At-Large Average Criteria (2023)

One of my last annual articles to do in a season is to look how the at-large teams compare, on average, with the previous years when it comes to the selection criteria. Specifically, I look at record, Strength of Schedule (SOS) and ranked results. Head-to-head and common opponent don’t really come into play here because these vary between the two teams being compared. Ranked results is a bit of a problem when I do this because I’m only looking at ranked record, which is not how the criterion is specified. To be clear, I’m not looking at the quality of wins and losses when I do this comparison. With that said, I do believe ranked record has been incorrectly used in the selection process in the past so why not add it to this comparison? My data goes back to 2017.

In 2017, the average for all of the at-large teams for record was 0.776. For SOS it was 0.603 and for ranked record it was 0.554. In 2018, the averages were 0.744 (record), 0.612 (SOS) and 0.526 (ranked record). In 2019, the averages were 0.778 (record), 0.610 (SOS) and .532 (ranked record). In 2021, the averages were 0.788 (record), 0.603 (SOS) and 0.579 (ranked record). In 2022, the averages were 0.744 (record), 0.620 (SOS) and 0.537 (ranked record). Remember, there was no NCAA Tournament in 2020. The data remains pretty consistent. Win ¾ of your matches, get your SOS to 0.600 and win over half of your matches against ranked opponents and you’ve got yourself an at-large bid.

Here are the 2023 numbers:

Record  0.761 SOS  0.604 Ranked Record  0.533

These numbers are similar to the 2019 numbers although slightly lower in both record and SOS. For this year, the SOS was a bit on the low end along with the ranked record.

This past season, we had 3 at-large teams that met or exceeded all three of these criteria averages. They were Babson, Emory and UW-Oshkosh. We had five teams that met or exceeded two of the criteria averages this year. They were Gettysburg, La Verne, Mary Washington, Mount St Joseph and Springfield. As is typical, all five teams missed out on the SOS number. This is the second year in a row where only 8 teams met or exceeded 2 or more of the criteria. In previous years, we averaged 10 teams. We still have never seen a team get bypassed for an at-large bid when they meet or match at least 2 of these 3 criteria averages. The only team of the 8 that I was worried about for an at-large was La Verne. Based on history, I shouldn’t have. I keep trying to remember to use this statistic when I do my projections but always forget.

Most of the at-large teams meet at least one of the criteria averages but there are one or two teams each year that fail to meet any that get selected. We had four this year, which is an increase from the three last year. They were Calvin, CWRU, Mount Union and Stevenson.

For the teams that met just one criterion, it was almost always SOS that they achieved. Of the nine  teams meeting only one average, eight met the SOS number. St Olaf was the exception as they met only Ranked Record. This is another indication that SOS is generally favored by the Selection Committee over the other criteria although it’s not specifically stated to be that way.

Some tidbits about our at-large teams:

• Tufts had the worst record at 0.583. Looking at the teams that were on the bubble based on regional rankings, all of them had a better record.

• La Verne had the worst SOS at 0.541. Nine of the ten bubble teams had a better percentage.

• Christopher Newport had the worst Ranked Record at 0.250. Eight of the ten bubble teams had a better percentage.

Looking at the bubble teams, and I define these as the teams next up in their regions, we see that only one team met one of the averages while the other nine met none. That team was UW-Platteville who I just did a Case Study on showing that they probably should have received a bid. This is the first year where only one team met a criterion. (Note – I really didn’t know who Region VIII had next. I assumed UW-Platteville.)

Here were the bubble team’s averages and how far that was from the at-large average:

Record  0.687 (-0.074) SOS  0.581 (-0.023) Ranked record  0.372 (-0.161)

Tidbits for the bubble teams:

• Colorado College had the lowest record at 0.621.

• SUNY New Paltz had the lowest SOS at 0.532.

• Centre had the lowest ranked record at 0.111.

It’s really interesting to me that UW-Platteville exceeded the SOS average of the at-large teams but didn’t get a bid. That’s pretty rare but happens when the other two criteria are really bad. That wasn’t the case for UWP with regards to record.

Not much else to write about this. I think it’s important for coaches to know the criteria thresholds that get teams at-large bids. It really shouldn’t work this way but history says it does. You can believe the written rules or how they are acted upon, I guess.

Here is the raw data used above: