Not sure if I said this before but that men’s 100 free A final at NCAC’s is apparently the deepest and fastest D3 conference final in history by quite a bit. 3 under 44, and 5 under 44.5 which I can’t remember the last time that has happened if ever.
I think Chicago's freestyle relays are worth a mention - Emory's slippery gutter edge didn't make for the easiest flip turns, and we still posted a leading time in the 400 free relay about 30 minutes after my 200 fly. Additionally, our freestyle core is still very young - Alex Schwartz, John Butler, Igor Benderskii, Rylan Kruep, and I all posted personal bests on the relays and are all first/second years, and we are adding on 1:37.3 200 freestyler Misha Kojanov next year.
I'm assuming that you are waiting for this week's conference meets before updating the SRS table data. One thing to note, if I have the rules right, is that each team is limited to 18 swimmers at NCAA and only 17 if they bring divers. It might make sense to compare teams SRS totals with that constraint in mind.
Until the invite sheet comes out, I won’t know which swimmers from which teams are actually going to Nationals. Restricting the data too early adds uncertainty and guesswork, likely introducing noise rather than improving accuracy. That’s why I’m not applying this restriction in the next SRS update—which raises the question of whether it should be done at all…
Because of the issue of small sample sizes. Narrowing the dataset can increase statistical noise…you know the story. SRS is a historical measure with intriguing but not fully defined predictive power—big-picture, probabilistic, and slightly enigmatic. All of it just means that limiting the sample size in admittedly logical ways may not actually improve its accuracy.
Once we have psych sheets a week out from Nationals, running a simulation based on confirmed times should be more reliable.
Not sure if I said this before but that men’s 100 free A final at NCAC’s is apparently the deepest and fastest D3 conference final in history by quite a bit. 3 under 44, and 5 under 44.5 which I can’t remember the last time that has happened if ever.
Hey Parker - good to hear from you. Appreciate the research and helpful post.
looking forward to racing Brayden in March - going for the 100 IM national record this weekend first though!
Good luck, you're a great swimmer and I heard you're an even better person.
You got this Cooper! Swims fast and have fun!!💪💪
I think Chicago's freestyle relays are worth a mention - Emory's slippery gutter edge didn't make for the easiest flip turns, and we still posted a leading time in the 400 free relay about 30 minutes after my 200 fly. Additionally, our freestyle core is still very young - Alex Schwartz, John Butler, Igor Benderskii, Rylan Kruep, and I all posted personal bests on the relays and are all first/second years, and we are adding on 1:37.3 200 freestyler Misha Kojanov next year.
Appreciate you sharing this. Looking forward to your swims the rest of the season.
I'm assuming that you are waiting for this week's conference meets before updating the SRS table data. One thing to note, if I have the rules right, is that each team is limited to 18 swimmers at NCAA and only 17 if they bring divers. It might make sense to compare teams SRS totals with that constraint in mind.
Mike, great point. A couple of thoughts:
Until the invite sheet comes out, I won’t know which swimmers from which teams are actually going to Nationals. Restricting the data too early adds uncertainty and guesswork, likely introducing noise rather than improving accuracy. That’s why I’m not applying this restriction in the next SRS update—which raises the question of whether it should be done at all…
Because of the issue of small sample sizes. Narrowing the dataset can increase statistical noise…you know the story. SRS is a historical measure with intriguing but not fully defined predictive power—big-picture, probabilistic, and slightly enigmatic. All of it just means that limiting the sample size in admittedly logical ways may not actually improve its accuracy.
Once we have psych sheets a week out from Nationals, running a simulation based on confirmed times should be more reliable.