UDFAs vs. Drafted Players, Part 1


via USA Today

Research on several decades of draft results and how players ultimately fared relative to their selection spot has revealed that in the National Hockey League  there’s less value in late first-round picks and more promise in the latter four of the seven rounds. At least that’s what Bill West, writing for the Tribune-Review, seems to think:

“By the 30th pick, you’re under a 40 percent chance of turning into an NHL player, and this is by my really modest definition of an NHL player, meaning you play in 100 NHL games,” said TSN analytics experts Scott Cullen. “The difference in value between having (that) first- and a second-round pick or two seconds is not massive, so it’s not as though you write off this year’s draft and think there’s no chance. It’d be nice for the Penguins if they had more than five picks, but such is the nature of winning the Stanley Cup.”

Cullen’s research seemed right in line with similar research in the NFL which indicates that below the top five-ten picks in the draft there is a rather precipitous decline in how likely the players are to be productive in the NFL. One would think that any pick in the first round is  more likely on average to have an NFL career, although taking Cullen’s “modest” assessment of about one season plus a quarter of another would make it look a lot closer.

But here’s where the research gets interesting:

Rarely is it a surprise for at least 50 percent of a draft class to fall short. But what statistical analysts determined in the past few years about which members of draft classes disappoint raised a few eyebrows.

Michael Schuckers, a professor at St. Lawrence and hockey analytics consultant, in 2011 published research that suggested the probability of landing an NHL lineup regular with the last pick of the first round was closer to that of the first 15 picks of the fourth round than to that of the first 15 picks of the first round.

His work also showed that the probability of a draft pick playing in more than 200 games barely fluctuated for players chosen over the final four rounds, and it barely dropped off from Round 2 to 3.

There are the same number of rounds in the NHL draft as the NFL. But there is one huge difference—the NHL has a “minor league” development process. The NFL does too, in theory—at least that’s what a lot of people consider college football to be. But the goals of a minor league system and of college football are completely different.

Because for a college football program, winning is everything. They aren’t interested in developing players for the NFL, except insofar as having a reputation for producing players who make it to the professional ranks makes your program look good. You run the style of offense which you think is most likely to produce wins, not the one which necessarily prepares players for the NFL.

The NHL draft (at least for American players) is open to players between the ages of 18-20. (There are exceptions for European players, who may enter the draft later if they have never played in the NHL.) But basically you’re taking players to develop at the age that guys enter a college program (or during their first year or so) in football. During their prime development years they are playing in a system which may not prepare them for professional football, and/or against competition which gives them a false sense of their abilities.

Perhaps this explains to at least some degree a really curious statistic I ran across recently.

Based on a break-down of NFL rosters from last year’s [2012] cut-down date (the day rosters are first trimmed to 53 players), there are almost 50 percent more undrafted free agents in the NFL as there are first-round picks, a margin of 412-277.

No team had fewer than nine undrafted players to start the year last year, and the NFL average is nearly 13 per team, meaning at least a few in each training camp are going to get lucky.

Chris Carlson, the author of the piece for syracuse.com. gave some of the possible reasons:

Teams often fill positions like long-snapper, punter, kicker and even fullback without spending a draft pick. And while there were no more than 48 picks in any round this year, the market is flooded with hundreds of undrafted players.

It still seems curious. For the 2012 season, according to their breakdown, the teams with the most UDFAs on the 53-man roster at the beginning of the season were the Packers, Rams, and Saints, each with 19. The Packers had nine first round picks on their roster, the Rams five, and the Saints eight. The four teams with the least UDFAs, nine, were the Chargers, 49ers, Giants and Texans. All of those teams except the Texans had more first round picks than UDFAs on their 53 man roster.

The Steelers were in 19th place with 12 UDFAs compared to 11 first-round picks on the roster. There were only five teams who had more first-round picks than UDFAs on the roster—the three mentioned above with the smallest number of UDFAs, plus the Jets and the Bengals.

And while one might assume the UDFAs were mostly special team guys, according to Carlson only one team did not have at least one starter who entered the league as a UDFA.

Another factor Carlson didn’t take into account was the number of UDFAs signed by teams in comparison to the number they pick in the draft. Seldom does a team have more than one first-round draft pick, and occasionally they don’t have any. In 2014 as of two weeks after the draft (the time in which most UDFA signings take place) there were 487 UDFAs signed, an average of  15.2 per team. The undisputed champion was the Texans with 25, one of whom was Chris Boswell. The team who signed the least was the 49ers, with seven.

In 2014 256 players were drafted altogether, meaning that more than one and a half times as many players as were drafted were signed as undrafted free agents. It’s still interesting, and indicates to me at least that perhaps some of the uncertainty about the development of one’s draft picks has to do with how little relation there is to what they do in their “developmental league” and what they will be asked to do in the NFL.

And as a final note, the player whose picture heads the article might never have had an NFL career had it not been for his developmental time in NFL Europe. Interesting to speculate, anyhow.



  • cold_old_steelers_fan

    Other than the CFL, which is a different game* in many ways, and college, what other developmental leagues exist for NFL football?

    * The leading tackler in the CFL is a slightly undersized linebacker who would be far too small to play in the NFL except as a slightly slow safety. A bigger field and ball changes the game at some very fundamental levels.


    • That’s my point. I think it is time for the nfl to develop their own players. College programs have baseball and hockey and so on and if a kid is drafted after college they still go right to the minors. They might not spend as long there, but they go through the team’s system.

      I realize this will never happen because it would cost the nfl a lot of money. But it seems to me there would be a lot of advantages to getting top prospects before they tear up their knees and get a ton of concussions in a college program.


  • Pingback: Meet the New Steeler: Marcus Tucker | Going Deep:

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s