Dynasty Disagreements, August 7

10 players we rank far above the dynasty consensus, 10 far below.

Our dynasty board against the dynasty consensus, across the 129 players both rank. Read this one carefully, because the two are not trying to produce the same number: our dynasty rank is a three-year production forecast, while a consensus dynasty rank is a market price that also carries contract situation, prospect romance and name value. A gap is often the two answering different questions rather than either being wrong.

We are higher

Our three-year forecast likes these more than the market does.

We are lower

The market is paying more than our forecast supports. Worth knowing that our dynasty board runs colder than the consensus overall, the median player here is ranked 34 places higher by the market, so this side of the list is longer by construction, not by conviction.

What this means

Start with Max Clark, because he is the cleanest case of a gap that is not an argument. Consensus has him 61st; our dynasty forecast has him 662nd. Nobody at either end thinks he is going to hit 30 home runs next April. The market is paying for the right to own six cheap years that begin at some point in the future, and our number is a three-year production total that mostly runs out before those years start. Colt Emerson works the same way, 162 against our 478. These are not opinions in conflict. They are answers to two different questions, and if your league pays for future surplus rather than counting stats through the window, you should ignore our rank on both and move on.

The genuine fights are the arms with track records, because there the production question is the only question. Strider at consensus 62 and ours 483 is the one I would argue about. A three-year forecast that buries a pitcher that far is not saying he is bad, it is saying the chance he throws a full, healthy, dominant three seasons is small enough that the average outcome collapses. Ragans, at 54 against our 365, carries the same shape. Schwellenbach at 69 against 497 is the harshest of the three, and that one I am less comfortable with: the market is buying an arm with real big-league performance behind it, and our number reads like a durability discount applied with a very heavy hand.

Here is the case against me, made properly. Pitcher forecasts that lean on health history have a bad habit of double-counting: the injury already showed up in the innings total, and then it gets charged again through the rate stats. Three of our ten lowest arms sit inside the consensus top 70, which is not a coincidence, it is a policy. If that policy is a touch too harsh, then Strider at 483 is not a read, it is a systematic markdown wearing a rank. The falsifier is specific and it arrives early: a full, uninterrupted starter's workload from Strider or Ragans over one season should pull that number hundreds of spots. If it does not, the model is not learning from innings, and you should stop trusting its pitcher tail.

On the other side, Sal Stewart at 8 against a consensus 127 is the boldest thing in the table, and it is bold in a way the Clark comparison makes interesting. Both are young. Only one of them is being credited with production inside the window. So the driver cannot be youth. It has to be that the forecast expects Stewart's bat to play at the top level immediately and at a high level, and the market is still pricing him as a prospect with a service-time path to clear. Drake Baldwin at 22 against 134 is the same story at catcher, where a bat that plays gets an enormous positional boost in any three-year count.

What I would actually do. Buy Stewart and Baldwin if your leaguemates are still valuing them off prospect lists, because you are getting a starter at a prospect price and the downside is that you waited a year. Sell nothing on Clark or Emerson, since our rank there is not evidence. And on Strider, I would not sell at consensus 62, I would sell at a small discount to it: the market price is closer to right than 483, but our forecast is telling you the distribution of outcomes is wider than the price implies. Being wrong on Stewart costs you a roster spot. Being wrong on Strider costs you an ace.

How to read this

Our dynasty rank forecasts three years of production. It does not price a contract, a service-time clock, or how much fun a prospect is to own, all of which move a market rank and none of which are production.

The cutoffs are percentiles of the gap spread rather than a fixed number of places. The whole spread is shifted, because our board runs colder than the consensus; a symmetric cutoff would have produced a long list one way and almost nothing the other, which would describe the shift rather than the disagreements.

A gap is not evidence. Our own research is explicit that disagreeing with the market does not by itself make a model right, and this piece reports the direction and size of the disagreement without claiming the market is wrong.

All articles · Redraft rankings · Dynasty rankings