On September 1, Turbo Stars CEO Alex Kozachenko, Odds Market's PM Allar Luks and Odds Market's CPO Sergii Mykhailenko joined Odds Market's first live panel, Choosing a Sportsbook Trading Model: Build, Buy, or Combine?, on sports trading and risk management. Over the next hour: where sportsbook volume actually comes from, who really bets on esports, why localization means more than translation, why Turbo Stars doesn't run its own trading desk, and whether the live-betting delay is protecting anything. Here is what came out of it.
Where sportsbook volume actually comes from
Sportsbook volume does not spread evenly across the market list. Alex: "The majority of volume is done on the top four or five markets: the winner, the handicap, the totals." Everything past that, including player props and derivative markets, is harder to build well and moves the needle for a narrower group of players.
Your product might be less appealing if you don't have something like micro markets, for a very small percentage of players. But again, it's a small percentage. The majority of bettors, from what I've seen, they're not that deep into these types of things.
The one exception is VIPs. A platform with a small user base can still need a wide market catalogue if a handful of high-value players are driving most of its volume: "If your VIP comes in and says, I want to bet on what color shoes Ronaldo will be playing his next football game, considering the volumes, you're best to make sure that you can cater this desire."
For a small or mid-size operator, that is the real question before adding more markets: is a wider catalogue serving your actual highest-value players, or just padding a feature list nobody asked for.
Who really bets on esports
The typical esports bettor is not a teenager. Allar: "It's probably a misconception that these players are young. There's actually a generation who have grown up playing these games." The profile: people with jobs and families who grew up on the original Counter-Strike and Battlefield roughly 25 years ago, betting on esports today more as a way of staying connected to something they grew up with than as a new hobby.
There is a business reason esports stopped being optional, too: players move between esports and traditional sport depending on what is live at any given time. Sergii remembered it from COVID, when Belarusian football kept running through lockdown and sportsbooks put it on the homepage out of necessity; Allar added that table tennis tournaments, played out of basements, briefly became prime content the same way. Combine that with players expecting something to bet on at any hour, in any time zone, and esports coverage stops being a nice-to-have.
The Gen Z acquisition story a lot of esports marketing is built on may be selling to an audience that mostly isn't there.
Why localization means more than translation
Same language does not mean the same market. Allar ran Estonia, Latvia and Lithuania under one operator: three markets outsiders tend to lump together as "the Baltics" and treat as interchangeable. "Everybody thinks very close countries, same interests, same people, but there are actually a lot of different priorities in terms of sports." Esports landed in Estonia's top three sports in some months, driven by a couple of world-class local players; cross into Latvia and interest concentrates around a completely different, much more local sport instead.
Personalization is the broader point (Alex):
Two people living in the same household, from the same family, with access to the internet, they're going to like very different things. It's literally up to us to make sure that they're provided with this personalized content, because this is something that makes those players feel like they're seen and they're heard. This is the one and only factor that's going to make sure you are going to come back to the same brand.
Language and currency are the easy part of localization. Knowing what the country next door actually watches is the part that's easy to skip. We measured what this looks like in practice in a recent personalization case study: same-player cohorts saw turnover per player per day rise 19% and bets per player per day rise 44% against a matched control that fell.
Why Turbo Stars doesn't run its own trading desk
Turbo Stars runs more than 60 operators globally and does not run its own trading desk. Alex: "We don't have a trading team, we have our own risk management team. We're taking control of the risk management, the limits, the delays, but we don't trade ourselves." The platform consumes odds feeds from specialist providers and puts its own team into monitoring and risk controls instead: "Those guys know what the hell they're doing, and I prefer those guys to take care of that."
Building a trading model in-house instead of buying feeds is a financial question more than a technical one. A proper model needs years of historical data, round-the-clock monitoring, and a data scientist on payroll at a salary Alex put at upward of 12,000 euros a month. And even with the budget, automation has a ceiling: ask a model to price a one-off world padel tournament, and it can't, not for lack of will but because the math has nothing to learn from yet.
Asked what would happen if a major feed provider disappeared tomorrow, Alex's answer: "I don't really think that anything will change." The skills needed to build a good football or basketball odds feed are not held by one company alone, so losing any single provider would not be the crisis it might sound like.
Live betting raises the same question in a starker form. Sergii pointed out that Pinnacle, a sportsbook known for being fully automated, still runs around 200 people trading live events out of Malta, and asked whether Turbo Stars would ever build a live trading team of its own. Alex's answer split the question in two: pre-match odds are close to fully automatable already, since the underlying data is public, but a live trading desk only makes sense if you are serving other operators rather than your own players, where regional taste in sport matters more than trading speed.
Before building an in-house trading team to look more self-sufficient, this is worth asking honestly: is it solving a real problem, or a perception problem. It is the same trade-off spelled out in managed vs turnkey sportsbook.
Is the live-betting delay protecting anything?
Every platform runs a 3 to 6 second delay on live bets as standard practice. Sergii: "I think the biggest issue with the experience is when a bet gets rejected." If a platform genuinely trusts the quality and speed of its own data, the delay itself may be protecting less than operators assume, and shortening it is a real option rather than something to treat as fixed.
If your data and infrastructure are as strong as you believe, this panel is a nudge to actually test that assumption rather than take the industry default for granted.
The full conversation runs close to an hour, and personalization is the one moment worth returning to: the reminder that even two people in the same house are two different markets.