You built execution, sizing, and exit rules. Under stress, the same brain overrides all three. Bias is not a soft topic—it is the leading cause of negative expected value after fees for forecasters who already understand the math.
By this point you have tactics without willpower mythology. Bias chapters exist because knowing Kelly does not stop clicking market buy at the top of a spike.
Why these markets amplify bias
Prices from zero to one hundred cents feel like precise probabilities. Politics and culture tie positions to identity. Feeds run twenty-four seven, feeding recency and availability. The book offers social proof. Binary payoffs become win-loss stories instead of calibration curves. Play-money courage does not transfer to dollars without a Brier check.
Probability fallacy chapters named the logic errors; here the focus is behavior that survives knowing the formulas.
The usual suspects
Overconfidence shows up as twenty-cent edges on liquid marquees when two venues cluster within a few cents—audit evidence, re-read resolution, add base rates. Recency treats the last poll as permanent; wait a news cycle, compare move to spread. Confirmation reads only bullish threads; steel-man the NO case for five minutes. Anchoring fixates on entry price; re-estimate probability fresh. Loss aversion hugs losers and sells winners—exit when remaining expected value is negative, not when P&L color hurts. Herding buys because everyone bought; real contrarianism needs a mechanism, not contempt. Narrative beats base rates on culture props. Sunk cost confuses research spent with forward edge. Illusion of control mistakes posting for moving odds unless you are the whale.
Overconfidence in numbers
You buy YES at 41¢ claiming 68% while Kalshi and Polymarket sit near 44–46¢ after fees. Twenty cents of edge versus two books is unlikely—maybe you misread rules or used inside view only. Adjusting to 56% still supports trade but shrinks size. The first ten cents of edge were phantom tax.
Kelly wanting more than five percent of bankroll on every third ticket is another tell—fraction down until calibration proves otherwise.
Recency and the cooling room
Headline hour zero: chase with a market order. Disciplined hour one: write pre- and post-mid; if move is smaller than twice the spread, ignore. Day seven: do not forget the move happened when recomputing. A cooling room rule—no trade for thirty minutes after major news when arousal scores four or five—filters many cascades that look like permanent repricing.
Confirmation and social proof
FinTwit threads that agree with you are not research. One strongest counter-article is worth more than ten supportive replies. Poll cherry-picks without comparing to market price repeat the poll-versus-market mistake from foundations material.
Contrarian versus herding again
Herding asks “everyone is buying.” Contrarian asks why the crowd is wrong with clean resolution and a limit at edge. Fading an active cascade without a catalyst is catching knives, not philosophy.
Loss aversion and disposition
Two positions: YES A up twenty-two cents unrealized, YES B down fourteen. Selling A to “feel safe” while holding B because “it can come back” is classic disposition. If A still has positive edge at the mark, hold unless portfolio heat demands trim; if B’s edge is negative, exit B regardless of comfort.
Doubling the loser to “get even” violates bankroll drawdown rules and turns sizing into therapy.
Platform and identity bias
Brand loyalty to one app, regulation halo on another, play-money heroics, and fee blindness on capped venues all distort venue selection you already studied. Rotate venues for signal, not tribe.
Controls that stick
Pre-trade checklist signed before submit. Cooling room tied to arousal score. Buddy pre-mortem. Monthly calibration from the journal. Bankroll caps trump conviction. Environment: no size changes on a phone in bed.
Willpower fails; friction and logging win.
Weekly bias audit
Pull the last ten journal entries. Tag overconfidence, recency, confirmation, or clean. On resolved trades, compare average stated probability on wins and losses to outcomes. If you are systematically five points optimistic, shave two points off future pre-trade probabilities for thirty days. Pick one trade you almost took emotionally and write why you passed or failed.
Identity and politics
When the contract is about a tribe you belong to, cut size automatically. Your brain will merge forecast with loyalty. There is no shame in smaller tickets on identity-linked events—there is ruin in pretending you are immune.
Screens and sleep
Phone trading in bed correlates with arousal scores of four and five. Move size decisions to a desk; use the phone for alerts only if you must.
Why this chapter sits late
Execution and sizing only work if you actually follow them. Bias is why smart people violate their own rules. Friction—checklists, cooling room, caps—beats motivation speeches.
Pre-mortem in one minute
Before submit, finish: “If this trade loses, it will be because ___.” If you cannot fill the blank with a factual risk—not bad luck—you are not ready. The sentence belongs in the journal row.
Key ideas
Bias is an execution bug with known patches. Overconfidence, recency, and confirmation dominate. Log arousal; steel-man NO; let bankroll caps veto conviction.
What comes next
Bias wins when memory rewrites history; logs keep you honest.
After a loss streak
Loss streaks trigger revenge sizing and narrative trading. The correct response is drawdown rules from sizing chapters, not “one good trade to get even.” Tag streak weeks in the journal; halve units until recovery metrics clear.
Misconceptions
“I am rational on politics” is the overconfidence talking. “Bias chapters do not apply to me” guarantees they do. “I traded small so emotion does not matter”—small trades train habits that scale. “Willpower is enough” loses to checklists every time.
Let journaling and caps be the adult in the room when headlines fire.
Next: Journaling and Tracking Your Performance