Dynasty content is part of the ELITE package for Footballguys Premium Subscriptions. We're making this preview available so you can see the edge these subscribers are getting. Sign up here.

See past Dynasty, In Theory articles:

Dynasty, in Theory: Most Comparisons Are Wrong

There's a lot of really strong dynasty analysis out there, especially when compared to five or ten years ago. But most of it is so dang practical-- Player X is undervalued, Player Y's workload is troubling, the market at this position is irrational, and take this specific action to win your league. Dynasty, in Theory is meant as a corrective, offering insights and takeaways into the strategic and structural nature of the game that might not lead to an immediate benefit but which should help us become better players over time.

Biases and Fallacies and Delusions (Oh My!)

It seems like all anyone wants to tell you these days is how you can't trust your own mind. Your reasoning is too flawed and fails in predictable ways.

For instance, you're too unwilling to make trades because of loss aversion (and anyway, you overvalue players on your roster because of the endowment effect, especially if you invested major resources to acquire them and therefore are falling prey to the sunk cost fallacy). You're too slow to update your rankings because of the anchoring bias (and also, you're too fast to update because of recency bias).

Here's the thing-- all of this is true. Our reasoning really is prone to failing in predictable and replicable ways. These fallacies and biases are easy to demonstrate experimentally; they've been some of the most robust findings in the social sciences, easily surviving the replication crisis.

But sometimes, the rush to condemn all mental shortcuts goes a bit too far. Each of these shortcuts was developed because it was the best strategy for survival in a particular environment. In some cases, we're no longer in that environment, and therefore that shortcut is no longer useful; it trips us up and costs us value. But this is a problem with that specific mental shortcut, not the concept of shortcuts altogether.

The reason we developed those shortcuts in the first place is that when they are well-suited to our current environment, shortcuts are tremendously useful. Not just because they save us time and energy (although let's not totally discount the utility of saving time and energy here), but because they lead to better, more accurate, more useful decisions.

Simple Heuristics That Make You Smart

To begin with, these mental shortcuts are known as heuristics (or, more commonly, "rules of thumb"). And while most psychologists have spent the last fifty years studying all of the ways that heuristics fail, a few have devoted themselves instead to studying all of the ways that they succeed.

Here's an interview with Gerd Gigerenzer, one of the foremost proponents of "fast and frugal heuristics". And he finds the situations where simple heuristics outperform calculate models the most tend to be the situations with the highest levels of uncertainty.

Consider chess. If I am playing my opponent, there is plenty of uncertainty involved. Mathematician Claude Shannon once famously estimated that there are more possible variations in a game of chess than there are atoms in the known universe. But the limits of that uncertainty are well-known. My opponent might move her king's pawn to e4, or she might prefer to move her queen's pawn to d4 instead, but she isn't going to pass Go and collect $200, and she certainly isn't going to move her Dreadlothian Battlecruiser to the Harkfasher Nebula to task the Grand Wizard Council with making humanity forget the concept of monarchy entirely, overthrowing my king and granting her a first-turn victory. We have a very clear idea of what is possible in chess. This is known as "bounded uncertainty".

Contrast this with investing in the stock market. I can study the business fundamentals of a cruise line, and it's possible that their current pricing structure might be insufficient to cover their costs, resulting in a loss of value. Or it might be possible that their new CEO's plan of building smaller ships helps them tap into a new market, resulting in an increase in value. But it's also possible that a zoonotic disease will jump from bats to humans in a wet market somewhere in China, and as a result, humanity will spend a year or more staying six feet away from each other, resulting in a decrease in value. Stock investing is a "game" where none of the participants know all of the possible "moves". This is, as you might have guessed, "unbounded uncertainty".

As it turns out, in situations with bounded uncertainty, complex models do exceptionally well. Deep Blue, the first computer model to ever beat a human grandmaster in chess, evaluated moves based on 8,000 different variables and had plans for 4,000 different opening positions and all endgame board states involving five or fewer pieces (plus many involving six). It was able to search 200 million chess positions per second and compare them to a reference database of 700,000 Grandmaster games.

But the more uncertainty increases, the more simple rules tend to outperform their more-complex peers. In the interview, Gigerenzer recounts the story of Harry Markowitz, one of the creators of Modern Portfolio Theory (or MPT), a mathematical approach to allocating investments in the stock market to maximize returns for a given level of risk. You can scroll through the Wikipedia entry to get a good idea of the math involved (and how quickly it scales as you increase the number of assets you're investing in).

Did Markowitz use MPT for his own personal retirement investments? No, he didn't; instead, he used a simpler rule called "1/N" (or "one over N"). For one over N, you identify all of the stocks you like, and you just put an equal amount of money in all of them. (Therefore, if you like N stocks, you will invest 1/N of your total portfolio in each.) This was good because, in most situations and time periods, 1/N has outperformed the more complex allocation optimizations.

Gigerenzer has studied many more instances where simple heuristics dominate complex models, and this tends to be the common feature: the more uncertainty is at play, the better simple rules tend to perform.

How Complex is Fantasy Football?

On a scale from "chess" to "the stock market", fantasy football lies... somewhere in the middle. Some elements are fairly certain. The NFL is probably going to play 18 weeks (unless another zoonotic disease makes the transition, I suppose), all of the fields will almost certainly be 100 yards long, touchdowns will continue to be worth six points with the option to go for an extra one or two after.

On the other hand, some elements are quite uncertain. A player might tear his ACL, his Achilles, or any number of other muscles or ligaments or tendons which are susceptible to tearing under the right combination of forces. Your first-round rookie pick might suddenly decide it's a great idea to drive 140 miles per hour in a 55-mile-per-hour zone. A promising young linebacker might decide to retire after a single year.

As a result, fantasy football, in many ways, lends itself well to complex models. But there also remains room for simple heuristics.

It's important to remember that simplicity is not in itself a virtue. All of those fallacies and biases in the introduction are simple; they also result in worse decisions and lost value. To the extent that we use heuristics, it's imperative that we monitor and test them to ensure that they're leading to better decisions in the long run.

But today, I just wanted to talk about why the trend towards more and more complicated analysis can sometimes lead us in the wrong direction and how a winning strategy can be built on a series of simple rules of thumb, zealously applied. And in the coming weeks, I'd like to take a look at some of those simple rules that I use for my teams, rules that have proven their value over time, and discuss why they tend to work so well.