This topic came up again on the Talking Baseball Live show on Wednesday night. There is another facet of the argument that tips the scales towards the “Melnick” position.

I never thought this topic would be so controversial, but the Farino position and Melnick position are really fighting it out. However, both sides have ignored an aspect that hasn’t yet been discussed in great detail yet.

One of, if not the prime, argument in the Farino camp is the idea that a strong closer in a head to head league will help enough in ERA and WHIP so that the mere number of saves is not the only consideration. This makes sense. But when hearing him defend this position today on the Fantasy Baseball Mafia show with Tony Cincotta I realized that no one addressed the marginal gain aspect. 

For example, let’s compare Papelbon and Borowski, since they were essentially the best and worst examples in ERA and WHIP.  Last year Papelbon had a 1.85 ERA and 0.77 WHIP in 58 IP (rounding for ease of use).  That is 12 ER and 45 Hits + BB. Borowski had a 5.07 ERA, with 37 ER in 55 IP, with 77 hits and 17 walks (94 H+ BB), for a 1.43 WHIP.

If we normalize these numbers to the same number of innings, let’s use 60 innings pitched, Borowski rates to have 86 hits plus walks and 40 earned runs. Papelbon will have 46 hits plus walks and 13 ER. So what we are fighting about from best to worst is the difference, or 40 H+BB and 27 ER.  With a mid range closer the difference might be half that.

Let’s ASSUME, and I use caps because it is a large assumption, that their records reflect their underlying skills. In reality the decision should absolutely NOT be based upon actual stats but on expected stats, but since I am examining the Farino position I will make all assumptions in favor of his position for the purposes of examination. In fact if we go by skills the difference between best and worst is far lower, only because the worst guy is generally unlucky (as Borowski was) and the best is lucky (as Papelbon was).

So what Todd is saying is that the difference between the two is what warrants the early selection. To put it another way, the marginal difference between the best and worst closers warrants one being taken in the fourth or fifth round and not being drafted at all.

When looked at in this light the pendulum clearly swings strongly to the Melnick position.  Why? Because now we must weigh the opportunity cost, essentially 15 rounds or more, between the two versus the marginal difference.  In roto leagues this marginal difference is essentially zero because of the sheer number of innings pitched.  If we look at it over an equal number of innings pitched we see that the marginal difference is negligible in roto leagues.  Melnick clearly wins the battle here. A difference of 40 H+BB in 1000 innings is very small.

What of head to head leagues? Tony Cincotta and his co-host were sympathetic to the Farino position in head to head leagues, admittedly as was I before I gave it some deeper analysis.  But I admit that I didn’t give it enough consideration.  In head to head leagues the marginal difference still isn’t enough to tip the scales.

The Melnick position essentially is that the sheer number of saves is not only the most important factor but cannot accurately be predicted so that the early round picks are not worth it. I am not sure if he would agree with my analysis or thought of it in this light, but I have to admit that I now agree with his conclusion.

With one exception of course!  That exception is when the individual owner decides that the risk factor trumps the marginal difference theory. For example, lets stay with this example, but lets look at what we could call an “expected number of saves.” We will define this merely as the number of saves multiplied by the percentage chance of getting that number. So, if Papelbon is 90% likely to get 37 saves that is approx 34 saves. If someone like Corpas this year is 25% likely to get 45 saves then his expected value is approx 13 saves.

Of course this is not correct since the actual percentage will be a weighted average.  But the principle is that one can take the risk into consideration in this manner and decide that that the safety of the Papelbon is worth the opportunity cost and risk compared to a lesser closer, and certainly over the long run this may very well be correct.

So, once again I can’t agree or disagree with either other than to say that judgment and skill are the paramount factors, and I hope that this at least elucidates the theoretical considerations behind how to evaluate these arguments. My verdict is Melnick by a length.