It being outfield week and all, it should be said that outfielders are great. Maybe they were not the best at handling groundballs and making quick, accurate throws growing up, but the reason the majority of them are still employed by MLB teams is because they can sure hit. Consequently, outfield is usually our most productive fantasy baseball position and the good ones are really good—of the 73 players (pitchers and hitters) to earn $22 or more in mixed leagues last year, more than a third (25) were outfielders. Because our teams are likely to contain multiple productive outfielders, we often consider mix when choosing between outfielders or top hitters for that matter. Getting production in each category is important in category leagues, but the more interesting question regarding player mix is how we should be balancing risk across our team. If you have been reading along, you know that our goal is going to be to select the players with the highest average return or the ones that give us the best chance at winning. However, we know that this goal is a tricky one to nail down and we can certainly benefit by discussing the intricacies of these decisions and why making the best decision can prove difficult. More specifically, we will discuss the implications of portfolio diversification and our perceptions of and attitudes towards risk.
Let us throw this over to Wikipedia to get us started,
“In finance, diversification means reducing non-systematic risk by investing in a variety of assets. If the asset values do not move up and down in perfect synchrony, a diversified portfolio will have less risk than the weighted average risk of its constituent assets, and often less risk than the least risky of its constituent.”
Using the above definition of diversification, we can view our fantasy baseball teams as diversified portfolios in that they are a variety of risky, independent assets. The reason our teams are diversified is because the production of one player (asset) is almost entirely unrelated to the production of any other player and, moreover, there are no uncertain, underlying factors that stand to impact a certain type or subset of players (there are such factors such as rule or equipment changes, but they are almost always happen in the offseason and are thus known and we can therefore adjust our valuations accordingly).
It is also important to note what diversification, as it relates to fantasy baseball, is not. It is not having players with different amounts of perceived of risk, different types of players (speed players, power hitters, balanced hitters, etc.), or players on different teams. Why not? Because there are no underlying factors that are going to only reduce the production of one type of player or one team (at least none that should be a part of our valuation). The most likely point of confusion surrounding fantasy baseball and diversification is that regarding players with different amounts of perceived risk. Yes, having Troy Tulowitzki and two “safe” players is less risky than having three Tulos, but that is impossible anyway. The real benefit of diversification is that owning any collection of assets, even three riskier assets (for example, Tulo, Carlos Gonzalez, and Ryan Braun), will lower the portfolio’s risk and will probably lower its risk below that of the least risky asset. Obviously, the package of three risky assets will still definitely have more risk than Tulo and the two safe players, but if people are underrating risky players because they already have risk in their portfolio or for any other arbitrary reason, do not hesitate to pounce when the price is right. Moreover, while risk is a part of our valuation calculation, we are really only interested in the end calculation, in investing in the portfolio with the highest expected yield.
To clarify, the point of this is not to say we should be disregarding risk in our valuations, but that when we choose less risky players in order to “reduce our team’s (portfolio’s) risk” instead of riskier players with higher expected returns, we are unnecessarily leaving value on the table. Additionally, we probably overrate our ability to accurately forecast risk, which makes placing inherent value on reducing risk an even less profitable strategy.
Attitudes Towardd Non-Systematic Risk
(Note: Non-systematic risk is simply risk the way we are used to thinking about it, which is player-specific risk.) We have all been there, the owner of a team that gets decimated by injuries. The funny thing about those teams is that they are almost never teams where we go into a season saying, “Wow, I chose a lot of risky players, I would not be surprised if this team gets decimated by injuries.” In other words, we often describe assets as risky in hindsight and this is strike one against our ability to properly factor risk into our valuations. Strike two is our difficulty in looking at the future probabilistically (it easier to say Player A is a risky proposition and Player B is a safe bet, rather than Player A has X amount of risk and Player B has Y amount of risk, and this is how it affects their values accordingly). Additionally, using black-and-white descriptors when dealing with risk makes decision making easier, but certainly does not make us better decision-makers. And lastly, our inherent risk aversion and preference for consistency further distorts our ability to properly weigh risk. Given all this, there is a pretty convincing case that we overweight risk in “risky players” and underweight risk in “consistent players.” Put differently, this means that players where risk is being built into their price are probably better bets than “super-safe” players, the “locks to play 150-plus games” where risk is being completely disregarded.
The takeaway here is not to target risky players or avoid safe players. Rather, the takeaway is not to avoid risky players only because they are risky or because we have other “risky” players. We know risk tolerance varies from person to person and we also know that there are certain times when we tend avoid risk. First, we tend to do so when we have multiple options that exceed our expectations. We also tend to avoid risk when we imagine that taking on risk could be easily criticized by others. As ridiculous as it sounds, we often allow being viewed as rational by our peers to matter in our decision-making (even though they are also our competitors and even though it is just fantasy baseball). When faced with these situations and fears in the future, the hope is that we can note that they are simply getting in the way of optimal decision-making and disregard them as distractions. Systematic risk, conversely, can potentially fall below our radar, while causing far greater damage to our teams than any misevaluation of a risky asset.
Systematic risk, also referred to as undiversifiable risk, almost never occurs in fantasy baseball (I cannot think of a midseason and thus unavoidable instance), and if it did, there would be nothing we could to about it. That said, if we want to be liberal with the definition, there is a systematic risk that we all run the risk of unnecessarily incurring—that being biases and oversights in our analysis and/or strategic process. If there are critical pieces of information being consistently overlooked or improperly valued (looking at stolen-base totals without caught-stealing totals, not taking shifts into account in our analysis, disregarding the depth of starting pitching, scouting the stat line, etc.), we run the risk of overvaluing or undervaluing a particular set of players, which causes multiple problems. First, if our analysis or strategy causes us to overrate a particular type of player, then we may end up with too many of a particular player at an inflated price. Conversely, if our analysis or strategy causes us to underrate a particular type of player, then we might miss out on those players even when they have fallen too far. Lastly, overrating or underrating a certain type of player necessarily causes us to respectively underrate or overrate all other players.
While the heads up is nice, pretending that this is something we can simply solve or erase is unrealistic. We will always have these biases and oversights in our process, which will never be perfect; consequently, we add another reason to the list of why we need to constantly be evaluating our own process. In doing so, we will certainly find results that look like patterns that end up only being variation, but occasionally we will find patterns that are a result of gaps in our process. Making changes to close these gaps is the way we can reduce systematic risk in fantasy baseball, not avoiding non-systematic risk for the sake of doing so.
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
What do you have to say about the playoff problem and systemic risk and injury?
Would you agree that, generally, the deeper the league, the more you should factor risk into your pricing?
Of course, at a certain point everyone becomes a worthwhile purchase, and solid veterans who haven't yet broken down can often be a good source of value, but in deep-only leagues, I've rarely seen a roster laden with injury risks provide the performance pay-off that's theoretically possible.
On an individual level, Torii Hunter, Victor Martinez, and David Ortiz have been solid 'stocks' to buy over the past few years.
Regarding your last two points, I saw several teams carried by Matt Kemp and/or Albert Pujols, two players where risk was overrated that turned a profit. Again, there will always be players on both sides, the key is not overly devalue risk and overly inflate the value of consistency.
I have noticed that the attitude regarding risk is quite different when it comes to injuries and rookies. Where injuries are concerned, there is a belief that a certain type of player can get injured more frequently than others and thus is a bad investment. So owners then shy away at any price and allow other owners to turn what may have been a bad investment into a good one relative to the risk. This type of analysis not only ignores the fact that the "injury prone" player might defy expectations and stay healthy but that other, so-called non-risky commodities can, in fact, get hurt. Baseball is a physical competition where everyone, theoretically, is prone to suffer and injury simply by being on the field.
Rookies are a different kind of risk. It is an unknown as to whether or not they will succeed in the short-term at the major league level, yet they are often (and with growing frequency) treated like proven commodities based on how they are priced. There is variability between how a rookie is ranked and how a rookie will perform (in both directions) yet the elite rookies frequently command prices that ignore all of the risk and solely focus on the reward.
While I agree it is an imperfect label, this is where Systematic Risk can apply. I have seen this frequently occur in keeper leagues where a fantasy owner collects a roster of players with no track record and the performance baseline is an entirely open question. These teams can succeed - of course- but frequently rookies are hoarded in the hopes that they will all become Mike Trout. This obviously cannot happen, but this won't stop an owner from undervaluing established production baselines in favor of unknown production.
I generally agree that risk is overblown. But in some leagues where rookies are concerned the risk is ignored almost completely.
In the draft, $260*#teams equals the dollars spent in total.
At the end of the year, $260*#teams equals the total value produced (e.g., if you run a retro PFM).
However, some value comes from undrafted players. If 80% of the value at the end of the year comes from drafted players, then 100% of the draft money bought 80% of the value.
That means, if you know with certainty that someone will produce $20 at the end of the year, you should be willing to pay $20/(80%)=$24 for that player in the draft.
So, my major questions are:
How much value at the end of the year comes from drafted players?
Are consistent players undervalued using a PFM style forecast?