keyboard_arrow_uptop

I was originally going to write this article on the impact of the winner’s curse on FAAB bidding, but in Google-checking my ideas before beginning to write them up, I came across this Erik Siegrist article. Am I the biggest fan of the name Eric(k) ending with a K? No. Is the article excellent and a must-read for those in keeper leagues? Absolutely. Does the article’s excellence make me feel self-conscious as a fantasy baseball writer? Yup (he pretty much wrote the article I wanted to write and wrote it better than I imagined I would). Anyhow, go read it.

The part I wish to expand on is midseason strategy. Come midseason, the FAAB market in keeper leagues tends to be markedly different than the market in the beginning of the season. Players who can help teams in the hunt almost always go above “keeper” value, leaving only players who will be of seemingly no use this year to potentially go at value or at a bargain. We have seen before that at this time of the year, when the goals of owners in keeper leagues diverge, discussion of strategy is important. Consequently, we will briefly discuss the winner’s curse and how that should impact the decision-making of owners depending on their competitive position.

Just in case you chose not to read Siegrist’s article, I will quickly introduce the winner’s curse. The winner’s curse is a phenomenon in which the highest bidder (winner) tends to overpay. Because crowd-sourced valuations tend to be more accurate than single-sourced valuations, the highest valuation in a set of valuations will tend to overvalue the subject asset. In fantasy baseball, this obviously means that the winning bidder will tend to overpay. Siegrist astutely points out that the extent of the impact of the winner’s curse is dependent on the number of teams bidding on a player. He writes,

“Winner’s curse has a stronger impact when more people are involved in the auction. The larger the pool of bidders, the more likely it is that the winning outlier bid is going to overshoot the true value of the player being bought, and by a wider margin. Conversely, the fewer people in on the bidding, the better the odds that you actually get a bargain, or at least acquire the player at or about his true value.”

The way in which we behave regarding mid-season FAAB bidding comes down to divergent goals. If we are trying to improve in the standings, then getting appropriate value per keeper price is not as important as getting the most production possible. If we are trying to improve our chances of winning next year, then the opposite is true. Most leagues already follow this phenomenon. By this time in the season (and probably even a month earlier) any useful or potentially useful player goes above keeper value as the teams competing are looking to improve their squad for this year only and are thus willing to pay higher prices. This phenomenon is compounded in fantasy baseball because winning a player not only improves your team; it also disallows (to an extent) other owners from improving their teams. We know who these players are: the productive player just traded from one league to another (Jeff Samardzjja, Jason Hammel), the un-owned recently promoted closer (Joe Smith), the un-owned called-up minor leaguer (Tommy La Stella, Jimmy Nelson). Again, this is what teams in the hunt should be doing, determining bid amounts through a this-season-production perspective rather than a keeper perspective. One other factor that these teams should be weighing is opportunity cost. Yes, Robbie Ray, while in the majors, is going to be an upgrade over Edward Mujica, especially when chasing strikeouts and wins. However, if paying for Ray cost you the opportunity to acquire Samardzjja or Hammel, then you are now most likely kicking yourself. To conclude, when bidding while looking to improve in the standing: (i) bid according to production rather than keeper values and (ii) consider opportunity cost.

That said, what are the rebuilding teams supposed to do? Rebuilding teams are going to want to target players that either (i) possess trade value or (ii) will have few bidders and thus have a chance of being a keeper the following year. Targeting players that possess trade value is great in theory, but is very league rule dependent. In a couple of the keeper leagues I play in, an owner cannot trade or cut a player acquired for $25+. Consequently, most players worth trading go for $25 or more at this time of the year in FAAB (there is a greater opportunity to grab players for the purposes of trading in the beginning of the season, before owners have truly set their sights on the finish line). If your keeper league does not include such a rule, then by all means grab players to trade if you can. The reason this advice applies to future minded teams and not this-season minded teams is because this-season minded teams will rarely have anyone to trade overpriced players to; future minded teams will not be interested in non-keepers (due to price) and they will have difficulty making trades with their direct competitors. Similarly, a future minded team will be able to trade such players to teams that are going for it.

The second group of players to target for rebuilding teams is the most practical, that being players that have a chance of being a keeper the following year. Assuming a non-shallow league, these players generally can be described as follows: they have played terribly this year. These are players that good teams cannot touch because they are likely to give them little in counting stats and to hurt them in ratios and averages. This immediately lowers the number of bidders, which lowers the impact of the winner’s curse. Worthwhile question: why do we want to pick up players who have played terribly? Just because a player has been terrible, does not mean he will continue to be terrible going forward. In deep leagues, all a player needs to build some trade value is a little bit of playing time, a couple “deep sleeper” predictions, and a couple of good weeks in spring training. Example time:

In my NL-only keeper league, Jace Peterson was available after his first call-up in June. He had not yet been verifiably terrible; thus, there were a handful of bids on him and he was won for an expensive $14. After being terrible, Peterson was released by the winning bidder (winner’s curse). Peterson was again available in the player pool after his call up last Saturday, but because he was terrible only two teams bid on him and he was won for a cool $6 ($1 greater than our $5 minimum bid).

Peterson is not a great player and will probably not hold any fantasy value in the future, but that is exactly why he is cheap. There is some small chance that he starts somewhere in the middle infield for the Padres next season and that makes him more valuable than rostering an untradeable asset such as a $10 Mark Ellis. If you can acquire a portfolio of these players throughout your rebuild, then you increase your chances of finding yourself some value that would not have come from having untradeable, over-priced players on your roster. Therefore, look for opportunities where the FAAB bids will be limited and go get yourself some potential future value.

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
You need to be logged in to comment. Login or Subscribe
Dodger300
7/18
Except at the end game, when other teams' rosters are complete or money is expended to the point that they can only make minimum bids, won't the winner's curse affect every player taken, at least to some degree?

In other words, it is already baked into the cake.

If we go to, say, an estate auction where a few parties are capable of buying up everything of value while the rest go home empty-handed, the winner's curse is certainly at play. The all-or-nothing result amps it up.

However, a Fantasy auction is more of a zero-sum game, because a person who spends heavily for one piece has fewer resources left to spend on the next pieces, resulting in a more level playing field. In other words, in the end everyone has to buy the same number of pieces, even in the same configuration of roles, as the next guy.

Since no one can dominate the spending at very many positions, the top end (where the winner's circle effect is most prominent), is more or less spread across the league. Thus, the effect of the winner's curse becomes by and large muted.
craneplace
7/18
Yes. To quote you "the winner's curse will affect every winner taken, at least to a degree." That said, the degree matters, and it often matters a lot. Depending on circumstances at the time of the bid, the winner's curse can play anywhere from a huge to a non existent role.

I don't think your response is contesting the above, but I would challenge the assumption (and this is an impotant distinction) - as far as FAAB goes - that "everyone has to buy the same number of pieces, even in the same configuration of roles". Teams are often left with leftover FAAB or left spending a lot (inflation) on players that would have gone cheaper earlier if only because it is the last opportunity to spend FAAB. That said, I think you are dead on when it comes to the actual auction. Combined with the fact that teams all don't put in their final bid (teams are often making/not making bids with future bids in mind) for each player, the affect of the winner's curse is distorted. I certainly hope that I did not indicate that the winner's curse extends to the actual auction; rather, I was applying the phenomenon to the FAAB process, the mid season FAAB process specifically.
Dodger300
7/18
Your distinction between the effect that the winner's curse has concerning the FAAB process vs. the pre-season auction is helpful. Clearly, I was conflating the two.

Thank you for the clarification.
craneplace
7/18
To clarify the last point, FAAB bids are also made with future bids in mind, but auction bids are often not a final assessment of a player. Because you can bid X knowing someone will then bid X + 1, X + 2, etc., and that you could even bid X + Y, a bid in the preseason player auction is often not as close to our actual valuation of a specific players as an FAAB bid.
craneplace
7/18
Thanks and anytime. I actually contemplated writing a "winner's curse and preseason auction" piece this offseason before going through this mental exercise with myself. I still think there is something there, but I need to think it through more. Your comment has actually helped me to do so a little. That said, thank you!
craneplace
7/18
Note that this is in response to Dodger300's second comment, not mine.
jonjacoby
7/18
Jeff, what is the purpose of your league rule that doesn't allow $25+ FAAB players to be traded? I've heard of other leagues that do it, but never understood what behavior the rule is meant to curb/eliminate. If anything it seems to hurt teams in the bottom of league that are trying to rebuild.
craneplace
7/24
I'm not really sure to be honest. In ways it can help the teams on the bottom, especially because players acquired for 25 or more cannot be dropped. So when a crappy closer becomes available, a whole bunch of teams will bid 24, the tiebreaker awards the player to the team with the worst record, and then that team can trade that asset. Ican definitely see the negatives around it though too.