By The Numbers

Now that the dust has settled on the MLB 2012 draft let’s take a quantitative gander back at the Orioles choices in the first 200 picks. The idea here is to apply research of several known variables to evaluate how well the Orioles did in drafting, given the talent available. I encourage the By the Numbers enthusiast to check out the background material, but essentially we will be using a regression model created by Sky Andrecheck that gives a baseline of how many Wins Above Replacement (WAR) you can expect from a given pick in the MLB. Sky’s model is shown in the graph. As you can see there is a major penalty for choosing a pitcher, especially early in the draft, and a smaller penalty for choosing a high school player.

Expected WAR Draft Regression Model

In the table below we’ll look at the Orioles draft picks in the first 200 players chosen. In addition to seeing how much WAR we should expect from each draft position, I have also included the Baseball America (BA) rank of each draft pick. We can substitute the BA Rank for the draft selection as an objective talent measure in the expected WAR formula. We can compute the ratio of (Rank Exp. War/Pick Exp. War) to get a very crude measurement of how much return the Orioles got on each draft choice, according to Baseball America. The list of Baseball America ranks that I have access to only goes to 200 (1-100, 101-200), so for Christian Walker and Colin Poche, who were not ranked in the top 200, I have used an estimate of 250.

Pick # BA Rank Name Pitcher? College? Pick Exp. WAR Rank Exp. WAR ROI Ratio
4 5 Kevin Gausman Yes (RHP) Yes 8.52 7.64 0.90
65 87 Branden Kline Yes (RHP) Yes 2.17 1.88 0.86
99 192 Adrian Marin No (SS) No 2.17 1.57 0.72
132 250 Christian Walker No (1B) Yes 2.31 1.69 0.73
162 250 Colin Poche Yes (LHP) No 1.01 0.81 0.80
192 148 Lex Rutledge Yes (LHP) Yes 1.28 1.45 1.13

Lex Rutledge checks in as the steal of the O’s draft. Rutledge had ascended to #148 on the BA Top 200 list before the draft and was drafted almost 50 picks later than his BA rank. I have no idea why he fell, but I am also not much of a prospect maven. Overall, very few college hitters were included at the top of the 2012 BA ranks. Recall, these are the players preferred by the regression model. In fact only one college hitter, University of Florida catcher Mike Zunino (who went #3 overall to the Chicago Cubs), was included in the top 13 players ranked by BA. Even if Zunino fell to the Orioles at pick #4, I doubt that they would have drafted him even though he would have maximized the execpted WAR in the model. The Orioles are currently strong at catcher and have enough other needs to address that they still would have drafted a starting pitcher.

As for the rest of draft picks (Kline, Taylor, Walker, Poche and Rutledge), it seems pretty clear that the Orioles wanted to continue to stockpile young pitching. As we have seen, the model does not like young pitching but it does prefer college arms over high school ones. Kline and Rutledege all meet this qualification while Poche does not. However, given the Orioles current lack of starting pitching, I think this draft strategy makes a lot of sense. The model dislikes young pitching because it is highly variable; young pitchers frequently don’t develop or face chronic injuries. The only way to actually end up with a sufficient number of effective arms is to draft them at a high enough rate to be able to sustain the expected number of failures.

In terms of Return On Investment (ROI) the Orioles draft was a bit of a disappointment this year. The only pick that achieved a ROI greater than ‘1’ was Rutledge and team ended with an average ROI of 0.86. Last year they achieved achieved a ROI greater than ‘1’ on two picks and finished with an average ROI of 1.15 which was good enough for 4th best in MLB. Of course grading any draft immediately after it occurs is a fool’s errand, but perhaps these metrics can give us some insight into the type of development we should expect going forward for the Orioles first six draftees in 2012.