Every year I look back on the season and say, “How did I miss on that guy?” Unfortunately, I always end up asking this question about multiple players after each season. I am going to venture a guess that I am not alone in experiencing this. That said, it is often a good thing that we miss on some out of nowhere players; to quote Kathryn Schulz’s excellent Being Wrong: Adventures in the Margin Error, “being wrong is often a side effect of a system that is functioning exactly right.” In other words, when dealing with an uncertain future, good process can still lead to misses—to bad results. Missing on Danny Santana posting a .405 BABIP or Michael Brantley posting a HR:FB rate nearly double his previous career high (by “miss,” we mean not paying a draft or auction day price for these breakouts) is actually a positive for our process rather than a knock against it. (Note: If we missed for predictable reasons, then that would be a knock on our process.)

However, these misses—the bad beats, the good process-bad outcomes—are not the misses I was talking about earlier. Rather, I was talking about the misses that should have been avoided. More specifically, the ambiguity effect causes us to miss out on players each year. Below we will take a look at the ambiguity effect, its different forms, and some strategies to battle it.

When Craig Goldstein and I put together our combined three-year outfield rankings, there was a particular discussion that happened repeatedly that relates to the previously mentioned misses and the ambiguity effect. That discussion always kind of went like this:

“Is Player X too low?”
“I just don’t know what to make of him.”
“Me neither.”

This actually worked out great because it forced us to dig into the player and make, hopefully, a better decision as to the player’s ranking. Why is this important to our fantasy baseball decisions? Because almost all of us do this; there are a lot of players to choose from and evaluating them all probabilistically is a daunting task. Consequently, we come up with valuations on the majority of players and ignore the ones that we do not know what to do with. You know these players, I know you do. Whether you make your own rankings, use rankings from the internet, or use some combination of the two, there almost always comes a time in every draft or auction at which you pass over (or consider passing over) the top player on your list or let a player go for less than your assigned auction value. We do this not out of a sense of charity to our leaguemates or masochism. Rather, this is probably just how we make decisions when facing uncertainty.

It took me a while to get here, but this is the ambiguity effect, a cognitive bias that shows our preference for more certain odds to less certain odds. The reason that this is considered a cognitive bias rather than just reasonable decision making is because we often pick propositions with worse, but more certain odds than propositions with better, but less certain odds. Here are a few players that I missed last year (and there were certainly many more) because of uncertainty that surrounded them, and I am again guessing that I am not alone: Melky Cabrera, LaTroy Hawkins, Charlie Blackmon, Phil Hughes. Again, the reason we pass on these players is not because we thought they were terrible; rather, we just did not know what to think. As a result, such players often end up being great values because they are being discounted for reasons other than their projected value.

There are several different ways an outlook for a player can appear ambiguous and with each way our valuations can be corrupted. Uncertainty about production, health, and playing time all cause us to overly devalue players. Regardless of the reason for uncertainty, our issue is not with a player’s entire range of potential outcomes (the upside is wonderful); rather, our issue is with the most negative outcome. We know we tend to overweight extreme-negative outcomes because of our risk-averse nature. The result of all this is that our valuations for players with seemingly ambiguous odds for future production tend to be attempts at removing all risk. For example, Jarrod Dyson went for about $5 in most AL-Only leagues because of uncertainty over playing time, even though he earned $14 in both 2013 and 2012, while also facing uncertain playing time. As we can see, Dyson got valued at his worst-case scenario because of the ambiguity effect and risk aversion.

Okay. So we have an idea regarding the how and why of the ambiguity effect as it relates fantasy baseball, but what can we do about it? Two pieces of advice:

1. Search your valuations for gaps and question marks
This solution is straight from the above list-making example. We analyze most players or at least have an idea why most players are valued at the rank or price they are valued. These players, however, are not the ones we miss because of the ambiguity effect. Therefore, identifying the players we are avoiding because of the difficulty in forecasting their production will help us avoid the situation where we pass up our top ranked player or let a player get acquired below our auction price. This is not to say we will not change the rank, but the first step in identifying a valuation we are uncomfortable with (and consequently performing further analysis to confirm or disconfirm the valuation), is finding these gaps and question marks ahead of time.

2. Put odds on outcomes
Put differently for fantasy baseball purposes, each player is a bet with different amounts of payoff and risk; thus, only looking at players as safe or risky is going to cause us to leave value on the table. Sure, a player may only have a 20 percent chance of holding onto a closer gig all year, but that 20 percent, while not optimal, still holds value. Yes, we should be factoring in odds for the worst outcomes, but we also need to be factoring in the odds for all other outcomes. There are plenty of bets (and players) that will not pay off the majority of the time, but they all have a price at which the risk becomes worth the potential reward.

Lastly, while we try to remove the negative consequences of the ambiguity effect from our own process, we would be keen to take advantage of its consequences on our competitors’ decision making. While every fantasy baseball participant understands that there is no absolute certainty in the game, many participants will take great measures to avoid risk as much as possible. The consequence of this is that there tends to be many profitable yet risky bets for the taking. When these bets fall to us, embrace the risk and take comfort in the cost and potential payoff.

Thank you for reading

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.

Subscribe now
You need to be logged in to comment. Login or Subscribe
A really helpful way to approach my draft, Jeff. Great article.
Thanks! Glad to hear it.
I am in a league that limits the number of free agent pickups per year. We have 4 total for the season, plus a two-round free agent snake draft at the all-star break. Every year, I have trouble evaluating the risky players. How would you modify your strategy above?
First, risky players are going to be less valuable in such a league. Second, we it is important to differentiate injury risk with performance variability as the former would hurt more in this league (I would think). I would be looking, like any league, if such players are being overly discounted/avoided. It would then be about finding when the risk becomes worth the reward. I would look back on the league (if possible) and see if there were discounts on perceived risky players or if owners were not taking risk/league structure into account enough.