Snowy Sunday musings. Among the two hundred and fifty comments below the Salvy injury report that went up on Friday, March 1, was a conversation about expectations for a first round draft choice. It was posited that "a productive career... a couple of all star appearances" was a reasonable expected outcome for a first round pick. That sounded a little high to me, but I wasn't certain. This caused me to question my own prior assumptions about first round June draft picks. So I started digging. Since I knew how to find the information I wanted in b-ref but not in Fangraphs, all quoted statistics come from there.

WAR as a proxy for a productive career seemed like a good fit, so I began by looking at what level of accumulated WAR was needed to find a couple of all star appearances. 18 to 20 showed up on numerous player pages for 2x All Star, so let's go with that. Then I had to decide the parameters for "first round pick." This was harder than one might think. To be fair, each year should be equal, and the set of years used had to give a fair chance to all of the players chosen. Since Christian Colon was specifically mentioned and I want to include him, I'm using an end point of 2010. This endpoint obviously doesn't capture anywhere near the full career of many players, but everyone from the 2010 draft that is going to play in the majors probably already has. Since the Royals started drafting based on order of finish in 1970, and I wanted an odd number of results, I decided to use that as my start point. Should I include supplemental picks as first round? Should the data set shrink as we get further back in time and the first round is smaller? Should we use average WAR for each slot, or each year, as a metric, or is the median result a better identifier? Do players who never made it to the majors count as a null result to be thrown out, or do they count as zero? Again, choices had to be made, so I made some.

Here's what I went with. "First round" means the first 30 picks in each draft for the 41 years from 1970 to 2010 inclusive. A player never reaching the majors doesn't mean the team didn't spend copious time and money on the prospect they drafted so no major league service time counts as zero WAR. Average doesn't work like I want it to. It skews the results for several slots. Pick 30 in our sample shows 316.2 bWAR for the 41 players selected for an average 7.7 per player selected. Mike Schmidt is responsible for over a third of the total WAR from pick 30. Sixteen players selected at pick 30 never played in the majors. Seven more accumulated negative bWAR. So the median #30 pick has zero WAR.

More digging shows that the median expected result for pick 1.1, the first pick in the draft, is Pat Burrell. Of the first overall pick in each of the 41 drafts I looked at 20 players had more bWAR and 20 had less. Two players in our time frame didn't sign but one was better and one worse, so the median result doesn't change. Now, I think it's fair to say that Pat Burell had a productive career. While he never appeared in an all star game he does have two WS rings, so we'll credit him with those and say he meets optimistic criteria. But round one pick one is the only point in the draft where the median outcome is "a productive career." The drop off after the first pick is precipitous. By the fifth pick and for subsequent picks, what we should expect from a first rounder is... nothing.

I hope that Christian Colon isn't finished accumulating major league service time, but even if he is he's currently the midpoint player for pick number four. Twenty better outcomes, twenty worse. Bubba Starling and Kyle Zimmer, if they make the major league club and actually accrue any positive bWAR at all, will be better than the median outcome for pick #5 over the 41 years I looked at. Of course, since the median is zero, if they are worse than replacement they'll be worse than we should reasonably expect. The midpoint bWAR performance result for the more than a thousand first round players drafted between 1970 and 2010 outside the top four picks is zero. There are some draft slots between five and fifteen where the median is positive, but overall, even in the first round, the median outcome is not just "didn't have a productive career" but "did not reach the majors."

Drafting is hard.