If you build it, they will come

If you quizzed a panel of venerable gambling industry execs for their views on key milestones in the sports betting industry over the past 25 years the majority would probably highlight betting exchanges, the explosion of in-play betting, the transition to mobile and the creation of cash-out and partial cash-out. On reflection, true game-changing developments and innovations have largely been few and far between.

Yet one potential – and arguably contentious – addition to that list of seminal developments is the request- a-bet personalised betting model pioneered by Sky Bet and copied by rivals with their own monikers for their services – #YourOdds (William Hill), #WhatOddsPaddy (Paddy Power) #PriceItUp (BetVictor) and so on. While these user-generated betting markets don’t have the broad appeal of in-play betting or cash-out, you can’t deny they are wildly popular with casual bettors.

Now, though, the natural evolution is automation with the emergence of à la carte betting products allowing customers to combine related contingencies via an algorithmically calculated price. Ordinarily, you can’t back a team to win and also win two-nil as a double because these bets are related. But with build-a-bet products, users can add or remove selections from a host of core and derivative markets to increase or decrease the odds for their bet.

For example, Lionel Messi to score, Luis Suarez to be booked, over 3.5 goals and under 10.5 corners could equate to odds of 25/1. Throw in Cristiano Ronaldo to score anytime and suddenly the odds

I jump to 50/1. It especially suits those punters who tend to bet multi-leg accumulators on a Saturday afternoon but there is one match on TV and they don’t want to bet on an improbable longshot like a central defender to be the fi rst goal scorer at 50/1. “There aren’t that many longshot selections out there, so the ability to build something up makes sense for those users,” says Propus Partners’ Mark Israney. “If you present them with a single 50/1-shot it just feels so remote, but if they are watching the Sunday game at 4pm, they can have a 50/1 bet without having to be on a defender to score first.”

According to Russell Yershon, former head of the UK sportsbook for Titanbet, it’s a “win-win for both bookie and punter”. “With the depth and breadth of markets already offered to punters and the live environment being filled with plenty of options, including minute markets, the request-a-bet product offers huge opportunity for the recreational and casual punter who wanted to place a small wager on a 50/1-plus bet on multiple events happening within a match.”

He continues: “Like cash-out, it was the established brands who released this first and now many of the UK sportsbooks have this product. Bet builder products are the same.”

Market forces

Earlier this year, bet365 unveiled its own in-house product, Bet Builder. Users can combine up to six selections to create their unique bet on any football match they choose. Interestingly, bet365 was conspicuous by its absence in the clamour to roll out a request-a-bet-inspired service. Now we know why; the online heavyweight was instead developing a proprietary version without the need for manual intervention.

Then, in April, BetVictor, which prefers to take control of its product and tech roadmap, unleashed its own in-house product, #PriceItUp Builder. Users are able to combine 10 individual selections into a single-match accumulator, while an ‘Edit’ feature means existing pre-match outcomes can be altered.

“There are obviously lots of challenges in the generation of these prices, given the highly related contingencies at play, and there’s a balance to be found between purity of pricing and speed of returning that price to the customer,” says Eoin Ryan, the Gibraltar-based operator’s head of sportsbook product.

For smaller players in the sports betting space who don’t necessarily possess the in-house tech fi repower or time to devote resources to developing this sort of product, it makes business sense to turn to third-party suppliers. Sportcast created their single- event accumulator service – BetBuilder – and have teamed up with Betgenius to help roll out to a number of operators, with more poised to launch in time for the World Cup.

“We wanted to remove the limitations customers face in betting – allowing for almost any bet or bet combination, and providing a new breed of interface to support that” says Sportcast MD Basil Hassan. “So, handling all of the contingencies was built in from day one, and pricing was key. We started with a mathematical modelling approach to ensure all of the contingencies are accounted for and the models are built correctly.

We also applied in-depth data science and machine learning to provide a scalable and accurate approach to pricing the more granular selections – in context to the expectations of a match – such as player shots or statistics. You need a level of data science and automation to drive this accurately and scalably.”

Sportcast also employs expert sports analysts to provide a “blend of human expertise over the top of the data science”. “They can finesse or account for scenarios that the maths can miss or sometimes get wrong,” Hassan explains. Because Sportcast is heavily reliant upon cloud technology, he says one of the biggest challenges was deciding on the right architecture to provide growth, both in terms of capacity and the product suite. “At the moment, we can support tens of thousands of requests dynamically priced per second. Google typically handle about 40,000, so we have the scalability solved because of the choices we made around architecture, and a framework that supports a fast and nimble approach to product growth.”

Cast your mind back

More than two decades ago, bookmakers introduced ‘scorecast’ (first goal scorer and correct score) football bets, with a grid on the back of in-shop football coupons displaying the odds for various outcomes. Scorecasts, and then ‘wincasts’ (anytime scorer and correct score), were rudimentary versions of build-a-bet products and a way for football bettors to back a longshot on a live match.

Much later, in 2013, BetVictor released iOS app Spincast, a football bet combining the match result, overs/unders and anytime scorer markets packaged in a slot machine-style UI. Nevertheless, it is somewhat surprising it has taken until now for these customisable single-game accumulators to arrive.

Israney, who has written the maths behind a yetto- launch build-a-bet product for an unnamed bookmaker, says he first had the idea when he worked as Ladbrokes’ sports product manager and then head of trading solutions between 2011 and 2013.

“Bet builder has been knocking around for a while, and it’s not just me who had the idea. I’ve spoken to other people and they said, ‘yeah we spoke about that, but we couldn’t do it’.” Many operators have been hamstrung by legacy platforms, outmoded back-end systems and a reliance upon third-party providers, while product roadmaps may have been focused on other tech projects and, in some cases, complicated company mergers.

Moreover, the industry has been so preoccupied with in-play betting and adding countless derivative markets, live streams and, of course, the ubiquitous cash-out feature, that investment in the pre-match product has, quite understandably, taken a back seat. Yet, as Ryan points out, advancements in trading technologies has played a significant role in BetVictor being able to create #PriceItUp Builder.

“Without advancements in technology trading, operators wouldn’t be able to price a 10-fold single-match accumulator with any confidence.” Calculating the pricing by figuring out the impact of one market on another isn’t the hard part; it’s telling the systems to deal with the new logic. Indeed, the industry has spent years coding systems to prevent punters from being able to place multiple related contingency bets.

“Many of the legacy systems still in use today defend the business against related bets being placed by attributing a unique ID to all selections within a specific class,” says Jeevan Jeyaratnam, headof trading and compilation at Abelson Odds. “When the system is asked to present all possible bet options, it checks this identifier before deciding if selections can be coupled. Recoding the base source for a machine is tricky and problematic work – the potentially unforeseen knock-on effects for a large back-end ecosystem are too great for many firms to consider it viable.”

For those looking to offer a build-a-bet product, there are different ways to compile and display the market. The less agile backend systems can only cope with hardcoded pre-defined selections. “Compiled pre-event and static in nature,” Jeyaratnam says, “these are essentially limiting choice, ruining customer UX and likely to be reliant on manual intervention to create and settle.”

Then there is the aforementioned method to create bets requested by customers. While the trader has access to algorithmic tools, it’s a slow, disjointed and labour-intensive process. And finally, there is the new breed of solution, allowing on-the-fly bet calculation and increased market variety, Jeyaratnam notes. “Some firms have realised their limits early and have effectively plugged – or are in the process of plugging in – a third-party solution,” he says.

“Integrated seamlessly, the punter is engaged with the third party’s cloud-based algorithm, which is fast and flexible enough to handle multiple requests and fire out prices in real time.”

BetVictor’s #PriceItUp bet builder product

Margin call

One reason why bookmakers love laying 10-fold accumulators is because of the juicy profit margin involved as it compounds with every leg that’s added to the slip. With build-a-bet products, it’s still a multiple bet but most punters won’t have a clue what the true odds should be or how much margin has been applied. It’s not as if they can compare the odds across the industry on the Oddschecker Grid or are likely to shop around.

Much like cash-out offers, the calculations are opaque, so your typical recreational customer won’t realise that the true odds for his build-a-bet creation should be 40/1 rather than the less-than-generous 20/1 displayed. One industry source suggests the margin on a six-selection wager with bet365’s Bet Builder is probably in the region of 10%.

BetVictor’s Ryan says the over-rounding applied to #PriceItUp Builder is “consistent with other products that a typical #PriceItUp customer enjoys”, while Betgenius describes BetBuilder as a “high margin product” in its promotional spiel.

However, the odds become almost irrelevant when you start adding three or more markets and the price shoots high enough to satisfy the customer’s craving for a big-priced winner, Jeyaratnam suggests. He adds: “The bettor isn’t in control; they are being guided along pre-defined, operator selected routes, most of which are likely to feature high-margin individual markets. With each additional selection, the operator’s margin increases and the gambler loses further sight of a fair price.”

Yet Israney takes issue with the notion that bookmakers are shoving in hefty margins. “It’s a bit of a misnomer,” he says. “The way these prices are calculated means they have to be done, by definition, to 0% [margin]. The operator has to calculate the probability of something to happen and then they can decide how much margin they want to put on.”

As competition heats up, it should drive down margins and lead to more competitive pricing. It is inevitable that these markets will be extended to a wide range of popular sports. Israney has written a pricing algorithm for horseracing and he says tennis is “mathematically pretty simple”. In fact, apart from sports like rugby, American football and cricket, he suggests most are suited to build-a-bet products.

“There are a lot of sports that once the back-end is sorted, the maths isn’t that hard.” A natural extension will be the ability to combine various outcomes across different sports. For instance, Manchester City to win, both teams to score and there to be over 4.5 goals, combined with Roger Federer to win in straight sets and serve at least 10 aces. “You can bet cross-sport multiples, so allowing cross-sport bet builders isn’t a big step at all,” Jeyaratnam states. Then there is the prospect of being able to place in-play build-a-bet wagers and then cash out winning positions. “In-play betting and cash-out is just around the corner but we won’t just load a pre-match product to in play,” Hassan says.  “With any product release we look to provide the best possible customer experience, considerate of customer trends, the opportunities of new technologies and the potential of partnerships.”

True innovation?

Gambling industry veteran Joe Saumarez Smith recently penned an op-ed for sister magazine EGR Intel in which he questioned whether there was a distinct dearth of product innovation in the run up to the World Cup. Yet a fully automated build-a-bet product could be a perfect addition for this summer’s tournament, especially when we reach the knock-out stages and there are only one or two games on match days and customers want the chance of a big-priced winner.

“I have no doubt that by the end of the year every major UK-facing firm will have delivered a bet builder of some description, whether developed in-house or taken from a third-party supplier,” Ryan says. For those companies behind the first wave of these products, the next step will be to iterate and refi ne the UX for mobile users.

On bet365’s BetBuilder, Jeyaratnam says: “Their technical range and capability is such that they developed their Bet Builder product in considerably shorter time than most. It isn’t the finished article and needs lots of work, but while it isn’t the best product yet, it is a good start. There is no doubt they will continue to improve its UX and scope.”

And with more and more granular statistics being supplied by the likes of OPTA, users will have even greater choice with compiling their bets, particularly around player-based markets. This will help considerably with user engagement during live sport. “For a small outlay, interest can go all the way from the first to the very last minute,” Yershon says.

As to whether these automated build-a-bet products can be classed as a game-changing development in the sports betting sphere, only time will tell as more operators and suppliers join the party. In the meantime, what these offerings certainly look set to achieve is to breathe new life into pre-match betting.