Sports betting risk in 2026: insights from Sergii Mykhailenko

A turning point is approaching for sportsbooks in 2026: customer acquisition costs have soared and the traditional approach of aggressive marketing and manual risk management is no longer effective. As costs rise, platform efficiency is no longer a ‘nice-to-have’; it is now a core requirement.

Sergii Mykhailenko is the Chief Product Officer (CPO) of OddsMarket, a company that has spent nearly a decade developing risk and trading tools used by some of the biggest names in the industry. He sat down with SiGMA News to discuss the shifting landscape of 2026: from the rise of professional betting syndicates to what the future holds for operators of all sizes: who will survive, consolidate, or exit the market.

Why risk management technology is becoming more sophisticated

What changes have you seen in risk management over the past few years?

Competition among bookmakers is fierce, and with traffic costs climbing, sportsbook companies are investing not just in marketing but in making their platforms more efficient, getting more revenue from the players they already have. These days, fewer operators run their own tech in-house. Most are moving to third-party platforms, but that space is getting increasingly competitive. Every year at iGaming conferences, I spot two or three new providers jumping in.

A platform is super complex, and a solid risk management system is a huge factor in how well it performs. At OddsMarket, we get to look at data from our platform clients, and it reveals significant disparities: two platforms in the same market can have trading margins and GGR that differ by 20-30%. And it’s not always about traffic quality.

Risk management tech has gotten a lot more sophisticated, too. The top platforms have been using machine learning to track player behaviour for a while; that’s well established. But now, with large language models being widely available and easier to implement in betting analysis, nearly any platform provider can build their own algorithms using big data.

On top of that, we’ve seen a growing number of specialised companies popping up, offering standalone solutions for player behaviour analysis. These can be run on a client’s secure servers or get betting data via a restricted endpoint for analysis, usually under NDAs and with limited data.

We are also developing a similar product at OddsMarket and plan to introduce it later this year.

At the same time, despite margins in the sportsbook business gradually declining, companies are investing more and more in fast data feeds to analyse a broader market. We’re seeing growing demand for our API products as operators look for more efficient ways to monitor odds and market movements in real-time.

What does the profile of a sharp bettor look like in 2026? Have they become more tech-savvy?

Absolutely, but they’ve also become more efficient. In many ways, they’re doing what platforms do, just from the other side of the market. Over the past four years, we’ve seen a clear trend: the number of individual sharp bettors is dropping, while betting syndicates are on the rise. People are increasingly teaming up. Most of the time, these groups form around a piece of software that, for a modest monthly fee, gives signals on where to place bets. In Latin America, for example, this ‘software’ is often just a Telegram group or bot sending arbitrage or value-bet tips.

Another trend is the rise of ‘betting academies.’ For a few hundred euros, people learn how to beat the bookies using subscription-based tools. The academy owners usually work as affiliates: they train users on a specific platform or tool and then earn ongoing commissions from the clients’ subscriptions. Many of these owners eventually develop their own sharp-betting software.

In countries like Italy, the UK, Greece, Albania, and Bulgaria, betting syndicates have become a serious force. They operate like real businesses: offices, scouts attending matches, in-house developers, and sizable budgets to get the fastest data feeds for real-time odds analysis. And of course, they always have reliable, ongoing access to new betting accounts.

Prediction markets as an additional channel for punters

We can’t skip over prediction markets. Are they really starting to influence sharp bettors and the betting market as a whole?

The idea behind platforms like Polymarket or Kalshi is to get the widest possible audience to bet on just about anything. Their approach is simple: mostly yes/no questions. In sports, that basically means moneyline-style bets – did it happen or not? Most of the action naturally centres on big events and top leagues.

There’s no doubt these platforms have huge liquidity and betting volume, which usually leads to very accurate pricing. The data is also open and often available via public APIs. But in reality, most sharp bettors aren’t making their money on moneyline bets for popular games. Their profits come from niche events or markets where calculating the correct odds is much tougher. So, in this context, prediction markets are more like a temporary, additional channel for punters and betting syndicates to squeeze out some extra profits, rather than a game-changer in how they beat bookmakers.

How platforms can better protect operators

In live betting today, it’s all a fight for milliseconds. Where are the biggest operators still losing the most money?

Here’s an unpleasant truth for any young sportsbook platform: syndicates and punter communities are constantly on the lookout for new operators running ‘raw’ software. For innovators trying to build their own platform from scratch in 2026, the hard reality is this. If you don’t prioritise a highly efficient risk management system and implement it as a turnkey solution for your operators, your first inexperienced operators will suffer major losses very quickly. And, of course, they’ll blame the platform for all their failures.

Fast data from a broad market is critical for the sportsbook business today. Sharp bettors won’t forgive even a few seconds of delay.

AI is increasingly being used in trading, but could it also introduce new risks? If multiple operators rely on similar models and data, could a single error in a feed trigger a synchronised collapse of margins across the market?

First, it’s important to clarify that widespread AI adoption doesn’t automatically mean more open data sharing between bookmakers. The infrastructure isn’t yet powerful enough for AI to independently parse websites where data updates at web-socket speed. If we’re talking about a trading feed provider using AI to adjust odds while analysing more parameters than before, I don’t see any major risks. In fact, more accurate odds generally lead to higher margins, no matter which part of the world an operator is in.

I’m more sceptical about the interactive chatbots that have recently become popular, the ones that ‘help’ players place bets by providing real-time match analysis or even direct betting tips. By the nature of the business, it’s not in a bookmaker’s interest to help players win, so these bots mostly create the illusion of assistance. They’re more likely to annoy the player, which is inevitable as soon as a bet recommended by an AI bot loses.

While I’m sure A/B tests of these bots often show an increase in sportsbook GGR, this is mostly due to a higher volume of bets, rather than any fundamental improvement. The bigger question is whether this approach could significantly reduce player lifetime value.

Many focus only on real-time data, but how does analysing historical odds, down to the second, help identify systemic gaps?

I’d start by splitting this into two areas: retrospective analysis of trading and retrospective analysis of player bets.

When it comes to trading, the focus is usually on things like:

• Average duration of open lines for a given sport or championship (an open line = potential risk from “bad” bets; a closed line = lost revenue).
• The margin at which an event was traded at any given moment, and how competitors behaved.
• How often were odds exploited for arbitrage against competitors’ lines, and who was “right” more often?

This kind of analysis isn’t particularly complicated, but it’s extremely revealing. It allows a trading department to correct past mistakes fundamentally. We run these types of backtesting analyses for our clients. To do this, you need fast data from the relevant sportsbooks recorded in an archival database with a granularity of at least once per second.

Retrospective analysis of player bets is more complex. Even today, most operators rely on manual limits, essentially cutting off players once they exceed a certain allowed profit threshold. Put simply, anyone who makes more than $1,000 gets limited. But can you really afford to do that in 2026, when player traffic and onboarding are both expensive and challenging? There are now plenty of tools for much deeper player analysis. At OddsMarket, we specialise in identifying arbitrageurs, value bettors, and late-betting players, giving operators a far more precise approach than blunt manual limits.

Sergii’s forecast for the betting industry

Looking ahead, what trends do you think will define sportsbooks, bettors, and the betting industry over the next few years?

The next four to five years will see further consolidation across the market, both among bookmakers and punters. Only the largest operators will survive. Bigger players will likely absorb mid-sized ones, while smaller operators will either grow into mid-sized players or disappear entirely. On the punter side, people chasing easy profits are increasingly handing over their accounts to syndicates, letting these groups earn money on their behalf.

Regulation is another interesting factor. In recent years, we’ve seen more developing countries introduce licensing frameworks in LATAM, Africa, and parts of Asia. It’s still unclear how white and grey operators will coexist in these markets. For now, the trend seems to favour grey-market growth, even as fewer countries remain without a formal gambling licensing process.

Fast data will play an increasingly important role in sharp betting and player protection. Platforms are slowly gaining the upper hand: the profitability of sharp betting is declining fast due to the systematic reduction of player lifetime value.

We’re also seeing the rise of so-called ‘crypto bookmakers,’ which can be grouped with prediction markets. Essentially, these are grey-market sportsbooks. Their growth isn’t accidental: players are migrating to these platforms, clearly driven by over-regulation and the increasing sharpness of regulated operators.

Finally, I expect a certain thaw in the traditional sportsbook market. Operators may start removing features that frustrate players. KYC funnels are a prime example: registering with a licensed sportsbook often requires filling out countless forms,  whereas a crypto bookmaker asks for only basic information. That gives a clear advantage in conversion and allows platforms to afford more expensive traffic. Of course, complex KYC is mandated by regulators, so I expect some lobbying and pressure to simplify these processes.

A new frontier rises beneath the skyline of São Paulo. From 06–09 April 2026, BiS SiGMA South America transforms LatAm’s gaming capital into a hub of innovation, bold talks, and billion-dollar opportunity. Don’t sit this one out.