Just about every year, one of my high school students asks me to be a faculty sponsor for a team that will participate in a stock market contest. I always refuse. My problem is that the strategy for winning the contest is the opposite of what I would recommend for a real-world investor.

A real-world investor should try to more or less match the market. But if you want to win the contest, you have to do much better than the average investor, which means you need a strategy that makes outrageous bets. I think if you entered me in the contest, I would put all of my money on out-of-the-money put and call options on the S&P 500. In the real world, you figure to lose all your money that way. But in a contest, if the market has either a good run or a bad run, you will win the contest. I am not saying that my strategy is absolutely the best for the contest, but I think it makes sense.

Anyway, speaking of such contests, this it the time of year for fantasy baseball. My thoughts below the fold.

1. I prefer auctions to drafts. I think that the auction poses more strategic opportunities and problems.

2. Here are some things that I am pretty sure are true:

a. Luck matters a lot. Often, key categories are ridiculously close, so the outcome can depend on relatively small factors. For example, in an auction in which I just participated on the Yahoo! site, if you use the Yahoo! projections, then 8 of the 12 teams will have between 220 and 232 home runs. Any confidence interval around the projections would be much, much wider than that.

b. Luck matters even more in shallow leagues, meaning that the ratio of fantasy owners to eligible major league teams is low. If you have 12 owners, all teams from both major leagues are available, and only 18 players’ statistics count, it is shallow. If you have 20 owners and 23 players’ statistics count, it is deep. If you have 12 owners and only the national league teams are chosen from, it is deep.

c. There is no unique optimal strategy, because if most owners follow strategy X, they will drive up the prices for players who fit that strategy and divide up the benefits of that strategy. In that case, your best chance is not to follow strategy X, even if strategy X makes great sense in theory.

d. The roster rules matter. For example, in Yahoo leagues, a batter on your bench has no value. However, with pitchers, you can rotate in pitchers who are more likely to appear in any given week, so pitchers on your bench have more value. This goes even more strongly for leagues that allow daily rotation of players.

e. The scoring rules matter. If a league has the usual 5 pitching categories but adds holds, then this totally alters the relative value of middle relievers and closers. Saves are diluted in value to the point where closers are nearly worthless. On the other hand, good middle relievers who would be worthless in standard scoring now become worth more than most closers.

3. Here are some opinions about which I am less certain:

a. Just as in a stock market game, a good strategy for maximizing your average result is more likely to help you finish 4th or 5th than finish 1st. In some sense, I would rather aim for 4th than aim for 1st.

b. I suspect that aiming for 1st you might buy players who the rest of the owners stay away from because of age, injury risk, or recent bad performance. Note that some players get bid up in spite of those factors–that is not what I am talking about. I am talking about players that nobody feels sorry about not owning. The pitcher who everybody thinks is over the hill, the hitter who has not played more than half a season in five years, etc. This year, Sabathia and Sizemore would fit that. I know I would be glad to let you have them.

c. You should go into an auction with a presumed value for every player, not just a few players that you have your eye on. You can find values at various fantasy baseball sites, which can make a reasonable background. During the auction, keep “score” of how many dollars over or under these values the players go for. Most of the time, early in the auction players go for more than reasonable value, which means that bargains who up later. I call this an “overbid auction.” In fact, the Yahoo “default values” generally reflect an overbid auction. Almost all the really top players are valued (even in terms of the Yahoo projected values) too high, in my opinion.

d. My main strategy assumes an overbid auction. I try to pick up very few players early in the auction and instead pick up most of my players in the late middle part of the auction. This strategy works really well in a deep auction. That is because in a deep auction there is more compression in value between the top and middle of player quality than there is between the middle and the bottom. For most of the auction, I nominate players that I do not want, in order to see other owners’ money taken off the table. For example, if I happen to land a catcher early, I proceed to nominate catchers for several rounds.

e. In a shallow auction, the overbid strategy does not work as well. That is because there is a lot of compression in value between the middle players and the bottom players (the difference between the 30th best outfielder and the 60th best outfielder is not so great). So you have to suck it up and overbid for some players. I prefer not to overbid for corner infielders or starting pitchers, because I think there is so much value in the middle at those positions. So if I see a real overbid auction developing, I would want to overpay at catcher, second base, and shortstop.

f. Closers represent an interesting issue. One strategy is to get lots of top-quality closers and a couple good starters. You hope to dominate three categories–saves, WHIP, and ERA, and just deal with what happens in the other two pitching categories. I understand the logic of this, but it is not what I do. I do the opposite–I prefer a portfolio of lesser closers for saves, and a lot of high-middle starters. But see point 2c.