Open Letter from the CEO of a DFS Operator

The recent barrage of TV ads by the two biggest DFS sites put the billion-dollar unregulated industry on the national radar and squarely in the crosshairs of regulators and politicians. Certainly the lucrative multi-million dollar contests and insider betting claims have not helped the industry’s image. The recent debate about the legality of DFS has raised important questions, but unfortunately much of what is reported and believed about DFS is misinformed and/or downright false. As the CEO of a DFS platform I’m not neutral, but my arguments will nevertheless be, I hope, if not impartial, objective. But I leave that to you, the readers, to decide.

According to Attorney General of New York Eric Schneiderman, DFS users “are clearly placing bets on events outside their control of influence, specifically on the real game performance of professional athletes.” But, if the outcome of an activity is random chance, then that should mean that no one can be a professional (there aren’t any professional craps shooters or lottery players). Given this, one should not argue that chance is a “material” element in a game by pointing out that only a few people win the vast majority of winnings. Yet, Mr. Schneiderman did exactly that when he pointed to the fact that less than 11% of players repeatedly win almost all the prizes. That doesn’t happen when users cannot influence the outcome. It looks like Mr. Schneiderman believes that just because a game is skill based it necessarily means that any given player has a better chance of winning. But the opposite is true; if a person does not have the requisite skill, he might be better off playing a game of pure chance.

Before NYAG Schneiderman’s crusade against DFS, the industry faced indirect attacks based on a comparison to poker. The basic premise of the attacks was that if poker is considered gambling, DFS should be too, because poker requires more skill. Most recently it was the two-time World Poker Tour champion Andy Frankenberger who argued that poker requires more skill than DFS. For the sake of argument I will concede the point that poker requires more skill. The concession, however, is meaningless. Their argument presupposes that a line can be drawn between pure skill and pure chance, upon which games could be placed and ordered. This would mean that skill and chance in a game together add up to 100%, but that is a false dichotomy.

The above table demonstrates that chance and skill in games are, in fact, independent of one another. Therefore, Frankenberger’s assertion that poker requires more skill cannot be used to extrapolate the element of luck.

If the element of chance cannot be extrapolated from how much skill is involved in a game, how do we determine it? The answer is in the definition of a game of chance: a game whose outcome is strongly influenced by some randomizing device. To determine the element of chance we need to consider whether it is overt or not. Poker is overt chance, the randomness is caused by an explicit randomizing device inherent to the game, shuffled cards. DFS on the other hand is non-overt chance, luck is an incidental consequence of the structure. This is not to say that non-overt chance can’t contribute significantly to the role that chance plays in a game. For example, the starting QB is injured in the first quarter. However, this type of randomness is more akin to what happens in trading than gambling (for example, the death of a visionary CEO can ruin a well-informed high probability trade).

Yes, luck is present in DFS. But I contend that game of skill doesn’t require the absence of luck; it can’t. Luck is just probability taken personally. A game of skill is one where the player has control over the probability of winning and losing. I therefore submit that a game should qualify as a game of skill if luck is not introduced via an overt randomizing device (e.g., shuffled cards, dice, random number generators which decide the outcome of trial in a game). Disagreement on this point means no game can be objectively deemed a game of skill. Even a game of perfect information like chess involves luck. In chess, one strategy can perform better or worse than another depending on how your opponent chooses to react to it. This introduces luck because (deep breath) your strategy’s probability of success is directly dependent on your opponent’s strategic reaction to it. Non-overt chance is present and “material” in chess, as well as in many other activities we do not consider gambling (e.g., trading).

Lastly, I want to touch on the insider trading scandal and cheating in general. A DraftKings employee published early data about the prevalence of particular players across all submitted lineups (drafted by percentages) for the site’s Million Maker NFL contest (this data is only posted after games commence and the lineups permanently lock). The same day he entered and won $350,000 for his 2nd place finish in a million-dollar NFL contest on a rival site (FanDuel). Allegedly, he used the early data he published to play on Fanduel. If he did this, he was wrong to do so. However, it is important to keep perspective when assessing the advantage of doing so. Generally, overvalued players will have lower drafted by percentages, while undervalued players will have higher drafted by percentages. Each site has a different roster format, scoring system, and salary structure making who is overvalued and undervalued also different. By extension, the drafted by percentages will also be different across sites. Therefore, the usefulness of DraftKings’ drafted by percentages for playing on FanDuel is questionable.

Let’s now assume a worst case scenario. The insider information is used to play on the site it was obtained from. Assume you know that very few people picked an athlete in a contest. Is it because he’s a hidden gem, or is it because he’s not a smart play? Clearly some of these players represent passed-over talent that can help you win, but you still need to be able to identify them. What about the reverse case, where lots of people picked a player? The only useful thing this tells you is that the performance of this athlete will not be the differentiating factor. Does this mean you shouldn’t pick him? Not necessarily, as he may perform very well, and even though he will not get you the win, he may be necessary to contend for it. For the above reasons, having access to this information when making your roster picks is far from sufficient to come out on top. Although it is necessary for operators to prevent any cheating from taking place (steps have already been taken), the dramatic splash the cheating incident made in the news and legal sphere is, in my opinion, overblown.

DFS has been in the news recently in a way no industry wants to be. The challenges and changes faced by the industry underscore the need for regulation that protects consumers and operators alike. However, the media frenzy is drawing issues out of perspective and thus exacerbating the tensions involved. I hope that by providing context we can bring back perspective and a sense of proportion that will allow for a well-reasoned solution to emerge.

Sincerely,

Anton Tzarfn

CEO | DraftOrPass