Former computer-assisted bettor Dana Parham on CAW wagering in today’s racing landscape

The man behind Odds On Racing details his former betting life and opines about the role of CAWs in today’s handle.

by Brett Sturman

Dana Parham is one of the larger names in harness racing through his Odds On Racing, through which he owns interests in close to 350 foals, broodmares and racehorses. But going back three decades, it was Parham and his associates also who were at the forefront of technology that would be known as computer-assisted-wagering (CAW). Leading a cutting-edge group that wagered tens of millions, if not more, on harness races annually, Parham is an expert in all aspects of wagering, including CAW wagering.

Parham stated clearly that he has not wagered professionally on horses since he sold his interest in the CAW business in June of 2016. While Parham can’t say precisely how mechanisms may be different in present time compared to nine years ago, much of what applied then remains completely applicable today.

Articulating CAW wagering at its core, Parham describes it simply as performing tasks, “Anywhere from placing your bets, to determining what those bets will be.” And whereas a human can reasonably handicap so many pools across so many tracks, “What a computer can do is bet into every pool, every race and at every track.”

Aside from the sheer volume of wagers that CAWs can support, the overriding edge is simply a matter of mathematics.

“The other part of it is being able to estimate what the closing odds are going to be which a computer through a direct program can really help you there,” said Parham. “You can get hints from looking at different pools or looking at money that’s recently come in, but the program is determining that through an algorithm that has been written and refined with countless tens of thousands of race data.”

For as good as the CAWs are at projecting odds, Parham doesn’t believe the systems are necessarily better than humans when it comes to handicapping the horses.

“I think a human that covers a track, or a few tracks closely can have a better opinion than what one could program,” said Parham. “The huge advantage for the program, however, is that everything is mathematically correct. I find a lot of handicappers might have a good opinion, but they’re mathematically inept.”

It’s that mathematical part with regards to closing odds that Parham takes issue with any negative CAW perceptions.

“How does one reasonably make a bet if they have no idea what the payoff is going to be?” asked Parham. “I don’t know if the average bettor really understands even his win prices because I continually hear that I bet this horse at even money and it went off at 1-5. In pari-mutuel wagering just because you made a wager when the odds showed something on a screen doesn’t make them entitled to that price.”

My contention to Parham was even with the above said, the public’s negative perception is that CAWs still produce such wild late swings in the wagering where it makes it that much harder – impossible even – for an average person to project out what those final odds will be not knowing the late CAW influence. To which Parham replied, “They’re absolutely correct.”

Parham then expanded on that topic of fairness.

“I was in this debate with Churchill maybe 25 years ago,” said Parham. “They’re trying to tell me that their job is to level the playing field. We’ve gone from wanting equal opportunity to now wanting equal outcome. But you can’t give a part-time player who isn’t math proficient an equal outcome with some smart people who are working 20 hours a day for seven days a week. When you bet the exacta, do you distinguish between Tuesday and Saturday? Most average bettors do not. So now, you’re either over-betting the pools on a slow day or under betting them on a busier day. So already, we’re starting with the premise that the bets are not equally aligned. It’s already a misallocation of capital on one day or both days. These are areas where the average bettor has no chance if he doesn’t understand the arithmetic behind the bets.”

One point where the CAWs became even more prominent in the spotlight was earlier this spring and its impact on handle at The Meadowlands. In April, The Big M announced that in addition to its reduction of post-time drag, according to a press release, the track would also be working to eliminate CAW activity in win pools less than two minutes to post. It’s been debated as to which components drove which parts of the results, but in any case, The Meadowlands’ handle immediately fell dramatically. Common handles of over $3 million fell to levels not seen in many years, including Saturday handles dropping to under $2 million and Friday handles in the mid $1 million range.

When asked to provide his best estimate on what percentage of The Meadowlands’ handle comes from CAW wagering, Parham said, “I can guess, and I’m going to guess 30 per cent.

“What The Meadowlands did from what I’m told – and the key word here is what I’m told – is that they made the CAWs bet two minutes ahead of post time. And many of them said ‘No, we’re not going to do that so we’re not going to bet at all.’ And so, all that handle was essentially cut off.”

Parham’s observations are supported by a note in a May 2 HRU column from Debbie Little in which it was written, “According to sources, Elite Turf Club has not been wagering as much at The Big M lately.”

Parham takes issue with the reasons cited by The Meadowlands for reinstating post-time drag following the drop in handle.

“The Meadowlands said it ran this post drag experiment and it didn’t work,” Parham said. “Well, they didn’t run the experiment. They cut off a large amount of handle and that part was left out of the story. It makes no sense to me that handle would fall so much because we said that here is post time and we stuck to it. There used to be no such thing as post-time drag, and handle is what it was. And now because everyone claims that it helps handle – well if that’s the case why don’t we drag each race for five hours? I can understand the reasoning for the small tracks that do it – but The Meadowlands is supposed to be The Meadowlands.”

I did note to Parham the inverse relationship observed where right as The Meadowlands’ handle was falling during the experiment, that handle at Woodbine Mohawk Park was suddenly increasing by almost the same degree. I asked if it was possible that change in handle was due to a shift in CAW wagering from The Meadowlands to Mohawk.

“I find that hard to believe,” Parham said. “No matter how good your algorithm is, you can only be at a certain percentage of the pool. It’s not like you can just take the handle from The Meadowlands and put it anywhere else or you kill your own price. Then, you’re taking a bigger risk to get a smaller return. These guys – every computer wagerer provided they have the bankroll and I’m sure they do – they’re betting the very maximum they can before they start to hurt their return. The math is the math, so that change in handle did kind of baffle me – perhaps part of it had to do with regular people.”

With the high percentage of handle that comes from CAW wagering not only at The Meadowlands but, in fairness, any track today that has enough handle to support CAW wagering, does it pose a risk to the industry to have so much handle coming from those sources? What would happen to industry handle is CAWs ceased?

“In a way, isn’t that true just with any individual like me or you?” asked Parham. “Look what’s happened over the years with normal people. We could actually go the opposite with these CAW guys and look at the percentage of their winnings that goes back into the pools. These guys are making $50 million, $100 million in a year, some of these groups maybe even more – what are they going to do with those numbers, not bet? I understand there is a group out there today betting $1 billion a year. Even if they lose at a 3.5 per cent rate but get a 13.5 per cent rebate, they are making $100 million dollars.”

What about changes in CAW technology over the last decade since Parham’s last direct involvement, understanding the rapid advancements since that time in the development of artificial intelligence? Parham still sees it coming down to the same basic principles.

“When we were doing it, and I think maybe since the beginning of racing, there are only so many horses that can logically win a race,” Parham said. “And so, whatever methodology one uses, it comes down to landing on those logical horses and having logical math. The methodology needs to be correct.”