Factors in evaluating trades

[reenum]

What sorts of factors do you guys look at when evaluating trades?

When I get a trade, I look at the past performance of the players, the likelihood of continuing success, and the relative values. These criteria are mostly subjective, and I'd like to get a more objective methodology in place.

So, if you have any stats or other metrics that have been successful in helping you evaluate players, then please share them.

[StlSluggers]

This is a loaded question, isn't it?

[Roger LaFrancois]

What would be cool would be if someone could create a simulator.

You would have to provide (1) your current roster (2) your traded for roster. The simulator could play the schedule say 100 times and come out with expected rankings either way.

Trades like 2 for 1 would be tricky because you would have to account for players waived and/or picked up. Plus it would have to be customizable for different league metrics.

Meh that would be quite complex. But it would be cool.

[StlSluggers]

There have been attempts at sites like that in the past.

They fall massively short in any keeper/contract league environment where simple production doesn't matter.

Also, the difference between roto and H2H has to be taken into account with extraordinarily streaky players.

[great gretzky]

Well before I worry about metrics and stuff too much, I consider the approximate player tiers for the guys in question. I know this sounds dumb, but I think of it like chess, where you don't get too in depth right off the bat when its your move. You consider your possible moves and whittle down until you have serious contenders, then do the extrapolation.

I say that to say this. I come from a different cut of the cloth than a lot here, but I really place a lot of emphasis on league economics. I.E. you can argue all you want that you "should be able to get player x, y,z" but distribution is similar in the fantasy world, but differs enough that you might have to overpay to get a certain player, or conversely the would be partner wants too big of a price back relative to values.

So when I offer a trade or look at one offered to me, I evaluate the type of trade in general -- power for speed, or lumber for arms. Then I think about the tiers of players that should be involved. Abject stud, b tob plus talent, spare parts, speculative plays, and then the junk transaction. To me it sounds obvious, but I frequently find that behavior doesn't reflect that. Then I scour the rosters and see what other teams could sync up in a similar way, and how players make up the unique universe of that league. And with that, I get a feel for how much I have to pay to get a need. A lot of times I find that you can run projections all you want and try to trade based off of that, but many times you only have a few potential partners for any given type of trade.

So I will let the stat guys give you in-depth look at metrics, but I generally only get involved in that if someone is either really green, or outlying what we generally conclude about said player. When we are dealing with established guys, I prefer to rely on the soft skills of trading and big picture. It doesn't bother me if I don't extract every last ounce of blood in a trade, if it adequately improves me where I need, and doesn't devastate the opportunity cost of player value. (Like if you trade a first rounder, if there are a couple of teams with a couple of player packages that would be satisfactory, I don't drive such a hard bargain that you sully the field. Maybe I'm a fool, but I have generally had good results with taking an overall stock, and working down from there, then relying on the more advanced stuff to break a hard decision, or shed light on a player i'm not overly familiar with.

[Dan Lambskin]

What would be cool would be if someone could create a simulator.

You would have to provide (1) your current roster (2) your traded for roster. The simulator could play the schedule say 100 times and come out with expected rankings either way.

Trades like 2 for 1 would be tricky because you would have to account for players waived and/or picked up. Plus it would have to be customizable for different league metrics.

Meh that would be quite complex. But it would be cool.

Sportsline has a crude lineup analysis tool when evaluating trades