NPI Simulator

[Kuiper]

The D3Datacast has created an NPI simulator for Men's Soccer.  They've been doing it for basketball since the NPI was introduced and have now extended it to other sports.  It is, of course, way too early for anyone to think about these types of things.  Not enough games played, very few conference games played, etc.  It's practically irresponsible to place any weight on it. 

Still, it is available, so I figured I might as well link it here so you can follow and rant about the craziness of it all (and, seriously, the AQ part is the most crazy thing about the simulator).  I won't post actual results until everyone has at least a few more games.

D3 Datacast Men's Soccer NPI Simulator

NCAA Division III utilizes the NCAA Power Index (NPI) to produce national rankings for all team sports. The final NPI rankings will determine who is awarded at-large bids and how teams are seeded for the 2026 NCAA Division III men's soccer tournament.

This is a replication of the NCAA's NPI calculations for Division III men's soccer. These numbers are unofficial but should be very very close to the official data used to select and seed teams for the NCAA Division III tournament in men's soccer. D3 Datacast provides NPI updates before and between official releases, which will be posted at stats.ncaa.org when available.

For simplicity's sake, this view generally designates AQ (automatic qualifiers, formerly called Pool A) as the the team with the highest conference winning percentage (using NPI as a tiebreaker). At-Large (formerly Pool C) then ranks the remaining teams by NPI. Officially, all automatic qualifiers will eventually be awarded by each conference according to league rules.

[ziggy]

Just to add some clarity for anyone who sees the word "simulator" used here and thinks this includes any forward-looking projection, that is not the case.

We are simply providing a daily tracker of how the NPI calculation stands as of current games played (with the necessary caveat that even though we are pulling results from the NCAA, errors and omissions can and do occur and may at times require cleanup on our end to correct the data set.)

Obviously, there are no real AQs until they are actually awarded but you can't get a view of the at-large cutline without first accounting for the AQs. What we have set up simply allows for identifying around where the cutline will be, which will make more sense and be more helpful later in the season.

[LibbyMoore]

This is so interesting, and will be great to follow along with! Thanks!

Just to add some clarity for anyone who sees the word "simulator" used here and thinks this includes any forward-looking projection, that is not the case.

We are simply providing a daily tracker of how the NPI calculation stands as of current games played (with the necessary caveat that even though we are pulling results from the NCAA, errors and omissions can and do occur and may at times require cleanup on our end to correct the data set.)

Obviously, there are no real AQs until they are actually awarded but you can't get a view of the at-large cutline without first accounting for the AQs. What we have set up simply allows for identifying around where the cutline will be, which will make more sense and be more helpful later in the season.

[Kuiper]

With most teams having 3-6 games under their belt and conference play getting underway, the NPI simulator is starting to be a bit more useful.  There are still small data anomalies, such as Olivet ranked the #17 At Large team with only 1 countable game against an NCAA D3 opponent (a 1-0 win over Anderson).  The Automatic Qualifier projections on NPI score can also seem optimistic.  For example, there are teams highly ranked due to only a couple of games such as Puget Sound being projected as the NWC AQ on the strength of 2 games, although the 3-0 win over Lake Forest suggests they could be competitive this year, or teams that have already started conference play off on the wrong foot, such as Chapman being listed as the AQ in the SCIAC despite winning only 1 of their first 4 games and tying their first SCIAC game against Caltech only after a desperate comeback that involved scoring two goals in the last two minutes.

Nevertheless, with only 22 at large bids, we can already see how some teams with great records against weaker non-conference schedules could be on the outside looking in if they don't win their conference AQs.  This group includes McDaniel (4-0-1), Haverford (5-0-1), Wabash (7-1), Hampden-Sydney (4-0-1), Salisbury (5-0-1), and Bard (6-0).  That will depend upon whether their respective conference schedules will help or hurt their SoS and whether their SoS improves because some of their non-conference opponents get into conference play and they improve their own SoS by amassing some wins and playing against higher ranked teams.