Base Rates: The Starting Point Every Prediction Market Trader Ignores

Here is a question. The Fed has not cut rates in the last four meetings. There is a jobs report coming Friday. The market is pricing a March cut at 22 cents. Should you buy or sell?

Most traders will answer this with whatever they think about the current economic cycle — hot inflation, resilient labor market, the latest dot plot. That analysis is fine. But it skips the most important first step: What is the historical base rate for this type of event?

The Fed has cut rates in approximately 18% of meetings where the trailing CPI was above 3%. If the market is at 22 cents and the base rate is 18%, the gap is only 4 percentage points. The market is not dramatically mispriced relative to history. You need a specific reason to bet against a 22-cent price on an event that has happened 18% of the time historically. You do not have an automatic edge.

Compare that to a different scenario: same event, market at 45 cents, base rate 18%. Now the gap is 27 points. The market is pricing something that historically happens 18% of the time at nearly 45%. That is worth investigating. Either something has fundamentally changed in the current cycle, or the market is overreacting to recent news.

The Base Rate Fallacy

In cognitive psychology, the base rate fallacy is the tendency to ignore statistical base rates in favor of specific case information. When a vivid story is available — the Fed chair's recent speech, a shocking inflation print, a dramatic bond market move — the human brain discounts the boring historical frequency data and weights the specific narrative too heavily.

Prediction markets are full of this. After a major news event, prices move based on recency and salience, not on calibrated probability. The traders who move prices in the immediate aftermath of breaking news are not running base rate calculations — they are reacting. That reaction creates the gap you want to trade.

The base rate is not the final word. History does not predict the future perfectly. But it is the rational starting point — the prior before you layer in what is specific and current about this situation.

How to Use Base Rates Correctly

Step 1: Find the relevant reference class. This is the hardest part. "Fed meetings" is too broad. "Fed meetings where CPI is above 3%" is better. "Fed meetings where CPI is above 3% and unemployment is below 4%" is better still. The more precisely you can define the historical situation that resembles today, the more useful the base rate.

Step 2: Anchor your probability to the base rate. The base rate is your prior. If the historical frequency is 18%, start at 18% before you analyze anything specific about today.

Step 3: Update for what is different. Now ask: what is different about this specific instance that should push the probability above or below the base rate? Not everything that is newsworthy is actually informative. A hot CPI print when CPI has been hot for two years is much less informative than a sudden spike in a stable environment. The update should be proportional to how much this specific evidence discriminates between the event happening or not.

Step 4: Compare your posterior to the market price. The gap between your base-rate-anchored posterior and the market price is your edge. If the gap is large, you have a trade. If it is small, you do not.

Where Prediction Markets Diverge From History Most

Some event categories are consistently over- or underpriced relative to base rates. Not always in the same direction — but reliably divergent during high-attention periods.

Fed decisions after major economic prints. After a hot CPI or strong NFP, the market consistently underprices Fed holds and overprices cuts. The base rate for holds in high-inflation environments is much higher than post-print market prices imply. This is one of the most consistent patterns I have found.

Election incumbent advantages. Historical incumbent re-election rates are around 68%. During periods of negative news coverage, markets sometimes price incumbents below 50%. That gap is rarely justified by the base rate.

Year-end asset price targets. The base rate for Bitcoin closing a calendar year above its January 1 price is 5 out of 10 years (50%). When Bitcoin is up 30% in Q1, prediction markets price December-close contracts well above 70%. The Q1 trajectory is not as predictive as the market implies.

Consecutive economic events. GDP contractions, consecutive Fed hikes, consecutive Fed holds — the base rate for any economic state continuing is usually high. Markets frequently underprice the status quo.

Sample Size and Data Quality

Not all base rates are equally useful. A Fed rate decision base rate built from 40+ meetings over 10 years is robust. A Bitcoin year-end price base rate from 10 years of data is directionally useful but not highly precise. A new market category with only 3 historical observations is barely informative.

When using the base rate scanner, look at the sample size. High-confidence base rates (25+ observations) are strong priors. Low-confidence base rates (under 15 observations) should be treated as weak priors that update easily with current evidence. Do not trade a small-sample base rate gap the same way you trade a large-sample one.

Run the Base Rate Scanner to see current gaps across 12 event categories.

The base rate is not a trade signal by itself. It is a starting point. It tells you where to look, not what to do. The trade is in the gap between the base rate, your updated posterior, and the current market price — all three together. Get that sequence right and you are trading with the weight of history behind you.