The Psychology of Fantasy GMs on Minor Leaguers

I am certain that you have seen all sorts of analyses on the failure of the first overall pick to turn into a star. There may be a reason for this that goes beyond analysis of the individual player, and implicates the decision making process that leads to the GM’s selection of the top overall pick.

The last twenty years has seen a proliferation of baseball analysts. There are certainly more draft and minor league experts around now than there were 20 years ago; and about this there is no serious debate. Normally, one would think that with more analysts comes more diversity of opinion. Common sense might lead one to this conclusion, and certainly conventional wisdom might lead one to this belief. But game theory offers a different perspective. Game theory, in a general sense, purports to study situations where multiple players or actors are involved in a strategic “game” where each seeks to maximize their return.

When faced with a large number of possible choices, what does one do?? To borrow from John Maynard Keynes, lets look at a beauty pageant, but lets assume there are hundreds of participants not just 50. The judges are trying to pick “the most beautiful.” But this is a loose standard, and is certainly not the same as picking the tallest or the one with the longest legs (desirable though those categories may be). So what occurs is that in an effort to cull through the masses, a focus is inevitably made on a distinguishing feature that separates one from another. What you end up with are models that are not “the most beautiful” but that have one feature that distinguishes them from the crowd. All of the contestants have some claim to “beauty” but how do we pick one over the other?

In baseball we have a similar situation when dealing with minor leaguers. The feature that draws focus is usually either fastball speed, college success or projectability, or some other factor. These factors may be correlated with major league success in a very general, non-specific way; however there is certainly no strong correlation. Picking the “best future player” is much more akin to picking “the most beautiful” than picking the tallest. Such uncertainty is fraught with peril. The fact that this one characteristic exists in a certain player not only draws attention to itself, but also may be the reason for success against inferior competition and will not hold up against more adaptable and talented opposition. This is a slender reed indeed upon which to base such a weighty decision.

The fact that there are more analysts than ever is also a significant variable; again looking through the prism of assuming that each is attempting to maximize their “return.” Much like stock analysts there are two general groups; the first group consists of those with a well-known reputation; their interest is in preserving their reputation. So if their opinion is vastly different from the consensus they better be correct. But the mere fact that their opinion on something as rigorously followed and publicized as the number one overall pick in the draft will be publicized means that there is less incentive to take a risk. So, for this group of analysts we see a “snowball” effect, and opinion will converge around one player so that each individual acting in his self interest winds up as part of the general consensus. This may (or may not) have little bearing on the actual skill of the player. If they are wrong they can comfort their detractors by pointing out that everyone else was also.

An example of this phenomenon is Tim Lincecum, who was regarded as possibly having the best overall stuff, but there were concerns about his size and workload. Given that there is extra perceived risk, by consensus, in taking this type of player he drops overall. No one wants to be the only guy who drafted or recommended Lincecum to go number one when no one else does and then see him get hurt or shellacked.

The second camp consists of exactly the type of analyst that WILL pick a Lincecum or a non-consensus pick. Generally this will be people looking to make a name for themselves or those outside the establishment. They may not be right, and probably won’t, but if they are they can make a name for themselves. Perhaps the greatest example of this was pollster John Zogby, who made a name for himself by predicting that the Clinton/Dole presidential election would be a lot closer than what the conventional polls predicted. He was right and has been living off of his one correct prediction for years.

Another factor is the fact that there is no benefit to be gained right now in being correct. But there is plenty of criticism that can be laid on the analyst or drafter right now. If the Rays drafted Matt Wieters first there would have been much obloquy cast upon them. So the GM who bypasses David Price (or Bryan Bullington or Brien Taylor) pays a current penalty now, namely a loss of reputation, if they are wrong. But the payoff to being right, if one even occurs, is two years away at least in most cases. By that time there are plenty more opportunities to prove oneself right, so there is little incentive other than pride or courage, to bucking the opinion of the masses of analysts.

So, if we look at things from the perspective of an individual GM who is acting to maximize their “return” we end up in a situation where the decision makers and analysts naturally become more risk-averse, and end up converging toward the conventional pick because they simply must start to separate the wheat from the chaff by distinguishing one player from another; generally by some defining characteristic. Of course, we do not know what is going on in an individual’s mind and this is the only way to really know what has happened. However, we can and do know how collections of people will act, be they baseball writers and analysts or not.

So what is the Fantasy GM to do? I am sure that many fantasy players are now curious about how or when they can acquire the recently drafted players, dreaming of getting next year’s Tim Lincecum or Chad Cordero. I once played in a league that allowed you to pick up any player at any time as long as they were under contract. That means that in that league you could pick up David Price as soon as he signed even though he never threw a pitch. In fact what happened was that the owners would snatch up these guys as soon as they were able. Occasionally this strategy worked, but more often it blew up in their face.

Even outside that league, too many fantasy GMs put undue emphasis on picking up minor leaguers. In one of my leagues this year, a 12 team mixed non-keeper league, head to head based on points, with weekly transactions. one of the owners stockpiled Phil Hughes, Yovani Gallardo, Homer Bailey, Tim Lincecum and Kevin Slowey, all in the late rounds of the draft. This type of league puts great emphasis on the two-start pitcher. So this owner has traded the potential of these pitchers for the actual, real return of two start pitchers in the first eight weeks. This is a disastrous outcome; and in fact this team is in last place.

Note that even in a keeper league this is usually a flawed strategy. In mixed leagues, most rookie performance is replaceable with a lot less risk. In AL or NL only leagues pursuing these guys as keepers is valuable, but even in these leagues it is of much less value than owners expect. Occasionally you will see an owner who stockpiles rookies win, but it is rare, and almost never happens against tougher competition. Why?? Fantasy baseball victory is about exploiting perception versus reality. You must simply be aware of this at all times, to fail to do see is the path to defeat. Minor leaguers are a prime area where the expert’s handling shines and others’ flounders.

Let’s take a look at Alex Gordon. Last year was my second year in my current high-stakes league. Year one I took over a disaster of a roster, and simply wrote the year off while I tried to rebuild. Year two I auctioned a group of minor leaguers, intending to trade most, and one of them was Alex Gordon.

Note that Gordon had no value to a team last year, aside from the potential to reap great profits this year. This is the kind of value that I am proud to let my competitors reap. The problem is that when a player is auctioned in a keeper league you generally can’t keep him forever (and if you can do this in your league these comments don’t apply).

From what I have seen, a reasonable projection consisted of a .280 BA, 20 HR 80 RBI 10 SB. This is a nice projection but note that it is laden with risk; Gordon is still a rookie and no matter how great his pedigree he probably had no more than a 50% chance to even be a league average player at his position as a rookie. Would you believe that last year Gordon netted me Justin Morneau in a trade? I still have Morneau in this high-stakes league while Gordon is struggling.

Gordon’s projection, even if he attained it, is replaceable in virtually all mixed leagues except the deepest mixed leagues, and in AL only leagues the perception of the value of this performance is much higher than it is actually worth in an auction. This allows you to trade up on almost all minor leaguers of sufficient reputation. You can get 90% of the projected production for probably 50% of the risk in virtually every 10 or 12 team mixed league. Yet, the perception of rookies is far greater than their usual performance. I know this first hand; as I was the proud owner of a $19 Jeremy Hermida last year.

Using Recurrent Neural Networks to Predict Player Performance
Technology is rapidly advancing possibilities in decision-making.

Exploiting the psychology that attaches to rookies is a key weapon in the expert’s arsenal. What is this psychology?? It is simply, that most players would rather prove themselves more knowledgeable than others by having the minor leaguer that becomes a star. Most owners will say that they value the return or payoff to their team, but if they did they would understand that most times you don’t get this payoff. You may find that this is what interests you. If so that is fine, but you will have a difficult road ahead against tough competitors and are likely to finish out of the serious money.

Print This Post

Comments are closed.