NCAA Power Index (NPI)

[Inkblot]

Last week at the convention, Division II approved a move to the NCAA Power Index (NPI) – which has been used in Division III since 2024–25 and in Division I hockey beginning this season. Starting in 2026–27, the NPI will be the sole criterion for selection and seeding (within each region). This will remove the inherent subjectivity of using a committee to evaluate multiple metrics and non-numerical criteria such as head-to-heads.

The NPI runs a calculation that's repeated as many times as necessary to stabilize every team's rating. Within the calculation, there are multiple factors that can be weighted as desired. Each sport's committee will choose settings ("dials") for these weights by this summer; these dials can be changed every 3 years or so. For example, one dial is the relative weighting of winning percentage and SOS; one sport might choose 20/80 and another might choose 25/75.

One key feature of NPI is that losses that would raise a team's rating are discarded, and wins that would lower it are discarded once a minimum win total threshold (one of the settable dials) is reached.

I wrote a detailed explanation of how the calculation works here (not paywalled): https://www.patreon.com/posts/148466259

[Buffalo/Islander Alum]

Now the game is will we know which dials and how much weight?

[Inkblot]

Now the game is will we know which dials and how much weight?

The dials should at some point be added to this table: https://ncaaorg.s3.amazonaws.com/cha...2_NPIQandA.pdf

[Redwing]

The dials should at some point be added to this table: https://ncaaorg.s3.amazonaws.com/cha...2_NPIQandA.pdf

So, does this mean each regional comm. can decide it's own dial and weighting or will there just be one "setting" for all of D2? I sure hope that it's consistent for all of D2 and not 4 different ones...one for each region. Uggh

[KleShreen]

My problem with this is that there needs to be human subjectivity in most D2 regions. Every region has conferences that are objectively better and objectively worse than the others. But under this, they'll all be treated basically the same.

[Inkblot]

So, does this mean each regional comm. can decide it's own dial and weighting or will there just be one "setting" for all of D2? I sure hope that it's consistent for all of D2 and not 4 different ones...one for each region. Uggh

Each sport selects their own dials. It'll be the same for every region within that sport.

[crixus]

My problem with this is that there needs to be human subjectivity in most D2 regions. Every region has conferences that are objectively better and objectively worse than the others. But under this, they'll all be treated basically the same.

Short and sweet without Legalese. Thank you.

[Buffalo/Islander Alum]

My problem with this is that there needs to be human subjectivity in most D2 regions. Every region has conferences that are objectively better and objectively worse than the others. But under this, they'll all be treated basically the same.

I agree with your statement for the most part. I think the "frustration" has always been how some final teams were chosen and why wasnt head to head used as a tiebreaker.

[Tony Nicolette]

I agree with your statement for the most part. I think the "frustration" has always been how some final teams were chosen and why wasnt head to head used as a tiebreaker.

Would need an example of where/when a team with H2H was left out or didn't "tiebreak" as I'm not recalling one. The thing to keep in mind is that H2H in the D2 selection process previously was ONLY a tiebreaker...and only in "certain situations" at that.

Most folks view H2H as the primary way to pick between two teams with identical records because that's what it is the NFL. But the thing to keep in mind is that the NFL's selection process has always been a math problem. It's a series of "if-then" statements from here to eternity to determine who gets each spot. The math runs everything, plain and simple. The D2 process has never been that. There are math/number-based qualities that are assigned to and/or are earned by teams which results in a "resume". The committees then review those "resumes" to determine a seeding of those teams. In most instances, regional committees would only consider H2H IF the two teams in question finished "next to each other" in the seedings that they determined upon the resume reviews.

As an example, if Team A and Team B finish the season 9-2 and they played each other, the only time H2H would typically have mattered is if their 9-2 "resumes" wound up putting them on bracket lines that are next to each other. If Team A's resume earned them the 4-line and Team B's the 6-line, even if Team B defeated Team A during the season they would not "jump" Team A. They would have to be (in this example) on the 4 and 5 or 5 and 6 respectively. I'm not saying it's right/wrong/indifferent...am just noting that for the most part that is how it worked.

So, again, I'm not recalling off-hand any instances where a team got left out of the bracket despite having the same (or better) record than a team that they beat that got in but I'm sure it has happened. With that said, because H2H wasn't the primary component of an "if-then" sequence that is pretty much why something like that could happen.

This new methodology will be interesting. I appreciate how the NFL does it, but also don't believe that it would be viable for college. The lack of subjectivity that this new format will present might be fine. Based on what I have seen in covering D2 for 20 years and watching it intently for longer than that, I definitely have some concerns. Guess we'll see sooner than we all think!

[Buffalo/Islander Alum]

Would need an example of where/when a team with H2H was left out or didn't "tiebreak" as I'm not recalling one. The thing to keep in mind is that H2H in the D2 selection process previously was ONLY a tiebreaker...and only in "certain situations" at that.

Most folks view H2H as the primary way to pick between two teams with identical records because that's what it is the NFL. But the thing to keep in mind is that the NFL's selection process has always been a math problem. It's a series of "if-then" statements from here to eternity to determine who gets each spot. The math runs everything, plain and simple. The D2 process has never been that. There are math/number-based qualities that are assigned to and/or are earned by teams which results in a "resume". The committees then review those "resumes" to determine a seeding of those teams. In most instances, regional committees would only consider H2H IF the two teams in question finished "next to each other" in the seedings that they determined upon the resume reviews.

As an example, if Team A and Team B finish the season 9-2 and they played each other, the only time H2H would typically have mattered is if their 9-2 "resumes" wound up putting them on bracket lines that are next to each other. If Team A's resume earned them the 4-line and Team B's the 6-line, even if Team B defeated Team A during the season they would not "jump" Team A. They would have to be (in this example) on the 4 and 5 or 5 and 6 respectively. I'm not saying it's right/wrong/indifferent...am just noting that for the most part that is how it worked.

So, again, I'm not recalling off-hand any instances where a team got left out of the bracket despite having the same (or better) record than a team that they beat that got in but I'm sure it has happened. With that said, because H2H wasn't the primary component of an "if-then" sequence that is pretty much why something like that could happen.

This new methodology will be interesting. I appreciate how the NFL does it, but also don't believe that it would be viable for college. The lack of subjectivity that this new format will present might be fine. Based on what I have seen in covering D2 for 20 years and watching it intently for longer than that, I definitely have some concerns. Guess we'll see sooner than we all think!

2007 Carson Newman was left out with a 10-1 record because Catawba beat them in the final seconds for their only loss and Catawba ended up going in..

Of course back then earned access from the HBCU schools (Shaw) hurt teams in that position a lot.

This year several people thought Western Oregon should have gotten in over UTPB due to their head to head win but lack of D2 games hurt them.

[Inkblot]

Head-to-heads will not have any special significance in the new system. I expect that to be one of the most controversial things about it.

[Buffalo/Islander Alum]

Head-to-heads will not have any special significance in the new system. I expect that to be one of the most controversial things about it.

Yeah the only times head to head was really significant was when a team was on the bubble line for the last playoff spot or "back in the day" when it made a difference on the 4/5 line for a home game before teams 5-7/8 were unseeded.

At least with this new system we should know with way more accuracy what to expect instead of seeing Head to Head, KPI, SOS, etc. applied more in some regions and not in others.

[Inkblot]

Yeah the only times head to head was really significant was when a team was on the bubble line for the last playoff spot or "back in the day" when it made a difference on the 4/5 line for a home game before teams 5-7/8 were unseeded.

At least with this new system we should know with way more accuracy what to expect instead of seeing Head to Head, KPI, SOS, etc. applied more in some regions and not in others.

Last fall there was someone on Twitter who called the NCAA a joke for selecting Northwood, and not Findlay, for the women's soccer regional, when Findlay was 1-0-1 against Northwood and finished higher in the conference standings...

It wasn't close in the slightest. Northwood had much better metrics and as a result got the #5 seed (out of 8). The conference standings might've been meaningful, except that the GMAC had an unbalanced schedule and Northwood had to play the three toughest opponents twice whereas Ashland played the bottom two twice. Not to mention that they had one common opponent in non-conference, whom Northwood beat and Findlay lost to.

[Predatory Primates]

Pitt and NW had /the same record last year and NW had to travel because of the HtoH. What would be the determiner under the new system?

[Inkblot]

Pitt and NW had /the same record last year and NW had to travel because of the HtoH. What would be the determiner under the new system?

I think any version of NPI would have Pitt higher due to having played a tougher schedule.