On trading

The reasons for trading can vary in a few ways. You can trade because you have a surplus, or you can trade because you have a need. Further, that need/surplus can either be at a specific position or (in a roto/category league) at a statistic. It's nice to be able to make a trade because of a surplus, but it's usually wiser to make a trade because of a need. If you're trading for a need, then you have a clear idea of what you need to target. If, for example, you don't have a lot of steals then you know to target a player who steals bases. If you need strikeouts, then you know to trade for a pitcher with a good strikeout rate. Or, alternatively, if you're getting little production from third base, you know to target a third baseman. If you're trading because of a surplus, you're not necessarily starting with a clear idea of what you want to get back. That's a recipe for disaster. You could end up trading a 30-steal player because you think you have enough steals, only to get a .270/80/80/15/10 guy back who isn't going to boost any one category enough to matter. You've given away your surplus for, effectively, nothing. After a bevy of trades last year and only a 50/50 success rate, I've come up with a few guidelines (not quite rules) I like to follow when trading.

1) In general, don't make a trade in April. Patience is a virtue all to absent in fantasy baseball. Those hot and cold starts are coming from a small sample size, and it's generally to early to know if those starts are an anomaly or a trend. For example, the mean batting average right now (end of play on 4/6/11) the average number of at-bats is 19.64 for the top 200 hitters. After rounding up to 20 for the sake of simplicity, someone who's "real" batting average is .300 should have 6 hits by now. However, if he's getting off to a slow start he may only have three hits (.150 average) by this point instead. That's not a significant difference. It's perfectly within the realm of normal variation to have a batting average half of what it "should" be this early in the season; it doesn't even mean he's in a slump. By the same token, someone who "should" hit .265 for the season could have an average of .500 at this point and still be in the realm of normal deviance from his season average. It's simply too soon to tell at this point how players are going to perform. Similarly, we don't know who is going to be injured at this point. You may have what looks like eight viable outfielders, but one centerfield collision and one pitch to the hand could fill up your DL, put a dead spot on your bench, and leave you scrambling to fill out a roster on Monday/Thursday when there are fewer games being played. This is always a risk, but it's especially true early when players haven't had much time to get injured.

There are, I think, two exceptions to this. The first is if you have the opportunity to buy low or sell high on a player. Some owners are impatient or swayed by small sample sizes, and that gives you the opportunity to get what is clearly excess value back in a trade. If someone is so enamored of Rickie Weeks' hot start that he'll fetch you Ichiro Suzuki, do it. For example, I noticed in one league an owner is attempting to remake his pitching staff (he added four new starters, and has no elite pitchers). He's also impatient, having already dropped Stephen Drew. Since I have an excess of pitching, I'm currently attempting to pry away the injured Longoria by dangling extra pitchers and Rollins for Longoria and a little extra. This only works because a) I have excess pitching, b) I'm only offering one of my top four starters, and c) I can replace Rollins with the aforementioned Drew. The other exception is if you purposely came out of the draft with an area of need. If you waited until the 22nd round to take a 3B, then 3B is almost certainly enough of a need to make a move if the opportunity presents itself.

2) When intiating a trade, target a statistic of need. Targeting a statistic allows you to quickly narrow down the players you want. You have to make sure you can start them, of course, but deciding you need steals allows you to set up a list of excellent base-stealers to work from. If they're not on the list, they're not worth considering. If you say "I need an outfielder" instead, you have roughly 100 players whose quality and contributions vary greatly. You may not end up getting what you need in return because you're deciding between Juan Pierre and Mike Stanton.

3) When filling out your end of a trade, pull from a position and not a statistic. The number one influence on fantasy is that you have specific constraints as to who you can play any given day. If you can, it's better to trade from a position where you have more players than you can start, or at least a surplus of players given the constraints. For example, you may be able to play three 3B (3B, CI, Util), but you only need one or two (one for the 3B spot, one for either CI or UTIL). It's going to be difficult for a positional surplus player to crack your lineup often, so trading from there means giving up the least production. If you have four 2B, one of them is likely always on your bench. That's (barring injury) a wasted spot, and can be better used by trading one of those players). Pulling from a statistic can lead to an unexpected shortage. Early last season, it looked like I was going to have surplus power and saves, so I traded those statistics for ones I needed. I ended up scrambling for home runs, and had to take on more batter Ks in the process. I also ended up having to use additional roster spots on temporary closers (such as Jose Contreras) to try and make up saves. It worked, but it meant I couldn't make other in-season adjustments easily.

4) Always walk through the ramifications of a trade. In many (if not most) trades, there will be a significant change to your roster that go beyond the players involved. If you have to drop or add a player to make a trade happen, then that player is effectively part of the deal. This must be accounted for. In one of my leagues I was recently offered Weeks, Ike Davis and Volquez for Robinson Cano (I didn't come close to taking it). In this case, I have to do more than subtract Cano, and add in Weeks/Davis/Volquez. I also have to consider which two players I would have to drop, and whether or not I liked those players more or less than Davis and Volquez. In this particular case, I'd have to consider if Davis was going to be more productive than Butler (my 1B) or Beltre (my Util), and the answer is really neither. In that case, Davis would simply be a bench player, and then I have to consider if I like him more than Brian Roberts (no), Logan Morrison (no), Miguel Tejada (yes, but Stephen Drew is out), or Russell Martin (yes, but I want to hold onto him while he's hot). I then have to go through a similar calculus with Volquez. Similarly, and whether or not there are drop/add/roster change considerations, a trade is always about marginal value. If I trade a .290/90/100/30/5 player for a .300/100/70/10/30 player, I'm not gaining .300/100/70/10/30. I'm gaining .010/10/-30/-20/25. That's the difference between the two players. If I'm trading (for example) Tim Hudson and Hunter Pence for Matt Holliday, I have to consider whether the difference in statistics between Pence and Holliday is worth the difference in statistics between Tim Hudson and whomever I replace him with. The real cost/benefit isn't Holliday's stats, or even Holliday's vs. Pence's, it's the difference between statistics of all the players involved.