The odds of success for a draft pick, Part 2
April 9, 2020
by Steve Thomas
I need to apologize to all of you right up front. Because I’m Steve, what started out as an innocent question about the draft in our comment section on the website has turned into 2 long columns and 13 different charts. Don’t ask me a data-related question when I have spare time on my hands unless you’re prepared to let me go overboard on the answer.
This column is the follow up to my analysis of the top 10 draft picks between 2000 and 2019. You can read that masterpiece by clicking here. Since that column was such a joy to write and had all kinds of great information in it, I thought it would be useful to expand the effort to three more categories: picks 11 – 20 and 21 – 32 of round 1, and round 2.
As I said in the previous iteration, this study was originally born out of my thoughts as to the chances of the Redskins actually drafting a 10 year, big-time, All Pro, Hall of Fame worthy starter with their #2 overall pick or another pick in the top 10 via a trade down. However, the same question lingers for players drafted later in round 1, and to a lesser extent, round 2. Therefore, I followed the same basic format for part 2 as I did for part 1: with the help of Pro Football Reference website, I pulled every player selected in picks 11 – 20 and 21 – 32 of round 1, and round 2 between 2000 and 2019, broken down by position group, and manually counted the number of players selected in 11 – 20 and 21 – 32 of round 1, and round 2 of each year’s draft who have been First Team All Pro selections, have been selected to at least one Pro Bowl, and only spent 1 or zero years as a full-time starter. The last category is a good indicator of players who were clear and obvious busts.
Being selected for an All Pro team or a Pro Bowl isn’t necessarily a conclusive indication of a “successful” top 10 draft pick, but it is an easy way to establish some sort of objective frame to the argument.
Study limitations
The most important limitation to this study is that the class of 2019 has only had one year of play time. It’s not completely fair to evaluate that year group, the All Pro and Pro Bowl selections earned in year one by those players are stats worth compiling. Therefore, I left the class of 2019 out of the “years started” category but counted it in the others categories.
My use of the draft classes of 2000 – 2019 was for no other reason than because it was an even 20 years, but it’s a totally arbitrary line.
Finally, these categories aren’t decisive; for example, Robert Griffin III was drafted #2 overall by the Redskins, made one Pro Bowl, and started multiple years, and thus doesn’t look like a bust by these evaluative criteria, but clearly was one. There are others in that situation as well. So, my system isn’t precise and is only intended to give a general idea of the success rate of drafting in the top 10 by position group.
Also, I didn’t separate these numbers by 3 – 4 and 4 – 3 linebackers, or by an “edge” category, because it became too much brain damage even for me in a project that started out as a quick look at some data but, in typical fashion, became much more than that.
With that having been said, let’s proceed.
Data
The first three charts show the (1) total number of players drafted between 2000 and 2019 in each position group in picks 11 – 20 and 21 – 32 of round 1, and round 2, respectively, (2) the number selected to the hall of fame in each group, (3) the number of players who earned multiple First Team All Pro selections, (4) the number of players who’ve earned at least 1 First Team All Pro selection, (5) the number selected to at least one Pro Bowl, and (6) the number who have only been a full time starter for just 1 year or zero full years, except for the class of 2019. It is sorted by the number of players drafted in each position group, from highest to lowest. I’ve also included the totals for all of the draft picks in each of the three groups (11 – 20, 21 – 32, and round 2), the median for each position group by category, and for the more statistically-oriented, the standard deviation for each category. Standard deviation is a measure of the average amount of variance of from the overall mean in each category.
Please note that only 3 punters and kickers have been drafted either rounds 1 or 2 since 2000, and none in picks 1 – 10. I left kickers and punters out of the calculation of the median and standard deviation in all 4 of the draft pick ranges so as to not skew the results for a position group that is historically never drafted even in the top half of the draft.
Fun fact of the day: in the history of the NFL, only 6 punters and 11 kickers have been drafted in rounds 1 or 2. The only two punters drafted in round 1 were Hall of Famer Ray Guy in 1973 and Russell Erxleben in 1979, who hung around for 5 seasons with the Saints and 1 for the Lions. Three kickers have been drafted in round 1, Charlie Gogolak, who was drafted by the Redskins in 1966 (55 games in 6 years with the Redskins and Patriots), Steve Little in 1978 (St. Louis; 33 games in 3 years), and Sebastian Janikowski (Raiders, 2000; 18 seasons but 0 First Team All Pro selections). There is never, ever a valid reason to draft a punter or kicker in rounds 1 or 2.
Picks 11 – 20
| Position | # drafted in 11 – 20 | # selected to HOF | # with multiple 1st tm AP selections | # selected as 1st tm AP tm once | # selected to a PB | # only started 1/0 yrs pre-‘19 |
| DB | 38 | 1 | 4 | 10 | 18 | 5 |
| OL | 36 | 1 | 4 | 7 | 14 | 0 |
| DE | 33 | 0 | 3 | 5 | 10 | 8 |
| DT | 27 | 0 | 3 | 5 | 9 | 2 |
| LB | 25 | 0 | 2 | 6 | 16 | 4 |
| WR | 16 | 0 | 0 | 0 | 3 | 0 |
| QB | 10 | 0 | 0 | 0 | 3 | 1 |
| RB | 9 | 0 | 0 | 2 | 5 | 2 |
| TE | 5 | 0 | 0 | 1 | 2 | 0 |
| P / K | 1 | 0 | 0 | 0 | 1 | 0 |
| TOTAL | 200 | 2 | 16 | 36 | 81 | 22 |
| MEDIAN | 20.5 | 0 | 1 | 3.5 | 7 | 1.5 |
| STAND DEV | 13.52 | 0.42 | 1.78 | 3.50 | 6.23 | 2.70 |
Picks 21 – 32
| Position | # drafted in 21 – 32 | # selected to HOF | # with multiple 1st tm AP selections | # selected as 1st tm AP tm once | # selected to a PB | # only started 1/0 yrs pre-‘19 |
| DB | 51 | 0 | 2 | 6 | 17 | 12 |
| OL | 37 | 0 | 2 | 6 | 10 | 7 |
| WR | 33 | 0 | 2 | 5 | 8 | 15 |
| DE | 24 | 0 | 2 | 3 | 4 | 6 |
| RB | 23 | 0 | 0 | 3 | 10 | 9 |
| DT | 21 | 0 | 0 | 1 | 3 | 3 |
| LB | 20 | 0 | 0 | 4 | 7 | 5 |
| TE | 15 | 0 | 0 | 1 | 7 | 2 |
| QB | 12 | 0 | 1 | 2 | 3 | 6 |
| P / K | 0 | 0 | 0 | 0 | 0 | 0 |
| TOTAL | 236 | 0 | 9 | 31 | 69 | 65 |
| MEDIAN | 22 | 0 | 0.5 | 3 | 7 | 6 |
| STAND DEV | 14.16 | 0.00 | 0.99 | 2.13 | 4.82 | 4.55 |
Round 2
| Position | # drafted in rd 2 | # selected to HOF | # with multiple 1st tm AP selections | # selected as 1st tm AP tm once | # selected to a PB | # only started 1/0 yrs pre-‘19 |
| DB | 146 | 0 | 3 | 9 | 22 | 43 |
| OL | 102 | 0 | 2 | 8 | 17 | 18 |
| LB | 90 | 0 | 1 | 6 | 19 | 30 |
| WR | 89 | 0 | 2 | 3 | 22 | 34 |
| DE | 60 | 0 | 0 | 2 | 9 | 29 |
| DT | 44 | 0 | 1 | 1 | 5 | 15 |
| RB | 49 | 0 | 2 | 3 | 11 | 27 |
| TE | 34 | 0 | 1 | 1 | 6 | 11 |
| QB | 21 | 0 | 0 | 1 | 3 | 11 |
| P / K | 2 | 0 | 0 | 0 | 0 | 1 |
| TOTAL | 637 | 0 | 12 | 34 | 114 | 219 |
| MEDIAN | 54.5 | 0 | 1 | 2.5 | 10 | 22.5 |
| STAND DEV | 43.08 | 0.00 | 1.03 | 3.17 | 8.10 | 12.78 |
The next chart shows the percentage of players drafted in picks 1 – 10 (as published in part 1), 11 – 20 and 21 – 32 of round 1, and round 2, respectively, by position group, who have been selected to multiple First Team All Pro teams, sorted from highest to lowest percentage, plus the total for all draft picks in each category, and the median and the standard deviation for each group (1 – 10, 11 – 20 and 21 – 32 of round 1, and round 2):
| Position | % to multiple AP tms from
picks 1 – 10 |
% to multiple AP tms from
picks 11 – 20 |
% to multiple AP tms from
picks 21 – 32 |
% to multiple AP tms from
Round 2 |
| LB | 19% | 8.0% | 0% | 1.1% |
| RB | 18.8% | 0% | 0% | 4.1% |
| DB | 12.5% | 10.5% | 4.0% | 2.1% |
| DT | 11.8% | 11.1% | 0% | 2.3% |
| WR | 10.7% | 0% | 6.1% | 2.2% |
| OL | 9.7% | 11.1% | 5.4% | 2.9% |
| DE | 8% | 9.1% | 8.3% | 0% |
| QB | 0% | 0% | 8.3% | 0% |
| TE | 0% | 0% | 0% | 2.9% |
| K / P | — | 0% | 0% | 0% |
| TOTAL | 10% | 8.0% | 3.8% | 1.9% |
| MEDIAN | 10.70% | 8.00% | 4.00% | 2.20% |
| STAND DEV | 6.82% | 5.34% | 3.63% | 1.37% |
The next chart shows the percentage of players drafted in picks 1 – 10 (again, as published in part 1), 11 – 20 and 21 – 32 of round 1, and round 2, respectively, by position group, who have been selected to at least one First Team All Pro team, sorted from highest to lowest percentage for the pick 1 – 10 groups, plus the total for all draft picks in picks 1 – 10, 11 – 20, and 21 – 32 of round 1, and round 2, respectively, the median by position group, and the standard deviation:
| Position | % selected to AP tm at least once
from pks 1 – 10 |
% selected to AP tm at least once
from pks 11 – 20 |
% selected to AP tm at least once
from pks 21 – 32 |
% selected to AP tm at least once
from round 2 |
| RB | 37.5% | 22.2% | 13.0% | 6.1% |
| LB | 33.3% | 24.0% | 20.0% | 6.7% |
| DB | 29.2% | 26.3% | 11.8% | 6.2% |
| OL | 25.8% | 19.4% | 16.2% | 7.8% |
| DT | 23.5% | 18.5% | 4.8% | 2.3% |
| DE | 20% | 15.2% | 12.5% | 3.3% |
| WR | 14.2% | 0% | 15.2% | 3.4% |
| QB | 8.8% | 0% | 16.7% | 4.8% |
| TE | 0% | 20.0% | 6.7% | 2.9% |
| K / P | — | 0% | — | 0% |
| TOTAL | 22% | 18.0% | 13.1% | 5.3% |
| MEDIAN | 23.50% | 19.40% | 13.00% | 4.80% |
| STAND DEV |
12.02%
|
9.71% | 4.83% | 1.95% |
The next chart shows the percentage of players drafted in picks 1 – 10, 11 – 20, and 21 – 32 of round 1, and round 2, respectively, by position group, who have been selected to at least one Pro Bowl, sorted from highest to lowest percentage for the pick 1 – 10 groups, plus the total for all draft picks in picks 1 – 10, 11 – 20, and 21 – 32 of round 1, and round 2, respectively, the median by position group, and the standard deviation:
| Position | % selected to at least 1 PB
from pks 1 – 10 |
% selected to at least 1 PB
from pks 11 – 20 |
% selected to at least 1 PB
from pks 21 – 32 |
% selected to at least 1 PB
from round 2 |
| DB | 66.67% | 47.4% | 33.3% | 15.1% |
| RB | 62.5% | 55.6% | 43.5% | 22.4% |
| OL | 54.8% | 38.9% | 27.0% | 16.7% |
| DE | 52% | 30.3% | 16.7% | 15.0% |
| QB | 50% | 30.0% | 25.0% | 14.3% |
| TE | 50% | 40.0% | 46.7% | 17.6% |
| LB | 47.6% | 64.0% | 35.0% | 21.1% |
| DT | 41.2% | 33.3% | 14.3% | 11.4% |
| WR | 35.7% | 18.8% | 24.2% | 24.7% |
| K / P | — | 100% | — | 0% |
| TOTAL | 51% | 40.5% | 29.2% | 17.9% |
| MEDIAN | 50.00% | 38.90% | 27.00% | 16.70% |
| STAND DEV | 9.60% | 13.98% | 11.10% | 4.31% |
The next chart shows the percentage of players drafted in picks 1 – 10, 11 – 20 and 21 – 32 of round 1, and round 2, respectively, who have only been a full-time starter for either just 1 season or zero seasons, sorted from lowest to highest percentage for the pick 1 – 10 groups, plus the total for all draft picks in the 1 – 10, 11 – 20, and 21 – 32 of round 1, and round 2, respectively, the median by position group, and the standard deviation:
| Position | % who started only 1 or 0 seasons
for pks 1 – 10 |
% who started only 1 or 0 seasons
for pks 11 – 20 |
% who started only 1 or 0 seasons
for pks 21 – 32 |
% who started only 1 or 0 seasons
for round 2 |
| RB | 0% | 22.2% | 39.1% | 55.1% |
| TE | 0% | 0% | 13.3% | 32.4% |
| OL | 3.2% | 0% | 18.9% | 17.6% |
| LB | 4.8% | 16.0% | 25.0% | 33.3% |
| DT | 5.9% | 7.4% | 14.3% | 34.1% |
| DB | 8.3% | 13.2% | 23.5% | 29.5% |
| QB | 8.8% | 10.0% | 50.0% | 52.4% |
| DE | 16% | 24.2% | 25.0% | 48.3% |
| WR | 17.9% | 0% | 45.5% | 38.2% |
| K / P | — | 0% | — | 50.0% |
| TOTAL | 8.5% | 11.0% | 27.5% | 34.4% |
| MEDIAN | 5.90% | 10.00% | 25.00% | 34.10% |
| STAND DEV | 6.35% | 9.37% | 13.40% | 12.06% |
Finally, the last chart shows the numbers of players drafted in rounds 1 and 2, by position group. Please note that the “# drafted per starter” columns is my attempt to normalize the number of selections in each position group on the basis of the number of starters at each to give some sense of how often teams are drafting these players versus the needs that need to be filled. I divided the total number drafted by the number of starters at the position on each team. For example, each team has one quarterback, so the calculation for round 1 was 56 / 1. I gave receivers 3 starting slots in recognition of the importance of slot receivers. I gave linebackers 3.5 starting spots due to the difference between the number of starters in the 4 – 3 and the 3 – 4. I counted running backs as 1.5 starters per team since some teams have fullbacks and some teams effectively have two starters. If this stat isn’t for you, that’s fine, but it’s there.
| Position | # drafted in rd 1 | # drafted per starter in rd 1 | # drafted in rd 2 | # drafted per starter in rd 2 |
| DB | 113 | 28.25 | 146 | 36.5 |
| OL | 104 | 20.8 | 102 | 20.4 |
| DE | 82 | 41 | 60 | 30 |
| WR | 77 | 25.67 | 89 | 29.67 |
| LB | 66 | 18.86 | 90 | 25.71 |
| DT | 65 | 32.5 | 44 | 22 |
| QB | 56 | 56 | 21 | 21 |
| RB | 48 | 32 | 49 | 32.67 |
| TE | 24 | 24 | 15 | 15 |
| K / P | 1 | — | 2 | — |
| TOTAL | 636 | — | 618 | — |
| MEDIAN | 66 | 28.25 | 60 | 25.71 |
| STAND DEV | 27.46 | 11.54 | 42.10 | 6.87 |
Analysis and Conclusions
I don’t want the analysis and conclusions to be longer than the data section, because there’s a ton of things one can take from this information, so I’ll try and keep this brief (hahahahaha!!) and fairly general.
First, in a broad sense, this data worked out in logical fashion. For the “good” categories – multiple First Team All Pro selections, at least one All Pro selection, at least one Pro Bowl selections – the percentages selected were highest in picks 1 – 10 and descended in each lower category. For the “obvious bust” category, the lowest overall percentage was for picks 1 – 10, and got progressively higher in each of the three other categories.
In every circumstance, the odds are severely against finding a “big time” player, even at the very top of the draft, but it appears as though there aren’t any big-picture flaws in the way that teams are ranking the players. However, the grades and/or expectations themselves are flawed – meaning, the rosy predictions each year about a player’s ability to transition to the NFL and become franchise-changing players are overly optimistic. It’s much rarer than the writeup in everybody’s favorite mock draft every year would make a person believe.
The data shows that it becomes much, much more unlikely to find an elite player in round 2 than anywhere in round 1 – only 5.3% of players at any position selected in round 2 went on to make even one First Team All Pro team, as compared to 22% for picks 1 – 10, 18% for picks 11 – 20, and 13.% for picks 21 – 32. Only 17.9% of 2nd round picks made even one Pro Bowl, as compared to 51%, 40.5%, and 29.2% for picks 1 – 10, 11 – 20, and 21 – 32, respectively. But it’s the “bust rate” in which round 2 picks really fall off the cliff: 34.4% only started 1 or 0 seasons.
Specifically, quarterbacks are drastically overdrafted. Of the 77 quarterbacks drafted in rounds 1 and 2 between 2000 and 2019, only 1, Aaron Rodgers, has earned First Team All Pro honors more than once, despite the annual predictions of greatness. That’s an almost unbelievable stat. Also, 56 quarterbacks were drafted in round 1, but just 21 were drafted in round 2, and the “bust” rate (as measured by my 0 or 1 years started metric) jumps from 8.8% and 10% in picks 1 – 10 and 11 – 20 to 50% and 52.4% in picks 21 – 32 and round 2. Also, not a single quarterback drafted in picks 11 – 20 was named to even one First Team All Pro list. It appears as though teams rush to draft quarterbacks at the top of the draft and throughout round 1 out of desperation without regard to their annual talent level. That conclusion is consistent with the commonly-held perception.
The data for the drafted wide receivers is all over the place. There’s no sense of logical order to the data as there is for the numbers across all positions groups. Just as an example, 15.2% wide receivers drafted in picks 21 – 32 earned at least one First Team All Pro honor in their career, but no one single receiver drafted in picks 11 – 20 has done so. Significantly more receivers drafted in round 2 have made at least one Pro Bowl (24.7%) than in either picks 11 – 20 (18.8%) or 21 – 32 (24.2%). I do trust that NFL talent evaluators know what they are looking at, for the most part, so this tells me that it’s just tougher for receivers to make the jump to the NFL than other positions. I didn’t examine objective data for this, but my hunch is that receivers tend to be drafted more for their athletic profile (i.e., speed), than route running and hands. Therefore, my recommendation to NFL teams would be to strike a better balance between speed on one hand and football intelligent, advanced route running, and hands on the other.
Running backs are a great bet to succeed if selected essentially anywhere in the top half of round 1, but have the highest “bust rate” in round 2 of any position group and are almost a full standard deviation above the mean for “bust rate” in the 21 – 32 pick range.
Finally, in looking at my “number picked per starting position” statistic, beyond quarterbacks, it’s defensive ends and defensive tackles that seem to get the most attention in round 1 (41 and 32.5, respectively). In round 2, it’s defensive backs.
What are your thoughts on this information? Is this just way, way too much data, and you recommend that someone take my laptop from me during the next quarantine? Leave me a comment below.