This post is way longer than I ever intended. Basically, I was curious if there was anything interesting statistically between the 15 and 16 offenses beyond the obvious surface stats. I ended up going down a bit of a rabbit hole, and maybe ending up with really more questions than anything else. Since I had some time on my hands this weekend, I figured Id write it up, and post what I found, and let everyone have at it. Its organized around some of the questions I started with. Most of these stats come from Pro Football Reference, while the short yardage and passes by distance come from the ESPN.com splits. A couple of stat notes the espn splits do not appear to include the final game of either season for some reason, or at least the passing ones dont. And there is a slight discrepancy (7 targets) between the total WR targets and the team passing attempts in 2015.
Would having a fullback on the roster really have made much difference?
One thing to keep in mind is that the Giants came out of the 15 season planning to upgrade the fullback position, which is why they signed Will Johnson. And they kept Whitlock on the roster as well. So it wasnt as though they had a plan to abandon the position altogether before the season started. So really the question I think is, once they were both hurt, would a generic off the waiver wire fullback really have made such a big difference to this offense?
In Reeses post season presser, he made the point that the offense doesnt use the fullback that much anyway the number he cited was 5-6 plays a game. As a general statement, its true, though hes slightly understating how much the Giants used the fullback in 15. Last year, Whitlock played 13+ games he got hurt in the Carolina game and missed the end of the season. Over those 13+ games, he played 132 offensive snaps, so about 9-10 plays a game, or about 11% of the total offensive plays, with no carries and no receptions. Did he make much difference to the running game? Big picture, its hard to see that he did. This year, in 2016, the Giants carried the ball 398 times for 1412 yards, averaging 88.25 yards a game, and 3.54 yards a carry.. Last year, through the first 13 games of the season, the complete games Whitlock played, they carried the ball 320 times for 1149 yards, or, 88.38 yards a game, and 3.59 yards a carry. The 2015 numbers are even slightly inflated, because Eli somehow managed to rush for almost 60 yards in the first 10 games. They were a more consistent running team they only had one really abysmal rushing day but by the numbers, with a fullback playing 10 plays a game, they were the exact same rushing team they were this year. Its the very big rushing totals at the end of the season, in the game against Carolina, and the finale against Philly , that made the season totals look much better.
NYG Rushing att yards y/g y/c
2016 (16 games) 398 1412 88.25 3.54
2015 (13 games) 320 1149 88.38 3.59
They were also not a better short yardage team. Last year, when facing 3rd and 2 or less, they rushed the ball 24 times, and got 14 first downs. This year, in the same situations, they attempted 21 rushes and also got 14 first downs, which means they actually were a better short yardage rushing team this year than last year. And as bad as it looked this year in only having 6 rushing TDs, they actually had fewer last year 5. And theyre run pass ratio was basically the same they ran the ball 39.27% of the time last year, and 39.95%% this year.
The are plenty of caveats to this. Just getting rid of Andre Williams, who somehow was given 90 carries even though he averaged less than 3 yards a carry in 15, should have made them a better rushing team. And you could argue that simply having a fullback who you could trot out when teams dropped into the cover 2 look probably would have been helpful. But basically they were as bad a rushing team last year as they were this year, fullback or not.
Was Eli a noticeably different QB in 16 than he was in 15?
The issue of whether Elis game is deteriorating is obviously a big question for the team. Its the drop off in the success of the passing game that was the most glaring difference between this season and last. Eli threw for 400 fewer yards and 9 fewer TDs (the overall team numbers greater because Nassib threw for about 70 yards and a TD in that Vikings game). Is it Eli?
Since McAdoo arrived, Elis volume and completion percentage has been almost exactly the same all three years. In 2014, 598 attempts, 377 completions, 63.04% completions percentage; in 2015, 618 attempts, 387 completions, 62.62% completion rate,; and this year, 601 attempts, 379 completions, 63.06% completion rate (or nearly exactly what he did in 14). He threw 14 interceptions in both 14 & 15, and 16 interceptions this season, so his interception rate ticked up a bit, but not that dramatically.
Manning att comp pct y/a yds tds int
2016 601 379 63.09% 6.73 4027 26 16
2015 618 387 62.62% 7.17 4432 35 14
2014 598 377 63.04% 7.34 4410 30 14
The big statistical change for Eli, besides the decline in TDs, was his yards per attempt. For his first 2 years with this system it hovered around 7.2/7.3, which is pretty much league average or slightly above. This year it declined pretty dramatically to 6.7 which puts him toward the bottom of the list of NFL qbs. (and is obviously consistent with throwing the same number of passes but for 400 fewer yards).
Given that his completion percentage has not changed at all, one possibility would fit with the game manager theory that hes attempting and completing as many passes, but more of them are closer to the line of scrimmage. However, if you look at his passing distance numbers the distance hes actually throwing the ball before its caught his 2015 and 2016 are almost mirror images in terms of completion percentage. (these are the splits that only cover the first 15 games of each season) He was actually slightly more accurate on his short passes this year (10 yards and less), but those throws actually made up slightly fewer of this throws this year compared to last year. And his completion percentages (and overall total attempts) on passes thrown between 11 and 40 yards were pretty much exactly the same this year and last year almost freakily so. (For example, this year he was 54 of 110 on passes that traveled 11-20 yards, in 15 he was 50 of 102). Interestingly, the Giants actually tried more deep passes this year than last. This year they tried 22 passes of 31 or more yards, compared to 12 in 15, and on those they were less successful as a percentage (4 total completions both this year and last). And he was actually arguably a better redzone qb on a per attempt basis, his TD rate was higher, he threw fewer ints, and he took fewer sacks. So I suppose if you wanted to look at an element of his game to worry about, it would probably be the very deep ball. But the sample size is pretty small. 3 more long completions and the completion percentage would be the same. Again, this doesnt mean if you were to look at his game on tape, you might not see issues with his decision making etc. But its hard to see a real decline in his skills based on the statistics alone.
One thing that did happen was that his connection with Beckham was less productive in 16. You can see it in Beckhams numbers. In 15, over 15 games, OBJ was targeted 159 times, and caught 96 balls, for 1450 yards and 13 TDs. In 16, in 16 games, he was targeted 169 times, and caught 101 balls, for 1367 yards and 10 TDs. it is worth noting that for all the talk of Eli forcing him the ball etc, OBJ averaged the exact same number of targets per game as the year before (10.5) had the exact same completion pct on passes thrown to him (60%), and had slightly more plays of over 20 yards (20 vs 19). He had slightly less YAC (516 in 16 vs 524 in 15). And he had 1 more official drop (6 this year vs 5 last year). So there really wasnt much of a divergence. But there were fewer TDs, and fewer yards, on more targets. If we think of receiving yards per target as the complementary measure to passing yards per attempt, then this is some of where we see the drop off in that measure. In 15, the Giants averaged more than 9 yards a play every time they targeted Beckham (9.17), and this year they averaged a full yard less (8.09). (If he averaged the same yards per target this year as last year, hed have had over 1550 yards receiving).
OBJ '16 v '15
Year g trgts rec yards trgts/g yds/t recptn% td YAC 20+ drps
2016 16 169 101 1367 10.56 8.088 59.8% 10 524 20 6
2015 15 159 96 1450 10.53 9.119 60.8% 13 516 19 5
Did the Giants miss Ruben Randle more than we realized?
This is a point a number of posters have mentioned that Randle as the other outside receiver was potentially more effective than most of us appreciated, and that the move from Randle and Harris to Cruz and Shepard did not do nearly as much for the offense as most of us expected. I do think theres some evidence for this, but its actually pretty interesting when you look at what happened.
If you had asked me at the end of the season based on what it felt like as a fan, Id have told you that the Giants simply did not throw as much to the other outside receiver as they had last year. But if you look at the numbers, its not the case. Randle was targeted 90 times last year. If you consider Cruz and Lewis together as the replacement for Randle, which is pretty much how I think the Giants used Lewis, in 2016 combined they were targeted 91 times, almost exactly the same amount, most of which was targeted at Cruz. The difference is obviously in the production - out of those 90 targets, Randle got 57 receptions, 797 yards, with 194 of that YAC and 8 tds. Out of their 91 targets, Cruz and Lewis only got 46 receptions, for 683 yards, though basically the same exact amount of YAC 196, and 3 TDs.
Whats interesting is that in 2015, on a per target basis, Randle was about as efficient as OBJ. He generated 8.85 yards a target, so only slightly less, and caught about 63% of the passes thrown his way (so slightly more.) Cruz and Lewis, as the stats reflect, were less efficient. Interestingly, Cruz was only slightly less so - he caught about 55% of the 72 balls thrown his way, and generated slightly more than 8.1 yards a target (which is essentially the same as 2016 OBJ). Lewis however was much less efficient he only caught about 1/3 of his targets, for a little more than 5 yards a target, which may explain why he was never made a more central part of the offense. Though its also probably true that he was more often a deep target, and that may explain some of the lower percentage.
Player(s) targets rec yds y/t recptn% td YAC 20+
Cruz/Lewis 91 46 683 7.505 50.6% 3 196 12
Rbn Randle 90 57 797 8.855 63.3% 8 194 13
But what about Shepard?
This year, Shepard became the main other target for Eli. He was targeted 105 times, for 65 catches, 683 yards, and 8 TDS, catching 62% of the passes thrown his way, and producing 6.5 yards a target. Compared to Harris last year, he basically doubled his production. In 15 Harris numbers were 57 targets, 36 catches, 396 yards, and 4 TDs, catching 63% of the passes thrown his way, and producing 6.9 yards a target. But Shepard was actually slightly less efficient on a yards per target basis, consistent with what seems to have been the case across the offense.
But if Shepard wasnt gaining all his targets at the expense of the outside receivers, where were they coming from? He basically got all the balls that that went to all the other random receivers in the Giants offense last year Nicks, White, Parker, etc and they were all much much less efficient. If you combine Harris and all those receivers 2015 production, you would get a single receiver who got 108 targets, producing 58 receptions, 608 yards, for about a 54% completion rate and 5.7 yards a target.
In fact, if you net out all the other Giant WRs besides OBJ and compare 15 and 16, it looks like this.
2016
Shepard, Cruz, Lewis, Harris, King
201 targets, 114 receptions, 1429 yards, 12 TDS, 7.10 yards a target, 454 YAC
2015
Randle, Harris, White, Parker etc
198 targets, 115 receptions, 1405 yards, 13 TDs, 7.10 yards a target, 390 YAC
Its actually a little bit crazy, but the other WR production was essentially exactly the same in both years. Did the Giants miss Randle? He was a more productive receiver than the players that replaced him, so in that sense yes. Had they gotten his same production from the position they would have been better. But from the overall production of the offense, at least statistically, it didnt make a difference.
What this does highlight, though, is something that surprised me in many ways, the TEs and the RBs will really the main culprit in the passing games decline, at least statistically.
However bad the TEs were in 2015 as receivers, and maybe the argument is that they werent as bad as they looked at first blush, in 2016, from an efficiency perspective, they were much worse. In 15, as a position, the TEs were targeted 129 times, caught 88 passes for 828 yards and 5 tds, catching 68% of the balls thrown to them, and averaging about 6.4 yards a target. In 16, they were targeted 113 times, for 79 catches, 609 yards, and 3 TDs, catching about 70% of the balls thrown their way, for about 5.3 yards a target. Its most noticeable in Tyes numbers in 16 he got more targets, for more catches, caught a slightly higher percentage, but for fewer yards. But its more than that - Donnell and the other TEs on the roster last year accounted for about 150 yards more last season than the Giants got out of their TEs this year.
And you can see the same thing with the RBs. In 15, the Giants RBs were much more efficient pass catchers than this year. Vereen was targeted 81 times, caught 59 balls for 494 yards, and 4 TDs for 6.4 yards a target. Jennings was even more efficient he got 29 receptions from 40 targets, for 296 yards, and 7.4 yards a target. This year, that productivity drop was the most pronounced Jennings got about the same number of targets 42 and caught more balls 35 but for almost 100 fewer yards, only about 4.8 yards a target. And the overall production of the RBs as receivers dropped in a similar fashion (though interestingly Bobby Rainey was actually the most efficient of all the RB receivers by nearly a full yard.) Outside of his 67 yard reception, Perkins in particular struggled as a target. And again, you can see it in the overall numbers. If you combine all the RBs, in 2015 they were targeted 129 times, for 92 receptions, and 828 yards, at 6.4 yards a target. In 16, it was 114 times, for 83 receptions, 622 yards, and 5.5 yards a target. And if you take out Perkins big play, its even worse under 5 yards a target.
And again, just as an interesting (or strange) statistical note of comparison in terms of how similar Giants 2015 and 2016 offenses were as far as passing distribution.
2016 TEs and RBs
TEs 113 targets, 79 receptions, 609 yards, 270 YAC
RBs 114 targets, 83 receptions, 622 yards, 733 YAC
2015 TEs and RBs
TEs 129 targets, 88 receptions, 828 yards, 338 YAC
RBs 129 targets, 92 receptions, 828 yards, 716 YAC
So what does this all mean?
Im not actually sure. The statistics do back up that they could use a better outside receiver to complement OBJ, but based on this, Id be very surprised if they invest a lot in either free agency or the draft in that position, unless someone fell into their lap in the draft. Instead I suspect theyll look for a veteran with enough speed to challenge DBs in a way that Cruz cant, but whos a more reliable route runner than Lewis (or Randle was) to complement whatever they think they have in Lewis and King.
I also think they probably missed Vereen far more than most fans, including myself, realized. He was the third most targeted receiver in 15, and he produced 4 TDs, all of them in the redzone. Its a very small sample size, but you can also argue that the offense was much more productive during the three games that he played in at the beginning of the year than they ever were after that. His injury obviously has to make him a question mark, but if they believe he will fully recover, Id be surprised if the Giants cut him. And regardless, this seems to be an area they must improve on. The whole point of the WCO is that you replace the running game with the short passing game. Your RBs have to be effective pass catchers. You can make the argument that for all the focus on the Giants inability to get the ball to Beckham against 2 deep coverages, their bigger problem was that when they did get the ball to their RBs and TEs, they simply didnt produce as much as they did in 15.
I think the biggest question, and the one that surprises me the most, is the decline in yards per target was everywhere in the offense. Beckham, the outside WR, Shepard vs Harris, the TEs, the RBs they all averaged nearly a yard less per target this season than last season, even though theres not any evidence of any wholesale change in where on the field the Giants were targeting their receivers. Is this a product of facing so much cover-2? Im not enough of an xs and os person to be able to say that definitively. If it were, youd think it would show up in a noticeable decline in YAC. But you dont really see that. Could it be offensive design? Or that the routes are the same but theyre all being slightly abridged because Eli is trying to get rid of the ball more quickly? Something else?
Also, I do think the pass distribution numbers are fascinating. Slightly more than 1/4 of the Giants passes go to OBJ, 1/3 to the other WRs, 1/5 to the TEs, and 1/5 to the RBs. And these percentages are pretty much the exact same in 15 and 16. I have no idea if these are standard percentages for NFL offenses, or WC offenses, or if other teams are as consistent year to year. But there you go.
Ditto what dep said..
Word.
If so, a lot of that is on McAdoo - who may have just stretched himself too thin this year.
I wonder how many different looks this offense gave last season compared to this one.
Great post, by the way. Hope to see more of them.
Clearly there was a lot in common and I agree that losing Vereen was critical to our passing game.
The most important stat is points scored, and we didn't failed the.
One last thing that would be interesting to look at is the comparison of the offense within the season. I think the offense we saw o over the last third of this season was a very different one than what we saw at the beginning of the season. If we break the season into quarters it might be telling to check the statistics again.
Thanks for doing this.
Agreed and as I wouldn't let DRC go, I wouldn't let Vereen go either
My takeaways:
How do you defend the WCO? Don't get beat deep and you must apply pressure with your front four. Force the offense to execute a 10-12 play drive, rally to the ball and tackle well. Force the offense to score TD's in the red zone.
Giant's offense:
Struggles to protect against a four man rush
Struggles to run vs a 7 man box. No explosive runs
Struggles to avoid penalties and negative plays making 10-12 play drives difficult to execute
Questionable red zone threats other than 13
In short: Need better players to execute the offensive system.
Reese:
Kudos for accomplishing this on D last offseason.
Can you do it again on O?
I know our scoring was down this year too; poor Red Zone execution, which ended up turning into FGs, a lack of ability to sustain drives which also resulted in a disparity in TOP, as well. I think you hit the nail on the head with two observations:
- a lack of Vereen on 3rd downs (helped sustain drives)
- lack of an Red Zone threat
- lack of a running game
If so, a lot of that is on McAdoo - who may have just stretched himself too thin this year.
I wonder how many different looks this offense gave last season compared to this one.
Great post, by the way. Hope to see more of them.
It's also partly on Reese for assembling a roster with no blocking TEs or FBs.
(TEs were average at best catching the ball too, so it's not like they "demanded" more playing time).
One number I would be extremely interested in seeing is the amount of attempts/targets 20+ yards down the field between 2015 and 2016. It's entirely possible that RR, for all of his faults, was simply better at getting open and being productive down the field than his 2016 counterparts. But it's also possible that the decline in o-line performance led to them taking less shots down the field for fear of poor pass protection. I think it's difficult to draw any conclusions from some of these numbers in a vacuum excluding the o-line. What we do know is they gave up less sacks in 2016 (they threw the ball marginally less), but with declining numbers in yards per attempt, 20 yard completions, and 40 yard completions.
Here's an interesting interesting stat for me (targets, INTs when targeted):
Beckham - 2 INT on 180 Targets
Sheppard - 6 INT (1 forced ball at end of half) on 114 Targets
Tye - 4 INT on 77 Targets
Cruz - 1 INT on 76 Targets
Donnell - 1 INT on 22 Targets
King - 1 INT on 20 Targets
Vereen - 1 INT on 19 Targets
Lewis - 1 INT on 19 Targets
If forcing the ball to Beckham was truly an issue, wouldn't we see a worse targets:INT ratio? Instead, he has the best on the team.
2 INTs I consider fluky.
- impact of no Vereen was big, simply based on Offense's performance in games 1-3 versus the others
- Eli was not as good and its on everyone
Lastly, stats from last year were padded in several cases from playing from behind or getting into a shootouts. Didn't occur this year.
Who the opponent was, and the game situation... did the 2015 team play a weaker schedule and/or were they in more games that were blowouts or so far behind that the other team played prevent and let the Giants score
Where the offensive drives started. In 2015 the Giants had many more turnovers created, and their average drive to score a td would be comparatively shorter.
Do the scoring ppg include defensive scores as well?
Just a few things off the top of my head since these stats compiled above are very very equal between 2015/2016.
Thank you for the hard work and for sharing. That is great information.
I think a small part of the analysis that is missing is the effect of the missing parts on formation and what flows from there is the resulting impact on production
I suspect that without a Fullback, and Vereen, that contributed to noticeably less variation in the formations -- it follows that defenses, in having to face less change in formation, were able to more successfully anticipate where the play was going and then clamp down on YAC
the other thing that seemed to happen is that the Right side of the offensive line wore down a little toward the end of the season, and as the runs are distributed across the line and the pocket collapsed a little faster - Eli faced more pressure on his throwing, thus forcing him to rely more heavily on outlet passes, and also thus contributing to his lower accuracy with longer passes
Perkins is also a rookie -- and he arrived in camp later than the rest of the rookie class and was worked into the rotation very slowly -- he had to run behind a suspect line too -- that probably impacted his development
I think universally, beyond the OL, we need a big target TE to eat up the middle of the field and cause attention.
RB wise, I think a year of experience for Perkins (as well as the return of Vereen) improves that area.
Hmmm.
The stat that jumped out at me was the deep ball - supports what we saw all year. Only 4/22 on passes that traveled beyond 31 yards in the air. That is disgusting.
Lack of having a decent pass-catching threat from TE and RB, allowed opposing defense to sit on two deep in the majority of the game. If a TE and/or RB would catch passes underneath and wreck havoc, safeties or whoever plays deep will have to drop shallower to help in support thus allowing more one on one situations for WR's with little to no help. Vereen's and Rashad's injuries/declining abilities didn't help. The TE production got worse from the prior season. We just need a competent TE and a 3rd down RB who can be a receiving threat.
I get the idea of a Johnson type player over a Whitlock type player.
What I don't get is thinking Will Tye or Larry Donnell can lead block, when they both exhibit weakness inline and downfield blocking.
There are situational mismatches that cumulatively add up to a significant advantage for the opponent.
With those two in the backfield there is literally zero chance you're handing off to the fullback on short yardage for instance.
I disagree with Area Junc. I don't think this supports Eli having a bad season. I do think it supports RB's staying in to block versus going out for passes and yardage per throw being shorter because the ball was coming out faster.
The team needs a better OLine....and a better threat at TE who can both block and catch...and keep Vereen or a healthier version.
The lack of production from the Tight Ends and Running backs is not surprising at all.
So add Brian Quick from the Rams, OJ Howard in round 1 and sign a couple of linemen.
Agreed! What a fantastic post - great job! I won't rush to make any predictions or recommendations on what it means that we should do to improve. In fact, it might actually suggest that we may have some regression to the mean in our pocket for 2017. But what a great analysis - please post more often!
Quote:
We heard defensive players and coaches from other teams even comment on how preparing for this offense was not that difficult since they spent so much time in the same personnel package.
If so, a lot of that is on McAdoo - who may have just stretched himself too thin this year.
I wonder how many different looks this offense gave last season compared to this one.
Great post, by the way. Hope to see more of them.
It's also partly on Reese for assembling a roster with no blocking TEs or FBs.
(TEs were average at best catching the ball too, so it's not like they "demanded" more playing time).
Just out of curiosity, did you read the OP, particularly as it relates to FB?
I don't want to speak for PN, but I would guess that it's just a conversational substitute for >30 since the NFL doesn't measure yardage fractionally.
Our TEs are demonstrably bad inline, are bad out of the backfield chipping. We didn't have a fullback who can lead block or take a single carry. We didnt have a QB who can extend the pocket. We didn't have a RB who has the savvy and speed to be a factor in the screen game. We didn't have a handsy, big WR to take bubbles or screens and make a play with his body.
You can't have a system in the modern NFL where your ability to block is only up to the 5 guys upfront. And most of this year I didn't see much situational help from the other parts.
We couldn't run the ball all year long nor could we pass protect consistently . When we won super bowls we had two
Outside receivers that could do damage ..
We missed Randle because he could get deep and separate
and he could win contested balls .
Then add in the lack of a TE and a screen guy .
We were never a threat to go deep this offense does not
fit Eli without a run game or proper weapons .
Of course something major happened to the offense between '15 and '16. And like everything else on offense it starts with the QB. Was it Eli, the OL the receivers or the running game? The eyeball test says it was not not the running game or OL which were equally bad last year. So it was the Eli or the receivers. Again the eyeball test, points mostly to Eli, but the stats are pretty convincing that OBJ was also affected, likely by opponent adjustments.
i did actually begin to look at qb comparisons, but that was it's own rabbit hole, because you do see variation across players and years. generally more accurate qbs are more accurate across the throws, and eli this season does not compare particularly favorably to brees, brady, ryan etc on the deeper throws. but interestingly his numbers almost match aaron rodgers exactly on the throws over 11 yards and beyond. eli's raw numbers on the splits at 11+ and above are 54-110, 13-44, 4-17, 0-5. and rodgers were 55-113, 15-47, 4-15, 2-10. which does make me wonder if it has something to do with the system and the routes.