Football Outsiders and the 49ers: Player Stats through Week 3

Welcome back for this week's look at Niner player statistics and rankings as published by Football Outsiders (FO). If you're unfamiliar with FO's player stats (i.e., DYAR, individual DVOA, and EYds) and line stats (i.e., ALY and ASR), or if you need a refresher course, take a look at my explanations here and here.

This week, QBs need a total of at least 23 passes to qualify as having valid FO stats, RBs need at least 24 carries, WRs need at least 10 targets, and TEs need at least 5 targets. The only Niner that qualifies this week, but didn't qualify last week is Glen Coffee, who carried the ball 25 times in Week 3 due to Frank Gore's (unfortunate) injury. Therefore, the players I'll be breaking down this week are Shaun Hill, Gore, Coffee, Isaac Bruce, and Vernon Davis. I'll also, of course, be presenting FO stats for the OL, the DEF front 7, and the secondary.

Based on the Vikings game, I'm particularly looking for answers to the following questions:

1. Was the OFF's 0-for-11 on 3rd downs a fluke or is it something we might end up seeing more of given Hill's stats through 3 games?
2. Given that Gore will be out for at least the next 2 games, is Coffee an "average" backup as it relates to Gore's DYAR?
3. Did VD's breakout game look as good to FO as it did to us?
4. With another game in the books, how culpable is the OL for the 49ers' offensive inertia?

After the jump, I'll present and discuss the stats, as well as give a pictorial representation of VD's statement to the league last week...

Before I begin, let me just remind you that opponent adjustments don't kick in until after this week's games. Therefore, like last week's, the stats I present today have been adjusted for all relevant factors (e.g., game situations) except what teams the Niners have played.

QUARTERBACKS

Here are Shaun Hill's stats through 3 games (bold = top 8 in the NFL; italics = bottom 8):

DYAR	Rank	Passes	Actual Pass Yards	Yds Per Pass	Rank
5	29	91	474	5.21	27
DVOA	Rank		EYds	EYds Per Pass	Rank
-10.1%	29		434	4.77	30

At first glance, it's apparent that, in every category, Hill ranks at the bottom of the league among starting QBs. However, when looking a little deeper, we find that nearly every stat in this table is an improvement over what his stats were through 2 games. Specifically, his DYAR increased by 15 yards, going from below average in value to just about average in value, his individual DVOA, although still suggestive of below-average efficiency, rose 6.4%, and his EYds per pass rose by about half a yard. The only stat that didn't improve was the discrepancy between Hill's EYds per pass and actual yards per pass, which nearly doubled from .27 to .44. As this is a yardage measure of situational efficiency, it means that Hill's passing yardage was less efficient against MIN than what NFL stats would otherwise indicate.

As a point of discussion, I'd like to focus on Hill's per-pass stats as they relate to 3rd-down inefficiency. Notice that Hill averages only 5.21 actual yards per pass and only 4.77 EYds per pass, good (bad?) for 27th and 30th in the league, respectively. The first point I'll make is that part of this week's .17 increase in Hill's actual-to-Eyds discrepancy is directly attributable to completing 3rd-down passes short of the 1st down marker. This is because, on 3rd down, "success" according to FO is making the 1st down. Obviously, by throwing all 8 of his 3rd-down passes to short routes (see here for my play-by-play) - regardless of whether it was 3rd & 5 or 3rd & 22 I might add - Hill contributed to a decreased likelihood that the play would be successful. In other words, Hill's passes were suggestive of inefficient 3rd down play. Now, whether that's his fault or should instead be blamed on Raye's play-calling is up for debate.

My second point regards what the reliance on short passes means for the future. As FO's stats accurately indicate, short-of-the-marker 3rd down passes are inefficient. Yet all 8 of Hill's passes were short of the marker. Not 4, 5, or 6 of 8; all 8. To me, this is suggestive of an overall philosophy ascribed to either by Hill vis-à-vis check-downs or Raye vis-à-vis conservative play-calling, and therefore indicates that 3rd-down inefficiency is ingrained in the 49ers' OFF. And as we all know, habits are hard to break. Obviously, we're only talking about 3 games, so it's not like this can't be fixed as the season goes on. Indeed, Mike Singletary said that 3rd-down inefficiency in the MIN game would be something that the coaching staff would analyze and evaluate. Therefore, what I'll be looking for as time goes by is whether the 49ers' OFF does, in fact, make the proper adjustment; namely, Hill throwing the ball beyond the marker on 3rd downs.

RUNNING BACKS

As Gore only had 1 carry against the Vikings, his FO stats remain basically unchanged from last week. Therefore, I'm not going to be discussing him until he comes back from his injury (aka out of sight, out of mind). Here are Coffee's stats:

DYAR	Rank	Runs	Actual Run Yards	Yds Per Run	Rank
-77	42	35	64	1.83	42
DVOA	Rank		EYds	EYds Per Run	Rank
-61.0%	42		1	0.03	42

FYI...42 RBs had enough carries to qualify for FO's rankings this week. That kind of puts Coffee's rankings in proper perspective. Obviously, they're dismal, and reflect pretty well the most common running play last week: "Coffee run up the middle for no gain."

We should kind of give him the benefit of the doubt, though, because (a) he's a rookie, and (b) he was going up against the "Williams Wall." Rather than pillorying him for the totality of his dead-last stats, I'd instead like to focus on his DYAR. Specifically, you'll recall that RB DYAR measures the yardage impact that losing RB X would have to Team A's rushing OFF if he was replaced by an average backup. Going back to last week, we saw that Gore's DYAR was 24, good for 13th in the NFL. So the question becomes, "Based on DYAR, is Coffee an average backup?"

Because both Gore's DYAR and Coffee's DYAR are in relation to the average backup, we can use simple subtraction to answer this question. However, it's a better idea to first convert DYAR into a per-carry stat. To this end, Gore's DYAR per carry was 0.63 before his injury, whereas Coffee's is -2.20 after having taken over for Gore. This means that the average backup RB would gain 0.63 fewer yards per carry when replacing Gore, but 2.20 more yards per carry when replacing Coffee.* Therefore, when we do the simple subtraction, we find that Coffee has gained 2.83 yards per carry less than the average Gore backup would have.

Clearly, then, Coffee is not your average backup, at least as much as that can be based solely on the MIN game. I'm confident that, with more playing time under his belt, and the effect that opponent adjustments are likely to have on his FO stats, Coffee's performance will improve in relation to an average backup RB. However, it's up to him to improve to the point that he's as good as or better than an average backup for Gore, specifically.

WIDE RECEIVERS

Here are Bruce's stats:

DYAR	Rank	Targets	Actual Rec Yards	Yds Per Target	Rank
-3	59	21	147	7.00	48
DVOA	Rank	Catch Rate	EYds	EYds Per Target	Rank
-14.3%	59	48%	129	6.14	60

Taken together, Bruce's FO stats didn't change much from last week: his individual DVOA is exactly the same, his DYAR is 1 yard worse, and his EYds per target is .05 yards worse. In terms of his actual NFL stats, his catch rate dropped 2%, but his yards per target increased by .19. When looking at the discrepancy between Bruce's actual yards per target and his EYds per target, we find that it grew from .62 last week to .86 this week. This suggests that his increase in yards per target was a situational mirage; he gained more yards, but in relatively easy game situations.

Given the totality of his stats, we continue to see that Bruce continues to not be a #1 WR by NFL standards. In fact, he's ranked at the bottom of #2 WRs (i.e., between 57 and 64). Some have said Singletary and Raye need to get Jason Hill on the field. I agree wholeheartedly. While they're at it, how about making Josh Morgan something other than a decoy? I mean, are Bruce's intangibles so valuable that the 49ers can't try to find out whether some other WR on the roster would play better than a borderline #2?

Whatever the case may be in terms of who's playing and who's not, one thing is clear. The Niners are almost unmatched when it comes to failing to spread the ball around in the passing game. Through 3 games, 4 teams have 4 WRs with 10 or more targets, 13 have 3 WRs with 10 or more targets, and another 13 have 2 WRs with 10 or more targets. For those keeping track, that's 30 of the 32 NFL teams, which means SF is 1 of only 2 teams in the NFL that have a lone WR with 10 or more targets. The other team? That would be HOU. The WR? That would be Andre Johnson. Last I checked, Johnson is pretty good. If I were Matt Schaub, I'd be throwing to him 10 times a game too. So really, when it comes to focusing on 1 unproductive WR in the passing game, the Niners' OFF really is in a league of their own.

TIGHT ENDS

Here are VD's stats:

DYAR	Rank	Targets	Actual Rec Yards	Yds Per Target	Rank
23	14	23	168	7.30	23
DVOA	Rank	Catch Rate	EYds	EYds Per Target	Rank
9.9%	22	61%	163	7.09	24

Guess the correct answer to last week's poll was "FISH!!! HOOK, LINE, AND SINKER!!!" All of the above stats are vast improvements over VD's stats through 2 games: He jumped a whopping 17 spots in the DYAR rankings, 9 spots in TE DVOA, and 7 spots in EYds per target. Comparing the specific numbers, he's now 23 DYAR more valuable, 17.6% more efficient per play, and gains 51.2% more EYds per target. In other words, to VD, the MIN game was just what the doctor ordered (pun intended).

Before we go anointing VD (with an ointment, perhaps?) as the next Dwight Clark or Brent Jones, let's just remember that one productive game does not a star TE make. Also, the wild improvement in VD's stats does owe something to the small sample size on which they're based; you know, just like a 4-hit game early in the season helps a hitter's batting average much more than one does at the end of the season. Nevertheless, one thing I'll point out is that VD's breakout (pun intended) came against the best-defending team against TEs that the 49ers have faced thus far. Prior to last Sunday, MIN's pass DEF was ranked 10th in DVOA vs. TE, while the Seahawks were ranked 13th and the Cardinals were ranked 21st. What might explain this potentially backwards opponent adjustment? How about this: Going into last week, SEA's DEF was ranked 4th against their opponents' #3 (or worse) WR and ARI was ranked 11th, whereas MIN was ranked 20th. So it could be that VD's breakout came against a pass DEF that either has trouble guarding the slot (where VD sometimes lines up) or has backup Ss who aren't very good in nickel or dime (when VD would likely be covered by a 3rd S on the field). Obviously, I'd have to consult Fooch's DVR to test my hunch. For now, though, it's about the only explanation I could find.

By the way, where do the Rams rank in DVOA vs. TEs and #3 WRs? 27th and 15th, respectively. So stay tuned to see whether these rankings translate into another good game this week from Mr. Pink. Oh, by the way...as promised, here's VD's statement to the league and to those - including me - who voted to cut bait last week:

OFFENSIVE LINE

Here are the 49ers' OL stats through 3 games:

ALY	Rank	Power Success Rate	Rank	ASR	Rank
2.46	32	25%	30	9.4%	30
Actual Yds Per RB Carry	Rank	Stuffed Rate	Rank	Actual Sack Rate	Rank
4.19	16	26%	26	9.8%	29

In terms of pass protection, the Niner OL's ASR actually got better (9.4% vs. 10.0%) despite their ranking getting worse (30th vs. 29th). Apparently, a lot of OL's got healthy in pass protection during Week 3. Regardless, whether we're talking about rankings or FO's stats, Hill is getting sacked way too often.

Pretty much everything in this table is in the same ballpark as last week. The only meaningful differences are that (a) the Niners' actual RB yards per carry dropped a full yard from 5.20 to 4.19, and (b) their power success rate was cut in half, from 50% to 25%. For both of these, think "Coffee run up the middle for no gain" if you want to solve the mystery. Clearly, the OL isn't doing a good enough job in the running game, and, when coupled with their ASR, it seems like they're pretty inept no matter the play call. Therefore, if I had to answer the question, "Who is most to blame for the OFF's inertia?" I'd have to go with the OL over Raye, Hill, the WRs, or the RBs.

Below is a chart displaying the OL's directional run-blocking performance:

Based on the percentages, it's evident that the 49ers ran up the middle against the Vikings at the expense of runs behind the tackles. Compared with the rest of the league, only 4 teams prefer running up the middle more than SF: the Buccaneers (78.8%), the Browns (69.8%), the Jets (67.0%), and the Jaguars (65.0%). You'll notice that TB, CLE, and JAC are a combined 1-8, NYJ are 3-0, and SF is, as we all know 2-1. So is SF's up-the-middle fetish more like the crappy teams or more like the Jets?

Well, a cynic would answer this question by observing that Raye was OC for the Jets last season under head coach Eric Mangini, who happens to coach the Browns this season. That's quite a game of "1 degree of Jimmy Raye-con." However, a non-cynic like me would probably wonder if the answer is related to killing time in the 2nd half with the lead; you know, the much ballyhooed "playing not to lose." Specifically, given the above teams' records, does the SF run game operate more like 3-0 NYJ when they have a 2nd-half lead (aka up the middle is a good thing), or like TB, CLE, and JAC when they're behind in the 2nd half (aka up the middle is a bad thing)?

Well, let's look at Week 3. Against the Titans, NYJ RBs ran up the middle (or behind guard) 6 of 8 times with a 2nd half lead, but only 1 out of 3 times while tied or behind in the 2nd half. Against the Texans, JAC RBs ran up the middle 5 of 7 times with a 2nd half lead, and 6 out of 10 times while tied or behind in the 2nd half. Against the Giants, TB never had the lead, but both of their 2nd-half RB carries while getting killed in the 2nd half were up the middle. Similarly, against the Ravens, CLE never had the lead, but 5 of their first 7 RB carries while getting killed in the 2nd half were up the middle. And then there are the 49ers, who had Coffee run 11 of 13 times up the middle with a 2nd half lead, and run 3 of 3 times up the middle while tied or behind.

So in answering the question, "is SF's up-the-middle fetish more like the crappy teams or more like the Jets?" we'd have to go with "more like the crappy teams" because, although both good teams (i.e., NYJ) and bad teams (i.e., JAC) ran the ball up the middle a lot with a 2nd half lead, the 49ers, like JAC, TB, and CLE, ran the ball up the middle too often while they were tied or behind. Granted, these are small sample sizes, so I don't want to draw too strong of a conclusion here. Nevertheless, it looks like, rather than the Niners being cautious while salting away a 2nd-half lead with up the middle runs, they instead seem to be too much in the mold of "3 yards and a cloud of dust" when they're tied or behind in the 2nd half. And as FO's research has taught us, good teams run because they're winning; they don't win because they're running.

The directional ALY stats make this "up the middle at all costs" running strategy look even more strategically absurd. Specifically, the Niners OL - and there's no nice way to put this - simply sucks at blocking in the very directions Raye seems to be calling plays towards the most. I've used this line before on Niners Nation, but it applies perfectly here: The definition of insanity is doing the same thing over and over again expecting different results."

So the bottom line here is that, if you're one of those who cursed Singletary and Raye this week for running the ball up the middle on 3 of the 5 Coffee carries at the end of the game, don't be. That's what most teams - good and bad - do when they have a 2nd-half lead. Instead, you should be more upset at the fact that they run the ball up the middle way too much while tied or behind, especially given the fact that their OL can't seem to block anyone on runs up the middle. To boot, even if you think the OL's ineptitude is due to their opponents knowing this obvious play-calling tendency, that shouldn't provide much solace. Rather, it just means the Niners' up-the-middle fetish is even more insane.

DEFENSIVE FRONT SEVEN

Here are the 49ers' DEF front 7 stats through 3 games:

ALY	Rank	Power Success Rate	Rank	ASR	Rank
3.35	6	100%	29	4.3%	23
Actual Yds Per RB Carry	Rank	Stuffed Rate	Rank	Actual Sack Rate	Rank
3.54	9	23%	13	4.4%	18

Compared to last week, the front 7 remains stout overall against the run. However, they still have yet to stop any running play this season on 3rd Down & 2 Yards or Less. That's more like a power failure than power success. As far as the other stats go, I'll give them a pass on their decreased stuffed rate and ASR. Facing a RB like Adrian Peterson, along with an OL like the one fielded by MIN, means that you're not going to be stopping many run plays for a loss and you're not going to be sacking the QB very much.

Below is the DEF front 7's directional run-stopping chart:

The Niners' front 7 continues to be amongst the best in the league at stopping runs to the outside. They're particularly good on runs to the defensive outside left. I have a sneaking suspicion that Nate Clements and Michael Lewis have something to do with this given that they both line up on the left side of the DEF and are anti-Deions when it comes to their eagerness and proclivity for open-field tackles.

What this chart also shows is that the weak links on run DEF continue to be the 2 OLBs and 2 DEs. At least a portion of their poor performance against runs behind tackle must be due to (a) OLBs having pass-rush-first responsibilities in a 3-4 DEF; and (b) Singletary and Greg Manusky wanting more pass rush in general, which necessarily means more pass rush from their DEs. However, given their shabby ASR, one wonders whether the emphasis on pass rush is having an unwelcome side effect when it comes to defending runs behind tackle. I guess as long as SF's opponents continue to ignore this glaring weakness, and the overall run DEF continues to excel, we can just rubberneck this one and keep on moving.

DEFENSIVE BACKFIELD

Finally, here's the chart showing stats for the secondary against various intended receivers:

The Niners' secondary now ranks in the top half of the NFL against every intended receiver except RB. This is certainly a byproduct of the MIN game, which had Nate Clements shutting down Bernard Berrian, but also had Sidney Rice scoring a TD, Visanthe Shiancoe picking up 1st downs with both of his catches, and Greg Lewis, well, we won't discuss what he did. As Clements was the black sheep of the secondary through 2 games, the MIN game brought him back into the family of efficient 49er DBs. Of course, you would have expected that if you read last week's post.

One other thing I'll mention here is that the day can't come soon enough when I can remove that "26^N" from the chart. It really pains me to see Mark Roman positioned in the chart almost exactly where he was on the last play vs. the Vikings. If only "38" would have stayed with Lewis just a little bit longer. If only.

BOTTOM LINE

OK, so through 3 games, we can draw the following conclusions about the 2009 49ers based on their player statistics thus far:

• Hill needs to start throwing longer passes, especially on 3rd down. However, it seems like this is endemic to the 49ers' pass OFF philosophy.
• After 1 game with meaningful action, Coffee has been much worse than an average Gore backup. However, playing more and not having to face MIN's run DEF again should improve his performance in the long run.
• Attention Jimmy Raye: Isaac Bruce is not Andre Johnson.
• A list of the top 15 TEs finally has VD.
• Evidence is mounting that the OL is a primary source of offensive ineptitude for the Niners. It's kind of hard to be efficient when your OL can't run block or pass block.
• When you watch 49er games, pray that their opponent runs a lot to the outside, but never runs behind LT.
• Thanks to Clements' performance against Berrian, the Niners' secondary is now in the top half of the league against all intended receivers except RBs. Given what we've seen in recent years from that unit, I think I speak for everyone when I say, "I'll take check-downs over beat-downs anyday."

OK, that's it for this week. Remember: FO begins incorporating opponent adjustments next week, so the stats and rankings will probably change a lot. Let's hope for the better. Go Niners!

*Keep in mind these comparisons to the average backup are in adjusted yards per carry, not actual yards per carry. In other words, I really mean the difference in yards per carry after eliminating the effect that game situation and other factors have on a RB's yardage performance.

**DVOA statistics used to produce this article were obtained from Football Outsiders.