If you had a chance to suggest how to use the 53 roster positions to Jason Garrett, what would you tell him?
The Dallas Cowboys, like all NFL teams, face two primary limits in building their squad. First, of course, is the salary cap. Despite a penalty this year (that still rankles), the Cowboys have, through the creative work of the front office, particularly Stephen Jones, done a very good job of having the money they have needed this year to build the club. Dallas, of late, has been very good at manipulating things to get the answer they need out of the cap.
The second number is much harder to manipulate. It is the 53-man roster. Figuring out how to put your various positions together to add up to that number is the decision that underlies the cuts in training camp, particularly the last few. Who makes the team as opposed to who looks for another job, or possibly goes to the practice squad, can come down as much to what position the man plays as his talent.
KD did a power ranking of what he judges to be the top 53 players on the field. His article got me thinking some more about this (I had already been mulling it over, to some degree, for weeks) and I decided it was about time to look at the topic in some depth.
Since KD essentially created a 53 man roster, I thought I would use that as a starting point. In ranking his top 53, here is the position distribution he came up with:
Obviously, he had to make several choices here. He dropped the third quarterback and the third tight end, went heavy on wide receiver and outside linebackers, and light on the offensive line. Pretty much by definition, his choices were talent driven. Based on what I understand about his evaluations, he did not take a lesser player who was at a position of greater need, but went strictly with the best 53 and let them fall where they would.
It's a pretty solid roster, but take a look at some other ways to approach it after the jump.
For contrast, take a look at how the team was constructed last season. This is a sort of average of the way Dallas used the 53 in 2011. There were changes during the season, primarily due to injuries, but this is the way the slots were used overall.
Obviously, there are several differences in the two. And many other ways to allocate the numbers.
This is going to be something that many of us will try our hand at once training camp gets started and the fortunes of players rise and fall. Deciding how to divvy things up is driven by the two considerations I mentioned above, talent and position of need. But all rosters will have certain minimum requirements. Every team has to have at least two quarterbacks. Dallas will have at least three specialists, a placekicker, punter, and long snapper (some teams use their regular center, or another lineman, for the long snapping duties; and occasionally, some possibly irrational person with more power than foresight will insist on both a placekicker and a kickoff specialist).
So I thought it might be useful for future discussion to set a minimum number for each position. Then the various prognosticators would have something to work with. First fill out your "gotta have" positions, and then you have the left over slots to allocate out.
Here are the numbers I think are the minimum at each position, followed by a little exercise to show you how you can use this to speed up building your own roster.
As you can see, if you buy the basic numbers I have here, there are only seven positions to play with in filling out a roster. I will say the numbers I came up with are very Cowboys specific. For instance, some other teams may go with only two running backs, but I cannot imagine Dallas doing that. Most of the other positions are pretty simple to derive: One starter and one backup in each position, except fullback (and when you just look at running backs and fullbacks as one total, then it does sort of work out the same). Wide receiver and offensive line are different, of course, with the third wideout being a "semi-starter" and the minimum calling for just two backups, and the offensive line reflecting a swing tackle and a swing guard for backups (the backup for center may be able to play guard as well).
Rather than do a whole new 53 to illustrate how this could be applied, I went back and "reverse engineered" KD's list. I went down his list and used the higher ranked names to fill all the minimum requirements I came up with. Applying my criteria, he could be said to have used his "extra" picks the following way:
- Clifton Geathers DL
- Kevin Ogletree WR
- Danny McCray S
- Adrian Hamilton OLB
- Isaac Madison CB
- Aston Whiteside OLB
- Kevin Kowalski G/C
Note that this does not include the last three names on the list. All of them would count as filling a minimum requirement. Just to show alternative positions some or all of those spots could have been used for, consider the following players he did not have on his list (I am not looking at rearranging KD's talent evaluations; for instance in this case I am accepting his judgement that Raymond Radway is not better than Ogletree, and therefore should not be on the 53).
- Lance Dunbar RB
- Stephen McGee QB (not that I'm arguing for including him, but Garrett does like three QBs)
- James Hanna TE
- Caleb McSurdy ILB (still the best football name on the team, IMO)
- David Arkin (still keeping Kowalski and going 10 on the OL)
So it's not all about talent. It is also about need. And I think this is a way to simulate the building of the real Cowboys roster. Jason Garrett and company are probably going to start with their own list of minimum positions and then start figuring how to juggle the last few slots to fill in the team. Then you look at the complications of the practice squad and signing some free agents late to fill a need (QB, WR and TE would be most likely). If you have done your own 53-man projections, you know how maddening this part of the process can be. And if you just flat out disagree on one of the numbers, it is very easy to make adjustments for your own use. Hopefully, this might be a handy aid to speed things up for you or simplify your methodology if you like that kind of thing.