Olympic Trials Predictions (With Math!)

by Andrew Mering 77

June 24th, 2016

The psych sheet is finally here, so we have essentially all the pre Trials information we are going to get. I constructed a rough statistical model of Olympic Trials. The two variables it uses are a swimmer’s long course times since September last year and their seed time on the psych sheet. I used this information to calculate an expected time for each swimmer, and a percentage chance of qualifying individually in each event.

A few notes on the predictions:

-The strongest favorite to make the team is Katie Ledecky in the 800 Free at 96%. For the men it’s Michael Phelps in the 200 IM at 85%

-The weakest favorite to make the team is Micah Lawrence in the 200 Breast at 39%. For the men it’s Connor Jaeger in the 400 Free at 45%

-The highest seed the model picks in the top 2 is 5th; Lilly King in the 200 Breast

-The highest seed the model picks in the top 8 is 19th; Clara Smiddy in the 200 Back

-The model ranks  every 1 seed in the top 2. The lowest it picks an 2 seed is 5th; Missy Franklin in the 100 Free

-Most of the predicted times are slower than seed. In 2012 only 19% of women and 32% of men beat their seed time. For example, the top 8 seeds in the 2012 women’s 100 free:

 Psych Final Missy Franklin 53.63 54.15 Natalie Coughlin 53.67 54.44 Dana Vollmer 53.94 54.61 Allison Schmitt 53.94 54.30 Jessica Hardy 54.14 53.96 Amanda Weir 54.14 54.41 Megan Romano 54.16 54.72 Lia Neal 54.35 54.33

Here are the model’s full predictions:

Men

50 Free:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 Adrian, Nathan 21.37 21.40 63% 21% 84% 2 3 Ervin, Anthony 21.55 21.67 15% 26% 42% 3 2 Dressel, Caeleb 21.53 21.69 13% 25% 38% 4 4 Schneider, Josh 21.8 21.87 4% 11% 14% 5 5 Jones, Cullen 21.83 21.88 4% 10% 13% 6 6 Chadwick, Michael 22.03 22.10 <1% 2% 3% 7 9 Powers, Paul 22.18 22.15 <1% 1% 2% 8 11 Copeland, William 22.25 22.21 <1% 1% 1%

100 Free:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 Adrian, Nathan 48.0 47.91 52% 18% 70% 2 2 Phelps, Michael 48.45 48.53 11% 14% 25% 3 5 Schneider, Josh 48.76 48.77 5% 8% 13% 4 4 Dressel, Caeleb 48.74 48.85 4% 7% 10% 5 7 Chadwick, Michael 48.87 48.88 3% 6% 9% 6 8 Lochte, Ryan 48.9 48.88 3% 6% 9% 7 3 Ervin, Anthony 48.71 48.89 3% 6% 9% 8 11 Conger, Jack 49.02 48.94 3% 5% 8%

200 Free:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 2 Dwyer, Conor 1:45.41 1:45.39 44% 22% 66% 2 1 Lochte, Ryan 1:45.36 1:45.8 29% 23% 52% 3 3 Rooney, Maxime 1:47.1 1:47.17 5% 8% 13% 4 4 Grothe, Zane 1:47.11 1:47.54 3% 5% 8% 5 9 Weiss, Michael 1:47.63 1:47.6 2% 5% 7% 6 7 Haas, Townley 1:47.55 1:47.61 2% 4% 7% 7 10 Klueh, Michael 1:47.73 1:47.65 2% 4% 6% 8 12 Smith, Clark 1:47.97 1:47.67 2% 4% 6%

400 Free:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 Jaeger, Connor 3:44.81 3:45.65 25% 19% 45% 2 3 Dwyer, Conor 3:46.09 3:45.77 24% 19% 43% 3 2 Grothe, Zane 3:45.98 3:46.46 16% 16% 32% 4 4 McBroom, Michael 3:46.69 3:46.5 15% 15% 31% 5 5 Smith, Clark 3:47.1 3:46.91 12% 13% 25% 6 6 Haas, Townley 3:48.69 3:49.45 2% 3% 5% 7 8 Sweetser, True 3:49.33 3:49.46 2% 3% 5% 8 7 Shoults, Grant 3:48.91 3:49.97 1% 2% 4%

1500 Free:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 Jaeger, Connor 14:41.2 14:44.4 53% 24% 77% 2 2 Wilimovsky, Jordan 14:53.12 14:50.53 25% 28% 53% 3 3 McBroom, Michael 14:56.17 14:54.96 13% 21% 34% 4 5 Smith, Clark 15:05.97 15:02.12 4% 9% 13% 5 4 Ryan, Sean 15:03.82 15:06.29 2% 5% 6% 6 6 Gemmell, Andrew 15:07.82 15:08.24 1% 3% 4% 7 7 Sweetser, True 15:10.73 15:12.3 <1% 2% 2% 8 13 Abruzzo, Andrew 15:15.99 15:12.65 <1% 1% 2%

100 Back:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 Plummer, David 52.51 52.46 42% 30% 72% 2 3 Murphy, Ryan 52.57 52.57 33% 32% 64% 3 2 Grevers, Matt 52.54 52.75 22% 28% 51% 4 4 Pebley, Jacob 53.57 53.52 3% 7% 10% 5 5 Godsoe, Eugene 53.96 54.11 <1% 1% 2% 6 6 Conger, Jack 54.09 54.30 <1% 1% 1% 7 7 Kaliszak, Luke 54.23 54.41 <1% <1% 1% 8 10 Mulcare, Patrick 54.5 54.56 <1% <1% <1%

200 Back:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 2 Murphy, Ryan 1:54.94 1:54.99 41% 24% 65% 2 1 Clary, Tyler 1:54.73 1:55.34 29% 25% 54% 3 3 Pebley, Jacob 1:56.29 1:56.02 14% 18% 32% 4 4 Lochte, Ryan 1:56.47 1:56.8 5% 9% 14% 5 5 Lehane, Sean 1:57.11 1:56.94 4% 8% 12% 6 6 Grevers, Matt 1:57.24 1:57.28 3% 5% 8% 7 7 Mulcare, Patrick 1:57.34 1:57.67 1% 3% 4% 8 9 Owen, Robert 1:57.96 1:57.74 1% 3% 4%

100 Breast:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 Miller, Cody 59.51 59.56 27% 19% 46% 2 2 Fink, Nic 59.52 59.67 22% 18% 39% 3 3 Wilson, Andrew 59.65 59.73 19% 17% 36% 4 4 Cordes, Kevin 59.7 59.95 12% 13% 24% 5 5 Tierney, Sam 1:00.15 1:00.2 6% 8% 14% 6 10 Prenot, Josh 1:00.66 1:00.4 4% 5% 9% 7 6 McHugh, Brendan 1:00.31 1:00.54 2% 4% 6% 8 11 Andrew, Michael 1:00.68 1:00.62 2% 3% 5%

200 Breast:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 2 Prenot, Josh 2:08.58 2:08.4 34% 24% 58% 2 1 Cordes, Kevin 2:07.86 2:08.46 32% 25% 57% 3 3 Fink, Nic 2:08.89 2:09.11 17% 20% 37% 4 4 Miller, Cody 2:09.08 2:09.8 8% 13% 21% 5 6 Licon, Will 2:10.02 2:10.36 4% 8% 12% 6 5 Wilson, Andrew 2:09.84 2:10.88 2% 4% 6% 7 7 Johnson, BJ 2:10.77 2:11.05 2% 4% 5% 8 9 Whitley, Reece 2:11.3 2:11.94 <1% 1% 2%

100 Fly:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 Phelps, Michael 50.45 50.69 56% 22% 78% 2 2 Shields, Tom 51.03 51.12 22% 27% 48% 3 3 Conger, Jack 51.33 51.47 9% 15% 24% 4 5 Lochte, Ryan 51.55 51.74 4% 8% 12% 5 4 Phillips, Tim 51.49 51.75 4% 8% 12% 6 6 Josa, Matthew 51.68 51.89 2% 6% 8% 7 8 Smith, Giles 51.92 51.93 2% 5% 7% 8 9 Nolan, David 52.15 52.25 1% 2% 2%

200 Fly:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 Phelps, Michael 1:52.94 1:53.71 65% 18% 84% 2 2 Conger, Jack 1:54.54 1:55.34 13% 23% 36% 3 3 Shields, Tom 1:55.09 1:55.77 7% 16% 23% 4 4 Clary, Tyler 1:55.42 1:55.88 6% 14% 21% 5 6 Kalisz, Chase 1:56.5 1:56.51 3% 7% 10% 6 5 Seliskar, Andrew 1:55.92 1:56.58 2% 7% 9% 7 8 Clark, Pace 1:56.84 1:56.85 2% 5% 6% 8 7 Whitaker, Kyle 1:56.67 1:57.45 1% 2% 3%

200 IM:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 Phelps, Michael 1:54.75 1:55.59 62% 23% 85% 2 2 Lochte, Ryan 1:55.81 1:56.61 24% 34% 58% 3 3 Dwyer, Conor 1:57.41 1:57.85 6% 14% 20% 4 4 Prenot, Josh 1:58.38 1:58.46 2% 7% 10% 5 6 Kalisz, Chase 1:58.73 1:58.74 2% 5% 7% 6 5 Licon, Will 1:58.43 1:58.8 1% 5% 6% 7 11 Nolan, David 1:59.4 1:59.32 1% 2% 3% 8 10 Bentz, Gunnar 1:59.19 1:59.36 1% 2% 3%

400 IM:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 2 Kalisz, Chase 4:09.62 4:10.09 42% 22% 64% 2 1 Clary, Tyler 4:09.03 4:11.52 21% 21% 42% 3 4 Lochte, Ryan 4:12.66 4:12.62 12% 15% 27% 4 5 Prenot, Josh 4:13.15 4:13.02 9% 13% 23% 5 3 Litherland, Jay 4:12.43 4:13.06 9% 13% 22% 6 8 Grieshop, Sean 4:15.67 4:15.44 2% 4% 6% 7 6 Bentz, Gunnar 4:14.16 4:15.56 2% 4% 5% 8 7 Weiss, Michael 4:14.85 4:16.35 1% 2% 3%

Women

50 Free:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 Kennedy, Madison 24.45 24.60 25% 18% 43% 2 2 Manuel, Simone 24.47 24.61 24% 17% 41% 3 6 Weitzeil, Abbey 24.72 24.71 15% 14% 29% 4 5 Vollmer, Dana 24.69 24.82 9% 10% 19% 5 3 Martin, Ivy 24.62 24.86 7% 9% 16% 6 4 Coughlin, Natalie 24.66 24.88 6% 8% 14% 7 10 Worrell, Kelsi 24.98 25.01 3% 4% 8% 8 8 Weir, Amanda 24.85 25.02 3% 4% 7%

100 Free:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 Manuel, Simone 53.25 53.41 28% 17% 45% 2 3 Vollmer, Dana 53.59 53.68 16% 14% 30% 3 5 Weitzeil, Abbey 53.77 53.80 12% 12% 24% 4 4 Ledecky, Katie 53.75 53.92 9% 10% 19% 5 2 Franklin, Missy 53.43 54.03 7% 8% 15% 6 8 Neal, Lia 54.01 54.03 7% 8% 15% 7 6 Coughlin, Natalie 53.85 54.24 4% 6% 10% 8 7 Geer, Margo 53.95 54.27 4% 5% 9%

200 Free:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 Ledecky, Katie 1:54.43 1:54.48 58% 19% 77% 2 2 Franklin, Missy 1:55.49 1:56.3 12% 18% 30% 3 3 Schmitt, Allison 1:56.23 1:56.48 10% 16% 26% 4 4 Smith, Leah 1:56.64 1:56.7 8% 14% 21% 5 5 Margalis, Melanie 1:57.33 1:57.41 3% 7% 11% 6 8 DiRado, Maya 1:57.7 1:58.04 1% 4% 5% 7 9 Manuel, Simone 1:57.9 1:58.11 1% 4% 5% 8 10 Runge, Cierra 1:57.97 1:58.2 1% 3% 4%

400 Free:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 Ledecky, Katie 3:58.37 3:58.88 75% 15% 90% 2 2 Smith, Leah 4:03.33 4:03.46 12% 30% 42% 3 3 Runge, Cierra 4:04.55 4:05.85 4% 13% 16% 4 6 Vrooman, Lindsay 4:07.16 4:07.28 2% 6% 8% 5 5 Mann, Becca 4:07.09 4:07.32 1% 6% 8% 6 4 Schmitt, Allison 4:06.88 4:07.55 1% 6% 7% 7 9 Flickinger, Hali 4:07.93 4:08.23 1% 4% 5% 8 7 Beisel, Elizabeth 4:07.46 4:08.34 1% 4% 4%

800 Free:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 Ledecky, Katie 8:06.68 8:07.05 87% 9% 96% 2 2 Mann, Becca 8:21.77 8:21.72 5% 25% 30% 3 4 Smith, Leah 8:24.74 8:23.74 3% 17% 20% 4 3 Runge, Cierra 8:24.69 8:25.85 1% 11% 12% 5 5 Peacock, Stephanie 8:25.89 8:26.02 1% 10% 11% 6 6 Vrooman, Lindsay 8:26.67 8:27.34 1% 8% 8% 7 7 Schmidt, Sierra 8:27.54 8:28.52 1% 6% 6% 8 13 Ryan, Gillian 8:31.97 8:30.12 <1% 4% 4%

100 Back:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 3 Smoliga, Olivia 59.41 59.41 24% 17% 41% 2 1 Coughlin, Natalie 59.05 59.57 18% 15% 33% 3 2 Franklin, Missy 59.38 59.69 14% 13% 27% 4 6 Stevens, Hannah 59.67 59.77 12% 12% 24% 5 7 Deloof, Ali 1:00.1 1:00.01 7% 8% 15% 6 5 Baker, Kathleen 59.63 1:00.04 7% 8% 15% 7 4 Adams, Claire 59.58 1:00.06 6% 8% 14% 8 8 Bootsma, Rachel 1:00.25 1:00.38 3% 4% 7%

200 Back:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 Franklin, Missy 2:06.34 2:07.47 44% 20% 64% 2 2 DiRado, Maya 2:08.19 2:08.39 22% 20% 42% 3 3 Beisel, Elizabeth 2:08.33 2:09.61 7% 10% 17% 4 6 Baker, Kathleen 2:09.36 2:09.7 6% 9% 16% 5 5 Pelton, Elizabeth 2:09.36 2:10.29 3% 6% 9% 6 13 Flickinger, Hali 2:10.6 2:10.51 3% 5% 7% 7 9 Voss, Erin 2:10.12 2:10.68 2% 4% 6% 8 19 Smiddy, Clara 2:11.15 2:10.76 2% 4% 6%

100 Breast:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 Meili, Katie 1:05.64 1:05.98 29% 21% 50% 2 2 King, Lilly 1:05.73 1:06 29% 21% 49% 3 4 Haase, Sarah 1:06.31 1:06.54 12% 14% 25% 4 3 Hannis, Molly 1:06.16 1:06.56 11% 13% 25% 5 5 Hardy, Jessica 1:06.51 1:06.95 5% 8% 13% 6 6 Lawrence, Micah 1:06.51 1:07.03 4% 7% 11% 7 7 Larson, Breeja 1:06.73 1:07.05 4% 7% 11% 8 9 Margalis, Melanie 1:07.26 1:07.39 2% 4% 6%

200 Breast:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 Lawrence, Micah 2:22.04 2:24.33 23% 16% 39% 2 5 King, Lilly 2:24.47 2:24.93 14% 13% 27% 3 2 Sogar, Laura 2:23.54 2:25.1 12% 12% 24% 4 3 Meili, Katie 2:23.69 2:25.21 11% 11% 22% 5 4 Larson, Breeja 2:24.16 2:25.27 11% 11% 21% 6 6 Margalis, Melanie 2:24.68 2:25.48 9% 10% 18% 7 7 Lazor, Annie 2:24.96 2:25.55 8% 9% 18% 8 8 Hannis, Molly 2:25.26 2:26.06 5% 7% 12%

100 Fly:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 Vollmer, Dana 56.94 57.01 49% 24% 73% 2 2 Worrell, Kelsi 57.24 57.31 30% 28% 57% 3 3 Stewart, Kendyl 57.82 58.00 8% 13% 21% 4 5 Donahue, Claire 58.03 58.31 4% 8% 12% 5 6 Lee, Felicia 58.14 58.52 2% 5% 7% 6 4 McLaughlin, Katie 57.87 58.53 2% 5% 7% 7 8 Merrell, Eva 58.58 58.79 1% 3% 4% 8 10 Moffitt, Hellen 58.86 58.85 1% 2% 3%

200 Fly:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 Adams, Cammile 2:06.33 2:07.04 38% 19% 58% 2 4 Flickinger, Hali 2:07.59 2:08.18 15% 15% 30% 3 5 Bayer, Cassidy 2:08.03 2:08.62 10% 11% 22% 4 2 McLaughlin, Katie 2:06.95 2:08.82 8% 10% 19% 5 6 Worrell, Kelsi 2:08.61 2:09.03 7% 9% 15% 6 3 DiRado, Maya 2:07.42 2:09.04 7% 9% 15% 7 7 Mills, Kate 2:08.89 2:09.51 4% 6% 10% 8 10 Saiz, Hannah 2:09.83 2:09.96 2% 4% 6%

200 IM:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 DiRado, Maya 2:08.99 2:09.91 34% 20% 54% 2 2 Margalis, Melanie 2:10.2 2:10.8 17% 16% 33% 3 3 Leverenz, Caitlin 2:10.35 2:10.92 15% 15% 30% 4 4 Eastin, Ella 2:10.54 2:10.97 14% 15% 29% 5 5 Cox, Madisyn 2:10.75 2:11.45 9% 11% 20% 6 7 Baker, Kathleen 2:12.09 2:12.61 3% 4% 7% 7 9 Henry, Sarah 2:12.25 2:12.83 2% 4% 6% 8 6 Small, Meghan 2:11.26 2:12.9 2% 3% 5%

400 IM:

 Predicted Place Psych Place Name Seed Time Predicted Time 1st 2nd 1st or 2nd 1 1 DiRado, Maya 4:31.71 4:33.84 34% 22% 57% 2 2 Beisel, Elizabeth 4:31.99 4:34.04 32% 22% 54% 3 3 Leverenz, Caitlin 4:35.46 4:37.12 9% 12% 22% 4 5 Ledecky, Katie 4:37.93 4:37.71 7% 10% 17% 5 4 Mann, Becca 4:37.04 4:37.73 7% 10% 17% 6 7 Adams, Cammile 4:38.97 4:39.26 3% 6% 9% 7 6 Henry, Sarah 4:38.88 4:40 2% 4% 6% 8 8 Eastin, Ella 4:40.7 4:40.31 2% 4% 5%

-The model was tuned using 2012 data. I would have preferred more years, but the older data isn’t as good. Trials performance from the super suit year in 2008 isn’t a good predictor of an un super suited year, and I don’t trust the completeness of in season data from 2004 and earlier.

-This is a probabilistic forecast, so it’s supposed to get some things wrong if it’s correctly calibrated. It should get around 28 of the 52 individual Olympic qualifiers right (28 of the people it says will get 1st or 2nd, actually make the team)

-The model doesn’t account for athletes’ specific taper history. Some athlete’s have a well established pattern of large or small tapers.
-The model only accounts for swimmers ranked 1-24. Swimmers out side that range are extreme long shots, but have a non zero chance of making the meet.

-The model doesn’t know which in season times were rested and which weren’t. All in season times are treated the same

-The model doesn’t account for likely scratches. For example, its pick for 2nd in the 100 free, Michael Phelps, typically does enough to get on the 400 Free Relay and scratches before finals. If you want an adjusted percentage after a scratch, allocate the scratched swimmer’s chances proportionally among the remaining swimmers.

-There is some evidence that top 3 seeds are a bit more consistent than lower seeds. The model doesn’t account for this which may explain some of the relatively low percentages for swimmer who are thought of as locks in their events.

-I only used in season times ranked in the top 6000 since last September. If a top 24 swimmer put up a time outside the top 6000, it’s not accounted for by the model.

-There are 19 top 24 seeds without a time in the event in the last year. Some of them will probably scratch the event (Connor Jaeger 24th 200 IM). Only 2 are ranked in the top 8 (Michael Klueh 8th 1500. Jack Conger 6th 100 Back).

-If you’re looking at this to fill out the SwimSwam time prediction contest, it’s worth noting that this model’s predicted 1st place times are better than the predicted time of its top ranked swimmer. For example in the women’s 400 IM, the model has Maya DiRado 1st in 4:33.84. It thinks she has a 50% chance of going faster than that time and a 50% chance of going slower. She only has a 34% chance to get 1st place. That means at least 16% of the time when she’s faster than the predicted time, it still won’t be good enough to win. This effect pushes the predicted winning time faster than the predicted time of the top ranked swimmer.

-To get a 95% confidence interval for the expected times, take approximately +/-2% from the predicted time.

-Edit: I wrote a follow up to this post.

In This Story

0 Followers

Most reacted comment
44 Comment authors
Know It All

These time predictions are too slow honestly, everyone will be faster

KeithM

No, not everyone will be faster. But the winning times in most races I think will be faster. And there will certainly be more PBs than this model predicts.

Kristaps Porzingis

D Rose is coming

Eric

Watch what you wish for. After seeing the results of several trials and the resulting Olympic times, sometimes the swimmer will swim super fast in Trials and end up having a mediocre Olympics. I know that you need to sometimes go uber fast to even make the Olympic roster, however, often times the athlete won’t do as well (in the Olympics) after having such a great swim-time combo in the trials. So, I hope that everyone swims great times but I hope they have what it takes to replicate and better their times at the Olympics as well.

Attila the Hunt

Eric,
Can you recall any particular top/medal favorite swimmers who swam far slower in the recent Olympics in comparison to their OT times?

SZ2016

Always underrating Cordes, simply because of his lack of recent competition in the US! Predict at least 2nd in 100 breast, 1st in 200

KeithM

I wonder how this would perform if the basic underlying formula of this model was applied to this year’s NCAA’s? How many of the records that were broken this year would it have predicted. I would hazard a guess not very many. But then these sort of models are probably rarely if ever going to predict records. It seems to want to reach a median between overpredicting and underpredicting. Thus neither predicting the breakout swims nor the disappointing swims.

KeithM – I think that’s a good analysis. Since it sort of works on the “average,” it’s unlikely to pick out big overperforms or underperforms (though, maybe an even more intense model, that also rolled in the past tendencies of these underperforms and breakouts?)

What I really love these models for, and really what the goals of ALL of our predictions are, is to give a baseline. What is an overperform or an underperform, if there’s never a baseline for ‘perform’ given? Leicester City wouldn’t have been such a great story if nobody knew what the expectations were (not much) for them.

KeithM

I think it’s an interesting model even with the limitations of the data. It will interesting to review it after the meet to see which swimmers stood out the most (for better or worse).

Andromeda

You can always use extreme value theory to make predictions about unexpectedly speedy or unexpectedly slow swims, if that ever becomes something of interest.

More from Andrew Mering
See All