Patience is a Virtue, Especially with Meaningless Off-season Spectacles

Last week the 2010 NFL draft doubled down on its already swollen media exposure. The first round was broadcast as a single event last Thursday, followed by rounds two and three on Friday night, and the remaining four on Saturday.

The NFL draft is easily one of the silliest events in the perpetual motion machine that is our modern sports consumption culture. I suspect the pomp and circumstance surrounding the draft has a lot to do with the comparative dearth of games in the NFL regular season. The NFL, in addition to being the most financially lucrative sports league in the world, provides a special kind of social cohesion to a large percentage of American males. It is not a stretch to suggest that, for many American men, it is easier to talk about football than to talk about anything else.

Yet there are only 16 games in the regular season, giving the football fan comparatively little actual sporting activity to talk about. Enter the overblown pageantry of mid-April’s selection of amateur football players by our favorite professional teams. Judging by the resources devoted to draft coverage by all the major sporting news outlets, one would think the outcome of the next 10 Super Bowls hinged on last week’s draft.

Data pulled from Pro Football Reference suggests a decidedly more patient perspective. Over the last 10 years, 463 individual players have either played or been voted to play in the Pro Bowl, the NFL’s all star game. The chart below looks at the round those players were drafted in against the average number of seasons it took to make their first Pro Bowl.

I’m assuming that making your first Pro Bowl effectively defines someone as an established above-average NFL player. One would expect that whatever latent combination of natural talent, work ethic, and luck that turns a college football player into a quality NFL player is more obvious for first round selections. After all, predicting a higher likelihood of success than other amateur players is by definition the reason why you would take someone in the first round (unless you’re the Raiders).

The crucial but neglected question at draft time is this: How long does it take for players to establish themselves? For 1st round picks, it appears to be 3 to 4 years on average.

Any undrafted free agent that lasts 7 years in the league is a huge windfall for their teams, whether they make the Pro Bowl or not. But the NFL draft is, in the words of Gregg Easterbrook, almost literally a lottery. Of course 1st round picks are better to have than 7th round picks. 1st round picks give you better odds on drafting a Pro Bowl player, but it takes years before you know if your gamble worked out.

Furthermore, picks are distributed more or less equally, so the idea that the outcome of any team’s season turns on that year’s draft is absurd. If you want to spend your offseason energy divining next year’s impact players, look at the 2007 draft class (featuring 2nd round pick Kevin Kolb, your new starting QB for the Philadelphia Eagles).

(Side note: I am much more likely to watch a show projecting the 2007 draft class for their impact next season than the three days of bloviating that covered this year’s rookies. The odds of a rookie making a substantive impact are empirically small.)

Obviously 1st round picks are more likely to make the Pro Bowl in their careers than 7th round picks. But not as obviously, undrafted free agents are as likely, if not more likely, to make the Pro Bowl as anyone drafted in the 3rd through 7th rounds.

The graph below shows the number of individual Pro Bowlers by draft position over the past decade. One could read this to mean that we need another two hours of ESPN programming breaking down the undrafted free agent signings, like we had last Friday night for the 2nd and 3rd rounds. Or one could read this as the draft being the crapshoot it mostly is, and stop wasting beautiful spring days trapped inside Radio City Music Hall to cheer your team once every hour.

This should convince you that intense scrutiny of the draft is a waste of time. Such a task wouldn’t be necessary if guys like this didn’t exist.

A few notes:

Players drafted after the 7th round before the draft’s modern format, but made their first Pro Bowl in 2000 or later, are counted as undrafted. I reran the data without special teams players, but the overall trends didn’t change much. The most notable exception is the number of undrafted free agents that make the Pro Bowl, because a considerable number of kickers, punters, and long snappers come from those humble ranks. The graph above makes clear the proportion of special team Pro Bowlers in each draft position.

The data includes outliers like former Cardinals,  Redskins, 49ers, and Lions guard Ray Brown, who went in the 8th round of the 1986 draft before making his first Pro Bowl 15 seasons later in 2001.

This entry was posted in Uncategorized and tagged , , , , . Bookmark the permalink. Post a comment or leave a trackback: Trackback URL.


  1. kleinOH
    Posted April 29, 2010 at 6:46 am | Permalink

    Forget where I saw this, but someone tweeted, re: Tebow, “he’ll flame out in three years and get elected to congress.” But damn, dude is cut.

  2. carl
    Posted April 26, 2011 at 9:31 am | Permalink

    You should include error bars on your “Seasons” graph.

  3. Daniel
    Posted April 26, 2011 at 9:37 am | Permalink

    Did you account for the sheer number of undrafted players there are each year for your number of pro bowlers per round graph? This is misleading considering there is a finite number of draftees per round each season (32), yet each year there are always so much more than that are undrafted.

  4. Nick K
    Posted April 26, 2011 at 11:27 am | Permalink

    I am afraid you have made some conclusions about the draft process that goes unwarranted. For example, you figure on bowl appearances by draft pick is totally misleading. The figure suggests that you might as well spend your time getting undrafted free agents, because after the first round they are just as likely or more likely to become above-average players than those selected in the 2nd-7th rounds. What the graph should depict is the proportion of players selected in each round (or not selected) that reached the pro bowl. The 180 pro bowl players from the first round represent 56.25% of the 320 players drafted in the first round since 2000. The 80 pro bowl players from the second round represent 25% of the 320 players drafted in the second round since 2000. How many players go undrafted each year? Many many more than 32. Say 100. That would mean that 80 undrafted individuals who made the pro bowl represent a measly 8% chance of an undrafted player becoming an above-average player.

  5. Andre Nyffeler
    Posted May 2, 2011 at 7:01 am | Permalink

    If anything, that probowl data shows that the draft is in fact NOT a lottery as each round produces steadily and predictably less and less probowlers. It is true that undrafted players have about as many probowlers as players in the second round, but the undrafted probowlers come from a player pool roughly ten times as big as the 32 drafted in the second round each year.

  6. Pete
    Posted June 20, 2011 at 7:36 am | Permalink

    Several people have raised more or less the same point, so let me do my best to address those critiques.

    First, I completely agree that the undrafted player pool is many times larger than that eligible for selection in any given round of the draft. I should have made that explicit in the original post. However, I faced a choice when conducting this analysis. The choice revolved around data quality. While it would be relatively easy to compute the percentage of eligible drafted players who made the Pro Bowl, it would have been prohibitively difficult for a blog post project to compute the same percentage for eligible undrafted players. For one, it is by no means obvious how one would define eligibility. Any non-drafted player that signs a contract with a pro-team? What if they received a tryout, but not a full contract? What if they were cut before training camp starts? What if they don’t get signed until their second year out of college? More importantly in my mind, I am not aware of an easily accessed database of undrafted players. In order to find the eligible pool of players, which undoubtedly includes large numbers of players who never saw the field in a regular season game, one would have to likely comb the archives of local sports coverage. That was quite simply not a task I had the time to undertake. So, my choice was to include undrafted Pro Bowl players with imperfect data, or not include them at all. I decided it would be more interesting and representative to report raw numbers that included undrafted players.

    Second, I disagree with the idea that the graph is in any way misleading. It is clear that it reports raw data, not percentages or other metrics that might be used to compare the value of a 4th round pick to the value of an undrafted free agent. I never claimed that undrafted players are as likely to make the Pro Bowl as 2nd round selections. I of course am grateful for the link Gregg Easterbrook gave this post (I’m sure that’s where most of the commenters came from). But it’s not always clear which of the two of us your issues with the interpretation of the data are addressed. My specific claim about this was that, “undrafted free agents are as likely, if not more likely, to make the Pro Bowl as anyone drafted in the 3rd through 7th rounds.” While it was not said as clearly and precisely as I would prefer in hindsight, I simply wanted to draw attention to the fact that it is quite possible to find a Pro Bowl player from the ranks of UFAs, yet they get comparatively little attention from the media. I suspect the gap in attention between UFAs and late round draft picks is more significant than the actual gap in the likelihood of them turning into a quality player. In this I agree with Mr. Easterbrook’s long-standing crusade in favor of increasing attention for UFAs. However, I admit that actually addressing that issue, as discussed above, was not possible given my data.

    There are many more interesting points to be made and discussions to be had about the topics covered in this post. An underexplored one is the marginal value of undrafted free agents, relative to the average signing bonus for draft picks. A very quick scan of data from the 2010 draft indicates 7th round picks got between $69K and $26K in signing bonuses. Finding equivalent information for undrafted players is difficult, but I know from following the Baltimore Ravens that their UFAs tend to sign for $8K-$10K. Maybe it is more valuable, especially for specialists, to go after UFAs instead of drafting them. Another layer in that discussion is, of course, the relative value of different positions.

    Finally, a quick note about data analysis and public presentation. I am truly grateful that people have taken an interest in this and offered feedback. Personally I intended more to draw attention to the first graph, since my main target was the perceived importance of a given draft for altering the competitive landscape of the season immediately following. As I collected data, I realized it would be possible to make something like the second graph, and it seems worthwhile and interesting to do so. But I did this in my spare time. Data does not appear out of thin air. I entered this data manually from because I thought it would make for an interesting study. In presenting it, I was careful to not mislead in any way. My only regret is not explicitly stating that the graph is not representing a comparison of Pro Bowl appearances as a fraction of all eligible players in each draft category. If you have an alternative opinion, I encourage you to conduct your own analysis and present it how you see fit. I’m happy to provide the data to anyone who asks. Perhaps they can supplement it with a comprehensive list of eligible undrafted free agents from the last 10 years.

Post a Comment

Your email is never published nor shared. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>