I acquired all of the Division I team data since 2002, and from that we can observe some fascinating trends and data relationships in the data. This is a multipart series exploring some of that data.
In the last 23 years the NCAA has invited 65-68 teams to its tournament. While many teams earned an automatic bid by winning their conference championship, others were invited to participate, nevertheless. If we look at those 351 teams, we begin to see a profile for what the NCAA typically includes.
Multiple regression analyses are designed to estimate a variable with an infinite range. When estimating a Boolean (True/False), there is a different formula, and the result is always a probability number ranging between 0.0 and 1.0.
If we look at the 351 teams' Points Per Possession (Smith Method) and Rebounding Differential, we can create a formula from which we can calculate a probability estimate of making the tournament based on the typical NCAA tournament team's profile. Plugging in the current teams' values gives us a better feel for the likelihood of these teams' inclusion in March Madness.
Following this post is the table of 351 teams this season, based on play thru the end of 2/22/24. Don't go to Vegas with the numbers in the table that follows, however the order of teams is pretty interesting, especially when you consider the conferences involved. Of course the top teams in the country should be closer to 1.0 and probably 200 teams should have a probability score of 0.0, but the order here is what is interesting. Also of note, the 68th-team cutoff is currently at 0.400, so pretty much teams above 0.400 are likely to make the tournament (with the exception of 2nd place teams in minor conferences, NET weights on good wins/bad losses, etc).
When the conference tournaments start, it will be especially interesting to watch these tournaments:
Next up: Possession Differential (Rebounding)
In the last 23 years the NCAA has invited 65-68 teams to its tournament. While many teams earned an automatic bid by winning their conference championship, others were invited to participate, nevertheless. If we look at those 351 teams, we begin to see a profile for what the NCAA typically includes.
Multiple regression analyses are designed to estimate a variable with an infinite range. When estimating a Boolean (True/False), there is a different formula, and the result is always a probability number ranging between 0.0 and 1.0.
If we look at the 351 teams' Points Per Possession (Smith Method) and Rebounding Differential, we can create a formula from which we can calculate a probability estimate of making the tournament based on the typical NCAA tournament team's profile. Plugging in the current teams' values gives us a better feel for the likelihood of these teams' inclusion in March Madness.
Following this post is the table of 351 teams this season, based on play thru the end of 2/22/24. Don't go to Vegas with the numbers in the table that follows, however the order of teams is pretty interesting, especially when you consider the conferences involved. Of course the top teams in the country should be closer to 1.0 and probably 200 teams should have a probability score of 0.0, but the order here is what is interesting. Also of note, the 68th-team cutoff is currently at 0.400, so pretty much teams above 0.400 are likely to make the tournament (with the exception of 2nd place teams in minor conferences, NET weights on good wins/bad losses, etc).
When the conference tournaments start, it will be especially interesting to watch these tournaments:
- America East: UMass (Lowell)/Vt.
- ASUN: Lipscomb/Stetson/E.KY - these are all dogs
- A10: Dayton/Richmond/
- Big Sky: Weber St./E. Washington
- Big South: High Points/Winthrop
- Big West: UC Irvine's to lose
- CAA: UNCW/Hofstra/Drexel
- ConfUSA: La. Tech/Liberty
- Horizon: Purdue(Ft.Wayne)/Youngstown St.
- Ivy: Princeton/Cornell
- Metro Atlantic: Marist/Quinnipiac/Fairfield - all dogs
- Mid-American: Akron's to lose (maybe Ohio?)
- Mid-Eastern: Norfolk St./NCCentral
- Missouri Valley: Indiana St/Drake/
- Mountain West: NM/Col St/NV/UT St/SDSU/Boise St. - all in the top 66
- Northeast: Cen Conn St./Merrimack - all dogs
- Ohio Valley: Morehead St's to lose
- Patriot: Colgate's to lose
- Southern: Samford's to lose, though UTChatt, W. Carolina, UNC-G all decent
- Southland: McNeese's to lose
- Southwestern: Southern U's to lose - a total dog
- Summit: S Dakota St/N Dakota/N Dakota St. - all dogs
- Sun Belt: JMU/App St/ Troy - very interesting trio
- WAC: Grand Canyon's to lose
- West Coast: Gonzaga/St. Mary's/San Francisco
Next up: Possession Differential (Rebounding)
Last edited:
