Writers of Pro Football Prospectus 2008

03 Feb 2013

Roster Construction, Super Bowl XLVII-Style

How do you build a good roster in the NFL? Let's start with this question in its most basic form, how can you build a roster of players? As I see it, there are five basic ways to obtain players in the NFL:

a. Draft them. If you draft players, you have them under your control for four seasons (five if they're a first-round pick) at a relatively reasonable price. When you see people put together lists of the best value for the money players in the NFL, look at how many of them are drafted players in the third or fourth year of their career. It'll probably be at least half of them. Of course, this potentially cheap ride only lasts for so long. Both starting quarterbacks fall into this category. Of course, as the Ravens know with Joe Flacco, once that period ends, you have to ...

b. Retain your own drafted players. If you draft good players, sooner or later they'll pass the period of draft control and want more money. You can pay them or not pay them. If they're really good and not asking for too much money, teams will very likely retain them. This is a good way to get (and keep) star players. Both starting running backs fall into this category. Note some teams pay players before they have to, though San Francisco's NaVorro Bowman is the only player in this game still in his draft control period to have received a second contract. Of course, in the era of the salary cap, you can't keep everybody.

c. Sign them as veteran free agents. If you don't retain your drafted players, when their control period is up, they can sign with anybody. Of course, you can sign everybody else's unwanted players as well. Players in this category range from the very good and expensive, like San Francisco's Justin Smith, to the not as great and cheap, like Baltimore's James Ihedigbo.

d. Sign them as young free agents. Of course, not every player makes it to four years with their drafted team. Some simply aren't drafted and can be signed as undrafted rookie free agents. Baltimore's Justin Tucker is in this category. Others are drafted, but get cut by their drafting team. San Francisco's Darcel McBath was a Josh McDaniels-era draft pick who didn't last long in Denver. This is a bit of an artificial split from the previous category, but I believe it's an important one because these players are cost-controlled in the same way draft picks are. Of course, teams may choose to pay players in this category to a greater or lesser degree, like the 49ers paid Alex Boone. Note I also included players acquired on waivers in this category.

e. Trade for them. Only one player playing in Super Bowl XLVII, Anquan Boldin, was acquired via trade.

So, how did the Ravens and 49ers build their teams?

Chart? Chart!

49ers Ravens
Draft Pick on First Contract 14 22
Re-signed Draft Pick 13 7
Veteran Free Agent 12 13
Young Free Agent 14 10
Trade 0 1

Impressions? Well, a few things stand out.

1. The Ravens have a lot more players from their recent drafts on their current 53-man roster than do the 49ers. One obvious guess is that the Ravens have had more draft picks over the past four years (plus) than the 49ers. This is not true. Over the last four years (plus the 2008 first-round), the Ravens have had 30 draft picks. San Francisco had actually had more, 33. True, Baltimore has had more picks in the first three rounds, but their advantage there is very slight (13 v 12). Yes, that I don't include NaVorro Bowman in the first category because the 49ers paid him matters, but the difference is still a big one.

2. Though most of the Ravens' recent draft picks are represented on their roster, the number of key players under draft control is basically the same for both teams. Using the conference championship games as our guide, the 49ers have 7 starters and 7 backups from recent drafts, while the Ravens have 8 starters and 14 backups. Different backups have different degrees of utility to their team, but this seems to indicate the difference in importance in recent drafts between the two teams is not as stark as the 22-14 disparity might lead you to believe.

3. The 49ers have retained a lot more of their draft picks from non-recent seasons than the Ravens have, keeping twice as many. One difference between what the two teams have done stands out. In re-signing their own players, the Ravens have re-signed only premium players. Of the seven players they've retained, punter Sam Koch is the only one who was signed to a deal with an Average Per Year (APY) of less than $6 million. The other players they kept are Ray Lewis, Haloti Ngata, Ed Reed, Ray Rice, Terrell Suggs, and Marshal Yanda. The 49ers have retained their own high-profile players, like Vernon Davis, Frank Gore, and Patrick Willis, but have also retained lesser-profile players and non-standout starters like Tarell Brown and Isaac Sopoaga.

What accounts for this disparity? One possible factor is that the Ravens are intentionally following more of a star-driven model, retaining a limited number of key players and letting the rest walk, while the 49ers are trying to build a more balanced roster. Another possibility is that Baltimore's more successful recent history has made their free agents more desirable in the eyes of the rest of the league, driving up their re-signing price and forcing them to have to make harsher decisions than the 49ers have had to make. This sort of explanation makes me uncomfortable, for several reasons, but I can't entirely dismiss it. Given what the 49ers have done the past couple seasons, one way to test this will be to look at how teams treat 49ers players in free agency next offseason-how aggressively will other teams pursue players like Dashon Goldson, and will they be stars or more disappointments like Adalius Thomas?

4. Both teams have dived into the free agent market, but mostly at the bottom end. In terms of APY, the top-paid player the Ravens acquired as a veteran free agent is left tackle Bryant McKinnie at $3.75 million. This is a reasonable starting salary, but it's not an excessive one. Eight of Baltimore's 13 veteran free agents were signed to contracts with an APY of less than $2 million. San Francisco has splurged a little bit more-Justin Smith was a big-dollar free agent acquisition, and Carlos Rogers' current contract has an APY of over $7 million as well. Beyond Smith and Rogers, though, their free agent salary is that of the Ravens, with a couple starters making in the $3 million range and a few players making less than that.

5. Some teams have a good number of young free agents on their roster. Other teams do not. Both the Ravens and 49ers are among the teams that do. Of course, late in the season these numbers look slightly different than they would in Week 1, but both teams have of necessity relied upon bodies in this pool. Unsurprisingly, players from this pool are mostly backups and other minor contributors, but they are almost all cheap. Eight of the ten Ravens players in this category have an APY of less than $1 million, while 11 of the 14 49ers players do.

6.  NFL trades of players are hard and rare. In the right circumstances, they can help your team, but don't count on them as a team-building strategy. I know, this is far from news, but trades involving players are still talked about a lot more than they happen.

What does this mean for your team?

Well, if you're a 49ers or Ravens fan, enjoy the game today. Super Bowl appearances are not birthrights, and just because you think your team is well-positioned to make a run back the next year doesn't mean they'll make it. Then again, both teams have been there before but not recently, so you probably know that. Also, winning championships is awesome. Watching your preferred team win a championship isn't quite as good, but it's still a recommended experience.

I don't think there's any great insight here. The core of both teams is similar: a mix of players the team drafted and retained, paying something like a good market wage to do so. To that they added another good complement of players in their draft control period. Together, those two groups account for just over half the roster and close to two-thirds of the key players for both teams. Veteran free agency can be used to supplement the lineup with a couple key starters. If you have enough good players without doing so, you can have a very good team without signing any big money free agents somebody else let go (Baltimore), but you can have success signing big money free agents as well, as long as those players really are that good (San Francisco). Both teams also have about four veteran free agents making starter-type money ($3 million). Diving into this part of the free agency pool can be a very useful method of filling draft and/or development holes, but the core of your team is built elsewhere. It's also very helpful to find a couple stater-type players as young free agents, but like veteran free agency is probably not an area you want to rely upon in building your team.

Posted by: Tom Gower on 03 Feb 2013

11 comments, Last at 06 Feb 2013, 3:54pm by bravehoptoad


by Guido Merkens :: Sun, 02/03/2013 - 2:35pm

By now it seems to be common knowledge that teams built for long-term success build through the draft, and that will probably be even more the case as the effects of the new rookie pay scale are felt. Having a pool of low-paid players on your roster is part of that, of course, but another major benefit that doesn't seem to be discussed is how drafting well allows a team's free agency money to be spent more safely.

One of the main dangers in signing high-priced free agents from other teams is the question of system fit. If a player was successful enough with his old team to command a free agency contract, chances are pretty high that he fit the system of his old team pretty well. However, it's difficult to project whether the player will be worth the same salary in the system of the team that signs him. There are exceptions, of course - Justin Smith is more valuable in the 49ers' 3-4 system than he was in the Bengals' 4-3 - but signing a player is taking a risk that he might not be worth the same money in your system.

When a team like the 49ers or Ravens considers signing a drafted player to a new contract, they already know how much production to expect in their system, because they've seen it for four years. The chance that their dollars will be wasted is very small. But when a team signs another team's players (insert Redskins joke), even if the players are young and proven, they might have wasted their money just because the signed player won't produce at the same level in a new system.

by Anonymous999 (not verified) :: Sun, 02/03/2013 - 3:13pm

I think you mean Alex Boone , not Adam boone. Good article, though.

by Tom Gower :: Sun, 02/03/2013 - 3:56pm

Yes, I did. Thanks. Fixed.

by David m (not verified) :: Sun, 02/03/2013 - 3:39pm

Great article, tiny nitpick: Larry grant left the 49ers for the rams, then was resigned, not actually retained.

by Tom Gower :: Sun, 02/03/2013 - 3:57pm

That completely slipped my mind. Thanks. Fixed.

by Jerry :: Sun, 02/03/2013 - 5:40pm

This would require a lot more research, but is some of the difference between the Ravens and 49ers where they are on the success curve? The Ravens have been good longer than the Niners (speaking currently, of course), which may account for some of how they acquire talent. You'd probably have to look at how all teams are constructed to have an idea of whether successful teams do something different, and what the success curve looks like.

by wiesengrund :: Mon, 02/04/2013 - 6:08pm

Speaking of success in recent years, I think it's also remarkable that the Ravens look like the more draft-based team even after they picked after SF in most of the last five years. SF had picks 7, 10, 11 and 17 during that timeframe, Baltimore never picked earlier than 18.

by theslothook :: Mon, 02/04/2013 - 12:36am

I think Tom is right on about many of the points it takes to be a superbowl team, but I firmly am a believer that the draft is and will always be 100 percent luck(unless you're an idiot like Matt Millen).

I've done research on this, there is no team, NO TEAM that has shown a consistent ability to hit on draft picks either by round or by position in the nfl. my research goes all the way back to 1980. It is all luck and circumstance.

To me, the best example of this are the seahawks. Think about it, they are known for their defensive backfield, but look where they got those players. Outside of Thomas, the others were low rounders that happened to develop. That isn't skill, thats serendipity.

by Turin :: Mon, 02/04/2013 - 6:00pm

Out of curiosity have you broken it down by coaches and/or GMs instead of by team?

by bravehoptoad :: Wed, 02/06/2013 - 3:54pm

Why would you break down success by team? It doesn't seem that would show anything at all, which is apparently what you're seeing. What you'd want to look at is GM.

by someguy (not verified) :: Mon, 02/04/2013 - 12:58pm

Would love to see that player chart for every team in the league. Maybe slip in a DVOA column too...