Legislatures in two states — Illinois and Tennessee — have now passed laws mandating that to-be licensed sportsbooks purchase only “official” league data if they wish to offer live in-play wagering.
Have you ever wondered, what the heck is official data, and how does it differ in origin and practice from “unofficial” data? Great. The goal of this article is to answer that — through a conversation with a pair of entrepreneurs who are developing an in-play wagering product for the sportsbooks in states that are giving operators the freedom of choice. (As opposed to the opportunity to have “official data” paywalled by the NBA and MLB, who are seeking a direct cut of wagers or additional fees as a precondition of receiving data.)
Politics aside, we have insights below from co-founders of Deck Prism. They are prolific poker writer (and player) Ed Miller (@EdMillerPoker) and poker player Matthew Davidow, who has a decade of experience in developing live wagering systems. Among the key takeaways? “Unofficial” data, while not as fast as a direct-from-league feed, may deliver a better bettor experience, may be more accurate, and depending on the provider, may be better suited for the flow of major U.S. sports and bettors for those contests.
Official league data versus unofficial sports betting data
Sports Handle (SH): WIll you both tell me what’s your background, work experiences, and what led to the formation of Deck Prism?
Ed Miller (EM): I’ve been in poker for a long time. I was a professional poker player and then I wrote a bunch of books. I transitioned to sports via daily fantasy sports, identified a pretty big opportunity to win, betting, basically, in DFS, and so I took it and started playing. And that’s actually how I reconnected with Matt because he was kind of doing something similar.
Matthew Davidow (MD): My background is originally in poker, and then I started a couple live modeling companies over the last 10 years or so.
SH: OK, let’s move to the data. Can you explain what so-called “official data” actually means? And explain the “game state” data component of sports betting data?
EM: From my perspective regarding the official data, as far as betting information goes, the main thing is “game state” data. You want to know what’s the score, how much time is on the clock, who has the ball. For football, what yard line is the ball on? What’s the down and distance? If it’s basketball, how many timeouts each team has, how many fouls? Now on top of that data could be other types of data. The amount of additional data beyond that is effectively limitless, but at the very basic to me is the game state information.
MD: One thing that I think is misconstrued is when people use the word “official.” To me that’s a legal word. Who owns the data is not something that we’re in any position to discuss. More importantly, any in-running model is going to use the data that Ed described. And what the leagues are offering that is valuable and anybody would love to have is the fastest possible game state data at the speed of what they’re selling. What we can do with Deck Prism, without having access to this, is going to be slower than what the leagues can provide.
SH: As far as the other component, the risk-management info that provides betting markets for a sportsbook to offer to bettors, can you talk about how that is created off of the game state information?
EM: Basically, our process is that we as a company take information about the game as our input. So part of the information that we take in about the game is this sort of nuts-and-bolts game state information that we describe in the data file. But there’s other information. When you watch the game, you can see things that are happening on the field that aren’t necessarily going to be reflected in a data feed, but it’s information that’s available. It’s happening, and you can see it. Such as a key offensive lineman getting injured.
So what our product is and what we do is, we take all the available information, both this basic game information and any other information, and we run analytics on that information to basically create a whole set of betting lines.
MD: The product we’re launching in August is going to cover two spreads, two totals, and the moneyline. That’s the basic product that’s going to launch in August. However, we have the ability through our models and through our processes to make literally hundreds of props at any time.
EM: Our focus as a company is on accuracy. We want to try to make sure that every single calculation we do is correct. And so early on, we’re making sure that what we put out is the highest quality possible, and once we’re on par with that, then we’ll expand what we offer, hoping to keep it at the same high quality.
SH: Question about the consumer experience. What I’ve gathered from you guys previously is that in-game betting on major U.S pro sports can be better, if not best enjoyed, when wagering during stoppages — whether that’s between quarters or during timeouts, for example. Why do you think that is?
MD: As a bettor in Nevada, I constantly find that when I press “go” on the app, there’s a delay that to me takes away from the experience. I don’t feel … I hate to use the word fair, but I don’t feel it’s fair that during the 10 or 15 seconds where there’s a delay at some places, that I cannot rescind my bet. Yet the operator is able to reject my bet, either because the line changed, the game state information changed, or maybe someone just wasn’t watching the screen. To me, the best customer experience is when: I see a bet. I make a bet. That’s my bet. And the large problem with what I call “betting while they’re running down the field” is that it’s very hard to do without a delay. Whereas with timeouts, everybody has the same information.
EM: The other big concern has to do with the broadcast. So customers are watching the game, probably on TV. There are different feeds. But all the feeds that are available today have substantial latency between when the event happens in the game, versus when you see it on your screen or on your TV. And often that delay could be 15 seconds or even more.
And so the problem is that if sportsbooks are making lines based on what’s happening in the game, up to the second, and they get this information from a data feed, then by the time the customer gets to bet on the event, the customer is betting on what happened 15-18 seconds ago, and the sportsbook already knows what happened, but the bettor hasn’t seen it on TV yet.
To me that’s also a bad customer experience. And that’s ameliorated by waiting for the stoppage or timeout. Because as soon as the consumer sees, “Okay, the game’s in a timeout,” now everybody has the same information.
SH: So in effect, if I’m watching the betting screen … if it’s a sportsbook using “official league data” and that’s the fastest, and they’re 15-18 seconds ahead of me, the moneyline I’m seeing for example may be telling me what’s happened in the game. A key stop on third down, muffed punt, whatever. Any drastic or I guess even slight change in price, that may be spoiling the game for me to some extent if I’m watching the betting screen, not the TV. That about right?
SH: OK, I’m not sure that everyone who may be reading this has experienced that “spinning wheel” for bet processing. Can you go over that process?
EM: So this is the process as it currently exists at many licensed sportsbooks in the U.S. In-play is basically presented in a separate part of the app from all the pre-game. So you would click on, “I want to play in-play.” So then, you reach the in-play menu.
Now you’re shown a bunch of bets. When you click on a bet it says how much do you want to bet. And you enter in, say, a hundred dollars. You click “Go.” And then what happens is a little timer starts ticking in the corner of your app. So the app, if it freezes up, it says “Waiting For Approval,” and the timer ticks. And the timer can take three, five, it might take up to eight seconds or more.
Now what this is often doing, the sportsbook is often just sitting on the bet. So the process is they say, “OK, Ed wants to bet a hundred dollars. Let’s wait for five seconds. And then we’ll see if there are any line changes that happen in that five seconds.” If the line is changed in that five seconds, by a substantial enough amount, typically the sportsbook either rejects the bet or presents a counter-offer to the bet.
That’s essentially the process. If the line doesn’t change in five seconds they typically accept the bet, and you move on.
MD: When it comes to operators, I believe they’ll find that when they remove the delays from the in-running, their volumes will increase by a lot. I think this will narrow the gap between the U.S. and Europe, as far as customers playing in-game as opposed to pregame. I mean the volumes will increase by a lot. And I think by using Deck Prism it will allow sportsbook operators to remove delays, which will increase volume and increase revenue.
SH: Looking to put some numbers and a timeline out there … We all have different providers. What is roughly the average cable TV delay, let’s say for an NFL broadcast?
MD: Interesting question. My experience from our NFL demo last year that used the NFL official feed is that we received the information that the play started approximately 8 to 10 seconds before the play started on my Cox Cable broadcast from ESPN, about 13 seconds before my Cox Cable broadcast for the network. I used FOX’s broadcast and NBC. And my DirecTV feed was then a further 4 to 5 seconds behind my Cox feed. Especially on the networks. So basically, depending on whether it’s cable or DirecTV, it was anywhere from 8 to about 20 seconds after the action started, that I was able to watch the play on my television in Nevada.
For Deck Prism, where we don’t have the feed directly from the league, we take in data in every legal way possible. Put that data together and we are going to make sure that the data that we are using for our products is accurate. Our number-one thing is accuracy. Accuracy to us is much more important than speed.
SH: OK, parting shot. Something we didn’t cover you’d like to add?
MD: It’s my belief that the product that we will offer in the fall will provide a much better and fairer customer experience, and that will promote a lot more in-play wagering on U.S. sports.
EM: For all the stakeholders here, I think their interests all align. I think everyone wants a clearer, faster, better, fairer experience for the customer because when that exists, you get happier customers, they want to bet more, they will watch those sports, they enjoy it more. Everything is driven by the customer experience, and the better you can make that customer experience across the board, the better everyone’s going to do it, and to me that’s the focus of why we started Deck Prism, that’s why we’re focusing relentlessly on the customer, because we think in the end what’s good for the customer is what’s good for the league. What’s good for the customer is what’s good for the betting operator, and so on. So that’s really where we’re at.