Almost every bad fantasy trade comes from counting the wrong thing. Managers add up the projected points on each side, see that their side is larger, and accept. But points on your bench are not points. The only currency that settles a matchup is points in your starting lineup, and that constraint changes which trades are good.
The unit of analysis
When you trade two players for one, you do not simply swap production. You vacate a starting slot, and something has to fill it — usually the best player left on your bench. The correct comparison is:
after = incoming player + the bench player promoted into the vacated slot
before = the two players you gave up
gain = after − before
Which means a 2-for-1 is good exactly when the incoming player beats the worse of your two outgoing players by more than that player beats your best bench option. Put like that, the reason most 2-for-1s disappoint is obvious: the gap being bridged is not between the star and your second-best player, it is between your second-best player and your bench.
When consolidating is right
Consolidation — trading quantity for quality — works under conditions that are specific rather than general:
- Your bench is strong relative to your starting slots. If the player promoted is nearly as good as the one leaving, the cost of the trade is small and the upgrade at the top is nearly free.
- The value curve is steep at the position you are buying. Two players a little above replacement level are worth less than one player far above it, because value is measured from the baseline and you only get to count the baseline subtraction once per starting slot.
- You are contending and want to reduce the number of ways the season can go wrong. Concentrated rosters have fewer weak slots.
And it is wrong when your bench is bare, when the position you are buying has a flat curve, or when you are behind and would be better off with the wider distribution that a broad roster produces.
The traps
Sunk cost on draft capital. What you paid for a player is not evidence about his value. It was your estimate, made with less information than you have now, and the market has probably updated. Managers hold second-round picks far past the point of reason and drop late-round picks far too readily, for reasons that have nothing to do with production.
Endowment. The same player is systematically worth more to the manager who owns him. It is one of the most reliably reproduced findings in behavioural economics and it appears in fantasy leagues in the purest form imaginable: offers that seem obviously insufficient in one direction seem obviously generous when reversed. A useful test — if the trade were offered to you the other way round, would you take it? If not, and you also would not take it as offered, one of the two judgements is wrong.
Recency. Three good weeks feel like new information and are mostly noise. In a sport with this much weekly variance, a short hot streak barely moves a well-formed prior. The manager selling a slow starter and the manager buying a hot one are usually both making the same mistake in opposite directions, which is why those trades get agreed so easily.
Name brand. Reputation lags production in both directions, and it lags longest for players who were excellent two or three seasons ago. Trading for the name is paying for a projection somebody else made a while back.
Trading to fix a losing week. The impulse to act after a bad Sunday is strong and almost always expensive, because it puts you in the market as a motivated buyer against someone who is not. If you would not have made the offer on Thursday, do not make it on Monday.
Treating a value chart as a valuation. Trade value charts are averaged across formats. Your league has its own scoring, its own slots and its own baselines, and the chart knows about none of them. Charts are a starting point for negotiation, not an answer.
A workable process
Project the rest of the season for every player involved — rest of season, not season totals, since games already played are not on offer. Work out your starting lineup before and after, filling any vacated slot with the actual player who would fill it. Compare the totals. Then check whether the change in variance suits your position in the standings, because a contender and a team chasing a playoff spot should evaluate the identical trade differently and correctly reach different answers.
Then, before sending it: ask what the other manager thinks he is getting. If you cannot construct a version of the trade that looks good from their side, either you have missed something or they will decline. Both are worth knowing before you press send.