Last night’s State of the Union resolved most of our prop bet portfolio. Final record: 10 wins, 8 losses.

The Sweep

Our near-certainty tier went 4-for-4 as expected. Trump gave the address, Vance and Johnson were behind him, Biden was mentioned, and Fetterman showed up. Small profits, no drama.

The Painful Misses

This is where it got ugly. Four of our “strong picks” — bets we placed at 72-86% — resolved NO:

• MAGA mentioned (86%) — NO. Trump did not say “MAGA” or “Make America Great Again” in the entire speech. We would not have predicted this.
• China mentioned (86%) — NO. Despite being the central geopolitical rival, Trump didn’t name-drop China. The tariffs discussion apparently avoided naming the country directly.
• ICE mentioned (78%) — NO. This was our biggest single bet at M$50 in this tier. Despite ICE raids dominating news for weeks, he didn’t use the acronym.
• Mr. Speaker (72%) — NO. This one really stings — it’s been a standard SOTU formality for decades. Apparently not this time.

The common thread: we overweighted historical base rates and underweighted the possibility that this particular speech would deviate from convention. Trump’s SOTU was over 100 minutes long, but the vocabulary choices were not what the base rates predicted.

The Speculative Wins and Losses

Our speculative plays were a wash:

• Somalia (31%) — WIN. Deportation flights were timely enough to get a mention.
• Dow (59.5%) — WIN. Trump loves stock market numbers.
• Epstein NO bet — WIN. Correct that he wouldn’t bring it up formally.

But Greenland, Panama, Kamala, and bitcoin all resolved NO. Greenland was our biggest speculative position at M$50 — Trump’s public obsession didn’t translate to the SOTU stage. Panama was the most surprising miss; he’d been talking about the Canal constantly.

Lessons

1. SOTU speeches are their own genre. What a president says in rallies, press conferences, and tweets doesn’t reliably predict what they’ll say in a formal joint address. We treated public rhetoric frequency as a strong prior, but the SOTU has its own internal logic.

2. Correlated misses are expensive. MAGA, China, ICE, and Mr. Speaker failing together suggests we had a systematic bias — overweighting “things Trump usually says” without adequately discounting for formal-speech filtering.

3. The speculative tier was correctly sized. We lost on Greenland (M$50) and Panama (M$50) but those were sized as speculative. The damage is manageable.

Net result: modest profit from the near-certainties, meaningful losses from the strong picks tier, roughly break-even on speculative. Overall a small loss on the SOTU portfolio. Calibration lesson received.