If you’ve sailed Gotland Runt, you know that Almagrundet means you’re nearly done.

Finish, Sandhamn, beer, Värdshusbiff and some well-deserved sleep.

Twenty years of conditioning.

The Baltic Sea Race course put that exact mark a third of the way into the course. With 430 nm of heavy upwind and tight reaching still to go. So when we bore away at Alma after 36 hours of racing, leading IRC Two, two-handed and FinRating, we could practically hear the music from Seglarhotellet.

Homer covered this three thousand years ago.

The sirens’ trick is that they sing your song. Odysseus knew willpower would lose, so he didn’t rely on it: wax in the crew’s ears, himself lashed to the mast, standing orders to ignore the captain when the captain begs. Behavioral economists later named it a Ulysses contract.

We didn’t have one.

But we had our car parked at Bullandö Marina.

We’ve since done a fairly brutal debrief; tracking data, weather archive, and an uncomfortable amount of psychology. Our conclusion: the decision was probably right for this race.

But the decision making process was wrong in ways that would bite much harder in a race that mattered more.

Here’s what we learned:

1. Decide what the race is for. Before the start. Asked separately how we’d actually treated BSR, we both said ~80 % “we’re here to win”. Asked how we wish we’d defined it: 40 % racing, the rest would be experiment and training. When the fleet collapsed and the best shorthanded teams retired, the 80%-frame made retiring rational.

2. Chart the sirens along with the course. Every course has places where quitting is cheap; a home harbor or a familiar marina near a mark. If the course rounds something your brain files under “finish”, expect the music. The kicker: we had identified this exact mechanism before the race…

3. The conditional pact IS the decision. Somewhere mid-morning we agreed: “if the forecast still looks like this at Alma, we’re done.” That sentence felt like postponing the decision. It was the decision. From that moment, every forecast check asked “do we get to stop?” instead of “should we?”.

4. Classify before you argue: RED, AMBER, GREEN. RED = safety, damage, dangerous fatigue: you stop, no philosophy required. AMBER = the sporting equation genuinely changed: analyse it. GREEN = cold, wet, bored, unmotivated, far from home: not a reason. That’s the sport you entered.

5. Double-handed consensus comes fast. Too fast. Both of you doubting privately, neither wanting to be the weak one. Then one careful “are you thinking what I’m thinking?” And the answer is relief. Two private doubts become one joint decision in minutes, and it feels like truth because you agree. Nobody made the case for continuing.

6. Nobody will hold you accountable. That’s the trap. We posted our RET with a screenshot of the leaderboard. 188 reactions: “wise”, “smart”, “love the honesty”. One single critical comment. Retiring while leading pays out socially like a win, at a fraction of the cost, which quietly teaches your brain to do it again.

7. Sometimes retiring is right. So decide in advance which races deserve the grind. Annie Duke’s research says most people quit too late; walking away at the top is rare and sometimes optimal. The skill isn’t grit – it’s aim. We now carry a one-page decision contract: the frame in percentages, what success looks like without a podium, the sirens marked on the chart, and a protocol for the moment “should we retire?” comes up.

Worth remembering how the story ends: Odysseus got to hear the song and sail past. That’s the actual goal. Not a crew that never retires, but a crew that knows why it continue and why it retires.

Decision contract — template v0.9

Built by the crew of J/99 BLUR after retiring from the RORC Baltic Sea Race 2026 while leading. Not a promise never to retire. A promise to know why we continue or why we retire.

Fill in BEFORE the start, per race. Ten minutes, both crew present.

Before the start

1. Frame (adds up to 100 %, both sign):
RACING ___ %
EXPERIMENT ___ %
TRAINING ___ %

2. Success without a podium. This race was worth doing even if we don’t win, provided: ______

3. What’s at stake. Qualification, season goals, anything else: ______

4. The sirens on this course. Where along the track is quitting cheap? Home waters, a comfortable harbor near a mark, a place our brains has filed under “finish”. Mark them NOW: ______

5. Retirement classes:

• RED = safety or capability: injury, illness, dangerous fatigue, critical gear failure, weather outside our margins. We stop. No discussion needed.
• AMBER = the sporting or strategic value of the race has genuinely changed. May be examined, but only through the protocol below.
• GREEN = cold, wet, bored, unmotivated, a long way to go, missing the shower. Never a reason on its own. This is the sport we entered.

During the race: when “should we retire?” comes up without a RED

a. No conditional pacts. A sentence like “if the forecast still looks like this at X, we retire” IS a retirement decision.

b. No consensus in the first conversation. Separately, in writing, five minutes: my reasons, each classed RED / AMBER / GREEN, plus “if I were deciding alone, my probability of retiring is __ %”.

c. Steelman both directions, out loud. The best possible case for continuing AND for retiring.

d. Goal test: “Does retiring affect what we set out to achieve in points 1–3?”

e. Decide no earlier than one full watch cycle after (b). Until then the plan stands; the default is continue. The burden of proof is on retiring: “What relevant change means this contract no longer applies?” If that question can’t be answered concretely, the answer is GREEN.

After the race

f. One line in the debrief: did the contract hold? Was the question raised, and was the protocol followed? What changes in this template for next time?

v1 — untested.

Further reading: Annie Duke, Quit

I actually listened to her first book, Thinking in Bets, just a few weeks ago when we thought about our decision system around weather models.

Should have read this one instead.

Annie Duke played poker for a living before she started writing about how people make bad decisions. Her main finding is that most people quit far too late, not too early. Leaving feels like admitting the whole thing was a mistake, so we don’t. Grit gets all the credit; knowing when to walk away is the harder skill.