Prediction Users Match Bettors’ Math Confidence, Not Test Scores

Sports bettors scored 9.56 out of 11 on an objective numeracy test in a new US study. Prediction-market users averaged 9.30. On self-rated numerical ability, the gap was much smaller: 4.75 out of six for sports bettors and 4.66 for prediction-market users.

The figures come from an American Gaming Association study released September 25. The final sample included 3,149 US adults.

Prediction Users Scored Close to Non-Gamblers

The objective score for prediction-market users was almost identical to the result among people who did not gamble. Prediction users averaged 9.30. Non-gamblers were at 9.28.

Their own assessment of their skills was further apart. Prediction-market users gave themselves an average score of 4.66, compared with 4.36 among non-gamblers. Sports bettors led the five groups on the objective test at 9.56.

The questions covered percentages, fractions, probability and numerical risk. The subjective section asked participants how comfortable they were working with numbers.

So the prediction-market group did not perform poorly. What stands out is the distance between how close it was to sportsbook users in confidence and how much smaller that similarity became once respondents were tested.

Sports Markets Accounted for 65% of Prediction Users

The prediction-market sample was also heavily exposed to sports. Of the 574 respondents in that group, 374 said they had traded sports-related markets. That is 65.2%.

Another 230 respondents said they used prediction markets only for sports, equal to 40.1% of the segment.

Those figures also show why the two groups should not be treated as completely separate audiences. Respondents could have experience with several gambling or prediction products. For the analysis, they were placed in the category they identified with most closely.

Kalshi was used by 258 respondents in the prediction-market group. Polymarket had 226 and Robinhood 173.

Gamblers Scored Higher on Financial Literacy

The broader financial-literacy results also favored the gambling groups.

• Gamblers averaged 3.93 out of five, against 3.72 for non-gamblers.
• 41% of gamblers were classified as highly financially literate, compared with 27% of non-gamblers.
• Sports bettors recorded the highest segment score at 4.18, with prediction-market users next at roughly 4.05.

That does not show that gambling itself improves financial knowledge. The groups were not identical on income, age or education, all of which can affect financial-literacy results.

How the Survey Was Conducted

The survey was conducted on Prolific from July 15 to 21. A total of 3,201 responses came in; 3,149 remained after screening. AGA lists a ±2 percentage-point margin of error at the 95% confidence level.

The association commissioned the study. AGA has separately opposed treating sports prediction contracts as financial products.