Wednesday, January 18, 2012
Can you really play it safe?Posted by Derek Ambrosino at 5:30am
It's well established that humans are poor at assessing risk; we constantly think actions that are, statistically-speaking, quite dangerous are safe and that acts that are actually quite safe to be dangerous. To borrow example from the Steves of Freakonomics, if your household has both a gun and a swimming pool, it is far more likely a child will die in the pool than by the gun.
In Ben Pritchett’s recent column about Mike Stanton, he mentioned the following:
According to the 2012 Baseball Forecaster, 80 percent of all players that will yield first-round value are found in the first four rounds of drafts. Less than half of the players drafted in the first round actually yield first-round value…Assuming that just five players picked in the first round actually retain first-round value, that leaves ten draft positions available to be filled by players drafted in the next three rounds. Using the 80 percent success rate discovered by the Forecaster guys, that means only seven of the remaining 45 possible players will deliver first-round-caliber stats.
This comment led to a brief discussion in the comment section about "playing it safe" when drafting. Commenter Simon, posed the question:
Perhaps one way of looking at the Baseball Forecaster stats suggests that playing safe in the first 4 rounds is not really the exact science we all like to think it is!?
Ben agreed, responding thusly,
Playing it safe is all relative. There’s really no such thing as safe in this crazy game we play. I would say there are players that can hurt you more than others, but that’s honestly relative as well.
I had actually been having a few thoughts of my own on this line and I want to offer some insights on this issue.
I don’t want to bog you down with more sob stories about my teams, but long story short, some of you might remember that I’m the proud keeper league owner of both Ryan Braun and Ryan Howard. Yeah, well, my fifth keeper in that league also happened to be Victor Martinez.
Basically, I was asking myself whether I could find five players on my roster reasonably capable of putting up first-through-fifth-round value in light of this most recent blow. And, my answer was yes. Troy Tulowitzki, Nelson Cruz, Jered Weaver, Drew Stubbs, and a partial season from Howard or Braun could conceivably yield value congruent with the rounds in which I was to keep those players. It’s not the most likely scenario, but it is certainly possible.
Every year, stud players suffer injury, players who were expected to make a leap flop, players written off as mediocre surge, outlying BABIPs, strand rates, or RBI opportunities lead to good players having fantastic seasons and great players merely having good seasons. Stuff happens.
In one respect, I agree with Ben and Simon, but at the same time it isn’t nearly that simple, though I’m not implying either of them made such the claim. When you are assembling a fantasy team you are accruing a portfolio of assets. These assets have different levels of risk and ranges of projected performance—in addition to injury risk. Some players have a history of less variance in performance, while others have higher ceilings, lower floors and less stability.
When you’re making your first four picks—and the rest of them—you are creating a complex matrix of probabilities, and not everybody builds their team the same way. As is the case with individual players, some teams embrace wider variance in performance with a boom-or-bust strategy while others focus on building a team to be “in the hunt” and then hoping to make a few key moves from there.
When advocates of early-round conservatism preach their philosophy, they are simply saying they prefer to take the known B+ student instead of the prodigy who also has a history of not showing up to class. The idea that more than half of the top 15 players each season will emerge from picks 16 to 60 is only tangentially relevant to the counterargument of early-round conservatism.
The reason why expanding your appetite to take on variance works so well in the later rounds is that the opportunity cost of a botched pick gets progressively lower throughout the draft/auction. Therefore, the relative value of a home run pick is also greater.
Selecting a player at pick 40 who winds up having top-15 value does not guarantee that I will have gotten value from the top end of my draft (though getting one top-75 player past round 20 pretty much ensures I got value from the bottom of my draft). I could easily botch a second pick as significantly as I profited on the first.
The underlying premise of early-round conservatism is that your top picks are not when you should be looking for value because with the opportunity for value comes added risk. And since the risk quotient for each pick/buy increases by almost unilaterally by nature as the player pool gets thinner, adding risk at the top of the draft can lead to compiling a team that is excessively volatile.
Early-round conservatives don’t contest the idea that there is a degree of risk with every player and that, anecdotally, allegedly safe top-round picks bust every year. However, the issue is about risk management: How much of your bankroll do you want to bet on the first few hands of the night? Do you want to throw away the pair of queens to chase the flush? Owners must make these decisions on their own and in consideration of their own preferred strategies and the settings of their particular league.
Here are two final thoughts. First, if you have a strong feeling about a player that you think is based on sound reasoning, act on it. You shouldn’t be unwilling to take a risk in the early rounds if it is one you believe in, but there are times when risks seem unnecessary. Stephen Strasburg could have a better season than Clayton Kershaw next year, but what sort of premium would you have to pay above Strasburg’s “normal” price to take him ahead of Kershaw, and what would be gained over Kershaw if he hit his 95th percentile season and Kershaw his 88th? Simply, is the risk worth the reward?
Second, at certain times, it may make sense to embrace the added risk in the grand scheme of things. The best players in the league have the lowest chance of busting, not just the highest chances of excelling. I’m reluctant, still, to trade Tulo to fortify the back end of my keeper roster because if I exchange Tulo for two lesser players in the 35-50 range, I’m also taking on additional risk.
There’s some probability that Stubbs reaches his full potential and has breakout year. There’s also a very high probability Tulo is one of those top 15 players. If I exchange that probability for a marginally increased probability out of Stubbs and a second player with similar probability at being top 50 (and a lesser probability of being top 10), in a very important way, I’m actually adding more risk to my portfolio. I would need one of my players to overperform to get my one top-15 player, and there would still be considerable risk that my back-end keepers don’t earn their draft price.
To switch sports for the closing analogy, I don’t want to pass up an open foul-line jumper for the chance to take two contested three-pointers. But, in the right context, a three-pointer can be a pretty sound percentage play as well.
Derek Ambrosino aspires to one day, like Dan Quisenberry, find a delivery in his flaw, you can send him questions, comments, or suggestions at digglahhh AT yahoo DOT com.