I Make 82% on my Money with This One Weird Third-String Point Guard!

A few days before Thanksgiving, I read this NY Post article titled “The Men Who Make 6 Figures  Off Fantasy Football“; I probably clicked on it for the same reason Buzzfeed exists. The piece showcased several dudes looking nouveau-riche, one sharing accounts of all the money he’s earned while posing with stuff he hasn’t bought. Mark Giundi poses in a pinstriped suit with an unlit cigar and a tumbler of whiskey; his caption reads: “Styling by Mindy Saad, Sienna blue stripe suit, $779 at … “, etc.:

“This eyebrow ain’t gonna pluck itself, people.”

The article described the relatively new gambling phenomenon of “Daily Fantasy Sports,” wherein entrants get to pick an entirely new team of fantasy athletes for each day’s slate of games, and to enter and close several bets each day. This contrasts with traditional fantasy sports gambling, which generally involves sticking with a team in a league over an entire season, a more time-intensive and likely less-lucrative commitment.

Daily Fantasy Sports, because of the in-and-out nature, has more of an appeal the way dropping 25 bucks on a game does when you happen to be in a casino for the night. It’s a lot closer to gambling the way most people are comfortable with gambling, rather than the hobby that traditional fantasy sports must be. Personally, I would never consider traditional fantasy sports, but Daily Fantasy Sports sounded like a reasonably fun way to spend an otherwise uneventful evening, just like a night at the craps table.

It is nothing like a casino chance game, however; the better analogy than craps would be poker. Daily Fantasy Sports is a skill competition, not an inevitable loss to the rules of probability. The Feds agree: the Unlawful Internet Gambling Enforcement Act of 2006 allows for online fantasy sports. (Though conversely, it does happen to outlaw online poker.)

On Thanksgiving, I put down seven bucks on FanDuel.com, currently the industry leader. I had to pick a quarterback, two running backs, three wide receivers, a tight end, a kicker, and one team’s entire defense. All the picks had to come from the three Thanksgiving Day games: Detroit vs. Chicago, Dallas vs. Philadelphia, and San Francisco vs. Seattle. The bet allowed for 1277 entries. Winnings would be paid out according to the following formula:

Somehow, I spent most of the day in first place; I got knocked down to fourth when San Francisco didn’t manage to score a touchdown the entire game and entrants who had selected Seattle’s defense jumped ahead of me. Still, I was intrigued, as any gambling novice or heroin addict might be when he scores the good stuff the first time out.

I wanted to know the actual returns I was capable of producing over a large-enough sample size. As a skill game, if one obeyed the math, could Daily Fantasy Sports be a more lucrative — or even safer — place for one’s cash than stocks, or that .3% yield you’re getting on a CD thanks to quantitative easing?

Two weeks later, the answer is an unsettling “absolutely.” My calculated expected return seems too good to be true, and my risk of loss improbably small.

I was not going to get a large enough sample size with football, so I focused on basketball, which I could enter nightly. Basketball also offers a much more predictable series of outcomes: while baseball has long received much of the large market vs. small market animus, most of it particularly directed at the Yankees, basketball has a far smaller pool of playoff teams and champions. You simply do not see a ragtag group of hustlers — like this year’s Kansas City Royals, which reached the seventh game of the World Series — getting a shot at an NBA championship. The NBA is significantly more money and star-driven: Prime-of-his-career LeBron, Shaq, or Kobe are around at the end of the season, and that’s that.

And, more relevant to Daily Fantasy Sports, LeBron is infinitely more likely to finish a game with 40 points than he is to finish with 0, whereas baseball MVP Mike Trout will put up plenty of 0 for 4 outings over the course of a season.

FanDuel.com offers several options for entering a contest: the higher the reward, the fewer winners. The safest bets are the “50/50” and the “Double Up” games. In a 50/50 contest, the top half of all entrants will win 180% of the entrance fee: a 10 dollar bet wins 18 dollars. In a “Double Up” contest, the top 44% win 200% of the entrance fee: a 10 dollar bet wins 20 dollars. In a “Triple Up” contest, the top 30% win 300% of the entrance fee: a 10 dollar bet wins 30 dollars.

The contests get much more lucrative — and risky — from there. Some bets will only reward any money to the top 10% or so of entrants, but you might walk away with $25,000 on a $1 bet.

I entered about 50 contests over the past two weeks, mostly $1, $2, or $5 bets, to see what expected return I could get from the different contests and what my average percentile finish was compared to all other entrants.

A FanDuel team consists of 9 players: two from each position, but only one center. (I don’t know why.) You get 60,000 “dollars” to construct a team, with the price of each player changing according to some mathematicians at FanDuel. For example, LeBron costs about $10,400; Festus Ezeli, $3500.

Over these 50 or so contests, on average I finished in the 58.1 percentile, meaning I performed better than 58.1 percent of entrants. I often entered several contests using the exact same lineup; per each lineup instead of per each contest, I averaged a finish in the 61.48 percentile.

I definitely improved significantly over the two weeks, my second week alone is closer to the 70th percentile.

Most importantly for figuring out a strategy, I finished in the top 50% of entrants 67.44% of the time, the top 44% of entrants 65.11% of the time, and the top 30% of entrants 60.5% of the time.

After doing the math with the help of frequent PJM contributor Charlie Martin, I end up with the following expected returns:

For every dollar I enter in a 50/50 contest, I can expect to make $1.21. That’s 21 percent on my principal every day the NBA plays.

For every dollar I enter in a “Double Up” contest, I can expect to make $1.30. That’s 30 percent on my principal every day the NBA plays.

For every dollar I enter in a “Triple Up” contest, I can expect to make $1.82. That’s 82 percent on my principal every day the NBA plays.

Wait. What the hell?

Obviously I cannot bet the entire wad every day, so this money is not going to be compounding like a CD would. I have yet to calculate an optimal amount of risk to place down every day, which will render those returns significantly less than that 82 percent. Maybe I can make 82 percent on, say, 1/5th of my money.

But yes, I started with about $30, and now have a few hundred. And the odds that I blow it all by losing several nights in a row seem rather low; I have yet to lose more than twice in a row.

I am certainly not using any sort of mathematics with which to pick my rosters (the NY Post article mentions one mathematician who quit his day job to do this, and is now earning a middle-class salary). I am achieving these returns using what appears to me to be the insight of an average lifelong sports fan, plus what I’ve learned from a couple weeks of trial and error with FanDuel. It takes about 30 minutes of research a day, done throughout the day whenever I have a few spare moments.

Here’s what’s working:

— First, I fill in the positions of point guard, center, and power forward. As a sports fan, I know point guards are likely to rack up points, assists, and steals; while the front court players tend to rack up points, rebounds, assists, and blocks. Shooting guards and small forwards are more likely to just give you points.

— I try to cram as many All Star-caliber players on the team as I possibly can, starting with whoever seems to have the easiest matchup. Usually I can afford to get five of them on my nine-man roster. For example, if the Cavs are playing a bad team like the Knicks, I’ll be sure to have LeBron and Kyrie Irving on my roster.

— I fill in the rest of the slots with as many starters as I can. More minutes on the court seems to be the best indicator of more stuff in the box score.

— The best way to find “inexpensive” players who will perform well is through that day’s injury report. For example, last week the Houston Rockets had both of their point guards injured. So veteran guard Jason Terry was moved up to be the starting point guard. He was simply going to be on the floor much more than he normally would, so he was a likely pick to far outperform his “salary.”

Really, that’s it. There is a little experience and knowledge involved — from being a casual basketball fan, I happened to know Jason Terry has been a great shooter his entire career — but not much.

How am I producing that rate of return? It appears to me that the ease with which one can enter a Daily Fantasy Sports contest, along with the comfort that the average person might have with placing a small bet just for fun as compared to the commitment required for a traditional fantasy sports league, is creating a situation where more than half of the entrants are the proverbial “sucker at the table.” Being a general sports fan and having about a week of experience on FanDuel is enough to remove one from the “sucker” category.

A subject from the NY Post article refers to Daily Fantasy Sports as the new “day trading,” that hobby which peaked during the dotcom bubble yet has had a bit of a QE-fueled insurgence. At least for now, as Daily Fantasy Sports is in its infancy, it’s a heck of a lot easier to figure out than a Bloomberg terminal.