Jonathan Bales is a DraftKings Pro and the author of the Fantasy Sports for Smart People book series, most recently Fantasy Football for Smart People: How to Win at Daily Fantasy Sports.
The word “Tao” (also spelled “Dao”) means “way,” “path,” or “principle.” I copied that from Wikipedia, which isn’t exactly the most Taoist practice out there, but it does show that I’m highly committed to my craft and willing to perform intense research.
So I got that going for me, which is nice.
“The Tao of Bales” is a weekly column I’m going to write—published every Thursday at noon—that will hopefully meld Taoism with daily fantasy football in that I’m going to do my best to help teach you “the way” that top DFS players make money, or at least what I think are the guiding principles behind a profitable daily fantasy football strategy.
Now I know what you’re thinking: “How am I going to get any work done in anticipation of the release of this article? Can’t you push it up so I don’t have to wait all week?” Look, I get it, but no.
This column isn’t actually going to be centered around Taoism—maybe a little, I don’t know—but I did like the title and Ethan approved it so now that’s what it’s called. The philosophy behind what I’m going to be writing is, well, just that: philosophy. With so much data and information readily available to DFS players, I think what’s really missing from the typical player’s repertoire is a sound process—the path of least resistance to success—that includes things like embracing uncertainty, utilizing game theory, projecting ranges of outcomes, and understanding how certain player pairings can increase risk/upside.
Instead of trying to answer every imaginable question, I’m going to instead be proposing a lot of questions—asking the ‘why’ instead of responding to the ‘what.’ Part of that process will include questioning everything—even widespread “common knowledge”—and, in typical Taoist form, emphasizing a process of constant change, improvement, and evolution.
The topic each week will be decided by me at approximately 11am on Thursdays when I sit down to write the column. My hope is to choose a concept that is relevant for the current week of games.
And in Week 1, there’s no better philosophy than embracing randomness and uncertainty, harnessing it to create the best possible DFS lineups.
How many fantasy points is Dez Bryant going to score this week against the Giants? I have no idea. Okay, I have some idea—I know it’s not likely to be negative, I know it’s improbable he scores one million—but really I don’t have an incredibly strong grasp on how well Bryant will perform in Week 1.
That doesn’t necessarily mean I’m a total idiot and can’t accurately project players. I might still be a total idiot, but there’s no one out there who is going to come within five points of Bryant’s projection probably even half the time.
I think there’s actually a pretty good philosophical argument to be made that no one will ever be able to do that just because football is filled with randomness. Bryant could get shut down all game and then break a tackle late for an 80-yard score. Or he could be wide open deep and lose the ball in the lights.
Shit happens. There’s a certain level of variance in sports outcomes that we probably just can’t overcome when making predictions, even at a theoretical level.
But randomness doesn’t equate to a lack of a competitive edge.
Ya hear that? Uncertainty—even total chaos—doesn’t make DFS unbeatable.
Let’s say you’re betting on the roll of a die—something that’s more or less random (it might or might not actually be determined beforehand, but for the purposes of predictions, we can consider it random). Can you possibly win in that scenario—on an event you know will result in each number falling 16.7% of the time (assuming a fair die)?
Yes, you can, if you can secure payouts that assume a lack of randomness. If you’re getting 6-to-1 on your money, then you’ll break even over the long run. There’s nothing you could do to not break even. If you’re getting 5-to-1 on your money, you’ll end up losing over the long run. Again, there’s nothing you could do to not lose money. This format is basically what happens at casinos.
But if you could acquire any payout greater than 6-to-1, then the bet would be +EV for you and you’d make money after a large number of rolls.
Football isn’t random in the same ways as rolling a die, but it’s probably more random than we’re led to believe. How many times have you spent all week creating projections, researching splits, analyzing matchups, and so on…only to say “Well shit” by, like, 1:15pm?
So if there’s more variance than what most believe, the question is if we can get an asymmetrical “payout” such that we can make money without beating people in terms of projection accuracy (which would then just be icing on the cake). And I think we can, both in terms of site pricing and tournament ownership.
Basically, there are two versions of a marketplace in daily fantasy sports. There’s the market created by DraftKings in which we’re trying to determine which players’ expected production exceeds their cost. Then there’s the tournament marketplace in which player value is set by the public in terms of ownership.
In the former, we can embrace uncertainty by identifying situations in which random factors might be unfairly influencing pricing. When a running back has a couple poor games due to something like fluky game scripts and his price plummets on DraftKings, that’s a perfect situation to buy low—to bet on certain random, low-frequency events not repeating.
In the tournament marketplace, we’re concerned with exploiting inefficiencies in the way others think about players. Again, “cold” players are a perfect example of how to do this as you’re buying in on something you think is more random than others—a player whose production is likely to regress toward the mean—in order to generate usable value. If all player production were theoretically random, you could still make money by properly predicting ownership and using the guys in the fewest lineups.
There are a number of similar examples I could use—and will throughout the season—but the fundamental idea is to look for pricing and ownership that is more the result of randomness than repeatable factors, then try to exploit public opinion in order to obtain a payoff that’s asymmetrical to the risk.
Before leaving you, I wanted to briefly discuss a couple situations in which uncertainty might affect your Week 1 strategy.
Talent/Opportunity > Matchup
Because of the holistic nature of defenses, their overall play tends to fluctuate pretty wildly from year to year. As we often see with a player like Revis, one change in personnel can significantly affect team quality.
There are certain defenses, like Seattle’s, that we know are going to be sick. But most defenses are sort of all near one another and it’s difficult to predict how they’re going to play, especially given the year-to-year turnover in the NFL.
It’s much easier to assess individual player talent, and certainly to project opportunities. I don’t know if the 49ers will be able to stop Adrian Peterson on Monday, but I do know AP is going to get fed the ball all night.
Thus, I think analyzing a player’s specific matchup is actually less important in the beginning of the year than it is near the middle and end, when we have more relevant data with which to work.
This follows from what I wrote about a few paragraphs ago, but if it’s more challenging for the crowd to accurately identify value early in the year—which I think it is—then it makes sense to be more contrarian in Week 1 than any other week. There’s a whole lot of uncertainty about player performance right now, so why buy in on the chalk unless it’s a very clear value, i.e. Davante Adams? More so than any other point during the season, I will go against the grain in Week 1.
Tournament strategy is about properly balancing predictability with ownership; the more uncertainty, the more public opinion matters relative to value.That’s the end of my column. Bye.