Polls, different questions
The table compares field dates, samples, findings and funding or commissioning information. Scripps used registered voters, Paradigm used a non-probability likely-voter sample, and POLITICO reported an understanding-screened subset. Their percentages cannot be treated as changes in one population over time. Sports use, putting money down and browsing odds are different activities.
Gambling or investing?
The AIBM/Ipsos poll found that 61% described prediction-market event contracts as closer to gambling and 8% as closer to investing. These are respondents’ perceptions, not legal classifications. UK readers should use the Gambling Commission and FCA context for legal questions about products offered in Britain.
Who says they use sports markets?
SSRS reports 5% of American adults using sports prediction markets, rising to 11% among men aged 18–34. The subgroup is narrower and its estimate should not be generalized to all men or all prediction markets. Other polls use “bet,” “trade” or “check a forecast” differently.
Why sponsorship belongs next to the number
The Paradigm/Echelon poll was commissioned by a crypto investment firm with an interest in the sector. That does not make the data automatically wrong; it does mean question wording, sample and sponsor should be visible together. A concise headline without those details can mislead.
The poll table.
Sort by column; percentages are not directly comparable across different questions and samples.
| Link | |||||
|---|---|---|---|---|---|
| AIBM / Ipsos | 27 February to 1 March 2026 | 2,363 US adults | 61% say closer to gambling; 8% investing Fieldwork 27 February to 1 March 2026. Sample 2,363 US adults. In the same poll about one in five said they were familiar with prediction markets. Q3 reports 3% using Kalshi and 1% using Polymarket in the past six months, not a combined 3% usage rate. Q1 records 21% very or somewhat familiar. | American Institute for Boys and Men | View poll |
| Scripps / Talker Research | 14 to 22 April 2026 | 2,000 US registered voters; double-opt-in online sample | 19% report betting; 63% doubt insider trading can be stopped | Scripps News | View poll |
| SSRS Sports Poll | Published 28 May 2026 | US adults 18+; article does not give the base or field dates for this estimate | 5% report sports-market use; 11% of men aged 18–34 | SSRS | View poll |
| Politico / Public First | 17 to 19 May 2026; published 22 June | 1,667 adults passing an understanding check, from 2,035 US adults | 19% positive about the rise of prediction markets | POLITICO / Public First | View poll |
| Paradigm / Echelon | 13 to 18 February 2026 | 1,008 US likely voters; non-probability online sample; favourability among those aware | 32% favourable; 20% unfavourable among aware respondents | Paradigm, a crypto-focused investment firm | View poll |
| Public First (Great Britain) | 29 May to 2 June 2026; published 11 June | 2,076 British adults | 58% never heard of prediction markets; 10% have used one; 48% say allow with limits | No external sponsor disclosed in the article | View poll |
| Morning Consult (US) | 17 to 22 March 2026 | 15,029 US adults | 81% say sports betting on prediction markets is gambling | Gambling Is Not Investing, a coalition campaigning against sports event contracts | View poll |
Methodology.
This is a comparison of published polls, not a pooled estimate. Sample frames include general US adults, an ongoing sports poll and likely voters. Some percentages are conditional on respondents’ familiarity. Funders and fieldworkers are disclosed alongside each row where established; check the linked methodology before comparing subgroups. Public First chart footers supply the British field dates. The SSRS annual sample of over 24,000 is not the base for this specific estimate. Paradigm’s detailed methodology gives 1,008 respondents, although its introduction rounds to 1,000.
Sources.
-
01
Americans view prediction markets as closer to gambling
Surveys & pollsIpsos / American Institute for Boys and Men, 17 March 2026
Funder: American Institute for Boys and Men / Ipsos.
Ipsos / American Institute for Boys and Men, 17 March 2026. Topline Q7: 61% closer to gambling, 8% investing. Q3: Kalshi 3%, Polymarket 1%, past six months; overlapping selections, not a combined usage estimate.
-
02
Scripps News poll: Most Americans doubt markets can stop insider trading
Surveys & pollsFunder: Scripps News / Talker Research.
19% reported placing a bet on prediction markets; 63% had little or no confidence insider trading can be stopped.
-
03
How many Americans use prediction markets for sports?
Surveys & pollsFunder: SSRS Sports Poll.
5% of US adults reported sports prediction-market use; 11% of men aged 18–34 reported it.
-
04
Poll: Americans draw a new line in the betting bonanza sweeping over Wall Street — politics
Surveys & pollsFunder: Politico / Public First survey; this citation is secondary reporting.
Original poll report: 19% positive. Fieldwork 17–19 May 2026; 1,667 respondents passing an understanding check from 2,035 US adults.
-
05
Paradigm February 2026 Poll on Prediction Markets
Surveys & pollsFunder: Funded by Paradigm, a crypto-focused investment firm; fielded by Echelon Insights.
32% favourable and 20% unfavourable among respondents who had heard of prediction markets; likely-voter sample, not all US adults.
-
06
Place Your Bets: How Brits Think About the New Wave of Prediction Markets
Surveys & pollsPublic First, 11 June 2026
Funder: None disclosed
Public First, 11 June 2026. Chart footers state fieldwork 29 May to 2 June 2026, n=2,076. No external sponsor disclosed.
-
07
Most Americans say prediction market sports betting could increase harm
Surveys & pollsMorning Consult / PR Newswire, 31 March 2026
Funder: Gambling Is Not Investing
Morning Consult / PR Newswire, 31 March 2026. Fieldwork 17 to 22 March 2026; 15,029 US adults.
-
08
AIBM / Ipsos prediction markets: topline and methodology
Surveys & pollsFunder: American Institute for Boys and Men; fieldwork by Ipsos.
27 February–1 March 2026, 2,363 US adults; probability-based KnowledgePanel. Overall sampling error ±2.2 percentage points; subgroup errors higher.
Follow the original documents for definitions, scope and subsequent revisions. Browse the full source library ↗