Predicting each of Fall 2024, Spring 2025, Fall 2025 and Spring 2026 using only races sailed before that season began, the site's rating model put 70.4% of boat pairs in the right order across 1,414,778 pairs in 16,070 races. The strongest baseline, Plackett–Luce, static rating, managed 68.8%; a coin flip gets 50%.
A rating is only useful if it predicts results it has not seen. The ratings on this site come from a Plackett–Luce model: every skipper has a strength, and a race finish is modelled as repeatedly choosing the next boat across the line with probability proportional to strength. The site's version lets each sailor's strength drift week to week. This page tests whether that model forecasts future races better than simpler ways of rating sailors.
| Model | Pairwise accuracy | Log-loss | Winner picked | Rank correlation | ICSA accuracy | ISSA accuracy | Returning sailors | With a newcomer |
|---|---|---|---|---|---|---|---|---|
| No information (coin flip) | 50.0% | 0.6932 | 10.4% | 0.000 | 50.0% | 50.0% | 50.0% | 50.0% |
| Average finish percentile | 65.3% | 0.6185 | 27.6% | 0.403 | 66.4% | 64.3% | 66.5% | 62.0% |
| Elo (multi-player, tuned K) | 68.7% | 0.5823 | 32.4% | 0.481 | 70.3% | 67.4% | 70.3% | 64.3% |
| Plackett–Luce, static rating | 68.8% | 0.5815 | 32.6% | 0.480 | 70.6% | 67.2% | 70.5% | 63.8% |
| Plackett–Luce, weekly, before the rating fix (previous site model) | 69.6% | 0.5704 | 33.5% | 0.502 | 71.1% | 68.4% | 71.3% | 65.0% |
| Plackett–Luce, weekly (site model) | 70.4% | 0.5676 | 34.4% | 0.518 | 71.8% | 69.2% | 72.0% | 65.8% |
The rating fix. Until September 2026 the model stopped fitting after 2,000 steps, long before the ratings had settled, which squeezed them together. Ratings are now fitted until they settle, with a pull toward the average on each sailor's rating when they start racing, which matters most for sailors with few races; its strength and the week-to-week smoothing were tuned on week-ahead forecasts from 2024 and spring 2025. On these season-ahead forecasts, pairwise accuracy went from 69.6% before the fix to 70.4%, and log-loss from 0.570 to 0.568. At championships, with ratings from the week before, the fix was worth +0.3 to +0.7 points of pairwise accuracy (95% interval).
Pairwise accuracy for each forecast season. Fall seasons bring a new freshman class, so a larger share of pairs involve sailors no model has seen race before.
| Forecast season | Ratings frozen at | Races | New sailors | Coin flip | Average finish | Elo | Static PL | Previous site model | Weekly PL (site) |
|---|---|---|---|---|---|---|---|---|---|
| Fall 2024 | 2024-W36 | 4,504 | 1,022 | 50.0% | 64.8% | 67.5% | 68.3% | 68.5% | 69.2% |
| Spring 2025 | 2025-W03 | 3,614 | 594 | 50.0% | 65.1% | 69.1% | 68.8% | 70.0% | 71.0% |
| Fall 2025 | 2025-W36 | 4,220 | 1,033 | 50.0% | 64.9% | 68.0% | 68.2% | 68.7% | 69.3% |
| Spring 2026 | 2026-W03 | 3,732 | 596 | 50.0% | 66.5% | 70.8% | 70.2% | 71.8% | 72.7% |
Championships are where ratings get the most attention, so this test scores only championship-level fleet regattas: ICSA national finals and semifinals, conference championships and Atlantic Coast Championship rounds, and ISSA national championships, district championships and district qualifiers. For each one, every model is fit on all races through the week before the event and then predicts it. Baseline settings come from that season's forecast tuning.
The site model ordered 72.5% of boat pairs correctly (95% interval 71.6–73.3%) across 166 regattas and 3,129 races; Plackett–Luce, static rating managed 71.3%. Resampling whole regattas, the site model came out ahead of Static PL in 100% of resamples, and the interval for its margin (+0.9 to +1.5 points) stays above zero. With ratings frozen at the start of the season instead, the same model scored 71.2% on these regattas, so results from earlier in the season add real information.
| Model | Pairwise accuracy | 95% interval | Log-loss | Winner picked | ICSA | ISSA | Site model better in |
|---|---|---|---|---|---|---|---|
| No information (coin flip) | 50.0% | 50.0–50.0% | 0.6932 | 8.4% | 50.0% | 50.0% | 100% |
| Average finish percentile | 67.0% | 66.2–67.8% | 0.6061 | 26.4% | 66.9% | 67.0% | 100% |
| Elo (multi-player, tuned K) | 71.0% | 70.2–71.9% | 0.5577 | 30.8% | 71.6% | 70.5% | 100% |
| Plackett–Luce, static rating | 71.3% | 70.4–72.2% | 0.5571 | 32.5% | 71.8% | 70.8% | 100% |
| Plackett–Luce, weekly, before the rating fix (previous site model) | 72.0% | 71.1–72.9% | 0.5451 | 32.9% | 72.4% | 71.5% | 100% |
| Plackett–Luce, weekly (site model) | 72.5% | 71.6–73.3% | 0.5367 | 33.0% | 72.7% | 72.3% | — |
Predicted versus actual team totals at every championship
Nationals are the hardest test for the ratings. Each table uses only results from before the week of the regatta: every skipper's rating going in, the standings those ratings predict, and how teams actually finished. Ratings are on the rankings scale, and “new” marks a skipper with no earlier races, who starts at the typical rating of their school's recent debut sailors (or the league's, for a school with too few). Predicted points leave penalties out, because adding them makes the totals worse; the chances below still include them. The chance columns come from simulating the whole regatta, every race and every penalty, thousands of times with those ratings: how often each team won, and how often it finished where it did or better.
The ratings picked the winner. Predicted totals were off by 27.8 points per team and finishes by 2.9 places, with a rank correlation of 0.76 between predicted and actual standings. Over 14 races per division, the ratings gave Brown University a 32% chance to win, and standings at least this far from the prediction came up in 18% of 2,000 simulated regattas.
| Actual finish | Team | Predicted finish | Beat prediction by | Chance of this finish or better | Win chance | Points | Predicted points | Skippers and ratings going in |
|---|---|---|---|---|---|---|---|---|
| 1 | Brown University Bears | 1 | +0 | 32% | 32% | 163 | 178.1 | A Guthrie Braun 492 B Blake Behrens 457 |
| 2 | Stanford University Cardinal | 2 | +0 | 43% | 23% | 174 | 187.1 | A Thomas Sitzmann 490 (14 races) B Reade Decker 418 (10 races) B Vanessa Lahrkamp 509 (4 races) |
| 3 | Yale University Bulldogs | 11 | +8 | 6% | 1% | 211 | 269.5 | A Morgan Pinckney 386 B Dorothy Mendelblatt 401 |
| 4 | U. S. Naval Academy Midshipmen | 7 | +3 | 18% | 3% | 223 | 245.4 | A Nathan Smith 422 B Henry Allgeier 406 |
| 5 | Dartmouth College Big Green | 8 | +3 | 30% | 2% | 225 | 245.5 | A Ryan Satterberg 413 B Chase Decker 415 |
| 5 | University of Pennsylvania Quakers | 10 | +5 | 16% | 1% | 225 | 266.1 | A Cole Woodworth 399 B Jackson Mcaliley 395 |
| 7 | Georgetown University Hoyas | 9 | +2 | 37% | 2% | 230 | 258.0 | A Enzo Menditto 397 B Peter Herlihy 410 |
| 7 | Harvard University Crimson | 3 | -4 | 89% | 21% | 230 | 190.9 | A Justin Callahan 494 B Mitchell Callahan 432 |
| 9 | Tulane University Green Wave | 5 | -4 | 82% | 6% | 242 | 222.5 | A Hamilton Barclay 461 (14 races) B Christian Ebbin 413 (10 races) B Kelly Holthus 395 (4 races) |
| 10 | College of Charleston Cougars | 4 | -6 | 89% | 7% | 243 | 218.1 | A Noah Zittrer 435 (12 races) A Pierce Olsen 407 (2 races) B Benjamin Dufour 445 (14 races) |
| 11 | Roger Williams University Hawks | 6 | -5 | 89% | 4% | 253 | 231.3 | A Carlos De Castro 466 (14 races) B Kyle Pfrang 392 (12 races) B Oliver Stokke 359 (2 races) |
| 12 | St. Mary's College of Maryland Seahawks | 13 | +1 | 57% | <1% | 302 | 289.9 | A Nathan Jensen 367 (12 races) A Raam Fox 352 (2 races) B Landon Cormie 388 (14 races) |
| 13 | Connecticut College Camels | 18 | +5 | 10% | <1% | 318 | 365.7 | A Henry Scholz 337 (10 races) A Rory Murray 274 (4 races) B William Hurd 294 (14 races) |
| 14 | Boston College Eagles | 12 | -2 | 91% | <1% | 321 | 275.8 | A Tanner Krygsveld 404 (6 races) A Alex Lech 376 (4 races) A Peter Busch 383 (2 races) A Peter Joslin 384 (2 races) B Caroline Sibilly 393 (6 races) B Cody Roe 360 (4 races) B Jack Redmond 404 (4 races) |
| 15 | Jacksonville University Fins | 14 | -1 | 76% | <1% | 337 | 310.8 | A Owen Bannasch 403 (14 races) B Patrick Igoe 301 (8 races) B Hank Seum 327 (4 races) B Cole Schweda 323 (2 races) |
| 16 | Bowdoin College Polar Bears | 16 | +0 | 63% | <1% | 343 | 345.6 | A Michelangelo Vecchio 334 (12 races) A Ryan Keenan 311 (2 races) B Kyra Phelan 329 (11 races) B Lucca Antonietti 299 (3 races) |
| 17 | Cornell University Big Red | 15 | -2 | 88% | <1% | 370 | 332.3 | A Winborne Majette 367 (14 races) B Gilda Dondona 321 (12 races) B Marcus Greco 244 (2 races) |
| 18 | Fordham University Rams | 17 | -1 | >99% | <1% | 380 | 355.5 | A Jacob Zils 333 (14 races) B Lucas Thress 321 (8 races) B Patrick Shachoy 286 (4 races) B Erickson Rankin 247 (2 races) |
The ratings picked the winner. Predicted totals were off by 26.2 points per team and finishes by 1.9 places, with a rank correlation of 0.86 between predicted and actual standings. Over 12 races per division, the ratings gave Stanford University a 68% chance to win, and standings at least this far from the prediction came up in 52% of 2,000 simulated regattas.
| Actual finish | Team | Predicted finish | Beat prediction by | Chance of this finish or better | Win chance | Points | Predicted points | Skippers and ratings going in |
|---|---|---|---|---|---|---|---|---|
| 1 | Stanford University Cardinal | 1 | +0 | 68% | 68% | 113 | 117.3 | A Vanessa Lahrkamp 477 B Sophie Fisher 397 |
| 2 | Yale University Bulldogs | 2 | +0 | 28% | 7% | 128 | 166.8 | A Dorothy Mendelblatt 388 B Carly Kieding 364 |
| 3 | Harvard University Crimson | 4 | +1 | 31% | 5% | 161 | 179.1 | A Zoey Ziskind 381 B Kate Danielson 344 |
| 4 | Bowdoin College Polar Bears | 10 | +6 | 13% | 1% | 164 | 216.6 | A Lauren Russler 356 B Kyra Phelan 289 |
| 5 | Cornell University Big Red | 6 | +1 | 38% | 2% | 171 | 195.1 | A Winborne Majette 340 (12 races) B Sophia Devling 359 (10 races) B Gilda Dondona 319 (2 races) |
| 6 | College of Charleston Cougars | 7 | +1 | 45% | 3% | 190 | 198.0 | A Bella Shakespeare 356 B Ashley Alfortish 329 |
| 7 | Tulane University Green Wave | 11 | +4 | 33% | 1% | 208 | 220.6 | A Ava Anderson 369 (12 races) B Gabriela Vassel 278 (8 races) B Lola Kohl 245 (4 races) |
| 8 | Brown University Bears | 3 | -5 | 87% | 8% | 214 | 172.6 | A Katharine Doble 381 (12 races) B Katherine Mcnamara 326 (6 races) B Laura Hamilton 394 (6 races) |
| 9 | Georgetown University Hoyas | 5 | -4 | 83% | 4% | 226 | 184.8 | A Emily Doble 347 B Kelly Bates 368 |
| 10 | Roger Williams University Hawks | 9 | -1 | 66% | 1% | 231 | 214.7 | A Lucy Meagher 367 B Tavia Smith 282 |
| 11 | Tufts University Jumbos | 12 | +1 | 54% | <1% | 236 | 238.8 | A Ella Hubbard 286 (8 races) A Sophia Hubbard 276 (4 races) B Maddie Janzen 316 (12 races) |
| 12 | Dartmouth College Big Green | 8 | -4 | 91% | 2% | 247 | 199.4 | A Bella Casaretto 363 (11 races) A Alders Kulynych-Irvin 242 (1 races) B Olivia Drulard 329 (12 races) |
| 12 | George Washington University Revolutionaries | 16 | +4 | 6% | <1% | 247 | 316.0 | A Arrieta Angueira Salbidegoitia 261 B Hayden Clary 160 |
| 14 | Massachusetts Institute of Technology Engineers | 14 | +0 | 72% | <1% | 258 | 263.6 | A Brooke Barry 269 (12 races) B Karya Basaraner 289 (9 races) B Emma Wang 240 (3 races) |
| 15 | Boston College Eagles | 13 | -2 | 94% | <1% | 276 | 249.8 | A Caroline Sibilly 397 B Kate Joslin 168 |
| 16 | U. S. Coast Guard Academy Bears | 15 | -1 | 87% | <1% | 316 | 288.6 | A Madeline Murphy 277 (12 races) B Ella Demand 218 (10 races) B Meara Conley 188 (2 races) |
| 17 | Jacksonville University Fins | 17 | +0 | 65% | <1% | 350.1 | 338.2 | A Kaitlyn Liebel 199 (12 races) B Fiona Froelich 188 (10 races) B Kaitlyn Anderson 8 (2 races) |
| 18 | University of Rhode Island Rams | 18 | +0 | >99% | <1% | 368 | 343.8 | A Ariana Schwartz 173 B Emaline Ouellette 175 |
The ratings picked Point Loma High School; Severn School won. Predicted totals were off by 46.2 points per team and finishes by 2.2 places, with a rank correlation of 0.89 between predicted and actual standings. Over 16 races per division, the ratings gave Severn School a 24% chance to win, and standings at least this far from the prediction came up in 48% of 2,000 simulated regattas.
| Actual finish | Team | Predicted finish | Beat prediction by | Chance of this finish or better | Win chance | Points | Predicted points | Skippers and ratings going in |
|---|---|---|---|---|---|---|---|---|
| 1 | Severn School Admirals | 2 | +1 | 24% | 24% | 167 | 214.8 | A Annie Sitzmann 358 B Harrison Szot 368 |
| 2 | St. George's School Dragons | 4 | +2 | 16% | 6% | 196 | 253.6 | A Gil Hackel 383 (16 races) B Miles Cundey 266 (12 races) B Amelon Rule 321 (4 races) |
| 3 | Mater Dei High School Monarchs | 3 | +0 | 36% | 9% | 203 | 239.7 | A Nickolas Lech 353 (16 races) B Kingston Keyoung 317 (12 races) B Colin Kennedy 383 (2 races) B Gage Christopher 377 (2 races) |
| 4 | Lucy Beckham High School Bengals | 6 | +2 | 32% | 6% | 213 | 264.9 | A James Pine 385 B Nathan Pine 262 |
| 5 | Point Loma High School Pointers | 1 | -4 | 88% | 40% | 221 | 196.7 | A Wyatt Kelly 373 (14 races) A Kevin Cason 360 (2 races) B Anton Schmid 386 (16 races) |
| 6 | Ransom Everglades School Raiders | 11 | +5 | 17% | 1% | 229 | 319.1 | A Ava Mc Aliley 297 (10 races) A Sander Block 291 (6 races) B Max Wolfensberger 274 (16 races) |
| 7 | Barrington High School Eagles | 7 | +0 | 49% | 3% | 304 | 281.5 | A Duffy Macaulay 359 B Ben Reuter 263 |
| 8 | Christchurch School Seahorses One | 5 | -3 | 75% | 6% | 313 | 260.6 | A Wylder Smith 355 (16 races) B Sam De Los Reyes 296 (14 races) B Elliott Lipp 304 (2 races) |
| 9 | Gulliver Preparatory School Raiders | 12 | +3 | 44% | 1% | 318 | 321.5 | A Connor Karr 333 (16 races) B Arturo Zizold 243 (14 races) B Danika Torres 124 (2 races) |
| 10 | Southern Regional High School Rams | 10 | +0 | 66% | 1% | 321 | 300.4 | A Jude Ryon 320 B Gannon Botwinick 274 |
| 11 | Key School Zags | 15 | +4 | 12% | <1% | 335 | 417.8 | A Trey Waters 257 (16 races) B Casey Burman 172 (14 races) B Ethan Purdon 107 (2 races) |
| 12 | Brunswick School Bruins | 9 | -3 | 80% | 4% | 360 | 296.7 | A Harrison Gandy 308 (14 races) A Sebastian Sheppard 358 (2 races) B William Whidden 286 (16 races) |
| 13 | San Marcos High School Royals | 8 | -5 | 96% | 2% | 365 | 286.5 | A Dylan Seawards 326 (16 races) B Sam Wells 310 (12 races) B Taylor Escola 224 (4 races) |
| 14 | Arrowhead High School Warhawks | 13 | -1 | 86% | <1% | 402 | 342.1 | A John Lieber 251 B Nicholas Berkowitz 285 |
| 15 | Bainbridge High School Spartans | 14 | -1 | 53% | <1% | 429 | 415.2 | A Cyrus Yan 189 (10 races) A Stone Dewey 244 (6 races) B Nelson Dorsey 217 (16 races) |
| 16 | Jesuit High School, NOLA Blue Jays | 20 | +4 | 9% | <1% | 429.7 | 519.3 | A Jack Meade 212 (14 races) A Liam Moore 48 (2 races) B Reed Gibbs new (12 races) B David Karcher 183 (4 races) |
| 17 | New Trier HS Trevian | 17 | +0 | 71% | <1% | 450 | 437.4 | A Nathan Finkelstein 227 (16 races) B Aiala Angueira Salbidegoitia 200 (12 races) B Ralph Lipford 32 (4 races) |
| 18 | Minnetonka High School Skippers | 19 | +1 | 66% | <1% | 459 | 464.3 | A Connor Jewett 223 (12 races) A Mark Yakovlev new (4 races) B Reese Kottke 191 (10 races) B Maggie Mcgary 111 (4 races) B Maxwell Kelley 100 (2 races) |
| 19 | Jones College Prep Eagles | 16 | -3 | 96% | <1% | 460 | 431.1 | A Nissa Berman 257 (14 races) A Jack Eskilson 41 (2 races) B Duke Diep 185 (12 races) B Quinn Frakt 89 (4 races) |
| 20 | Olympia High School Bears | 18 | -2 | >99% | <1% | 539 | 456.9 | A Alan Timms 228 (16 races) B Simone Reck 163 (10 races) B Kaden Kim 65 (6 races) |
The ratings picked the winner. Predicted totals were off by 25.1 points per team and finishes by 3.9 places, with a rank correlation of 0.59 between predicted and actual standings. Over 7 races per division, the ratings gave Stanford University a 35% chance to win, and standings at least this far from the prediction came up in 24% of 2,000 simulated regattas.
| Actual finish | Team | Predicted finish | Beat prediction by | Chance of this finish or better | Win chance | Points | Predicted points | Skippers and ratings going in |
|---|---|---|---|---|---|---|---|---|
| 1 | Stanford University Cardinal | 1 | +0 | 35% | 35% | 58 | 86.7 | A Thomas Sitzmann 472 B Vanessa Lahrkamp 489 |
| 2 | U. S. Naval Academy Midshipmen | 8 | +6 | 8% | 3% | 74 | 130.1 | A Nathan Smith 414 B Henry Allgeier 393 |
| 3 | Dartmouth College Big Green | 9 | +6 | 9% | 2% | 91 | 135.9 | A William Michels 392 B Chase Decker 396 |
| 4 | Yale University Bulldogs | 3 | -1 | 45% | 11% | 98 | 108.4 | A Jack Egan 433 B Stephan Baker 448 |
| 5 | Harvard University Crimson | 2 | -3 | 69% | 20% | 107 | 97.9 | A Justin Callahan 501 B Mitchell Callahan 419 |
| 6 | Brown University Bears | 5 | -1 | 54% | 9% | 113 | 114.2 | A Guthrie Braun 450 B Blake Behrens 411 |
| 7 | Tufts University Jumbos | 13 | +6 | 27% | 2% | 119 | 144.4 | A Ben Mueller 395 B Kurt Stuebe 364 |
| 8 | George Washington University Revolutionaries | 18 | +10 | 2% | <1% | 132 | 181.2 | A Tyler Wood 323 B Jedidiah Bechtel 304 |
| 9 | College of Charleston Cougars | 4 | -5 | 76% | 8% | 135 | 113.1 | A Noah Zittrer 449 B Benjamin Dufour 415 |
| 10 | Boston College Eagles | 6 | -4 | 81% | 6% | 136 | 115.2 | A Peter Busch 425 (7 races) B Jack Redmond 436 (6 races) B Michael Kirkman 417 (1 races) |
| 11 | University of Miami Hurricanes | 14 | +3 | 45% | 1% | 148 | 147.5 | A Atlee Kohl 408 B Aidan Dennis 340 |
| 12 | Bowdoin College Polar Bears | 16 | +4 | 40% | <1% | 149 | 156.9 | A Thibault Antonietti 348 B Sam Bonauto 367 |
| 13 | Georgetown University Hoyas | 10 | -3 | 77% | 1% | 151 | 136.0 | A Piper Holthus 383 (3 races) A Enzo Menditto 381 (2 races) A Mateo Di Blasi 434 (2 races) B Peter Barnard 389 (6 races) B Diego Escobar 398 (1 races) |
| 14 | University of Rhode Island Rams | 12 | -2 | 78% | 1% | 159 | 141.9 | A Kerem Erkmen 461 (7 races) B Tyler Nash 311 (4 races) B Christopher Chwalk 292 (3 races) |
| 15 | Tulane University Green Wave | 7 | -8 | 94% | 3% | 161 | 124.8 | A Kelly Holthus 412 (7 races) B Hamilton Barclay 409 (4 races) B Christian Ebbin 418 (3 races) |
| 16 | St. Mary's College of Maryland Seahawks | 11 | -5 | 93% | 1% | 168 | 137.5 | A Owen Hennessey 426 (7 races) B Landon Cormie 346 (4 races) B Charlie Anderson 368 (3 races) |
| 17 | Hobart and William Smith Colleges Statesmen | 17 | +0 | 85% | <1% | 193 | 165.9 | A Juan Carlos Lacerda Jones 335 (4 races) A James Kopack 295 (3 races) B Jj Klempen 362 (7 races) |
| 18 | Massachusetts Institute of Technology Engineers | 15 | -3 | >99% | <1% | 206 | 156.3 | A Sam Bruce 399 (7 races) B Julius Heitkoetter 312 (5 races) B William Kulas 332 (2 races) |
The ratings picked Yale University; Stanford University won. Predicted totals were off by 35.6 points per team and finishes by 2.4 places, with a rank correlation of 0.85 between predicted and actual standings. Over 16 races per division, the ratings gave Stanford University a 37% chance to win, and standings at least this far from the prediction came up in 64% of 2,000 simulated regattas.
| Actual finish | Team | Predicted finish | Beat prediction by | Chance of this finish or better | Win chance | Points | Predicted points | Skippers and ratings going in |
|---|---|---|---|---|---|---|---|---|
| 1 | Stanford University Cardinal | 2 | +1 | 37% | 37% | 198 | 182.7 | A Vanessa Lahrkamp 480 (16 races) B Ellie Harned 355 (8 races) B Sophie Fisher 374 (8 races) |
| 2 | Tulane University Green Wave | 5 | +3 | 14% | 4% | 209 | 252.6 | A Samantha Gardner 392 B Ava Anderson 332 |
| 3 | Harvard University Crimson | 6 | +3 | 17% | 2% | 223 | 264.3 | A Cordelia Burn 356 B Zoey Ziskind 350 |
| 4 | Yale University Bulldogs | 1 | -3 | 93% | 47% | 250 | 175.8 | A Emma Cowles 428 (8 races) A Mia Nicolosi 470 (8 races) B Carmen Cowles 404 (16 races) |
| 5 | Cornell University Big Red | 3 | -2 | 58% | 5% | 255 | 244.2 | A Bridget Green 425 (12 races) A Winborne Majette 352 (4 races) B Sophia Devling 332 (16 races) |
| 6 | Boston College Eagles | 12 | +6 | 13% | <1% | 277 | 327.5 | A Caroline Sibilly 364 B Sara Schumann 247 |
| 7 | Georgetown University Hoyas | 4 | -3 | 78% | 4% | 284 | 248.2 | A Piper Holthus 377 (16 races) B Emily Doble 359 (8 races) B Kelly Bates 349 (8 races) |
| 8 | Bowdoin College Polar Bears | 10 | +2 | 37% | 1% | 287 | 315.2 | A Kyra Phelan 318 B Lauren Russler 311 |
| 9 | Brown University Bears | 7 | -2 | 60% | <1% | 294 | 292.2 | A Katharine Doble 334 (16 races) B Katherine Mcnamara 332 (10 races) B Laura Hamilton 324 (6 races) |
| 10 | Dartmouth College Big Green | 8 | -2 | 61% | <1% | 297 | 305.3 | A Sarah Young 366 (14 races) A Bella Casaretto 331 (2 races) B Olivia Drulard 283 (16 races) |
| 11 | College of Charleston Cougars | 11 | +0 | 60% | <1% | 307 | 317.4 | A Emma Tallman 336 B Emily Alfortish 291 |
| 12 | George Washington University Revolutionaries | 17 | +5 | 9% | <1% | 314 | 409.9 | A Avery Canavan 232 B Arrieta Angueira Salbidegoitia 245 |
| 13 | Massachusetts Institute of Technology Engineers | 9 | -4 | 83% | 0% | 320 | 308.2 | A Brooke Schmelz 323 B Lucy Brock 316 |
| 13 | Northeastern University Huskies | 15 | +2 | 46% | <1% | 320 | 361.4 | A Eva Ermlich 281 B Lucia Loosbrock 277 |
| 15 | Roger Williams University Hawks | 14 | -1 | 87% | <1% | 371 | 332.2 | A Lucy Meagher 350 (16 races) B Tavia Smith 274 (10 races) B Katherine Mcgagh 217 (6 races) |
| 16 | University of Pennsylvania Quakers | 13 | -3 | 95% | <1% | 388 | 328.4 | A Sofia Segalla 343 B Adra Ivancich 267 |
| 17 | University of South Florida Bulls | 16 | -1 | 85% | <1% | 426 | 386.1 | A Kay Brunsvold 309 (15 races) A Kailey Warrior 208 (1 races) B Kalea Woodard 220 (14 races) B Heidi Hicks 183 (2 races) |
| 18 | Tufts University Jumbos | 18 | +0 | >99% | <1% | 453 | 420.3 | A Maisie Macgillivray 204 (6 races) A Meredith Broadus 216 (6 races) A Kiana Beachy 200 (4 races) B Sophia Hubbard 247 (16 races) |
The ratings picked the winner. Predicted totals were off by 38.2 points per team and finishes by 1.8 places, with a rank correlation of 0.92 between predicted and actual standings. Over 20 races per division, the ratings gave Point Loma High School a 60% chance to win, and standings at least this far from the prediction came up in 86% of 2,000 simulated regattas.
| Actual finish | Team | Predicted finish | Beat prediction by | Chance of this finish or better | Win chance | Points | Predicted points | Skippers and ratings going in |
|---|---|---|---|---|---|---|---|---|
| 1 | Point Loma High School Pointers | 1 | +0 | 60% | 60% | 266.6 | 206.7 | A Ian Nyenhuis 415 B Anton Schmid 420 |
| 2 | Antilles School Hurricanes | 3 | +1 | 26% | 12% | 270 | 295.3 | A Tanner Krygsveld 430 B Cobia Fagan 284 |
| 3 | Ransom Everglades School Raiders | 6 | +3 | 21% | 2% | 294 | 324.9 | A Griggs Diemar 392 (20 races) B Sebastian Van De Kreeke 293 (16 races) B Ava Mc Aliley 237 (4 races) |
| 4 | Severn School Admirals | 4 | +0 | 48% | 5% | 301 | 300.6 | A Harrison Szot 342 (12 races) A Alex Baker 333 (8 races) B Annie Sitzmann 368 (20 races) |
| 5 | Christchurch School Seahorses | 5 | +0 | 59% | 4% | 307.8 | 303.6 | A Bo Angus 370 (16 races) A Madeline Janzen 256 (4 races) B Wylder Smith 354 (20 races) |
| 6 | Lucy Beckham High School Bengals | 9 | +3 | 24% | 1% | 326 | 380.9 | A James Pine 362 B Nathan Pine 244 |
| 7 | Mater Dei High School Monarchs | 2 | -5 | 94% | 16% | 327 | 260.0 | A Tate Christopher 406 (20 races) B Brady Kennedy 348 (10 races) B Noah Stapleton 350 (10 races) |
| 7 | St. George's School Dragons | 11 | +4 | 28% | <1% | 327 | 392.6 | A Gil Hackel 313 (20 races) B Amelon Rule 292 (18 races) B Kai Watters 159 (2 races) |
| 9 | Tabor Academy Seawolves | 10 | +1 | 52% | 1% | 345 | 388.1 | A Peter Herlihy 344 (20 races) B Jack Spillane 265 (16 races) B Abby Wei 209 (2 races) B Sander Skaane 202 (2 races) |
| 10 | The Hotchkiss School Bearcats | 7 | -3 | 73% | 1% | 351 | 364.3 | A Pierce Olsen 343 (20 races) B Thomas O'Grady 269 (11 races) B Fynn Olsen 300 (9 races) |
| 11 | Southern Regional High School Rams | 12 | +1 | 52% | <1% | 407 | 421.4 | A Turner Ryon 274 (16 races) A Gannon Botwinick 221 (4 races) B Jude Ryon 296 (20 races) |
| 12 | Corona del Mar High School Sea Kings | 13 | +1 | 59% | <1% | 443 | 432.0 | A Michael Sentovich 276 (20 races) B Maddie Nichols 293 (12 races) B Siena Nichols 270 (6 races) B Jonah Moore 132 (2 races) |
| 13 | Christian Brothers Academy Colts | 8 | -5 | 92% | <1% | 444 | 377.0 | A Christopher Small 321 B Cole Buczkowski 289 |
| 14 | Jones College Prep Eagles | 16 | +2 | 33% | <1% | 445 | 518.1 | A Nissa Berman 230 B Grace Renz 213 |
| 15 | Arrowhead High School Warhawks | 14 | -1 | 77% | <1% | 547 | 473.7 | A John Lieber 252 B Nicholas Berkowitz 245 |
| 16 | Lake Forest High School Scouts | 15 | -1 | 73% | <1% | 559 | 516.3 | A Mason Keane 233 (13 races) A Grady Strothman 266 (4 races) A Owen Kohut 343 (3 races) B Keegan Chatburn 224 (14 races) B Maddie Rode 104 (4 races) B Jackson Schwartz 61 (2 races) |
| 17 | Wayzata High School Navy | 19 | +2 | 21% | <1% | 565 | 616.7 | A Dominik Moncur 224 B Stonewall Anderson 76 |
| 18 | Olympia High School Bears | 18 | +0 | 51% | <1% | 603 | 616.5 | A Alan Timms 173 B Liam Taylor 138 |
| 19 | Clear Lake High School Falcons Green | 17 | -2 | 88% | <1% | 621 | 576.4 | A Sydney Small 245 B Casey Small 118 |
| 20 | Roosevelt High School Rough Riders | 20 | +0 | >99% | <1% | 643 | 635.2 | A Roan Olson 146 B Ethan Lee 135 |
Predicted versus actual standings at every championship
The model predicts each race from ratings alone, as if every race were a fresh draw. Real regattas are not like that: a sailor who beats their rating in one race tends to beat it again. To measure it, each championship skipper's result in every race was scored against their rating going in (places better or worse than expected, as a share of the fleet), then compared race by race, with up to 4,491 sailors behind each pair of races. If races were independent every cell below would be about zero.
Every pair of races is positively correlated, 0.16 on average. Back-to-back races are the most alike (0.24) and the link fades with distance, to 0.09 for races 11 apart, so part of it is form or conditions that change through the regatta. Races also come in pairs: back-to-back races within the same pair (1–2, 3–4, …) average 0.27, while back-to-back races across a pair (2–3, 4–5, …) average 0.19. Rotations show why. Back-to-back races sailed in the same boat correlate at 0.27, back-to-back races in different boats at 0.18. The boat matters even when races are far apart: 2 races apart, the same boat gives 0.34 and a different boat 0.17.
So yes, winning race 1 means something. Across 266 championship divisions, the race-1 winner went on to average 4.4 in later races. Their rating alone predicted 5.2; updating on race 1 predicts 4.8.
Using it. If each sailor's strength is nudged by how they have sailed so far in the regatta, the order of later races is easier to predict. Tuned on 2024–25 championships and tested on 2025–26, pairs ordered correctly rose from 73.0% to 74.5% and log-loss fell from 0.532 to 0.513. That is a bigger gain than the site model's whole edge over Elo.
In the simulations. The championship simulations and the regatta analysis pages now build these correlations into every simulated regatta instead of treating each race as a fresh draw. Each sailor gets a good or bad regatta, form that drifts from race to race, and the speed of whichever boat the rotation puts them in, shared with every team that sails that boat. The three sizes were fitted to the correlations above. How big the swings should be depends on how well a sailor's rating is known, because they also cover the rating being wrong. For skippers with more than 200 races before the event they are 0.7 times the fitted size; for 61–200 races 0.8; for 1–60 races 1.3; for newcomers 2.0. Those sizes were chosen on the 2024–25 championships and checked on 2025–26 (the newcomers still finish worse than their starting rating suggests, which bigger swings cannot fix). Before both changes, team totals needed much smaller swings than individual sailors did: championship winners are mostly experienced skippers with well-known ratings, and extra swings were also flattening favourites whose ratings were already too bunched. With both, team totals chose 100% of the fitted size. See how likely was each result.
When the model says a sailor has a 70% chance of finishing ahead of another, that should happen about 70% of the time. Points on the diagonal mean the stated probabilities can be taken at face value. Up to about 85% confidence the points sit within 3.4 percentage points of the diagonal. Above that the model is slightly overconfident: when it says 97%, the favourite finishes ahead 93% of the time.
| Confidence bin | Mean predicted | Observed | Pairs |
|---|---|---|---|
| 50–55% | 52.5% | 52.2% | 161,007 |
| 55–60% | 57.5% | 56.2% | 159,138 |
| 60–65% | 62.5% | 60.9% | 157,519 |
| 65–70% | 67.5% | 64.9% | 150,788 |
| 70–75% | 72.5% | 69.5% | 149,369 |
| 75–80% | 77.5% | 74.3% | 142,424 |
| 80–85% | 82.5% | 79.1% | 138,842 |
| 85–90% | 87.5% | 84.1% | 131,278 |
| 90–95% | 92.5% | 89.1% | 125,315 |
| 95–100% | 97.0% | 93.1% | 87,375 |
An OCS or a DNF says little about boat speed, so boats that did not finish are left out of each race's finishing order. They never count as losing to every boat that finished, and they are left out of the accuracy numbers above. Penalties are predicted separately: each skipper gets a per-race chance of a start penalty, a DNF and a DNS, pulled toward the league rate by an amount tuned on the previous season, so a few penalties in a handful of races do not brand a sailor. The tenth of race entries it rated most penalty-prone had a penalty 3.0% of the time, against 2.1% for everyone else.
| Outcome | Share of race entries | Log-loss vs league rate | Championships: share | Championships: vs league rate |
|---|---|---|---|---|
| Start penalty (OCS) | 0.26% | +0.1% | 0.32% | +0.3% |
| Did not finish (DNF, RAF) | 0.70% | +0.9% | 0.42% | +3.4% |
| Did not start (DNS) | 1.20% | +1.5% | 0.72% | +8.0% |
| Any of these | 2.15% | +1.6% | 1.46% | +5.3% |
“Log-loss vs league rate” is how much better the penalty scores predict than giving every skipper the league's average rate, season-ahead and at championships. Penalties are rare and only loosely a habit, so these gains are small; did not start is the most predictable.
The trainer holds out every 20th regatta (339 regattas, 3,681 races) and prints its accuracy on them. Re-scoring that split with this evaluator should reproduce the trainer's numbers, which confirms the metric code. These figures are optimistic for the two Plackett–Luce models: the weekly model smooths ratings across time, so a held-out race's rating is informed by races that came after it. Elo and average finish only use earlier weeks here.
| Model | ICSA accuracy | ICSA log-loss | ISSA accuracy | ISSA log-loss |
|---|---|---|---|---|
| No information (coin flip) | 50.0% | 0.6932 | 50.0% | 0.6932 |
| Average finish percentile | 62.5% | 0.6401 | 66.8% | 0.5995 |
| Elo (multi-player, tuned K) | 66.8% | 0.6004 | 69.6% | 0.5701 |
| Plackett–Luce, static rating | 70.3% | 0.5596 | 72.7% | 0.5357 |
| Plackett–Luce, weekly (site model) | 70.8% | 0.5529 | 73.8% | 0.5213 |
| Trainer's own printout (weekly PL) | 70.84% | 0.5529 | 73.74% | 0.5214 |
| Forecast season | Tuned on | Elo K | Avg-finish shrinkage | Static PL L2 λ |
|---|---|---|---|---|
| Fall 2024 | Spring 2024 | 64 | 3 | 3 |
| Spring 2025 | Fall 2024 | 64 | 20 | 6 |
| Fall 2025 | Spring 2025 | 128 | 3 | 3 |
| Spring 2026 | Fall 2025 | 64 | 20 | 6 |
python3 analysis/plackett_luce/evaluate_models.py
Generated 2026-09-30 from 103,570 races and 26,409 sailors (2008-W38 to 2026-W39). Full run 100 minutes.
