‘Round this time last year, we took a behavioral look at lineup setting. It was pretty good, but there is one point briefly touched on that deserves more attention and more explanation as we are in the middle of setting our playoff lineups. That point follows: Our role in lineup decisions does not matter as much as we think.
To start, we think we have more control over the outcomes of our decisions than we really do. Some people call this bias the illusion of control. We start with a healthy chunk of words from the aforementioned article:
“When you take into account that there is probably not much difference between your alternatives (they might perform much differently on any given week, but this is due to random variation, the nature of small samples, and probabilistic futures, not the actual players or the decision maker), then you really find how little the impact our sit/start decisions carry.
I mention this not to point out the insignificance of the decision, but to point out that our role in the decision is largely overblown…thinking that we can make a genius sit/start decision that will be the deciding factor for our team could lead us to making sub-optimal rather than optimal decisions.”
The things we want to touch on today (and that were unfortunately not touched upon in last year’s article) are the ways that illusion of control can negatively affect our decision-making processes.
1. Seeing outcomes instead of probabilities
When choosing between top-50 Pitcher X and not-top-150 Two-Start-Pitcher Y, we are usually not seeing the weighted range of possible outcomes for each choice; instead, we “see” or think something like the following: “Pitcher X is most likely going to give my 15-25 points with a chance of giving me 30+ if he pitches a gem. On the other hand, Pitcher Y has a good chance of pitching at least one dud, making him less likely to score as many points as Player X in his other start, but he is probably more likely to give me 25-30 if he pitches well in [insert pitchers park] and holds his own against [insert a seemingly struggling offense].”
This is seeing outcomes instead of probabilities. It is how our brains work. We work on expected outcomes. The issue with this, as we have discussed on this internet before, is that expected outcomes are often not properly weighted. Instead, our expected outcomes overweight the vivid, recurring, negative, choice-supportive, and contrastable among other things. This is bias dressed up as analysis. This is taking the path of least resistance when faced with complexity and uncertainty. It works fine for choosing an EZ-Pass lane or deciding whether to sort or filter in Excel, but when it comes to decisions where our pride and earnings are on the line, we are better off with actual analysis (which is still easier typed than done).
We need to stick to more statistical, historical, and/or process-based projections. We can use the experts projections or our own or some combination of the two. Most of us use this blended approach (it is fun and feasible), adding in our own information or rules or intelligence to industry standards or expert projection in a way that we believe adds value, but we need to make sure we are applying that process across all players, otherwise we risk making the previously mentioned bias-dressed-up-as-analysis mistake.
2. Solving an easier problem
When we face an uncertain future, when the probabilities are fuzzy, our brains often try to reframe our decisions in a way that makes the decision we face easier to solve. We do this when setting our lineups when we say things to ourselves, co-workers, domesticated cats, whoever such as “Player X allows is a lock for at least 10 points,” “there’s no way my team can win unless some of my players blow up (figuratively obviously), so I gotta go with Player Y’s upside”, “If I go with Player A and B, then I have one safe play and one upside play,” or “I’m going with Player C because I am already starting Player D in two other leagues.” All of these are ways of solving an easier problem by adding in a criterion that is easier to decide on (risk as opposed to probable outcome).
What we are really doing when we solve an easier problem is helping ourselves to make a defensible decision. Yes, we can rightfully ring the defensive decision making alarms. With the best option often being unclear, we make a decision we can defend should the outcome go south. We are now in a second order of non-value-added criteria in our decision-making process—another juncture in which we can skew our decisions instead of focusing on the odds.
We need to properly frame our decisions if we are going to be at our most successful (note: this is a goal, not a solution). To achieve this, we have to reframe our decision, because our brains are already trying to incorrectly frame it either through ease, defensibility, or both. There are a bunch of ways to properly reframe decisions (none that work as much as we would hope), but my two favorite are (i) choosing the option you would choose if you had to run each decision through a Monte Carlo simulation and (ii) choosing the option that you would advise someone else choose. The first method forces us to think probabilistically, while the second—from Decisive: How to Make Better Choices in Life and Work by Chip Heath and Dan Heath—allows us to “attain distance before deciding,” which helps us battle defensive decision-making.
Lastly, the only way we get better is by measuring, reviewing, and improving our process. It is some of least-sexy advice out there, but it might be the most helpful. Good luck.
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