West Regional Rankings Reviewed

The NCAA has released the first regional rankings of the season, which you can find here. Lots of interesting tidbits to be found but it’s probably best to go through it region by region. Before we do that let’s go over the basics. The regional rankings are used for NCAA Selection at-large bids. Teams that win their conference Pool A bids will be ignored as will any team that receives the Pool B bid. These bids are done first and only when we know those teams does the committee worry about the Pool C (or at-large) bids. When selection starts, the first remaining eligible team in the list is compared against the seven other regions that have identified their first eligible team. Not to bog down into this process, but you can see how important it is to be high on the regional rankings. The rankings are done by the Regional Advisory Committee (RAC) for each region using the five primary selection criteria. Prior to release the rankings are approved by the national selection committee, which consists of the RAC Chairs from each region and NCAA representatives. In theory, the national committee should be making sure the RACs follow the criteria and hopefully provide consistency between the regions. (It would be bad if one region evaluates the criteria one way while everyone else does it another way.) There is information on this in the FAQ page and you can always access the Pre-Championship Manual from there which has information on the selection criteria.

At the high level, first glance we see that the New England Region has dropped from 12 to 11 on the number of teams they can rank. New York has dropped from 11 to 10. I’m getting old and can’t remember the NCAA deciding to do this. I do remember talk of reducing the number of ranked teams but the argument against was that there was no proof that it was unfair to the smaller regions, which was a lie or a joke. Small step done here but it’s in the right direction. Regardless of the number a region is allowed to rank as long as the selection committee properly evaluates the quality of opponents, it really doesn’t matter. What does matter is that year after year they have shown they are incapable of doing this.

I want to go through everything I can but let’s start with the West Region. Here are the rankings with my prediction in parenthesis:

1. Colorado College (1)

2. Claremont-Mudd-Scripps (3)

3. Trinity (2)

4. Mary Hardin-Baylor (4)

5. Chapman (7)

6. Whitworth (6)

7. UT-Dallas (5)

8. Cal Lutheran (8)

So I got the teams correct, which actually was pretty obvious except for Cal Lutheran, but the ordering is slightly off. Taking a look at the differences, we see that CMS gets the second spot over Trinity. One thing to mention is that I had Trinity with an extra win. You have to remember that the NCAA only considers DIII wins/losses in the primary selection criteria. Where I got sideways is that Trinity has played JWU (Denver) and St. Thomas (TX) twice each and they don’t count until their third year in the division. I guess I missed one of those matches as I had the other three correct. In my write-up from Tuesday I said that CMS had the advantage in record but Trinity had the advantage in SOS. Well, the selection committee uses a different calculation for SOS than what they release on their statistics page. Looking at their datasheet provided with the rankings we can see that Trinity has an official SOS of 0.593 compared to CMS SOS of 0.584. The RACs use a standard deviation somewhere between 0.020 and 0.030 and since these numbers fall within this range they are considered statistically even. With that event happening, I can agree that CMS should receive the second spot.

The other difference is the ordering of UT-Dallas, Whitworth and Chapman. I had them in that order while the RAC has them reversed. Let’s use the official data to confirm:

Teams	Record	SOS	H2H	C. O.	R. W.
UMHB*	.750	.580	UTD	?	na
UT-Dallas	.818	.545	na	?	na
Whitworth	.778	.570	na	?	na
Chapman	.857	.541	na	?	na

* I added UMHB for a reason, which I’ll explain below.

Chapman has the best record of the three followed by UT-Dallas and then Whitworth. We don’t know the exact standard deviation the RAC is using so to determine the SOS advantage is difficult. I personally would give the advantage to Whitworth and call UT-Dallas and Chapman a wash. I’d like to note that when I was updating the official SOS numbers I saw most teams drop from what the NCAA publishes through their statistics website. The teams with the biggest drops in SOS were Cal Lutheran, Colorado College, Southwestern and UT-Dallas. What this means is that they were hurt when factoring in their opponent’s opponent records, which counts as 1/3 of the overall official SOS number as opposed to just the record of their opponents.

For this first week we are not using Ranked Wins and there were no head-to-head results between these three teams (ignore UMHB for now). Each of the teams has a common opponent in Pacific Lutheran but each of them came away with wins. Using these official numbers we can see that Chapman should be ahead of UT-Dallas. I would consider the Whitworth/Chapman comparison to be a tie, which means the RAC may have looked at secondary criteria, which I will not go into here. Whether it was this secondary criteria or the fact that the RAC favored record over SOS, I don’t know but they decided to rank Chapman over Whitworth. I will say, however, that whatever logic they used to make this decision should also have been applied to UMHB. Looking at the chart above, it seems to me that if you are going to put Chapman 5^th then you should really have them 4^th ahead of UMHB for the same reasons you have them ahead of Whitworth. (Note – It is conceivable that the secondary criteria favored Chapman over Whitworth and UMHB over Chapman. Maybe they leaned heavily on results in last 25% of their matches, which is a secondary criteria. Sorry if you have tired head.)

The UT-Dallas/Whitworth comparison is also a tie if the standard deviation isn’t 0.025 or higher in which case UT-Dallas would be ranked ahead of Whitworth. So, since they weren’t, we can assume that if a standard deviation is being used it’s 0.024 or less. I can only assume in this case that the secondary criteria favored Whitworth over UT-Dallas because otherwise they favored SOS over record which would be inconsistent from the previous examples.

Overall, we learned that an SOS in the West Region outside of 0.024 is not considered the same. That’s something. Other than that, we don’t really know how these ties were broken but giving the benefit of the doubt to the RAC, they used secondary criteria. What a nightmare for them in week 1 of this process. Just know that the difference in secondary criteria wouldn’t have been much of anything so they are really splitting hairs here and it just means that Ranked Wins could turn this ordering on its head next week.

I will say that now that I have seen the official numbers I am more concerned over the at-large chances for both Chapman and UT-Dallas. The SOS numbers for both teams is at the low end of what is normally accepted for an at-large team. They both need to keep winning is all I can say.

I’m going to look at the other regions and probably put together a podcast tomorrow on all of this so stay tuned!