This series ran across three regattas and 17 races: Dartmouth v Harvard Lights (3), Princeton Invitational (7), and The Dual (W) 2026 (7).

The top 6 were separated by just 28 points, and two of the coin-flip races the preview flagged in advance turned out to be exactly what separated the pack.

Final standings

Rank Player Points
=1 Mapes's 510
=1 Joseph Koyfman 510
=3 fergusmainland 509
=3 mmmaurel 509
5 Fraser Innes 483
6 Finn F 482

The shared base - where everyone agreed

All six players took essentially the same approach:

  • Harvard Lights sweep - Lights 1V (34), 2V (39). All six had both.
  • Princeton Invitational - Brown M 1V (34), Cal M 1V (28), Stanford M 1V (29). Everyone in.
  • The big contrarian call: Virginia W 1V - the preview had this one 55/45 as "a mild upset brewing" and all six top players had it. It paid 45, the biggest single points haul of the Princeton card.
  • Washington Dual sweep - Washington W 1V, 1V 4+, 2V, 2V 4+, 3V 4+, 4V. The 4V alone paid 51, the biggest prize of the whole series. Every top-6 player had these.
  • The Washington W 3V surprise - the one Washington boat that didn't win. Washington W 3V scored 0. Five of the top 6 still backed Washington here and took the zero. Only Finn F broke ranks and picked California W 3V (29 pts).

What separated 1st from 3rd - one coin flip

The preview warned Syracuse W 1V vs Brown W 1V was on a knife-edge at 51/49. That race is literally the 1-point gap between rank 1 and rank 3:

  • Mapes's and Koyfman went Syracuse W 1V → 30 pts
  • fergusmainland and mmmaurel went Brown W 1V → 0 pts

That's the entire 510 vs 509 gap. One race, one lane-call, one point.

What separated 1st from 1st - nothing, via different routes

Mapes's and Koyfman tied at 510 despite picking differently on two Princeton races:

Race Mapes's Koyfman
Rutgers/Princeton/Cornell W 1V (three-way) Rutgers (0) Princeton W 1V (29) ✓
Syracuse v Penn M 1V Syracuse M 1V (29) ✓ Pennsylvania M 1V (0)

Both ate one zero and landed one winner - same net, different paths. A neat illustration that there's more than one way to climb the leaderboard.

What separated the top 4 from 5/6 - the V4 trap

The gap from 509 to 483 is 26 points, and it comes almost entirely from one race: Dartmouth v Harvard Lights V4, flagged in the preview as a 51/49 coin flip.

  • Top 4 all backed Harvard Lights V4 → 27 pts
  • Fraser Innes and Finn F both backed Dartmouth V4 → 0 pts

That single wrong coin-flip is almost exactly the gap. Fraser's book is otherwise identical to Koyfman's —> 510 - 27 = 483.

What separated 5th from 6th - the W3V / W2V 4+ trade

One point, one decision. Finn was the only top-6 player to back California W 3V against the Washington sweep - and he was right (+29 vs Fraser on that race). But he paid for it by also picking California W 2V 4+ (0) where Fraser took Washington W 2V 4+ (30). Net result: −1 for Finn, 483 vs 482.

The takeaway

Four of the six top players ran almost-identical books. The series wasn't won by finding hidden value, it was won by losing fewer coin flips. Specifically:

  1. Harvard Lights V4 - the coin flip that decided who was in the top 4 vs who wasn't
  2. Syracuse W 1V vs Brown W 1V - the coin flip that decided 1st vs 3rd
  3. Virginia W 1V - the one consensus upset everyone correctly called; skipping it would have buried you

The preview called the two races that mattered (V4 and Syracuse/Brown) both as 51/49, and they ended up being worth a combined 27 points of separation. The margins were as tight as the predictions said they'd be.