Everybody loves an underdog. And this time of year, the underdogs are undrafted free agents (UDFAs). Consider what they've already been through and what they have yet to face to make an NFL roster.
There are about 6,000 draft-eligible players per year. Only about 250are actually drafted, and even for them it's in no way guaranteed that they'll make a roster. The other players who think they have a shot at the NFL sit through three days of the draft and don't hear their names called. Some are crushed at not being selected, some didn't expect to be selected in the first place, and all end up in the UDFA pool.
NFL teams occasionally find players like Kurt Warner, Tony Romo and Miles Austin among the ranks of UDFA, and these are great Cinderella stories. But most UDFAs lose any chance of ever playing in the NFL sometime between rookie minicamp and the final roster cuts.
The Cowboys have signed 18 undrafted free agents (UDFA) so far this year.
Based on historical data, what are their chances of making the team? And are they better off with the Cowboys than with other NFL teams?
The Cowboys have had their fair share of Cinderella stories with Romo, Austin, Cliff Harris, Nate Newton, Drew Pearson and many others. Gil Brandt has a recent list of the 75 best undrafted players of all time. Look it up if you're interested, you won't find any lists like that in this post.
A few days ago we looked at late round draft picks. By popular request, today we look at UDFAs. All stats are based on Pro-Football-Reference.com data.
The Raw Numbers: UDFAs in the NFL
There is no way for me to find out how many UDFAs initially signed a free agent- or tryout contract with all NFL teams, only to be subsequently cut before the season started. Based on some of the UDFA signings this year, I would estimate that every team signs between 15 and 20 UDFAs on average, or between 500 and 700 per season. Over the last 10 years that would be between 5,000 and 7,000 UDFAs who tried out for NFL teams.
The number that I do have is this: 1,437 UDFA players over the last 10 years (2000-2009) appeared in at least one regular season game for an NFL team. This is an important number for the rest of the post, because this is the base for what I will consider as 'making the team': Appearing in at least one regular season game means that a UDFA has either been signed to the 53-man roster at the start of the season, has been called up from the practice squad or has been called up off the streets. In all cases, the UDFA has signed an NFL rookie contract (even if only for one game).
If we use the 6,000 as the total number of UDFAs trying out for a team over the last 10 years, the 1,437 means that about a quarter of all UDFAs play in at least one NFL game in their career.
Survival rates
The first target of an UDFA is to make the team, in whatever capacity. Usually this means initial playing time on special teams, or waiting to be called up from the practice squad in case of injuries. The second target should be to make the team for a second year, and then a third and so on. Here's an overview of the survival rates per year for UDFAs:
| 2000 | 2001 | 2002 | 2003 | 2004 | 2005 | 2006 | 2007 | 2008 | 2009 | AVG | |
| Year 1 (no. of UDFAs) | 140 | 148 | 165 | 124 | 155 | 142 | 134 | 136 | 138 | 155 | 144 |
| Year 2 in % | 66% | 55% | 58% | 60% | 61% | 51% | 59% | 59% | 64% | - - | 60% |
| Year 3 in % | 41% | 39% | 39% | 40% | 41% | 33% | 40% | 45% | - - | - - | 40% |
| Year 4 in % | 31% | 27% | 33% | 29% | 33% | 25% | 32% | - - | - - | - - | 30% |
| Year 5 in % | 24% | 24% | 23% | 23% | 27% | 21% | - - | - - | - - | - - | 24% |
On average, between four and five UDFAs play an NFL game per team in their rookie season. Every year about a third of the UDFA rookie class gets cut (60-40-30-24).
The average number of games the UDFAs play in also goes up with tenure. In their rookie season, the UDFAs appear on average in 7.0 games, 10.6 games in the 2nd year, 12.1 in the 3rd, 12.5 in the 4th and 13.0 in the 5th.
Ok, so you're an UDFA and you're suiting up for almost every game, what's next? You want to be a starter of course.
Starters
For the purpose of this analysis, I'll define 'starter' as any UDFA who starts in nine or more games per season. Because of the sheer amount of data points, I cannot weed out players who are starters but missed playing time due to injuries, suspensions or something else. Also, special teamers do not count as starters, which means that UDFA kickers and punters do not show up in the stats for starters.
| Rookie Season | 2nd year | 3rd year | 4th year | 5th year | |
| Roster (1 or more games played) | 1,437 | 757 | 453 | 303 | 208 |
| Starter (Started in 9 or more games) | 28 | 69 | 68 | 74 | 61 |
| % | 2% | 9% | 15% | 24% | 29% |
Over the last 10 years, only 28 UDFA players became starters in their rookie season. By the second season 69 of 757 players who made it into a second season were starters. Note that the 29% in the 5th year does mean that 29% of all UDFAs become starters in the fifth year, merely that 29% of the UDFA players remaining on the roster have earned a starting spot.
In the overall numbers game, getting 70 starters out of 1,282 UDFAs (2000-2008, jury is still out on 2009) is a fairly low number: Only about 5.5% of all UDFAs become starters on their teams for at least one season. Again, keep in mind that this excludes special teams players. Which brings us to the next point, how do the individual positions differ?
Longevity by position
To understand longevity by position, I calculated 'Roster Years', the average number of years players at each position remained on the roster. Also, to get a halfway accurate read on the longevity, I only looked at UDFAs from 2000-2006, so that each UDFA draft class has at least 4 NFL years to its credit.
| T | QB | WR | TE | RB | DE | DT | DB | G | LB | K | P | C | FB | |
| No. of Players | 47 | 20 | 132 | 71 | 109 | 64 | 61 | 191 | 43 | 153 | 31 | 25 | 41 | 20 |
| Avg. Roster Years | 2.2 | 2.3 | 2.4 | 2.6 | 2.6 | 2.7 | 2.7 | 2.9 | 2.9 | 2.9 | 3.2 | 3.5 | 3.6 | 4.0 |
Particularly striking is the relatively low longevity of the offensive skill position players, with QB, WR, TE and RB ranking at the bottom of the table in terms of longevity. Tony Romo and Miles Austin notwithstanding, looking for offensive skill positions (and tackles!) among UDFAs seems like an uphill struggle. Fullbacks, kickers and punters on the other hand look like surer bets. Somewhat surprisingly, centers are also a position where UDFAs seem to have had a fair amount of success.
UDFAs by NFL teams
In the table below I've summarized how each NFL teams ranks against the following four criteria over the last 10 years: No. of UDFAs on the roster, survival rate after year 3, overall longevity of UDFAs and number of starter seasons.
Note on the data method: The team where a UDFA gets his first NFL game experience gets the full credit for that players' career. This is not necessarily the same team that signed that player to his first NFL contract, nor is it necessarily the team where the player plays the bulk of his NFL career. Two examples:
- Former Cowboys like Danny Amendola and Matt Moore initially signed with the Cowboys, but then moved on to other teams where they played their first regular season NFL snaps. In these cases, the Rams and the Panthers get the credit for these two players.
- Similarly, a player like FB Jason McKie initially signed with the Eagles in 2002. He was signed from the Eagles practice squad to the Cowboys' active roster, played in one NFL game and was waived in 2003. The Bears picked him up and he eventually became a starter there. In this case the Cowboys get the credit. Why? Simple: because I can't track the NFL career of 1,437 players accurately.
If you're an UDFA, the best NFL team for you to sign with are the Dallas Cowboys. Not only have the Cowboys developed UDFAs like Romo and Austin to bonafide NFL stars, the Cowboys have historically taken on board more UDFAs than any other team, have one of the highest survival rates and longevities for UDFAs in the NFL, and they've churned out more UDFA starter seasons than most other franchises:
UDFAs by NFL team 2000-2009 (click column header to sort)
| Team | No. of UDFA on roster |
Rank | Survival rate after 3 years | Rank | Average longevity | Rank | No. of starter season | Rank | Avg. Rank |
| DAL | 67 | 1 | 43% | 2 | 2.8 | 3 | 22 | 4 | 2.5 |
| PHI | 41 | 21 | 49% | 1 | 3.0 | 1 | 26 | 1 | 6.0 |
| GB | 53 | 5 | 40% | 6 | 2.7 | 7 | 19 | 7 | 6.25 |
| NYG | 52 | 6 | 40% | 5 | 2.6 | 9 | 15 | 10 | 7.5 |
| BUF | 51 | 7 | 39% | 7 | 2.6 | 10 | 14 | 11 | 8.75 |
| BAL | 48 | 10 | 38% | 9 | 2.5 | 14 | 23 | 3 | 9.0 |
| TB | 46 | 15 | 37% | 11 | 3.0 | 2 | 18 | 8 | 9.0 |
| PIT | 41 | 22 | 39% | 8 | 2.7 | 5 | 19 | 6 | 10.25 |
| IND | 56 | 3 | 29% | 19 | 2.3 | 18 | 26 | 2 | 10.5 |
| SD | 50 | 9 | 42% | 4 | 2.6 | 12 | 10 | 19 | 11.0 |
| MIN | 46 | 14 | 37% | 10 | 2.6 | 13 | 12 | 14 | 12.75 |
| DEN | 38 | 28 | 42% | 3 | 2.7 | 6 | 10 | 18 | 13.75 |
| WAS | 59 | 2 | 24% | 25 | 2.1 | 26 | 21 | 5 | 14.5 |
| NYJ | 47 | 12 | 30% | 18 | 2.4 | 17 | 14 | 12 | 14.75 |
| NO | 39 | 26 | 36% | 13 | 2.7 | 4 | 10 | 17 | 15.0 |
| CAR | 38 | 27 | 37% | 12 | 2.6 | 11 | 13 | 13 | 15.75 |
| NE | 40 | 24 | 35% | 14 | 2.7 | 8 | 9 | 20 | 16.5 |
| CLE | 42 | 20 | 24% | 24 | 2.2 | 23 | 16 | 9 | 19.0 |
| KC | 47 | 11 | 21% | 29 | 2.1 | 25 | 11 | 15 | 20.0 |
| ARI | 55 | 4 | 24% | 26 | 2.1 | 27 | 8 | 24 | 20.25 |
| ATL | 46 | 13 | 30% | 17 | 2.2 | 24 | 5 | 27 | 20.25 |
| CIN | 42 | 19 | 24% | 23 | 2.1 | 29 | 11 | 16 | 21.75 |
| DET | 39 | 25 | 23% | 27 | 2.4 | 15 | 8 | 21 | 22.0 |
| CHI | 36 | 29 | 33% | 15 | 2.3 | 21 | 8 | 23 | 22.0 |
| SF | 41 | 23 | 27% | 20 | 2.2 | 22 | 7 | 25 | 22.5 |
| TEN | 43 | 18 | 19% | 32 | 2.3 | 20 | 8 | 22 | 23.0 |
| OAK | 32 | 31 | 31% | 16 | 2.3 | 19 | 5 | 26 | 23.0 |
| JAC | 50 | 8 | 24% | 22 | 1.8 | 32 | 1 | 32 | 23.5 |
| SEA | 36 | 30 | 22% | 28 | 2.4 | 16 | 3 | 29 | 25.75 |
| STL | 43 | 17 | 19% | 31 | 2.1 | 30 | 4 | 28 | 26.5 |
| MIA | 45 | 16 | 20% | 30 | 2.0 | 31 | 3 | 31 | 27.0 |
| HOU | 28 | 32 | 25% | 21 | 2.1 | 28 | 3 | 30 | 27.75 |
Notes on the data:
- UDFA on the roster: Number of undrafted players between 2000 and 2009 who had at least one regular season game with the team in their rookie season.
- Survival rate after 3 years: Percent of UDFAs still on the roster in the 3rd season. Data limited to 2000-2007, years in which all UDFAs have at least 3 years NFL experience.
- Average longevity: Average number of seasons a UDFA is on the roster. Data set based on 2000-2009, so the averages are lower than in the 'longevity by position' table further up the post.
- No of starter seasons: Total number of seasons in which an UDFA started in 9 or more games.
- Houston: The Texans only joined the league in 2002, so their numbers are naturally lower than those of other teams, but I couldn't be bothered to correct for that.
Observations:
Three NFC East teams are in the top four, the Cowboys, Eagles and Giants. The Eagles appear to have taken a historically slightly different approach to UDFAs, going for a lower quantity but ensuring higher quality. The Cowboys appear to have gone the high quantity route while still maintaining a high quality, the Giants are somewhat in the middle.
The differences between the teams are huge, and this may be a reflection of their UDFA strategies. The numbers suggest that it is likely that some teams regard UDFAs as little more than a cheap source of camp bodies (either that or they are really bad at evaluating talent). Other teams seem to view UDFAs as an extended talent pool from which to further shore up the talent level on their teams.
Also, as you look at the top ten teams on this list, most of them are teams where you would argue that the talent level is already pretty high, yet these teams still get more out of the UDFA pool than teams with lesser talent where you would expect any UDFA who can halfway play to make the roster.
Finally, imagine if you're competing with other teams for a player, say a tight end out of New Hampshire. Showing numbers like the above will certainly give your promises a lot more weight than if you're one of the meat packing plants who find themselves towards the bottom of the list.
The Cowboys have signed 18 UDFAs so far this year. Competition for the 53-man roster is likely to be harder this year than ever before. Chad Reuter of cbssports.com is very high on SS Barry Church to make the Cowboys' roster this year. How many UDFAs do you think make the roster this year and who will it be?