How Computer Betting Took Hold in Horse Racing
Admitting it may come slowly for old-school players, but the reality is computer betting will be a part of horse racing for as long as the sport shall live. Whether that relationship proves toxically crippling or even lethal to the game is a whole nuther matter, one that has fueled discussion, debate and frustration in lockstep with its exploding influence.
Computer-assisted wagering even has spawned a polarizing acronym. Pronounced letter by letter or barked as one syllable, CAW is on the tip of a lot of potentially influential tongues.
“In 2003, CAWs were roughly 8 percent of total handle but, last year, probably about one-third of all betting on U.S. Thoroughbred racing,” research expert Pat Cummings, now of the National Thoroughbred Alliance, told the Jockey Club’s annual roundtable last summer. “When you adjust these figures for inflation, the concern becomes quite obvious. Handle from all non-CAW customers is down by nearly two-thirds in just two decades.”
Key Players in the CAW Debate
The California Horse Racing Board dived in last month when it hosted a 65-minute public discussion on CAW. It featured Scott Daruty. He is the president of Elite Turf Club, a joint venture of The Stronach Group’s 1/ST Racing and the New York Racing Association. Like Velocity, which is the comparable operation owned by Churchill Downs Inc., it caters to a private clientele of high-end computer players from around the world whose big-money wagers are rewarded not only on their winning merit but with significant rebates from the racetracks.
Marshall Gramm, an economics professor at Rhodes College in Memphis, Tenn., is a successful horseplayer who won the 2020 Breeders’ Cup Betting Challenge. He was part of Cummings’s panel last year at Saratoga Springs, N.Y. In addition to being a breeder, Thoroughbred owner and even an investor in Horse Racing Nation, he is not ashamed to admit he has become a computer player.
Having grown into a bettor who pushes millions of dollars through the tote system, Gramm has not forgotten the roots of his playing days or the frustrations that persist for those who anachronistically are known as the $2 player.
Gramm knows firsthand the advantage of batch bettors who make thousands of wagers in microseconds. He also knows that edge is perceived more and more as a big foot that is squashing the potential profit not to mention all the joy out of the game for the little guy.
Against that backdrop, Daruty and Gramm appeared together last week on Horse Racing Nation’s Ron Flatter Racing Pod to compare and contrast notes. Highlights of their conversation follow, including that very basic premise that Cummings offered last summer.
How Much of the Handle Comes From CAW?
In terms of computer-assisted wagers, what percentage of the actual betting population do they make up? And what percentage of the handle do they make up?
Marshall Gramm: For me, that would be just a guess. Scott would actually probably have a more precise number. It also really depends on how you define computer-assisted wagering. There’s the sort of strict teams that have accounts with Elite or Velocity that are basically doing computer-robotic wagering with complex models, but then you have other individuals who are approaching it maybe more small time but are using computer and automated uploading in their wagering. My guess is the big chains are probably 15 to 20 percent of the handle, and there may be another 10 or 15 percent of people who are using sort of heavily involved computers in their everyday wagering. That’s just a guess.
Understood. But Scott, this doesn’t sound too much different from what you presented to the CHRB in terms of the numbers that you have derived in your work with Elite.
Scott Daruty: You’re right, it’s very consistent. One of the things I think Marshall points out, and it’s an important point, is when we’re talking about these really complex issues, definitions matter. I’ve had people approach me and say, “Well, CAW used to be X percent of the industry, and now it’s Y percent, and that’s this huge growth.” Some of that has come from growth, but some of it has also come from definitional differences. If you go back in our industry 10 years, there were people who were playing who today we identify as CAW players. Ten years ago when they were playing, they didn’t fall in that category. If you look at that, and you say that’s growth in CAW, well, no. That was a definitional change. That wasn’t a substantive change. So it really does matter how you define it.
Do you want to go ahead and pursue some of the numbers you gave to the CHRB so that we can frame this a little more with some of the quantifying of this, Scott?
Daruty: We have a very broad industry. We have, I don’t know, a couple hundred racetracks we’ll run over the course of the year at different points in time, and so there isn’t a one-size-fits-all answer. There’s a couple of ways to approach it. One is to look at a blend across all tracks in North America. I think a more helpful way to look at it is a specific track-by-track basis. With 1/ST Racing and our premier racetracks Gulfstream Park and Santa Anita, we would like to see, as a reasonable target, CAW play constituting about 20 percent of the pools. In certain pools on certain days, it can be higher. Certain pools, certain days, it can be lower. I’ve heard one of the CAW commentators out in the industry reference a day at Gulfstream where there was a pool, and, “It was 40 percent by my calculations.” Well, yeah, maybe that’s the case. I could point out one where it’s 5 percent. Really, what matters is over the course of year at a given track, what is the management of that track trying to accomplish? And are they properly utilizing the tools that they have at hand to keep the CAW in the range they think is appropriate. So that’s for the tracks that we own and control. Another track might decide that’s too high, and they want the percentage lower. Another track might decide they’d like more liquidity, and maybe 30 percent is OK for them. So I don’t profess to speak for everybody in terms of what they’re trying to achieve, but as for Santa Anita and Gulfstream, that’s kind of our target.
Rebates, Takeout and the Edge for CAW Teams
Targets also, in the case of players, they’re looking to cash at a certain percentage rate. Marshall, what about for you in terms of what you’re looking at either as a player yourself or in terms of the studies you have made of computer-assisted wagers? What’s the target that they’re looking at, and how much is dependent on what the tracks are giving for rebates?
Gramm: In some ways, don’t we want everyone to have access to computer technology? Then what ultimately do they have? They have the ability to batch bet, right? And they have large rebates. I think it would be great if more players had access to both of those things, whether it be lower overall takeout. Either the takeout needs to come down overall, or we’ve got to lower the margin between the casual player, the professional player and the teams. I’d like to see that difference shrink. And then I’d like to see CAW-type tools available to regular horseplayers. I know that 1/ST group, Stronach through Xpressbet, have more tools available. There are other ADW (advanced-deposit wagering platforms) that are creating tools as well. But the ability to create more efficient tickets with quick upload processing, I think those are things that could help the run-of-the-mill weekend warrior compete on a level footing. How big should Elite and the teams be relative to the overall pools? I don’t have a good answer for that, right? We don’t really know. But I believe what would benefit everybody is if we could figure out how to encourage the pie to grow, encourage more handle through lower takeout. I think that would benefit everybody. It would benefit the teams, because there’d be more liquidity for them to bet in. And it would benefit the weekend warrior, the casual player, to allow them to churn more.
Scott, there’s a counterintuitive factor here, though, because it was pointed out at the CHRB meeting that higher takeout actually can create more CAW traffic, because that then turns into bigger rebates. Do I have that right?
Daruty: Well, I don’t think that’s exactly correct. I think that it ends up being a wash. If there’s a 20 percent takeout, and let’s say I’m getting an 11 percent rebate. If the takeout goes up to 22, and I’m then getting a 13 percent rebate, mathematically, in the way the models work, those things are essentially similar. The difference between the takeout and the rebate is what the player has to overcome through handicapping and through the selection of the correct horses. If the player can overcome that, then the rebate's going to make up for the increase in takeout and still get the player overall 100 percent, recognizing that that’s the objective of not just CAW players, I guess, but every player, right? You’re sort of betting, because you want to end up with more than you started with. For the CAW players, they’re trying to end up with more than they started with, but they all recognize they’re not going to win outright against the takeout. They may, in my example that I gave to the CHRB, and this is a number that’s backed up I think by pretty good math at least at the tracks that that 1/ST owns and represents, a typical payback for a CAW player on all their wagering activity over a period of time would be 90 percent. Meaning if they’re getting an 11 or 12 or 13 percent rebate or more than that, they’re coming out ahead. If they’re getting a 10 percent or less rebate, they’re not going to move money, and eventually they are going to go to something else.
Sharps vs. Squares in the Pari-Mutuel Era
There’s a delicate balance here between the CAW players and the so-called retail players, the everyday players. Is this like sharps vs. squares that we’ve seen in Las Vegas and other gambling centers forever?
Gramm: Yes. That’s what I would say. This is just another evolution in horse playing. Whether it be the Sartin methodology or speed figures or trip handicapping. This is just the modern innovation in playing the horses using computer technology and modeling and handicapping. The only thing I think that makes it different and unique from just an evolution in horse playing is the fact that they can batch bet, and it can affect the pari-mutuel system and the payout. The biggest problem that I have with CAW, that a lot of people have with CAWs, is not so much that they win. There’s going to be winners and losers. Everyone, if they play by the same rules, that’s just the way the game is going to turn out. We can’t be bitter at them because they’re winning. The problem is what they do to the final odds and their ability to batch bet and their ability to change the odds. I know that the industry, NYRA, the Stronach Group, has done things to prevent them from dramatically shifting the odds, but we see it all the time. The pari-mutuel system was never really built to where someone could make so many bets in the last minute. We think about when Pierre Oller developed the system, he was a (19th century) perfumer in Paris, to eliminate the risk that bookmakers faced when taking bets, no one could imagine that someone could come in the last minute, put a lot of money and basically establish the market prices. So that’s where we are today. I think some of the stopgaps that have been used to help the win pool, especially what NYRA has done (stopping batch wagers two minutes before post time) and to some extent what 1/ST and Stronach has done have been helpful. I really wish ultimately we could move away from the pari-mutuel win pool to an exchange-based system for win betting. It would benefit everybody. The teams would love it, because they would get to creating markets. Horseplayers that get their version of fixed odds priced correctly could work out. If not that, really push the teams into trifecta, superfectas and the pools where they’re definitely winning, and they’re winning a lot. Those are pools that they’ve really dominated. As a recreational player, I’d be very wary of betting Pick 4s or the superfecta. But the teams are exceptionally good there. Give them an incentive to play there, and then push them further away from the win pool and the exacta pool, pools where the money movements really affect people’s play in the game. It’s not only the odds drops that are problem. If I bet a 4-5 horse and it goes to 1-5, I’m frustrated. But just as well, if I am looking at a horse and thinking I don’t want to bet the horse 4-5, and it goes to 7-5, that’s money I wouldn’t bet otherwise. I think that it’s the opposite movements in both directions are problematic.
Managing Late Money and Odds Movement
Pat Cummings believes what Marshall just said. But Scott, I know you believe that the two-minute cutoff that NYRA has imposed on win bets, that you can’t make a batch bet after the two-minute mark, hasn’t had the influence that maybe the everyday player would hope it has.
Daruty: I don’t think it has. I applaud their efforts to solve the problem, they being NYRA. We’ve made similar efforts at the 1/ST tracks. There’s no perfect solution that I’ve figured out yet or been told about yet. We acknowledged the problem of late shifts is a problem for everybody. By the way, it’s a problem for the CAW players as well, because they spend a lot of their time and energy trying to guess and predict where the final odds will end up. They can’t see it in the pools any better than the guy sitting at the racetrack can see it in the pools two minutes out, because there’s not enough money there. So it is a problem fundamental to the pari-mutuel nature of our game. We would love a solution. We’ve not been able to figure one out. If anybody has any great ideas, tell me, and we’ll implement it. Different ideas have been kicked around. What if you close the pool when the first horse loads? That way the money will be in by the time the race starts. OK, but have you really solved the problem? If you have to bet when the first horse is in, then you’re just backing up the problem. You’re still going to have all the flow of money in the last minute or two. It’s just that minute or two will expire when the first horse loads instead of when the gate opens. Now it may solve the perception problem of the odds boards moving during the race, but it doesn’t really help the player understand the odds better when he or she is betting. That’s just one example of one solution that’s been proposed. You can go through and propose all kinds of different solutions, and we’ve taken external ideas. We’ve had internal discussions. Every idea here has its own set of problems with it. For example, if you were to say CAW has to stop at two minutes to post or two minutes to off time or however you want to measure it, it has to stop before the pool closes. The win bet is the bet in which this is this problem is most visible. It also happens to be the bet in which it is least necessary to have quick betting capabilities. Even the most sophisticated computer player may only be betting, I don’t know, two or three bets on a winning pool. They may be big bets, but you don’t need to put in 1,000 different combinations like you do on a Pick 6. It’s a win pool. You just need to pick a horse or a couple of horses and put your bet down. By closing the pool at two minutes to post, have we prevented that player? We’ve prevented him from betting through the CAW channel where every CAW player is identified by their own unique tote code, so we can track their play and study their play, and if there ever were a problem, have complete transparency. We’ve taken them out of that environment, and we’ve sent them to the racetrack either personally or with a bet runner to go up to a window and place it at the last minute or to do so through a TVG or Twin Spires or Xpressbet retail account. Again, I applaud NYRA’s efforts. I just don’t know that that solution has been all that successful.
Let me run another idea by you at the risk of throwing clay pigeons in the air here. What if you change the percentage of the rebate the closer a bet is made to the start time of the race? You reduce it if you wait until the last second.
Daruty: I think that’s a great idea. I think it’s a much better, more sophisticated plan, a version of the plan that the 1/ST tracks adopted. What the 1/ST tracks adopted was rather than telling the players you have to bet before two minutes to post, and once that time period occurs you can no longer bet, we didn’t take that. We took an approach that said if you bet prior to two minutes, your price that you have to pay the racetrack goes down, which means your rebate goes up, and we try to incentivize through that higher rebate people betting early. Conversely, if you don’t want to bet early, and you’re a CAW player, and you want to wait until the end and bet, your price that you have to pay the racetrack goes up materially, and thereby your rebate goes down. You’re effectively paying for right to wait and bet later. Why I say yours is a more complex and probably a better variation than what we did, we just had sort of one break point. Before it you get a higher rebate. After it you get a lower rebate. I hear what you’re saying almost to be like a sliding scale three minutes out, two minutes out, one minute out. That becomes very complicated. You’ve got to recognize we have people coming through different totes into the pool at different times. If somebody were betting in through a United Tote platform, for example, it would be complicated, because you wouldn’t actually see exactly when they’re placing your bet. You’d see when United Tote is passing that bet from United Tote, so it gets very complicated to try to set exact times. Yours is better in theory. I just think it would be very hard to implement. However, what we did find out, and I mentioned this at the CHRB, I said we adopted this policy, and I said it hasn’t been wildly successful. What we found is originally some of the players were betting early and taking advantage of the bigger rebate. But as time went on, virtually all the players abandoned the early wagering and started waiting later. Why is that? We think it’s two factors. The primary one is that these players are also trying to bet against the competition. They need to know what the pool is, and the earlier they bet, the less definition they have, the less knowledge they have about where the pools actually going to close. By having them bet early, they’re betting blind just like the public. I think ultimately, one of the challenges we have is as more and more players move away from brick-and-mortar environment, whether that’s at a retail OTB (off-track betting) facility or whether it’s at a racetrack, more and more players are moving to online wagering. By the way, even players who are sitting at a racetrack are now on their phone and are betting through an account-wagering platform. The more the industry goes that direction, the more money comes in late. Whether that’s retail money or whether it’s CAW money, more and more is coming in late. I think eventually we’re going to get to the point where your pools might be zero at two minutes to post, and all the money comes in at the end. Will some of that problem be a CAW problem? Yes. But a lot of it will be a reflection of late retail money as well.
Technology, Pricing and the Future of CAW
Which brings me back to you, Marshall. Technology. The establishment of horse racing is still catching up in a lot of ways. Computer players, I think there’s a certain edge that they always will have. How much is technology a challenge on both ends of this?
Again, computer analytics, however we want to call it. Modeling. It’s everywhere. It’s definitely dominating sports betting, too. It’s definitely involved in in everyday business. ... It’s definitely in the future. I want to see a world where more of us use CAW technology. Where the pricing is more equitably distributed. Scott even mentioned at the CHRB meeting that the increase in effective takeout, as a result of the CAW teams, is just a couple of percentage points. How do we easily solve that? We drop the takeout by a couple of percentage points. That solves part of the problem for our recreational players. It can be done. The takeout in California on multi-horse wagers is 23.68 percent. That can be done fairly easily. That’s onerous relative to sports betting and all the competition that exists for our wagering dollar now. I know that our sport is tied up in regulation, and it’s hard to change these numbers, but I just would like it to price more effectively. I think that would help us solve part of the problem. Hopefully, the price discovery part of the problem, which is big, which is the late money movements making it difficult for people knowing what prices they’re going to get, that’s a bigger challenge. I’d really like to see us move to exchange wagering. It’s something that might solve this. That’s one that, again, if we can get the price down that we pay, if we can get the takeout down, get the difference between the teams and the recreational players down, I think that’s at least a step in the right direction that can be done.
Scott, you could address what Marshall just said on takeout. But let me double down on this and also ask about the rebates. How often do you review what you’re giving back to the batch bettor?
Daruty: That’s reviewed almost on a continual basis. It’s reviewed and adjusted. The CAW players are getting this rebate, because it is so directly related to how much liquidity they bring. It’s not uncommon for retail players to get some rebate as well, but it’s typically much smaller. If you were going to go in, and you were going to say let’s close the gap, as Marshall said. You want to close the gap in a way that’s fair to the consumer and creates more turnover, but you want to do that in a way that ultimately brings more revenue to the industry, not less revenue. That becomes a complicated mathematical analysis in its own right. If you were to go to a regular consumer, and you were to give him a 2 percent rebate or a 4 percent rebate or a 6 percent, where are the break points? How much more do they pass at each of those levels? I don’t know for the regular retail consumer that it’ll be as directly correlated as it is for the professional player. We can almost predict, I don’t want to say perfectly but to a very, very high degree of accuracy, what a professional player’s handle will do based on if we give him a 2 percent more rebate or a 2 percent less rebate. We can very accurately predict what that’s going to do to his turnover. Which then means it’s very accurate to say, OK, at that new turnover number, how much revenue comes from that vs. a retail player. You might give one guy a 4 percent rebate, and his handle goes up 50 percent. You might give another guy a 4 percent rebate, and he bets the exact same amount. They don’t have the same motivation as the professional players do. The idea is a good one. We’re certainly open to that. We’re 100 percent supportive of the notion of giving more technology and better tools to the retail player both in terms of helping them analyze the field and select their horses as well as formulating their wagers and getting those wagers in promptly. That’s something that we do focus on, are 100 percent supportive of, and if anybody has any ideas in that regard, we’re totally open to how can we provide better tools to the retail player?
Gramm: All those things sound great. I’m thinking more in terms of the takeout relative to other betting opportunities. If we think about pricing, the landscape has changed. Horse racing was a monopoly 30 years ago and in some ways was priced better 30 years ago to what it is now for recreational player. We have such a tiered system. I actually get substantial rebates, because I’m lucky enough to live in the great state of Tennessee. I don’t have a track, so I don’t pay any source-market fees. If you’re probably the same level of bettor as me but in California or Minnesota or Virginia, with source-market fees your rebates are much lower. In fact in certain states, you can’t even bet in the pools, because it’s doesn’t make sense for the ADW to take you, because they lose money on your bets. I just wish there were more equitability across pools for players where they don’t have to be like me and be in the correct state. The rebates do level the playing field. I don’t think that you can be a winning player without rebates, and we need more people with at least the illusion of winning or at least the confidence to think they can win, to invest time in the game, that more learn more. I went from being a mid-five-figure player, got access to rebates, and I’m an eight-figure player. That’s all because of the ability to start winning or at least being able to compete competitively at a lower price. I think there are a lot of people like me out there that don’t get the chance.
Conflicts of Interest and Perception
Scott, I don’t want to let you get away without addressing the questions that will arise about a conflict of interest with racetracks owning betting platforms and even now, as was pointed out by the California board, 1/ST has gotten into bed with MyRacehorse, so it’s even owning racehorses. I know you’ve said that this has been the way racing has operated for a long time. Do you want to go ahead and advance that?
Daruty: I’m going to leave the MyRacehorse part out of it, because that’s not my area of expertise. Obviously, horse ownership is a key part of what drives our industry, but I’m focused on the betting side in my role. You go back to the start of horse racing, when the only place you could legally place a bet was at the racetrack where the race was being run. It was the track who conducted the wagering. Then you move into the OTB expansion and simulcasting, and again, it was always the tracks who controlled the wagering. When ADW launched, you got Twin Spires owned by Churchill, and you got NYRA that’s with its own platform, and you got Xpressbet owned by the 1/ST company. Even in the retail ADW space, it’s the racetracks that operate the wagering. I feel like the ownership of Elite Turf Club being divided up between NYRA and 1/ST, and then our competitor Velocity being owned by Churchill Downs, I think that is a natural extension of the fact that the betting in horse racing has always been owned by the racetracks, always operated by the racetracks. I don’t see it as a conflict. I believe it’s beneficial for the industry, meaning the racetrack owners and the horse owners. I think it’s beneficial to that group that the CAW platforms are owned by the racetracks, because it gives us a visibility and a transparency that might not otherwise exist if it were a third-party platform. On balance I’d see it as a positive thing. I’m not blind, though, and I understand why there would be an appearance of a conflict. I understand that people raise that point. I just think that, on balance, it’s better that we own the platforms.
Gramm: I’m not worried about it, but I do think that it leads to conspiracy theory. There are people who sort of raised alarm bells over CAW believing that the industry itself is profiting from them, or they own the teams, which is not correct. ... I don’t believe the conspiracy theory that given the teams have access to higher-up officials at Stronach and NYRA that they can help some of the decision making, that maybe that’s why we end up with a blind maiden race in the middle of the force-out in a Pick 6 sequence. I don’t think that’s true. I hope it’s not true. But you can see where the conspiracy theories would come from.
Perception then is important, Scott. Perception can be reality for a lot of the public. Is the pressure then inevitable to make changes to, for instance, the rebate structure for CAW, especially with the fact that the California Horse Racing Board did call you in to answer questions about that?
Daruty: Perception is critical in in any business but especially in the gambling business. Perception is critical. People need to have confidence that the game is fair. Knowing what I know, and seeing it day in and day out, I have every belief and every confidence that the game is fair. Marshall mentions there are a lot of conspiracy theories out there. None of them really have any validity to them. Still, it’s something we have to address and try to instill confidence. I don’t know exactly how to do that other than to have conversations like these. It’s why I appreciate the opportunity to talk with you and Marshall, who I respect tremendously. It’s why I show up at CHRB meetings and try to spread the truth. We’ve got a lot of challenges in this game. It is a tough competitive market now that sports betting is legal in 35 states or whatever the number is today (actually 38 plus D.C.). We’re in a competitive environment, and we have to do things that instill confidence in our customers. I just don’t know that there are any silver bullets, at least not that I’ve seen. We’ll continue to work day in and day out to try to improve things, and as I said, always open to suggestions and ideas, because we certainly think a lot about these and haven’t come up with solutions for some of these problems yet on our own.
What the Next 5–10 Years Could Look Like
Marshall, I’ll let you have the last word here. I think we would all agree that racing in the 21st century under at least the economic model that is extant cannot survive without computer-assisted wagering. What is the next step? How do you see maybe the next five to 10 years and the advancement of CAW vs. retail betting?
Gramm: I don’t believe rebates are the problem. I think that rebates are very important. I think it’s important that we reward our high-volume players. Whether they be computer teams, whether they be individuals. I think it’s important. It’s one way that we can bring the price down. I would hope we could bring the price down for everybody. I believe that there is the issue of will there be a short-run gain. I don’t know. But I’m confident in the long run that there’ll be benefits to it. Ultimately, the pricing of rebates isn’t the problem. The problem is the price discovery and what it does to late odds movements. What does the next 5-10 years bring us? I hope it brings us lower pricing. I hope it brings us a rise in both computer play and casual play. I hope that some of these things start to dissipate. If we can, I believe, price our product competitively and bring back some of the casual players and weekend warriors and compete with sports betting, I think we’ve got one of the best gambling products out there. The gambling price is built for people with short attention spans. A race is over in 72 seconds. You bet on football, it’s like getting tortured. Or basketball, where we can have a lead, and then your team starts fouling, and you lose your bet. I think that what we have is a great gambling game for the 21st century. I just hope we can figure out how to make it work. I am pleased that we have people like Scott working on it. People like Pat Cummings out there who think about these things. Maybe there’ll be some creative solutions where the game can really thrive.
