: (2 games) 1-1
#1 Seeds: (13 games) 5-8
#2 Seeds: (7 games) 6-1
#3 Seeds: (5 games) 3-2
#4 Seeds: (5 games) 4-1
--
Totals (30 games) 18-12

#3 Tulane
National Seeds: (0 games) 0-0
#1 Seeds: (6 games) 1-5
#2 Seeds: (2 games) 0-2
#3 Seeds: (3 games) 2-1
#4 Seeds: (0 games) 0-0
--
Totals (11 games) 3-8

#4 Southern
National Seeds: (0 games) 0-0
#1 Seeds: (1 game) 0-1
#2 Seeds: (2 games) 1-1
#3 Seeds: (0 games) 0-0
#4 Seeds: (0 games) 0-0
--
Totals (3 games) 1-2

Brian

<center><img src="https://forumeus.com/images/2002tiguemoorefield(450).gif"><br><b>"Geaux Louisiana"</b></center>