The Pitfalls of Emotional Betting

The 2026 FIFA World Cup provided a revealing test of American betting behaviour. Dimers’ survey of 2,000 regular US bettors found that 51 per cent backed Team USA even when the odds and predictive models suggested otherwise, while bettors relying on data substantially outperformed those following instinct. For Adam Fiske, CEO of Dimers, the findings do not mean emotion has no place in betting. Instead, the challenge is helping recreational bettors distinguish between betting for entertainment and identifying genuine value.

Adam, was the emotional betting you identified during the World Cup uniquely about Team USA?

We see it in every sport. It could be NFL home-team bias, your favourite player or your national team. It’s very difficult to talk someone out of that. From a recreational perspective, it’s part of the fun. The question for us is how data can help improve the decision rather than somebody simply putting their heart ahead of their head.

The World Cup was particularly interesting because soccer is still relatively new to a lot of US bettors. You had the whole nation behind the USMNT and 51 per cent backed the US in their match against Belgium even when the odds were against them. There was clearly patriotism involved.

Is the issue that bettors don’t trust data, or that emotion overrides it?

One in six respondents were new bettors. They were likely soccer fans, either Americans caught up in the tournament and the USMNT story, or fans from other backgrounds supporting international teams, who haven’t been drawn into betting on US sports, so there are different factors at play. More broadly, adoption of data is growing. The US has had daily fantasy and fantasy sports for a long time, so people understand statistics.

What’s perhaps missing is the predictive element. A statistic or trend doesn’t necessarily mean something is going to happen again. We take a huge number of data points and try to predict an outcome. There’s an educational piece around the difference between forward-looking predictive information and just saying, “This happened historically, therefore it will happen again.”

Dimers’ predictive model generated a 14 per cent return across 123 World Cup bets. Where did it find the biggest edges?

Our best value came from moneyline bets, where we generated a 46 per cent return on investment across the 35 plays identified by our model. We were clearly finding more edge around underdogs, which was interesting because the four favourites ultimately made it to the latter stages, but as the ROI suggests, we the model found some really good opportunities that the human eye may have missed.
Futures were strong as well. We had teams such as Mexico, Switzerland and Canada progressing at good plus-money prices, which is a good example of using data to make a decision rather than your heart.

Did emotional betting extend to individual players?

Definitely. You had Messi, Mbappé, Kane and Haaland attracting enormous attention from recreational bettors alongside USMNT’s Christian Pulisic. Our player prop bets didn’t perform as well as our match models during this tournament. One reason was that these props markets carried higher levels of vig, which limited betting opportunities and left us with a relatively small sample.

One in six respondents placed their first-ever sports bet during the World Cup. Can operators retain those customers?

We don’t have enough data to answer that yet, but we’ll find out quickly with the NFL and college football seasons. The question is who those people were. Were they general sports fans placing their first bet because the World Cup was such a big event, or primarily soccer fans? There’s an obvious soccer retention strategy around introducing somebody who bet on the World Cup to MLS or the Premier League. The other question is whether another segment now migrates towards traditional US sports.

Prediction markets have become a much bigger part of the US conversation. Does that encourage more probability-based thinking?

We’re huge believers in probabilities and percentages because that’s effectively what we do. We take the likelihood of an outcome and compare it with the probability implied by the available odds. That’s where betting opportunities exist.

Prediction markets have arrived at an interesting time. Legal sports betting has been maturing for five or six years, and prediction markets have given the wider industry another boost in mainstream attention. People may perceive these products as closer to financial instruments or financial trading, which is true in some ways, but potentially something to watch out for in others. We need to be responsible about that as well.

Is social media encouraging bettors to chase eye-catching wins rather than make better-informed bets?

There’s some great content being produced and good people giving good information. But I think the incentive structure can potentially become imbalanced. Is the incentive attention? Is there another financial gain involved? Parlays are a good example. During the World Cup you’d see three-star players put together as goalscorers and, when it happened, social media posts highlighting the big win. But it wasn’t necessarily a mathematically strong bet.

Interestingly, social media-led bettors were wagering slightly more – around $55 per match compared with the $52 average – while potentially relying on worse information. Incremental gains aren’t sexy. Everybody is looking for the big hit. What we’re trying to teach people is that it’s a long game and finding edges over time is more sustainable than chasing one-off parlays.

How much responsibility should sportsbooks have for helping customers understand value?

I think there’s some responsibility around showing how much vig there is in a market, but operator responsibility extends beyond that. Building better player protections is something the industry needs to continue doing, and regulators should hold operators responsible for that.

Are American bettors still learning how to bet?

When we launched in 2019 and 2020, I thought this was a completely new space and nobody would know what they were doing. That was naive. Betting has existed in the US for a very long time, whether that’s Las Vegas, Atlantic City, offshore sportsbooks or local bookmakers. There has always been a betting culture.

The regulated market has matured quickly, but there’s still work to do around educating people about data, and making sophisticated information accessible. We’ve seen our own audience change significantly. They’re no longer necessarily casual players looking for basic information. They’re arriving with some knowledge already and looking for something more sophisticated.

Will NFL bettors repeat the mistakes you saw during the World Cup?

The NFL is different because it’s a very efficient market and there’s an enormous amount of historical data. The mistakes are more likely to come from things like putting three touchdown scorers into a parlay and trying to win the lottery on day one rather than following a sustainable betting process. One thing we see is that there’s often more value earlier in the week. By the end of the week the market has become more efficient. When lines first go up, that’s often where there’s an opportunity to find a genuine edge.

What should sportsbooks take from the World Cup findings?

There are opportunities to responsibly reward patriotism or fan engagement in ways that give users value without constantly asking them to bet more. We’ve seen odds boosts around the USMNT, for example. Rather than reducing the odds because everybody wants to back the US, you’re improving the price.
For a recreational bettor approaching it as entertainment at low stakes, that can create engagement. The important thing is keeping stakes responsible and encouraging people to only bet what they can afford to lose.

Is there a responsible gambling message in encouraging better-informed betting?

For most people, betting is a form of entertainment and we should accept and support that responsibly. What we can do is help make a player’s money last longer, help them have a good experience and have them come back when it suits them rather than pushing them to bet when it doesn’t. There are obviously thresholds where customers need to stop and we need better protections around those situations. We see our responsibility as helping people make better decisions that suit their financial position and make the experience more enjoyable.

Finally, as AI and predictive analytics become more sophisticated, what should bettors be wary of?
Machine learning and reinforcement learning have been part of this field for a long time. What’s changing is the emergence of more advanced techniques and natural-language AI. I’ve seen people asking Claude or ChatGPT for betting advice, which is absurd. For established businesses, AI can help build on an existing foundation of accurate, trustworthy data.

But I’m also seeing products being built very quickly using AI that haven’t necessarily been back-tested properly. They can have sharp marketing and look great, but the substance underneath isn’t necessarily in the bettor’s best interests. In some respects, that risks becoming the new social media tout.

We’re big believers in AI and it’s an important part of our technology stack, but education is going to become increasingly important so bettors aren’t simply asking an AI what they should bet on and assuming the answer is reliable.