There are two ways of approaching the stream and live data disconnect in esports, said PandaScore CEO Flavien Guillocheau, but neither are ‘magic’ solutions.
As a follow up to this analysis of the relationship between latency and stream delay in esports, he went even further into the concept’s ‘crazy’ betting climate, where players are seeing action on their screens that is up to five minutes against the bookmaker odds.
First of all, he revisited the difference between latency and delay, as it is an important distinction to make – particularly from the perspective of an esports data provider.
Ordinarily, the latency or time taken for PandaScore to process and share the data is negligible, whereas the stream delay is a conscious process from the broadcaster.
As Guillocheau explained, there are good reasons for this, but it can significantly detract from the end user experience – both in-stadia and for those watching or betting online.
“The trouble with esports is that there is a symmetrical amount of information between the two teams in a game like Counter Strike, where you have to fight five players against five players and you’re trying to plant or defuse a bomb, or kill the opposing team.
“You don’t know exactly where every player is on the map, because someone might be hiding behind the corner. So, what’s happened before is that someone in the stadium that can see both teams at the same time has tried to share information with the players.”
Of course, the threat of this is negated by showing the spectators a stream that is three to five minutes behind the real action, but you must also consider the reaction of the players.
We understand that the guidelines are now for players not to react (as it would give away what’s to come next on the screen), but isn’t the in-stadia user experience significantly worse if you can’t celebrate with the players in real time?
“I agree,” replied Guillocheau. “And there are solutions to this where you might soundproof a booth for the players so noise isn’t shared with the crowd. But that can be more complicated if you’re organising smaller tournaments. Another option would be to have referees present in the room to prevent information being shared out to potential exploiters.”
Just to explain the threat of exploitation briefly, betting operators are at risk if a customer or group of customers becomes aware of the disparity between the stream and the odds data, and either has access to the data or is watching the event happen in real time.
Now, we’ve talked a bit about the in-stadia experience but the online consumption of esports from a betting perspective is perhaps more where Guillocheau’s ‘crazy’ comment comes in.
Because if you go back to the difference between latency and stream delay, there is pretty much no delay in the bookmakers getting the data reflecting what’s happening live (and pricing this up for the punters), it’s just that the stream follows subject to broadcast delay.
In reality, then, there’s no value in watching the stream to guide your betting. “Right now, you’re not watching the stream to bet, because you know you’re going to get screwed,” Guillocheau confirmed. “So the betting experience right now is just basically looking at numbers and saying, okay the odds move here so I can anticipate this happens.”
Obviously, there is a balance to strike between providing potential exploiters with too big a window to get hold of data, and giving enough time to match the stream to the live odds. Is holding back the data to tally with the live stream a complicated process?
“No, I mean, it’s easy to stick in a piece of code,” said the PandaScore CEO. “You can say don’t share the data for five minutes or whatever the delay is. You can even work around it, and potentially adjust the delay according to your integrity concerns.”
If one option to reduce the disconnect is applying a data (and live odds) delay, for example to four and a half minutes where there is a five minute live stream delay, let’s just clarify the other option. Well, for Guillocheau, this is essentially the current approach.
“What we are currently doing is protecting ourselves against the minority. We’re protecting against the exploitation of data by putting it straight out there, regardless of how it aligns with the stream. But we’re screwing the betting experience for the majority of punters.”
You might surmise, therefore, that he is in favour of syncing up the data with what you see on-screen, but he admits there is no ‘magic’ solution to the problem.
“Both options have pros and cons. It’s not like one is the magic solution. But I’d rather have a fair betting experience for the majority and have to handle the people that are exploiting and trying to beat the system, rather than providing a bad experience and unfair experience for everyone just to protect against the minority.”
We’ve covered off his preferred approach to harmonising the esports betting experience, but what about PandaScore? What can his company do as a key industry stakeholder?
“Our role is to educate as many people as we can on how it works, and what the impact is of delaying the stream. And how we can still protect ourselves from those cheating. There’s a big educational part that needs to be done. And that’s, for me, the first thing.
“Second thing, and it’s kind of a reaction. And that’s basically implementing the solution we talked about. But another parameter that is important for anyone to understand, especially the bookmakers, is that no one in the space will always offer you fast data. It’s impossible.
“If you work with one provider you will have exclusivity for a tournament, and that would be for different tournaments if you work with another provider. Working with everyone comes with a huge cost that most bookmakers can’t afford. So, usually, you will be without the fast data for a majority of tournaments. So, how do you react when you don’t have it?
“The first thing PandaScore acknowledged when we started is that it’s impossible for one player to get access to all of our data, it’s just not going to happen. So as a company, we want to build the best workflow and the best tools so that even if we don’t have access to the fast data we can perform.”
Wrapping up the conversation, we asked Guillocheau about the short to mid-term future for PandaScore. “We have spent a lot of energy in the past year to 18 months in developing the most robust and stable products,” he said. “As a newcomer, it’s crucial to have this.
“But now that everything works perfectly, and it’s been more than a year since we had any problems, we’ve started to focus more on new features – a few months ago, we launched player props on Counter Strike (CS:GO). We are very happy with their performance – it’s in the five biggest markets for CS:GO. And the margin is really good.
“We are also starting work on the BetBuilder. So it’s more about product development, new features, and a new exciting experience for PandaScore. With more data, you can deliver more exciting markets. So now we are entering this phase.”