Upside Down

There’s been a lot of material on players with upside. Recognizing and properly valuing players with upside potential is a key to winning in fantasy. However, chasing upside is dangerous and, in a sophisticated league, does not come for free.

Take an abstract league and baseball world with no injuries and no match-up considerations. In this world, you’d start your 20 starters for as long as you don’t have a bench player worth replacing the starter with. Midway through the season you may feel that your starting catcher is putting in a worse season than your blossoming backup catcher and so you switch them, starting the latter. In this way, your starters are sort of like stocks in the stock market while your bench players are like call options.

When you own a stock, like Ford, you own a piece of the company’s performance. When it does better than expected, you make money or a “profit”. When it does worse than expected, the stock goes down relative to the price you bought it at, and you lose money. Players that you draft to be starters in fantasy are more or less the same.

A call option on Ford stock is different. You pay a much smaller price for the option than you would for the stock itself. In turn, an option protects you from the downside—you make money if Ford does well and lose nothing if it does poorly. There is no explicit downside risk from buying a call option. In this sense bench players are similar to call options—if they end up as stars, you get the upside by putting them into your starting lineup. If they stink, you can cut them at virtually no loss.

Of course, nothing is free—in the stock market, you do have to buy this option from someone willing to take the risk. In the fantasy world, the cost is the price of the player in an auction or the opportunity cost of not drafting a different player in a draft league.

Risk is a spread in potential performance. Increased risk means higher highs and lower lows in performance, keeping the same average or expected performance. Increased risk is mostly bad for starting players. Everything else equal (including average performance), you’d rather take the steadier player – this is one reason why the best pitchers are valued lower than the best hitters.

Increased risk may be very good for options and bench players though. If your starter gets a season ending injury rather than just a 15 game DL stint, that’s a big deal. If your borderline bench player does, you don’t care, as you’d probably drop him in either case. With bench players, you don’t care how low the low is, but a higher high is great. Upside players may be on average just as good as journeymen players, but you only care about the jackpots. So, for the same price and expected performances, you should take the player with upside over the steady journeyman.

The thing about options is that since no cares about how bad the downside is, all the option’s (or player’s) value comes from beliefs about how high the upside is. Consider these three players: Player A is your starter and is guaranteed to hit exactly 25 home runs; Player B is has a 90 percent chance of hitting exactly 10 home runs and a 10 percent chance of hitting 24; Player C has a 5 percent chance of hitting 30 home runs and 95 percent chance of hitting 10.

Should you draft player B in this simple world? No, since he has zero chance of out-performing your starter, Player A. Player C is the one with more upside value since he has a chance of being better than A, despite the fact that his average performance is worse than B’s. Note that all of this came only based on your expectations of the probability of relatively low-chance events. Small changes in expectations can cause large changes in value. So Upside-only players are the ones where inflated expectations can cause you to overpay by the most.

However in an auction or draft you’re going to a pay a price to buy this upside, frequently giving up the chance to buy a steady player with a better average outcome. This may or may not be a risk worth talking.

Allocating at least some roster spots for upside players is an important part of any winning strategy. In fact, one could argue that modern versions of the LIMA strategy are all about upside starting pitchers. Still, I want to caution against too much risk taking. Capturing upside performance in a bottle is tricky. Let’s consider one likely sources of upside: young players with little performance history.

Players with little or no major league experience present two problems. First, their upside potential is much more uncertain and so harder to value. Rookies are the stuff that dreams are made on for fantasy owners. But the low-chance events that upside-rookies’ values are based on are the ones that are hardest to learn about with small samples of data.

Second, during the season you are also dealing with small samples of data. Say your rookie, bench outfielder has a hot April and one of your starting outfielders is struggling. Do you pull the switch in a weekly league where you’ll be stuck with your choice for at least a week? Many players revert to their average, expected performance. This is one of the toughest decisions an owner has to make and chances are you’re not going to get the timing just right. That’s fine, but when you’re valuing these upside players, you should adjust accordingly for the timing mistakes.

Print Friendly
« Previous: The root (Part 1)
Next: Today at THT »

Comments

  1. Beanster said...

    Great article.  I’ve been struggling with this question because unlike the real world where call options cost a fraction of the underlying stock, here you’re paying for the stock itself even though I fully agree with the option analogy.  For a $1 player you can stash on the bench it’s an easy call but how do I value, say, Matt Wieters or Travis Hafner?

    As a result I’ve been trying to approaching this from the opposite direction: if I can properly value each bench and DL spot, factoring in needs and risks of my starting lineup, I could theoretically determine how much “capital” to allot to high upside players.  In practice this has typically meant being outbid for the next Evan Longoria, which is fine but I’d prefer to have a better handle on whether I have a good valuation method or if I’m being too conservative.

  2. Wickethewok said...

    Well, I think the big issue is how you know whether and when you are hitting that 5% 30 HR upside.  A mark against high upside/low probability players is that if you want the upside, you have to play them.  If they perform poorly, that is still a big deal.  Just because they cost little to get doesn’t they can’t still cost you in terms of playing time.

  3. Todd said...

    “Player A is your starter and is guaranteed to hit exactly 25 home runs; Player B is has a 90 percent chance of hitting exactly 10 home runs and a 10 percent chance of hitting 24; Player C has a 5 percent chance of hitting 30 home runs and 95 percent chance of hitting 10.

    Should you draft player B in this simple world? No, since he has zero chance of out-performing your starter, Player A. Player C is the one with more upside value since he has a chance of being better than A, despite the fact that his average performance is worse than B’s.”

    That was exactly what I was thinking when I drafted Tom Brady in the early 2nd round and waited until my final pick to grab Brady Quinn as his backup during my football draft last year. You can guess how well that turned out…

    I actually think Player C is the better option, simply for the fact that the majority of Player C types will go undrafted and can be picked up in mid-April should that 5% potential become reality. Do you really want to leave yourself open to the possibility of being forced to actually play Player C, who has a 95% chance of being a detriment to your team? To borrow a term from poker, these are not good “pot odds.” There is a greater than 5% chance that this blows up in your face with the starter getting injured than of Player C blowing up.

  4. http://www.myscarygames.com/profile/view/609496.ht said...

    WONDERFUL Publish.thank you for share..much more wait around .. …

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Current day month ye@r *