Playing For Keepsby Jonathan Halket
December 04, 2008
These past weeks, I’ve been writing about the trade-offs that a league should consider as it (re)designs its rules. Last time, I wrote about how reverse draft orders distorted incentives to compete at the end of the season and should not be used in most draft leagues. This week, I’m tackling trade-offs of luck versus skill and specifically about using keepers in your league. Keepers, if you’re not familiar with the term, are used in your fantasy league if not all players are available for draft or auction after the first year—some players are kept by owners (perhaps at a cost) and not released to the general pool.
Randomness is inherent in fantasy baseball, as it is in life. In a competitive league, it is hard to win without getting a little lucky—the winner will likely have rather more players exceeding expectations than under-performing. This is not to say that skill isn’t important. Indeed, they are substitutes: The more skillful you are relative to your opponents, the less luck you should need to win (on average).
As a commissioner of your league, you should seek to limit the role luck plays in determining the winner. You’ll never be able to remove it entirely, but a league where luck plays a large role is a league where your players can be inattentive or outright buffoons and still win. Plus, a ridiculously lucky winner leads to bitterness among the other owners.
Keeper rules increase the scope for both luck and skill in leagues, so I cannot offer a clear "thou shalt not." But I offer a warning: Be careful how you design your keeper rules and know your owners before you decide that keepers might be a sensible addition.
The upside of keepers is that they offer a richer set of components to a player’s value. Not only do projections of performance for this season matter, but you have to forecast future seasons as well. Keepers increase the value of younger players as their abilities are likely to increase by the most in the future. Keepers, therefore, reward owners with skills in forecasting ability.
Keepers also expand the possibility for end-of-season trades. Some way out-of-the-money owners can trade their "talent now" for players with "talent later." A third place team can mortgage its future a bit by selling a Matt Wieters for an A.J. Pierzynski upgrade at catcher this season. Furthermore, without keepers, there’s hardly any reason to look at minor league players in most leagues, so keepers add an element of "realism," too.
However, adding keepers means adding more luck to your league. To see this, consider a league with a deep minor league bench, one where you can keep through the off-season, say, 10 players who still have minor league eligibility—but where you don’t have a bunch of owners that follow the latest news from the fall and winter-ball leagues. In this case, each team will mostly have 8 or 9 players that the owner has no idea about. These unknowns may pay off big time or they may flame out. The flame-outs the owner can cut practically painlessly once the season starts but the superstar rookies may yield a ton of value that the owner received entirely due to luck.
More generally, most leagues with keepers require an owner to pay a cost in order to keep a player—in auction leagues, you have to pay a set (perhaps complicated) amount; in draft leagues, you’d have to give up a certain draft pick. So keeping a player is only smart if the cost of keeping him is lower than the player’s perceived value next year. Indeed such a system is nearly necessary if a league wants to have keepers.
As an example, an auction league may let you keep any free agent acquisition the following year for a price of $1 and any drafted player for $5 more than the price that you drafted him at this year.
I think these types of rules are incredibly dangerous. In essence, keepers end up being players that vastly outperform expectations (and are expected to go on outperforming in the subsequent year). For instance, a kept drafted player in the above auction league would have to outperform his value by $5.
I would propose that, in most leagues, owners obtain huge out-performers due to luck and not skill for two reasons. Firstly, it is hard to forecast such out-performance. I don’t have data to offer for this, though. Secondly, moreover, if everyone in the league forecasts that this player would outperform his value, than his price would obviously rise in an auction (or he would be drafter soon in a draft league), which would remove the player’s excess value. So, not only does an owner have to be able to forecast this out-performance, but he has to out-forecast (through skill and not through luck) his fellow owners.
So, if you’re in a league with a bunch of keen-eyed, skilled forecasters with sharply differing opinions and forecasting tools (see Derek Carty’s article yesterday for a possible example), then you may want to include keepers. Otherwise, you’re mostly rewarding luck with possibly huge windfalls.
A final, personal reason why I dislike keepers—I like the draft/auction part of fantasy. Why diminish it by removing a bunch of players from the table?
If you have a question for the Roster Doctor email here. Emails in simple text with players' full names properly spelled are much more likely to get responses. Also be sure to include your league's player pool (mixed, AL-only, NL-only), number of teams, scoring format (roto, head-to-head, points, etc.), categories, whether or not it's a keeper league, and any other pertinent information.
<< Return to Article