Based on my 3-part season review (found here, here, and here), I've identified 5 positions that the 49ers need to upgrade, none of which should surprise you that much:

• Quarterback - Although Shaun Hill was clearly an upgrade over JT O'Mulligan, it's hard to escape the fact that (a) Hill had the 28^th best QB DVOA in 2008, and (b) the Niners' pass offense DVOA from Weeks 10-17 was ranked 22^nd. I think both of these stats will improve now that Mike Martz is gone, but that will only take him (and the 49ers) so far. Also, what happens if Hill gets hurt? Do you trust David Carr...I mean Alex Smith...to lead the pass offense for weeks at a time? Or how about the fact that the Niners' only have 2 QBs on the roster at the moment. All of these issues point to the fact that the QB position, as a whole, needs an upgrade. I think a desire to upgrade at QB - either through free agency or the draft - is one reason why Mike Singletary has been reluctant to name Hill the de facto starter for 2009.

• Fullback - At least 4 stats from 2008 suggest the 49ers would benefit from having a lead blocker for Frank Gore: They were 31^st in 3^rd-and-short DVOA, 32^nd in 1^st down rush DVOA, 29^th in red zone rush DVOA, and 32^nd in power running success. And here are 3 non-stat reasons: (a) the new OC, Jimmy Raye, loves FBs; (b) the new RB coach, Tom Rathman, was a FB; and (c) Frank Gore had a blocking FB in 2006, his best season thus far.

• Right Tackle (aka "The Turnstile") - Like many other NFL teams, the 49ers' RT is supposed to be a better run-blocker, whereas its LT is supposed to be a better pass-protector (See file labeled, "Reasons why Kwame Harris Kept his Roster Spot for so Long"). Indeed, the Niners ran right on 24.4% of their RB carries in 2008, as opposed to the 18.5% of carries that went to the left. If that trend is going to continue, they'll have to improve their 29^th and 22^nd rankings on runs behind RT and outside right, respectively.

• Pass Rusher Extraordinaire - It's not a secret that the 49ers' pass defense woes during the Nolan era are at least partially, if not primarily, attributable to a deficiency in pass rushing. Case in point: Their ASRs from 2005-2008 were ranked 32^nd, 20^th, 25^th, and 26^th, respectively.

• Free Safety - Everyone knows by now that Mark Roman is on the chopping block due, in part, to his lack of playmaking ability (1 INT and 0 FFs with the 49ers). As Barrows points out (although his numbers actually were a little off), FS Ed Reed has had 38 more INTs than Roman over the past 7 seasons. Add to that the 2008 Niners' 26^th DVOA ranking when it came to covering #3 WRs, as well as their perennially low-ranked pass defense DVOA, and an upgrade at FS seems necessary.

With the free agency signing period beginning on February 27^th, I'll be spending the next two weeks previewing the 5 urgent position needs detailed above. Obviously, I'll be focusing primarily on the statistical side of free agent evaluation, rather than the organizational (aka "Can the 49ers afford him?"), adversarial (aka "Is he worth the market value?") or availability (aka "Is he going to re-sign before Feb. 27?") sides. However, because positions like pass rusher and FS are limited in the quantity and quality of available stats, I'll freelance a little bit more with non-statistical stuff when I preview those 2 positions.

All in all, my goal for these free agent previews is to give 49er fans some statistical bases for preferring one free agent over another at the most urgent positions of need for the team. In each preview, I'll begin with an argument for evaluating only a select group of free agents that fit a certain profile. Then, I'll present the available stats, and conclude by offering my opinion about which specific free agent I think the Niners should target at that position. First up, the QBs...

IDENTIFYING THE TARGET(S)

By some accounts, the 49ers are on the verge of becoming a playoff contender. Rather than arguing the merits of that claim, I'd rather just assume that, whether or not the claim is true, the 49ers' goal is to make the playoffs. With that said, the question becomes, "Where do playoff QBs come from?"

From 2006 to 2008, there have been 24 different QBs that have led their teams to the playoffs. Before previewing the free agent QBs, we should first determine whether it's even useful to sign free agent QBs as a way to make that playoff leap. Below is a table listing the 24 playoff QBs over the past 3 seasons, along with the year and manner in which they were acquired by their team. Also displayed are the age at which QBs were acquired, if via free agency or trades, and the round in which each QB was selected, if drafted by their playoff team.

After the jump, I'll discuss the table, identify 4 free agent QB targets, and pick the one whose stats best fit what the Niners are looking for...

A couple of housekeeping issues before I discuss what this table suggests with respect to identifying the 49ers' free agent targets. First, Jeff Garcia and Chad Pennington each led two different teams to the playoffs the past 3 seasons. I've listed them twice because they each represent two acquisitions. Second, I gave credit to Garcia for PHI's 2006 playoff appearance due to the fact that (a) he was the QB for their 5-game winning streak to end the season, without which they never would have made the playoffs; and (b) he started both playoff games. Third, although Eli Manning was acquired by NYG, technically, via trade, it was a draft-day trade that had been worked out prior to his selection. Essentially, he was going to be a NYG whether SD traded the 1^st pick to NYG (which they didn't) or selected Manning themselves prior to trading him (which they did). Finally, and this is sort of along the lines of my Eli Manning reasoning, I listed Tony Romo as being drafted even though he was technically an undrafted rookie free agent. Depending on market value, DAL was either going to draft Romo (which they didn't) or wait until shortly after the draft to sign him (which they did). It was highly unlikely on draft day that he was going to end up with a team besides DAL. OK, now that the furniture's dusted and the laundry's folded, let's move onto identifying the target(s).

Looking at how each of the playoff QB starters was acquired, 15 were drafted by their playoff team, 8 were signed as free agents, and 3 were obtained via trade. Converting these numbers into total playoff teams from 2006-2008, 24 of the 36 teams were QB'ed by players that were drafted by the team, 8 were QB'ed by players that were signed as free agents, and 4 were QB'ed by players that were acquired via trade. Clearly, if 2006-2008 is any indication, playoff teams are more likely to have QBs that they drafted. Based on information in the last column of the table, we can specify this conclusion to say that playoff teams who drafted their QB are more likely than not (19 out of 24) to have drafted them in the 1^st round. In other words, the Tom Brady's and Tony Romo's of the world are exceptions rather than the rule.

This is a free agency preview, though, so what can we say about the 8 players who, after being signed as a free agent, QB'ed 25% of the playoff teams since 2006? Well, we can come to 3 conclusions. First, except for Jake Delhomme, all of these QBs had been multiple-year starters with other teams prior to being signed. Second, except for Kurt Warner and Kerry Collins, all of the QBs reached the playoffs in their 1^st or 2^nd year with the team (Delhomme made it in 2003 even though the table only covers 2006-2008). This makes sense on two fronts: (a) Teams that sign free agent QBs are most likely looking for that last (or next-to-last) piece of the puzzle; and (b) Warner and Collins were delayed by the arrivals of Matt Leinart and Vince Young, respectively. The final conclusion about playoff QBs who were acquired via free agency is that, except for Drew Brees (and perhaps Delhomme), they all were over the hill (Aside: Being 31 right now, I don't much like the idea of saying 32 is over the hill.).

In summary, the stats suggest that NFL playoff teams are typically QB'ed either by a player they drafted in the 1^st round or an old, multiple-year starter they signed as a free agent. Also, teams that reach the playoffs after signing a free agent QB were typically only a player or two away from playoff contention prior to the signing, and reached the playoffs within 2 years of the signing.

THE LITTER

Here's how I would apply the above stats to the 49ers' current situation:

• Scenario A: If Mike Singletary and Scot McCloughan (McSingle? Is that on the dollar menu?) believe that the Niners are more than a player or two away from the playoffs, then they should go the 1^st-round QB route.

• Scenario B: If McSingle thinks the Niners are on the brink of playoff contention, then they should go the over-the-hill free agent QB route.

It's certainly open to debate whether the 49ers are closer to Scenario A or Scenario B. The obvious, and much-speculated, result of Scenario A is Mark Sanchez being taken with the 10^th pick in April. The result of Scenario B is not so obvious, so here are 3 free agent QBs who meet the age and multi-year starter criteria, and 1 who's close on age but meets the experience criterion:

Kurt Warner, age 38

Jeff Garcia, age 39

Kerry Collins, age 36

Rex Grossman, age 29

As I said earlier, my main goal here is to tell you how these free agents rank according to the relevant stats. Factors like availability and sign-ability have been (and will be) accounted for by other people. For the purposes of this evaluation, I'll be judging QBs using the following stats, most of which you've seen in my previous articles: DYAR, individual passing DVOA, team passing DVOA, and the discrepancy between actual yards and expected yards (EYds). For all stats, I'll be displaying each player's 3-year average. However, for DYAR, individual DVOA, and EYds, I'll also be displaying a weighted 3-year average, which gives priority to seasons in which a QB threw more than 100 passes.* In reading these tables, if a QBs weighted 3-year average is different from his regular 3-year average, then his weighted average is the more reliable stat.

Without further ado, here are the stats for Warner, Garcia, Collins, and Grossman:

Based only on 2008, you'd have to conclude that Warner was - statistically - the best QB of the 4. He was 3^rd in DYAR among NFL QBs, had the 6^th-best individual passing DVOA, led the 7^th-best passing offense to the Super Bowl, and would have shattered the NFL record for passing yards in a season had he played in game situations of average difficulty (See EYds). However, when 2006 and 2007 are included, the gap between Warner and Garcia narrows considerably. In fact, while Warner has the better 3-year DYAR average, Garcia has the better 3-year DVOA average. Similarly, while Warner has the best single DVOA season of the 4 free agent QBs, Garcia is the only one with 3 consecutive seasons of highly efficient play.

Here's a different way of looking at it. Remember that DYAR is an indicator of a player's value to his team because it measures total passing yardage above what a league-average replacement player would have thrown for in an identical set of situations. In contrast, DVOA is an indicator of a player's efficiency because it measures the success of each play compared to league average performance in that specific situation. Therefore, what the stats in the table suggest is that, over the past 3 seasons, Warner has been the most valuable QB in the free agent litter, whereas Garcia has been the most efficient.

As far as the other 2 QBs in the litter go, both had at least 1 season in which they threw fewer than 100 passes, so their weighted 3-year averages are what we want to look at. According to the stats, then, Collins has been slightly more valuable than the average replacement QB over the past 3 seasons, and basically average in efficiency, while Grossman has been neither valuable nor efficient.

PICK OF THE LITTER

My distinction between the interpretations of DYAR and DVOA is relevant for the 49ers because Singletary has made no secret of the fact that he wants a QB who simply manages the game. Translation: He wants an efficient QB. You need look no further than "game management" QB incarnate, Shaun Hill. Certainly, then, Singletary's desire for an efficient QB disqualifies Grossman from consideration.

Among the other 3, I'd look at it this way. Warner's been more valuable than efficient. Garcia's been more efficient than valuable. Collins has been equally mediocre in value and efficiency. In terms of consistency from year-to-year, Garcia's been the most consistent by far. However, Warner and Collins had to wait out their teams' flirtations with 2006 draft picks, so I can't really tag them with the "inconsistent QB" rap...unless, of course, their performance renaissances in 2008 were merely a byproduct of impending free agency. Then again, who knows for sure? Unfortunately, "who knows," is a major problem for a Singletary-led team. Above all else, Singletary needs to know what he's going to get from his QB, and what he better be getting is consistent efficiency. With Samurai Mike, the operating principle at QB is "No alarms, no surprises." Therefore, this is how I rank the 4 players that fit the old(ish), multiple-year starter standard for successful free agent QBs...if McSingle decides to go the free agency route:

1. Jeff Garcia
2. Kurt Warner
3. Kerry Collins
4. Rex Grossman

On Thursday, I'll be previewing the RTs and FBs. TO BE CONTINUED...

*Football Outsiders' advanced stats for seasons in which a QB threw fewer than 100 passes are less reliable because of the variability associated with smaller samples. Basically, the fewer number of passes thrown by a QB, the more likely it is that his performance on those passes was atypical for him. This leads to a stat that has a larger margin of error, and is therefore relatively untrustworthy. To remedy this, I calculated the proportional amount of sampling error added to below-100 QB stats by virtue of their relative number of attempts, and weighted the below-100 season's stats accordingly. The end result is that, for a crude calculation of weighted 3-year averages, above-100 season stats counted 3 times as much as below-100 season stats.

**DVOA, DYAR, ALY, and ASR statistics used to produce this article were obtained from Football Outsiders.