Exploring Top Prospect Lists: Draft Position

So far in our series on top prospects, we’ve introduced our Baseball America dataset and examined differences in performance between ranking groups, hitters and pitchers, age groups, and newcomers and repeaters on the list. Now, we’ll finish the series by looking at performance based on the player’s draft position.

Starting off, there’s an obvious caveat here: not all prospects are drafted, period. We’ll include undrafted free agents as a separate category for analysis. (Note that “undrafted” doesn’t mean the same thing in baseball as it would in the NBA or NFL; the MLB draft does not include international prospects, so Dominican, Cuban, Venezuelan, and Japanese players, among others, join the league under different rules.)

Within the set of drafted players, we need a way to separate them into usable groups. As we did with changes in ranking, we’ll go with a squares-based approach: picks 1-4, 5-16, 17-64, and 65 and up. (For current players, this is basically top of the draft, top half of the first round, late first through second, and rounds 3-plus.)

As always, we’ll start with the basics: how many players in each group? Columns are draft position, rows are prospect ranking:

Rank	UDFA	1-4	5-16	17-64	65-plus
1-10	55	62	51	31	31
11-25	60	48	89	78	70
26-50	125	31	118	154	147
51-75	133	30	97	138	177
76-100	119	17	84	169	186
Total	492	188	439	570	611

Fully a third of the top-4 draft picks in the sample were also top-10 prospects, while barely 5% of top-100 players taken outside of the top 16 picks could say the same. Which is an eminently sensible result – you’d expect high draft picks to be better prospects than low ones. It will also inevitably lead to some sample size issues, which are worth keeping in mind as we explore the results.

Average career WAR by draft position:

Rank	UDFA	1-4	5-16	17-64	65-plus
1-10	24.2	29.4	21.9	20.1	15.3
11-25	14.6	13.5	18.2	14.5	11.7
26-50	11.3	13.3	13.5	9.4	11.2
51-75	9.6	4.8	9.7	8.8	6.5
76-100	8.1	4.5	8.7	6.2	6.5

So, high picks who are also top-10 prospects do well – but that advantage doesn’t appear to persist once you leave the top 10, which is potentially interesting. Let’s break out the percentile table to go into more detail, starting with outcomes for the aforementioned top-10 prospects:

Pct	UDFA	1-4	5-16	17-64	65-plus
90	59.8	78.3	50.6	56.3	39.8
80	44.9	54.4	41.8	28.2	26.7
70	37.9	35.3	28.1	22.6	14.5
60	22.4	29.8	23.2	19.3	13.4
50	17.1	18.8	15.9	15.2	11.3
40	9.0	12.8	10.2	10.9	9.3
30	5.7	8.9	5.5	7.4	8.9
20	2.6	5.5	3.3	1.2	4.9
10	-0.3	2.2	0.0	0.0	-0.1

The outcomes for high draft picks are some of the most impressive we’ve seen so far. A top-4 pick who also grades as a top-10 prospect has something like a 20% chance at a Hall of Fame-caliber career in this dataset. Meanwhile, the lower picks still become useful players at a high rate, but can’t keep up with the top of the draft. (The undrafted free agents also do quite well, particularly when accounting for the fact that only three of our 23 seasons of data had an undrafted prospect at #1: Andruw Jones twice, and Daisuke Matsuzaka.)

Really, though, this is not a shocking outcome; high picks who do well on prospect lists should get good results. Let’s see what happens across the next group, prospects 11-25:

Pct	UDFA	1-4	5-16	17-64	65-plus
90	34.4	34.8	49.8	46.2	29.6
80	28.8	19.8	33.6	28.2	26.8
70	19.9	16.7	23.8	16.4	15.1
60	15.6	13.1	19.5	10.5	11.5
50	9.0	9.2	10.5	5.9	6.8
40	5.4	7.7	6.0	3.4	2.7
30	1.5	3.0	2.9	0.9	0.4
20	0.0	0.6	0.0	0.0	-0.1
10	-0.3	-0.4	-2.0	-0.4	-0.3

This one is a bit more unexpected. Even barely outside the top 10, not only has the advantage for picks 1-4 vanished, they suddenly do worse than the players picked right behind them, as well as the upper-percentile internationals.

Does that trend continue? Let’s take the remaining groups in order: 26-50, then 51-75, then 76-100:

Pct	UDFA	1-4	5-16	17-64	65-plus
90	32.4	39.9	39.4	30.4	26.2
80	22.8	20.0	27.2	15.5	18.0
70	13.8	12.2	15.5	9.4	12.7
60	8.3	8.4	10.8	6.3	9.4
50	2.8	5.4	6.5	2.4	6.0
40	1.6	3.7	4.2	0.9	2.9
30	0.2	1.4	0.0	0.0	0.8
20	-0.3	0.1	-0.3	-0.3	0.0
10	-0.9	-1.1	-1.5	-0.8	-0.3

The 1-4 group gets closer, but still largely runs behind the 5-16 set. Also, the 65-plus group, which had been pretty uninspiring so far, does at least reasonably well here.

Pct	UDFA	1-4	5-16	17-64	65-plus
90	32.2	15.3	26.6	28.2	19.7
80	22.5	10.3	20.2	17.0	11.2
70	10.7	4.8	12.8	10.9	6.3
60	6.0	3.4	6.3	6.3	3.3
50	1.5	1.1	3.3	2.7	1.4
40	0.3	0.4	0.3	0.2	0.1
30	0.0	-0.4	0.0	0.0	0.0
20	-0.8	-1.2	-0.1	-0.3	-0.3
10	-1.5	-1.3	-0.7	-0.5	-0.9

Yeah, that’s not good for the high picks. For everything median and up (which is to say, all of the useful-player outcomes), picks 1-4 are either the lowest of the five groups here, or virtually tied for lowest.

Pct	UDFA	1-4	5-16	17-64	65-plus
90	28.0	12.5	24.2	20.3	20.4
80	13.8	6.3	20.0	8.9	11.2
70	7.9	2.0	11.1	4.8	8.1
60	4.6	0.2	6.4	1.6	3.2
50	1.2	0.1	2.7	0.4	1.5
40	0.0	0.0	0.0	0.0	0.3
30	-0.1	0.0	0.0	-0.1	0.0
20	-0.7	-0.7	-0.1	-0.4	-0.3
10	-1.3	-1.3	-0.7	-1.0	-1.0

Same result, but even more extreme for prospects 76-100. Yes, the sample sizes are small (remember, picks 1-4 have 31, 30, and 17 representatives in the last three groups over our 23 years of data). But that’s a combined sample of 78, which is a bit above three players per year. And in each of those sets, both undrafted players and picks 5-16 are significantly more successful than top fours.

My best guess is that we’re seeing two subgroups of the top picks here: the actual top players in the draft and the signability picks. The latter group may be less common these days thanks to the changes in the draft in recent years, but they do still happen, and they tend not to be what you’d call wildly successful. It’s also possible that BA is subconsciously over-ranking the signability picks (they got picked high, maybe our evaluation was wrong), leading to them underperforming their counterparts in the same groups. Whatever the cause, top-4 draft picks who don't make the top 10 in the BA prospects list do not have promising outcomes compared to their similarly-ranked peers.

So, coming down to the end of this series (barring any other ideas occurring to me), what have we learned? Some of the results have been clearer than others, but let’s summarize article by article:

Higher-ranked prospects tend to do better than lower-ranked ones, which makes sense. This result came with our original percentile table, which I like enough to include it again here at the end of the series:

Percentile	#1	#2-5	#6-10	#11-15	#16-25	#26-50	#51-75	#76-100
90	76.3	58.5	57.2	50.8	31.8	34.5	27.0	21.9
80	56.7	39.3	37.9	36.8	20.3	19.9	15.9	12.5
70	51.1	32.3	26.8	28.2	15.1	12.6	9.2	7.3
60	47.2	20.7	17.7	21.5	9.7	8.4	4.8	2.9
50	35.7	16.7	12.9	16.5	5.7	4.9	1.6	0.8
40	23.5	10.6	8.6	8.7	2.6	1.8	0.2	0.0
30	14.7	7.4	5.6	6.5	0.2	0.1	0.0	0.0
20	9.0	3.2	2.7	1.1	-0.1	-0.2	-0.3	-0.5
10	1.3	0.1	0.0	0.0	-1.2	-0.9	-1.1	-1.1

Moving on to the complicating factors: Hitters outperform pitchers in basically every group. (BA has probably adjusted to this, as the number of pitchers in the last few top-100 lists has decreased. The original result was severe enough that I would expect hitters to still be favored even post-adjustment.)

Young prospects have higher ceilings, old prospects have higher floors; no huge surprise here.

It is (probably, slightly) preferable to have someone who’s climbed in the rankings as opposed to someone who’s held steady, among returning prospects in the same range.

And finally, a top-4 draft pick is wildly promising if evaluators agree he deserved the spot; otherwise, run far, far away.

Hopefully somewhere among the incessant tables of numbers, this series contained a bit of perspective on what outcomes you might expect from the top-100 prospect your team is hyping up. If he’s a 20-year-old shortstop picked in the top 4 and ranked in the top 10, look out! If he’s a 23-year-old pitcher drafted in the fourth round and ranked in the 80s, well... look out.