keyboard_arrow_uptop

Not sure if you all have noticed yet, but I am pretty fascinated, maybe even obsessed, with how we choose to make the decisions we make. Writing about fantasy baseball works great for me because participants are forced to make numerous “meaningful” decisions about an uncertain future. The interesting part about this point of the season is that while the decisions being made are certainly less impactful than those made in the offseason (when selecting the players for our teams), there are fewer relevant tools owners can use to aid their decision making right now. Whereas projection systems, rankings lists, chatrooms, and software tools can be used and adapted for offseason drafts, auctions, keeper decisions, and trades, the in-season projections and ranking are often less easily adapted to the decisions we face at this point of the season. They are less easily adapted or all-in-all unhelpful because (i) we are weighing far more variables (promotion dates, injury return dates, category rankings, playoff considerations, etc.) when making these decisions and (ii) league structure and dynamics are so varied at this point of the year that the availability (be it through trade or free agency) and market cost of a particular player will range greatly across leagues. Consequently, we are often left to our own accord to make these roster decisions. Put differently, in season—more than any other time of the year—is a time when we must make decisions using our own forecasts.

This is the fun part, but it is also the terrifying part. This is also the part that makes us most likely to cast sound process and analysis aside, thus allowing intuition (and thus biases) to creep into our decision-making. Luckily, some smart people have written about ways we can improve our decision making when facing an uncertain future. An excellent article titled “Outsmart Your Own Biases” in the May 2015 edition of the Harvard Business Review, takes a look at many things regarding our biases, but the part we will take a look at in this article is how we choose to employ our intuition (as opposed to a more structured process) and ways we can help correct such sub-optimal decision making. The first two paragraphs from the article are below and they explain why leaning on our intuition too often is not just a mistake, but a mistake in logic:

“Suppose you’re evaluating a job candidate to lead a new office in a different country. On paper this is by far the most qualified person you’ve seen. Her responses to your interview questions are flawless. She has impeccable social skills. Still, something doesn’t feel right. You can’t put your finger on what—you just have a sense. How do you decide whether to hire her?

You might trust your intuition, which has guided you well in the past, and send her on her way. That’s what most executives say they’d do when we pose this scenario in our classes on managerial decision making. The problem is, unless you occasionally go against your gut, you haven’t put your intuition to the test. You can’t really know it’s helping you make good choices if you’ve never seen what happens when you ignore it.”

You all can see the parallels to making fantasy baseball decisions in the above quote. Replace deciding to hire someone in the first paragraph with accepting a trade offer, deciding when to “sell” in a keeper league, or determining which player to prioritize first from the waiver wire, and the second paragraph becomes something we can all relate too. Sure we do our research and preparation for any decision we make in fantasy baseball, but the factors we are weighing when (i) doing that research and (ii) making the decision are often dressed up ways of sticking with our intuition. As mentioned before, this wonderful article (go read it) not only points out the problem, it also provides some solutions. My hope is to translate some of these solutions to fantasy baseball so that we can improve our in-season decisions that are based on projecting the future.

(Please note that the methods below (make three estimates, think twice, use premortems, and take an outside view) are all taken directly from the above article and I by no means intend to pass them off as my own ideas).

Make three estimates

“To improve your accuracy, work up at least three estimates—low, medium, and high—instead of just stating a range. People give wider ranges when they think about their low and high estimates separately, and coming up with three numbers prompts you to do that.”

It seems obvious that we should think probabilistically about a probabilistic future, but the nice part of this method is it makes you think of both positive and negative outcomes for every player we are analyzing. We have discussed the downside of trying to evaluate players we have never seen fail, such as rookies who show success upon their call up, and this method at least tries to expose our blind-spots if we have them for a given player.

Think twice

“A related exercise is to make two forecasts and take the average. For instance, participants in one study made their best guesses about dates in history, such as the year the cotton gin was invented. Then, asked to assume that their first answer was wrong, they guessed again. Although one guess was generally no closer than the other, people could harness the “wisdom of the inner crowd” by averaging their guesses; this strategy was more accurate than relying on either estimate alone. Research also shows that when people think more than once about a problem, they often come at it with a different perspective, adding valuable information.”

This is essentially the crudest Monte Carlo we can do. I love it. Sure it is incredibly flawed (and a disrespectful comp for Monte Carlo), but it is better than just going with a single, personal “forecast” (this is different than using an actual projection system, which is essentially thinking thousands of times). “Think twice” also makes us consider time, which is an important and often overlooked part of decision making in fantasy baseball. Sitting down at the computer, doing some research, and then making your best decision is probably going to be overly influenced by recency bias and potentially other biases. By thinking about a decision and then thinking about it at a later date or time, we are able to do some analysis of our analysis, which has the benefit of exposing potential flaws in our original process.

Use premortems

“In a premortem, you imagine a future failure and then explain the cause. This technique, also called prospective hindsight, helps you identify potential problems that ordinary foresight won’t bring to mind… Thinking in this way has several benefits. First, it tempers optimism, encouraging a more realistic assessment of risk. Second, it helps you prepare backup plans and exit strategies. Third, it can highlight factors that will influence success or failure, which may increase your ability to control the results.”

This method should help us when thinking about larger strategic moves, such as big trades. It is easy to get caught up in “am I getting the best side of this deal” or specific player analysis and thus easy to get away from the important question, which is “is this the action that will most help me win (or achieve whatever other goal we are trying to achieve)”. It also gets us to think about other oft-overlooked factors such as competitive response and uncertainty.

It is very important to know our tendencies when using this method because this can easily be used as a way to confirm our preexisting nature. To combat this, think of the downside of acting in accord with your tendencies. So if you are an overeager trader, spend more time doing a premortem of making a trade. Conversely, if you are too conservative with your FAAB (you tend to always have left over FAAB at the end of the year), spend more time doing a premortem of not spending your FAAB.

Take an outside view

“It’s important to have an expansive mindset about your objectives, too. This will help you focus when it’s time to pick your most suitable options. Most people unwittingly limit themselves by allowing only a subset of worthy goals to guide them, simply because they’re unaware of the full range of possibilities.”

Making ourselves take an outside view can be very difficult in fantasy baseball. I have found the easiest way to do so is having someone to bounce problems and ideas off of. Whether it is the bat signal, a co-owner, or a friend that understands our league, being able to see how someone else views the decision we are facing is invaluable. At, best they will see angles that we overlooked and at worst they will give unhelpful advice that we should be able to easily recognize and disregard. Not every decision warrants such discussion, but for those ones over which he have agonized, such resources can help us refocus on what matters the most.

We end with the disclaimer that we will never be perfect decision-makers, but we can always keep improving and now, as we head towards trade season, is a great time to improve our decision-making processes.

Sources:

Soll, Jack B., Katherine L. Milkman, and John W. Payne. "Outsmart Your Own Biases." Harvard Business Review May 2015. Harvard Business Review. 01 May 2015. Web. 09 June 2015.

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
ErikBFlom
6/10
One thing that scientists do is ask the a question that use independent techniques of analysis to obtain an answer. If the answers agree, then there is a good chance the answer is correct. For example, if you can measure something chemically (chemical reaction) and spectrographically (shine light through it and measure the light), getting the same answer from both tests is comforting that you are avoiding an artifact in the measuring technique. So ask your two questions totally different ways. Scouting and stats. Individual performance, versus team and role. Look at the gem from each facet. The flaw may be really visible from only one.