What you Don't Know Can Hurt you (or Help You)

Humility, Uncertainty and Anti-fragility: Foundational Principles of Probabilistic Gaming

The late, great David Crosby once said of Kanye West; “He can’t sing, he can’t play, he can’t dance… (and he says) he’s the world’s greatest rockstar.”

If I was to distil the purpose of this newsletter into one sentence: Thinking About Thinking will help you beat your league-mates in fantasy football, without evaluating players better than they do.

Akin to Mr. West (in this way alone I hope), I am not the most knowledgable ‘football guy’ in the industry, nor am I the best modeller or number-cruncher. Any advantage in pure player evaluation I hold vs. an average dynasty manager, is primarily a result of time spent. Despite this, I’m a consistently profitable player. Where my edge derives, and the edge I hope to provide you through this newsletter, is an understanding of probability, risk, markets, and incentives.

By understanding the market forces which drive your game environment, and your opponents, you can position yourself to minimize the detriment of bad luck, and maximize the impact of good luck. Over this off-season (and more) I will outline the core tenants of the dynasty market in great detail. Today serves as the first layer upon which we will continue to build.

Leveraging Humility

The first edition of my old “Thinking About Thinking” column, was titled ‘Prospecting Through Humility. I won’t regurgitate it all here, but recommend reading it for the apple analogies alone.

In my view, the first step to success in any probabilistic game is humility. You need to have a realistic conception of what your strengths are as a player, what your weaknesses are, and how much is out of your - or anyone’s - control.

There are several ways to action a ‘humility-based’ approach that will lead to wildly different results.

A data-driven process shows humility by not over-estimating their ability to identify outliers, and instead attempting to win by stacking marginally higher-probability bets on aggregate. Someone choosing to input film into their process may argue it shows greater humility by acknowledging there are aspects to the game data can’t measure. My approach is based on the notion that any evaluative process is destined to have a certain degree of failure, so the emphasis ought to be on choosing bets which maximize the hits to outweigh the inevitable misses.

In any case, humility is eye of the beholder. So how do we define it, and action it?

Known Knowns, Known Unknowns, and Unknown Unknowns

Humility is in essence, the acknowledgement of what we don’t or can’t know. And this is a key distinction. Each fantasy roster has several different types of risk.

A ‘known known’ risk is one you know exists, and can reasonably foresee with specificity.

For example, will Trey Lance start for the San Francisco 49ers in 2023?

There is a high degree of uncertainty around Lance, but much of it is tied up in this question. How to value Lance hinges to a large extent on whether you are more or less optimistic on Lance’s odds to start than the market. If you choose to roster Lance, you can build with the knowledge of this risk in mind, or decide at which cost you’d prefer to remove that risk from your team. (We’ll return to this example soon)

Known Unknowns exist everywhere in fantasy as well. These are the risks you know generally exist, but cannot attribute to any single source. For example, you know injuries happen in fantasy football. But you can’t state with certainty which players will get injured, or whether your team is more or less likely to suffer from injury risk than your opponents.

You can’t predict how or when these risks will manifest, but we can prepare for them in the aggregate. You can build a team which is more or less prepared for the realization of injury risk to your assets, and a team which is more or less prepared to benefit from injuries to players not on your roster. While you can’t guarantee injuries will benefit that specific team, you can build a portfolio of teams which is likely to benefit on net when injuries, or any other source of random shock, occurs.

Lastly, unknown unknowns are sources of risks we know exist somewhere, but we don’t know how to define them or prepare for them until they reveal themselves. The best way to respond to unknown unknowns is to acknowledge the existence of risks beyond those we anticipate or explain, and plan accordingly.

In other words, Shit Happens.