January 16, 2017
First Basemen, Roto Categories, and Total vs. Complementary Production
Good Monday and, more importantly, good Martin Luther King Day to everyone.
Unrelatedly, we go from the catcher position, the least productive offensive position (likely a result of all the ways catchers can impact the game defensively), to first base, one of the most productive offensive positions, likely a result of the limits on the way a first baseman can impact the game defensively. Talk about pointing out contrast because I didn’t know how to get rolling on this article.
Anywho, as mentioned last week, while the catcher position’s lack of fantasy prolificness often provides us with a somewhat clear cut best option at any given time, first base often provides us with multiple, potential best options at any given time, which can make choosing a first baseman (in either a draft or auction) somewhat difficult. In theory, when it comes time to make our draft pick or bid on any given player, we would have perfectly conceived valuations for all players and know exactly who to pick or how much to bid. In reality, with every decision that lacks an obvious choice, we struggle; and the more options available, the greater the struggle. And the worst part of all of it is that knowing that we struggle ain’t enough. To improve, we need to know the sub-optimal decisions this struggle leads to in order to better avoid them. So, that said, let’s get on with the knowing.
When faced with complex, difficult choices, our brain uses certain shortcuts to help us out, many of which we have discussed previously (such as the default effect). In last offseason’s first base article, we took a look at the representativeness heuristic—about how we will tend to gravitate towards first baseman whose production looks more like what we expect for the position (home runs, RBI), while therefore underrating the production provided by first baseman in non-traditional categories (runs, stolen bases). As mentioned previously, this is an issue because it lowers our chances at winning because we are not taking players with the highest expected value.
Today we will look at another way we choose players based on something other than expected value, that way being when we select a player because of fit or complimentary production to the rest of our team. What does that mean? That means, for example, choosing First Baseman X (1B X) who is expected to hit 40+ home runs, but only with a .215 AVG because your current hitters lack elite high-end home run production over First Baseman Y (1B Y) who is expected to be better than 1B X overall, but with only 20-25 home runs. We see this happen a lot when justifying our decisions; heck, we see this all the time when experts break down their rosters from mocks, drafts, and auctions. We hear explanations like, “I felt really strong in the power department with my outfielder, plus I have Arenado at third, so while I slightly prefer Carlos Santana to Adrian Gonzalez in a vacuum, I grabbed the latter to avoid taking the hit in AVG.” When I used to read an explanation like that I would shrug, thinking it made sense; heck, I made plenty of decisions for similar reasons.
Make no mistake, though, choosing a player because of their means of production as opposed to their total production is almost always a bad idea. Why? Mainly because we (people) are not as good at predicting the future as we think we are doing so (we are overconfident). When we pick 1B X over 1B Y from the above example, we do so thinking that we are already likely to do very well in the non-home run categories and that it is worth taking a slightly lesser player to try and improve in an area we see as a weakness. The first way we err here is by thinking we know how well we are going to do any given category with any large amount of certainty. Between over-performance, underperformance, injury, and suspension, our teams will likely produce much differently in season than we expect, if we expect them to perform in a very specific way (a way specific enough to get us to pass on the best player available). The next way we err is in thinking we will know what production will make a difference in any given season. Depending on the performance of our competitors, the fifteen home run difference 1B X and 1B Y could be worth anywhere from zero to, reasonably at most, six rotisserie points; and the same can likely be said for the advantage 1B Y provides in the other categories and likely more. Put differently, we are not likely going to be specifically good enough in certain categories so that we are left with no room to improve in the standings with a certain type of player versus another while being specifically bad enough in one category so that we are left with a ton of room to improve in the standings with certain type of player versus another. Sure, a team that is the best in a rotisserie category cannot gain you any more points in that category, but it is usually overconfident to think we will certainly be at or near the top of a given category (especially in redraft leagues).
More than anything, though, we are mistaking what it is that causes teams to win. Look at teams that win leagues and they usually roster more of the best players than other teams, they usually have more good players than the average team, and they usually have less bad players than other teams. It sounds simple, but the best way to construct a team that performs the best across your league’s rotisserie categories is to consistently pick the player with the highest expected value. Players with higher expected value—which takes into account our best guesses at the probabilities of all possible production outcomes for a player—are, by definition, more likely to be the players we see on winning teams than are the players with lower expected value (the players we pick based on fit).
Choosing complementary production over total production seems like a mistake the more we dig into it, and it often feels wrong when we make the decision to do so, but we do so not because we lack skill or intelligence. Rather, it is because these are, as mentioned, hard decisions and we lack a better way to frame the decision. When we get stuck with a tough decisions, such as the best first baseman to pick, let us therefore try to frame our decision around building a winning team, which in rotisserie leagues is not a team that is in first in every category, but rather a team that does as well as possible in all categories because it rosters the best players possible.