Who Got Faster? Improvement at D1 Men’s Conference Meets

All the D1 men’s conference meets have been completed. We know who won the meets. We know who lost. But many swimmers and programs don’t measure success by how their team placed at the meet or if they gained 50 points over last year. Many people just want to go faster. Maybe their top swimmer is injured. Maybe their recruiting class was down this year. Even for top teams, simply dropping time and performing well is enough to make a meet a success. To that end, I’ve attempted to measure who dropped the most time.

Last week I did a similar analysis for D1 women’s conferences.

Methodology

So how do we measure who dropped the most time? It’s a more complicated question than it seems. First the methodology I used.

1. Grab every conference and nationals time from the last 4 years
2. Find a swimmer’s previous best time in an event at a conference or national meet
3. For times at conference meets this year, compare to a swimmer’s previous best conference/nationals time. If there is no previous best time move on

There are pros and cons to this method. By limiting previous best times to conference and nationals, we can be pretty sure that the swimmers baseline time was a rested swim that they were focused on.  This means we’re not giving credit to swimmers who pick up a new event for getting way faster when the main explanation for the improvement was a change of focus. (for example, I switched from the 500 to the 200 IM my senior year in college. My 200 IM at conference was a 5.5 second PR, but most of that improvement was because I’d never swum it rested before. My team’s improvement percentage shouldn’t benefit from this. If I had tapered it before, the improvement would have been much smaller. Therefore, this time isn’t included). This method has the downside of leaving out best times achieved at mid season meets, but that’s a small price to pay for relatively clean data.

There are always a few swimmers  who enter the 50 and swim obscenely slow times. The most egregious tanking was tossed as outliers. For example, Penn State’s Gunter Cassell swam a 25.57 50 free. His previous best was 21.64, an add of 18%. Clearly something wasn’t on the level here. Either an injury, or saving energy for other events seem plausible (edit: apparently in this case, he was swimming breaststroke). Penn State’s ranking shouldn’t be punished for either, so it’s tossed.

Another flaw is that freshman performing extremely well makes a team do worse in this metric. Fast freshman times mean that swimmer’s baseline times are better, so it’s harder to drop time from them. Also because high school and club times aren’t included, team’s don’t get credit for freshman PR’s. Similarly, slow freshman times followed by fast times later will make a team look good. This isn’t a fatal flaw but it is worth keeping in mind while reviewing this data. Until there’s a single unified database for all swim times, I’m not sure there’s an easy fix to this issue.

Results

On average swimmers dropped .14% from their conference/nationals PR’s at conference meets this year (standard deviation 1.7% median .21%). There were 4357 data points. 2429 were faster or 56%. The data followed a pretty symmetric normal distribution (if anything there’s a slight skew to the right. positive is slower, negative is faster):

The top performing team was Mt St Mary’s which dropped an average of 1.54% on 7 swims. They are a new program with a small number of repeated swims, so an extreme result isn’t surprising. Next was a pair of major conference teams, Texas A&M and Virginia. A&M dropped an average of 1.53% on 37 swims, 84% of their swimmers were faster. Virginia dropped an average of 1.25% on 42 swims, 88% faster. The only other major conference team in the top 10 was NC State who dropped 1.04% on 33 swims, 70% faster. Making these team’s performances more impressive is the fact that among power 5 conference teams in the women’s improvement data Texas A&M were 2nd, NC State were 3rd, and Virginia were 5th.  NC State and Virginia each have the same head coach for their men’s and women’s programs, but Texas A&M do not. Jay Holmes coaches the men and Steve Bultman coaches the women. The top men’s and women’s programs did not always line up. The 4th women’s power 5 team, Georgia Tech, were 3rd to last overall among the men with an average add of 1.02%.

Among smaller conference teams, Monmouth, Fairfield, St Bonaventure, GWU, and UCSB were in the top 20 for both genders.

Three of the top eight performing men’s teams came from power 5 conferences. This is in contrast to the women where the top 9 performing teams were from smaller conferences. After the top 8 15 of the next 17 top performing men’s teams were from small conferences, so the trend of those team dominating the top of the rankings is mostly intact here.  This trend isn’t surprising as there are quite a few more small conference teams.

While the vast majority of teams fully rested for their conference meets, there are teams (ex Texas) who haven’t fully tapered yet, so this data doesn’t paint a full picture for them.  It’s worth noting that Texas added an average of .65% to their conference/nationals PR’s at their conference meet. At their conference meet last year they added an average of .14%. It appears they had a bit less rest this year than last year.

Data

Negative is faster. Positive is slower