Welcome to Part 4 of my free agency preview. In Part 1, I showed that, if McSingle thinks the 49ers are a playoff contender, then signing an old free agent QB has done the trick for several teams the past few seasons. Of the old free agent QBs that are available, I identified Jeff Garcia as the best option because he's been highly efficient (and consistently so) over the past 3 seasons.
In Part 2, I showed that most starting RTs on NFL playoff teams were drafted, not signed as free agents. However, if the Niners are going to sign a free agent RT, overwhelming evidence suggests that they should sign a player who's started at RT for a previous team, rather than signing a LT and moving him to RT. Originally, I identified Vernon Carey as the pick of the litter based on the Dolphins' performance on RT and RE runs. Since I wrote Part 2, however, Carey's been re-signed by MIA, so Mark Tauscher is currently the best available free agent RT based on his stats.
In Part 3, I showed that playoff teams have an all-or-nothing philosophy about a lead-blocking FB, and that most starting lead-blocking FBs were either drafted late or not drafted at all. In addition, those that were acquired via free agency were signed when they were inexperienced no-names. Although none of the available free agents this offseason fit that profile, I identified Tony Richardson as the best - and most likely - option for the 49ers.
Today, in Part 4, I'll be taking a look at the free agent pass rushers. As I said during Part 1, I'll be freelancing a little bit more (aka not being as stat-heavy) for pass rushers and free safeties (Part 5) because there aren't a lot of good stats out there for individual defenders. Fortunately, as you'll find out, I'm not going to need much in the way of stats to evaluate the available free agent pass rushers that fit the profile of what Singletary and his 3-4 defense will be looking for this offseason.
IDENTIFYING THE TARGET(S)
For the past several seasons, anyone who has regularly watched 49er games has realized that the most glaring weakness on defense has been the pass rush (or lack thereof). Indeed, the stats back up our suspicions: The Niners' adjusted sack rate (ASR) was ranked 32nd in 2004, 32nd in 2005, 20th in 2006, 25th in 2007, and 26th in 2008. In other words, the 49ers' defense has been in the bottom quarter of the league 4 of the past 5 seasons with respect to sacking their opponent's QB. Obviously, you don't need a statistician to tell you that (a) that's not good, and (b) it needs improvement.
But how should the 49ers go about improving it? Should they sign a free agent or draft a pass rusher? If they draft a pass rusher, what position should he have played in college and how early should they draft him? If they sign a free agent pass rusher, what kind of experience should the player have, and how old should he be? How important is their switch to a base 3-4 defense for determining which route they should take? Here's another thought: What if the 49ers' elusive "sack artist" is actually on the roster already? First things first; let's try to remove some of this uncertainty
For this article, I'm defining a "pass rusher" as any player who recorded 4 or more sacks in 2008. I'm using 4 sacks as a threshold because, hey, if you can't sack the QB more than once every 4 games, then you're probably not a "sack artist," or even a "sack artist in the making." Below is a table showing the position breakdown of every sack in 2008 that was recorded by a player with 4 or more sacks:
|
3-4 |
Players |
Sacks |
Average |
|
4-3 |
Players |
Sacks |
Average |
|
LDE |
3 |
18 |
6.0 |
LDE |
19 |
113.5 |
6.0 |
|
|
LDT |
-- |
-- |
-- |
LDT |
4 |
19 |
4.8 |
|
|
NT |
2 |
12 |
6.0 |
NT |
-- |
-- |
-- |
|
|
RDT |
-- |
-- |
-- |
RDT |
8 |
43.5 |
5.4 |
|
|
RDE |
2 |
15 |
7.5 |
RDE |
25 |
180.5 |
7.2 |
|
|
LOLB |
7 |
46.5 |
6.6 |
LOLB |
2 |
10 |
5.0 |
|
|
LILB |
1 |
8 |
8.0 |
LILB |
-- |
-- |
-- |
|
|
MLB |
-- |
-- |
-- |
MLB |
0 |
0 |
0.0 |
|
|
RILB |
2 |
9 |
4.5 |
RILB |
-- |
-- |
-- |
|
|
ROLB |
9 |
91 |
10.1 |
ROLB |
3 |
13 |
4.3 |
|
|
Total |
26 |
199.5 |
7.7 |
Total |
61 |
379.5 |
6.2 |
Immediately, a couple of things should jump out at you. First, only 35% of the "pass rusher" sacks were achieved by teams that use the 3-4 as their base defense. Don't let this alarm you with respect to Singletary's move to a 3-4, though. In the NFL, 9 of the 32 teams - or about 30% - employ a 3-4 base defense. This means that pass rushers in a 3-4 actually sack the QB more than you would expect given the ratio of 4-3 to 3-4 teams in the league. This fact is borne out in the sack averages for each type of defense: Individual pass rushers in a 3-4 average more sacks than pass rushers in a 4-3.* Also, although it's not detailed in the table, 3 of the top 4 individual sack totals in 2008 were achieved by 3-4 OLBs (DeMarcus Ware, Joey Porter, and James Harrison).
After the jump, I'll continue to discuss this table, identify free agent criteria based on current 3-4 pass rushers, make an argument that might be a bit controversial, and tell you what the stats (and my opinion) say the 49ers should do to address their pass rushing deficiency this offseason...
The second, and more important, thing that should jump out at you is where the vast majority of sacks come from in each defense. As is clearly depicted in the table, 4-3 defenses get the lion's share (77.5%) of their sacks from pass rushers playing DE. In contrast, 3-4 defenses get the vast majority (68.9%) of their sacks from pass rushers playing OLB. I don't think we need any more proof of the tactical difference between the two defenses than the fact that (a) pass-rushing DEs in a 4-3 out-sacked pass-rushing DEs in a 3-4 to the tune of 294-33, and (b) pass-rushing OLBs in a 3-4 out-sacked pass-rushing OLBs in a 4-3 by a count of 137.5-23.
A couple of other things, though, might not jump out at you from the table. One is that, although 4-3 DEs and 3-4 OLBs have the greatest share of sacks in general, RDE (in a 4-3) and ROLB (in a 3-4) are the most prolific pass rushers specifically. This evidence is akin to an NFL proof of concept: NFL defenses position their best pass rushers on the right side, so LT is the most important pass-protecting position for NFL offenses. Want 49er-related proof of concept? Parys Haralson, SF's best pass rusher in 2008, played ROLB.
Another subtle point in the table is in line with an argument I made when I reviewed the 49ers' defensive stats: NT and ILB are clearly not pass-rushing positions in a 3-4. So, as it relates to the 49ers, Patrick Willis, Takeo Spikes, and especially Aubrayo Franklin should not be evaluated by their sack totals. This last point is my main reason for thinking that - in agreement with some of you - signing Albert Haynesworth would be a waste of money for SF. Haynesworth has starred at a position (4-3 RDT) where he's both a run-stopper and a pass-rusher (RDTs have the 3rd-most 4-3 sacks). Why should the Niners pay market value (read "A lot of money") for Haynesworth when they're only going to use half of his skill set (i.e., run-stopping)? Conversely, why would Haynesworth sign with SF to toil away in NT anonymity when he can continue getting accolades as a total-package RDT? To me, that signing just wouldn't make sense...from either party's perspective.
So it's pretty clear that, in the 3-4 defense, OLBs are the primary pass rushers. This means that, if the 49ers want to improve their pass rush in 2009, they should do so by upgrading at OLB (Aside: This poses an interesting problem, but I'll get to that a little bit later.). Therefore, it's time to answer the question: "Where do pass-rushing OLBs come from?" As I said earlier, there are 9 teams in the NFL that play a base 3-4, which makes the total number of starting OLBs 18. Of these 18, 16 had more than 4 sacks in 2008. I'm pretty sure you can guess one of the teams that had a starting OLB with less than 4 sacks (Hint: Their nickname rhymes with, and is a synonym for, "Sporty miners"). Below is a table showing the 16 players who had 4 or more sacks in 2008 from the OLB position in a 3-4 defense:
|
Player |
Team |
Draft Team |
Sacks |
How Acq |
Draft Round |
Age Acq |
College Pos |
2008 Pos |
|
Greg Ellis |
DAL |
DAL |
8.0 |
Draft |
1 |
-- |
DE |
LOLB |
|
Bryan Thomas |
NYJ |
NYJ |
5.5 |
Draft |
1 |
-- |
DE |
LOLB |
|
DeMarcus Ware |
DAL |
DAL |
20.0 |
Draft |
1 |
-- |
DE |
ROLB |
|
Terrell Suggs |
BAL |
BAL |
8.0 |
Draft |
1 |
-- |
DE |
ROLB |
|
Kamerion Wimbley |
CLE |
CLE |
4.0 |
Draft |
1 |
-- |
DE |
ROLB |
|
LaMarr Woodley |
PIT |
PIT |
11.5 |
Draft |
2 |
-- |
DE/OLB |
LOLB |
|
Matt Roth |
MIA |
MIA |
5.0 |
Draft |
2 |
-- |
DE |
LOLB |
|
Shaun Phillips |
SD |
SD |
7.5 |
Draft |
4 |
-- |
DE |
LOLB |
|
Jarret Johnson |
BAL |
BAL |
5.0 |
Draft |
4 |
-- |
DE |
LOLB |
|
Parys Haralson |
SF |
SF |
8.0 |
Draft |
5 |
-- |
DE |
ROLB |
|
Jyles Tucker |
SD |
SD |
5.5 |
Draft |
ND |
-- |
DE |
ROLB |
|
Mike Vrabel |
NE |
PIT |
4.0 |
Free Agent |
3 |
26 |
DE |
LOLB |
|
James Harrison |
PIT |
PIT |
16.0 |
Free Agent |
ND |
26 |
OLB |
ROLB |
|
Calvin Pace |
NYJ |
ARI |
7.0 |
Free Agent |
1 |
27 |
DE |
ROLB |
|
Joey Porter |
MIA |
PIT |
17.5 |
Free Agent |
3 |
30 |
DE |
ROLB |
|
Adalius Thomas |
NE |
BAL |
5.0 |
Free Agent |
6 |
30 |
DE |
ROLB |
By now, you should be familiar with most of the abbreviations and classifications in the table. All I'll remind you of is that (a) "Draft" players with ND in the draft round column were signed as undrafted free agents by their current team, and (b) a player's primary position in college is listed first if he started extensively at multiple positions. OK, back to the entree...
Among the NFL's 3-4 defenses, 11 of the 16 pass-rushing OLBs were selected by their current team in the draft. Of these 11, 5 were 1st-round picks, and 9 were chosen in the 4th round or earlier. In addition, 15 of the 16 pass-rushing OLBs played DE as their primary position in college. This is most likely due to (a) differences between the height/weight requirements for OLBs in college vs. the NFL, and (b) the overwhelming prevalence of 4-3 defenses in college.
Of the 5 pass-rushing OLBs that were signed as free agents, 3 were drafted by PIT, 4 are ROLBs, all 5 were acquired by the age of 30, and all 5 were signed away from teams that used a base 3-4 defense during the season immediately preceding their acquisition. This last point is similar to what I found out about starting RTs in the NFL (i.e., they had recent experience as a starting RT), and suggests that - despite opinions to the contrary - NFL teams prefer signing players who have previous starting experience at the position of interest. In other words, the NFL does not - for the most part - engage in free agency metamorphosis. They rely on other teams to figure out what a player's skill set is, and to position him accordingly. When they sign a free agent to be their starter, it's usually with the intention of keeping him at that specific position.
So what does the information in the table suggest about the origin(s) of pass-rushing, 3-4 OLBs? Well, most were acquired through the draft by teams who selected them in the early rounds, and converted them from college DE to NFL OLB. If a team signed a free agent, pass-rushing OLB, they did so when the player was relatively young and had recent experience in the 3-4 defense (BAL and ARI dabbled in the 3-4 during Pace's and Thomas's final seasons with the team). Also, that free agent signing was likely an attempt to fill the ROLB (aka primary pass-rusher) position.
In summary, then, 3-4 defenses sack the QB more than you would expect given that there are only 9 of them, and 3-4 pass rushers average more sacks than 4-3 pass rushers. In the 3-4, OLBs are the primary pass rushers, and ROLB is usually where teams position their best pass rusher. Most 3-4 OLBs were drafted (usually early) by their current teams as college DEs. Of the 3-4 OLBs that were signed as free agents, all were under 31 when they were signed, all had recent experience as an OLB in the 3-4, and almost all were signed to play ROLB.
THE LITTER
Given that they'll be playing a base 3-4 defense this season, here's how I would apply the above stats to the 49ers' current situation:
- They should probably draft a "sack artist" rather than sign him as a free agent. If they choose to do so, they should target a college DE in the early rounds, with the intention of converting him to OLB.
- If they choose to sign a free agent pass rusher, that free agent should be young, experienced in the 3-4, and acquired with the intention of starting him at ROLB. They should not (and probably don't want to), for example, sign a starting 4-3 RDE, and try to move him to ROLB when he arrives in SF.
These recommendations present an interesting problem for (a) the 49ers' foray into OLB free agency, and (b) my preview of OLB free agents. Simply, there aren't any free agent OLBs who fit the evidence-based criteria. Special thanks here go to BAL for franchising Terrell Suggs. Of course, it's much more important for the 49ers to adjust to this problem than it is for me to adjust to it. Well, actually, what if perhaps this is actually a blessing in disguise for the Niners? Personally, I tend to think it is; here's why.
I just mentioned Terrell Suggs on purpose. In their pre-franchising Suggs-lust, it seems to have slipped past many Niner fans that Suggs only had 8 sacks in 2008 (See the above table) as a ROLB, which is equal to the number recorded by - wait for it - Parys Haralson. To boot, Haralson's 8 sacks came in 10 starts, whereas Suggs' 8 sacks came in 16 starts. So basically, and keeping in mind I'm only talking about one season, Haralson averaged about a sack per game, while Suggs averaged a sack every 2 games.
Here are a couple of other interesting tidbits of information related to the Haralson/Suggs comparison. As I mentioned in the comments section of my Niner defense review, Singletary started the same 11 players in a 3-4 defense for 5 of the last 6 games this past season. It turns out that Haralson got 3.5 of his 8 sacks during those 5 games, all of which he started at ROLB. So, obviously, the main pass-rushing beneficiary of switching to the 3-4 seems to have been Parys Haralson. The other tidbit you asked? Well, Haralson will be 25 at the beginning of the 2009 season, while Suggs will be 26.
The point of my Haralson/Suggs comparison is to suggest that, perhaps, the 49ers already have a young, blossoming, "sack artist" on their roster who has experience in the 3-4, and, therefore, that they don't need to go out and sign one in free agency. This is especially true now that BAL used the franchise tag to take Suggs off the market. But really, even if Suggs was available, why spend the money when you already have a player on your roster with similar (arguably better) 2008 stats, a later birth date, and an apparent acumen for registering sacks when starting at ROLB in the 3-4 defense?
So my vote (and that of the stats) is clear: Screw signing a free agent ROLB. If anything, the 49ers might be better off signing a LOLB to pair with Haralson on passing downs. After all, LOLB is where most of the non-ROLB sacks come from in the 3-4, right? Well, signing a LOLB in free agency poses additional problems. First, there aren't any young, 3-4 LOLBs on the market either. Second, the Niners don't really use a LOLB much on passing downs. Instead, they go to one of their nickel packages, which sends LOLB Manny Lawson to the bench, has Willis and Haralson at LB, and adds either an extra CB (i.e., Sub package) or an extra S (i.e., Big Sub package) to the field. So basically, the benefit of signing a pass-rushing LOLB would be attenuated by the Niners' substitution tendencies.
Taking all considerations into account then, the 49ers should probably stay away from free agency in their quest to find a "sack artist." Therefore - and sorry if this disappoints you - I'm not going to evaluate any free agent OLBs.
PICK OF THE LITTER
Let me just recap my logic here:
- The 49ers are going to be using the 3-4 as their base defense in 2009.
- The stats suggest that sacks in a 3-4 defense come predominately from pass-rushing OLBs, especially the ROLB.
- The stats say that OLBs for 3-4 defenses are acquired predominately through the draft. Those that are signed as free agents tend to be young and have recent experience in the 3-4.
- This offseason, no available free agent OLBs meet the evidence-based criteria.
- The 49ers already have a starting ROLB who is young, experienced in the 3-4, had as many sacks in 2008 as the once-coveted Terrell Suggs, and blossomed as a pass-rusher when Singletary switched to the base 3-4.
- The 49ers substitute a CB or S for their LOLB on passing downs, so signing a pass-rushing free agent LOLB (of which there are none), wouldn't benefit them much. Basically, it's kind of pointless to sign a pass-rushing LOLB when the LOLB doesn't play on, you know, passing downs.
Based on this logic, I'll tell you what I think the 49ers should do to upgrade their pass rush this offseason, rather than ranking free agent OLBs who meet certain criteria (of which there are none):
• Start Haralson at 3-4 ROLB for the entire season.
• Draft a college DE in the first few rounds with the intention of converting him to pass-rushing OLB. This gives the Niners depth if Haralson gets hurt, and flexibility if they want to play less nickel on passing downs. Right now, Lawson doesn't give them that flexibility.
• Be more aggressive in their defensive play-calling. In other words, "release the hounds!"
On Thursday, I'll be previewing the free safeties and summing up all 5 free agency preview articles. TO BE CONTINUED...
*I performed an independent samples t-test to see if there was a statistically significant difference between the sack averages of 3-4 and 4-3 pass rushers. The p-value for this test was .06, which just misses the traditional standard for statistical significance (p < .05). As the traditional standard of "statistical significance" is arbitrary (e.g., I could choose p < .10, rather than p < .05, as being the cutoff for statistical significance), the result of my t-test was meaningful nevertheless. Basically, it means there's a 94% probability that 3-4 pass rushers actually have higher sack averages than 4-3 pass rushers, or, alternatively, that there's only a 6% chance this difference is due to statistical dumb luck (aka random chance). Also, with only 26 pass rushers in the 3-4 sample, it's highly likely that the difference would be statistically significant if the sample size was increased to 61 (aka equal to the sample size of 4-3 pass rushers).
**ASR statistics used to produce this article were obtained from Football Outsiders.