Does Offensive Line Matter in Madden 17?

This seems like one of the hottest topics of Madden every year (besides how you would define "cheese" or why you shouldn't pick the Seahawks every time in Head to Head, you gutless coward). How much do the five fat guys up front matter? Is it worth it to be boring in your draft and take solid linemen instead of flashy receivers? Does a reliable Left Tackle mean as much in Madden as it does in real life? These questions are especially relevant to me since I took a gamble in our fantasy draft league by not taking an offensive lineman until the 32nd round.

My thinking: The difference between a scrub OL and an elite OL isn't that big. I should just collect playmakers, and I'll be fine with a bunch of no names protecting me. 

The verdict? I could be in for a long season.

Setting up the Test

I made custom rosters updating the following attributes for my OL:

  • Strength
  • Awareness
  • Pass Block
  • Run Block
  • Impact Block

I set up each roster with all starting lineman at either 70, 80, 90, or 99 in each of the above traits. I used the Eagles, so for example, there'd be one roster where Jason Peters and crew each had 70 STR, 70, AWR, 70 PBK, 70 RBK, and 70 IBK. I didn't change any of their other traits.

All practice mode plays were run against the Houston Texans in All-Pro. I wanted to play against a team that has a solid line along with one elite player to see how he handled each tier of OL. 

Testing the Pass Rush

This was a simple test: I load the custom roster, open up practice mode, hike the ball, and see how long it takes before Sam Bradford hits the turf (without me pressing anything else). 

(Full disclosure: Part of me wanted to run this test to see how good OL is, part of me wanted to run it just to see Sam Bradford get sacked 500 times in a row.)

I called Empty Eagle Dbl Slot Cross every time so there would be five wide receivers out running routes and no players in to help block pass rushers. 

I ran 20 plays each against the following defenses: 

  • Cover 3 Sky
  • Storm Brave 1
  • Cover 2 Man
  • Edge Blitz 3
  • Cover 4 Drop 2

This gave me 100 dropbacks for each trait level and a decent assortment of man, zone, blitz, and 3-man rush. 

Practice Mode

Seconds Allowed before Sack, by Defensive Play Call and OL Trait

OL Zone Man Blitz 3-Man ALL
70 4.19 3.63 3.84 6.21 4.34
80 4.57 4.23 3.99 8.1 4.97
90 4.66 4.03 4.23 8.58 5.14
99 5.74 4.62 4.68 7.94 5.53

The chart above shows how many seconds each group gave the QB before he got sacked. It actually goes pretty much as you'd expect across the board: Good players are good and do what they're supposed to; bad players are bad and don't do what they're supposed to. What a novel concept!

The good OL were giving the QB two full seconds more against a 3-man rush, and well over a second more on average against all plays. The elite offensive linemen seemed especially effective against the blitz, when a guard was expected to loop around the edge to defend against a blitzer. The lower tier OL would frequently just stand in their designated spot with their thumbs up their butts. The higher tier OL would almost always shuffle over and easily pick up the blitz.

Seconds Allowed before Sack, by OL Trait in Percentage of Snaps

OL 2+ 3+ 4+ 5+ 6+ 7+ 8+ 9+ 10+
70 100% 82% 48% 24% 11% 9% 6% 3% 0%
80 100% 88% 58% 36% 19% 14% 10% 7% 5%
90 100% 92% 66% 35% 22% 15% 12% 11% 9%
99 100% 95% 78% 58% 30% 19% 12% 6% 3%

This chart feels especially damning of the crappy offensive linemen. It shows the percentage of snaps where the QB had that many seconds before getting sacked. As you can see, every play across the board gave the QB at least 2 seconds. But then the gap starts widening almost immediately until you get to 5+ seconds, where 70 OVR offensive lineman give you that much time only 24% of the time, compared to 58% for 99 OVR offensive lineman. 

J.J. Watt felt especially effective against the 70s, getting though the line with just a single move and giving the QB less than three seconds about 1 in every 5 snaps. 

The numbers for the 99s dips down in the 9+ and 10+ second range, but I'm going to chalk that up to some fluky Madden things happening that drove up the percentage in a small-ish sample size.

This testing pretty much settled it for me: Yes, a good offensive line is important and will give you more time in the pocket. I am already feeling pretty stupid about my drafting strategy.

Testing the Rush Rush

So Offensive Line seems important for the passing game, but what about the rushing game? I've actually heard from guys in my league that OL is only important or most important for running and not as important for passing. 

I ran similar testing against the Texans on All-Pro, but this time I would hike the ball, press nothing else, and see how far the running back made it before going down. Ryan Mathews was my running back, and thankfully for him, injuries don't exist in Practice Mode. I called the following plays:

  • Single Back Doubles -> HB Dive
  • Gun Trey Open Offset -> Inside Zone
  • I-Form Pro -> Inside Zone

I thought this gave a decent cross section of runs from under center, shotgun, and out of boring I-Form. All runs were from the opponent's 30-yard line, so the longest a run could be was 30 yards. (Yay, math!)

Yards per Carry, by Formation and OL Trait

OL Single Gun I-Form OVR
70 8.9 8.2 4.9 7.3
80 8.4 9.1 6.1 7.8
90 7.8 5.2 11.2 8
99 9.9 6.8 7.2 8

The results of this test were not as overwhelming as the passing test. There is still a general trend upwards: As the OL get better, you seem to get more rushing yards. But the difference seems negligible, and there are a lot of moving parts on running plays that make it harder to test. 

The chart below helps draw a little clearer picture about the difference that you can see:

Yards Gained on Each Carry by OL Trait

OL <0 3+ 5+ 8+ 10+ 15+ 20+ 30
70 10% 70% 48% 37% 27% 12% 10% 8%
80 3% 82% 63% 33% 27% 12% 10% 8%
90 3% 82% 58% 33% 28% 12% 10% 10%
99 3% 87% 60% 42% 25% 10% 7% 7%

The 70 OVR OL let defenders stop the runner behind the line of scrimmage 10% of the time. This is by far the worst. You can see a stark difference between 70 OVR OL and everyone else at the Trent Richardson Line (3 yards) and even the 5 yard mark, but there is pretty much no discernible difference across the board between 80, 90, and 99. Those numbers, for all intents and purposes, are identical. 

Simming Seasons

Offensive Line obviously makes a difference for the passing game in live games and can probably kind of be argued to make a little difference in the rushing game, but not as much as you'd think.

So how about in simmed games?

To test this out, I loaded the custom rosters, turned off injuries and trading and roster filling (to ensure the guys would start and play every game), then simmed 5 seasons for each group.

The takeaways from the sims are pretty clear: You will get sacked a lot more with a bad OL, and offensive line is completely, totally, 100% irrelevant to the running game.

Attempts per Sack

OL Rtg Att/Sack
70 6.8
80 11
90 15.9
99 35.5

This shows you the passing attempts per sack for each tier. The amount of times a QB gets sacked goes down significantly as the OL gets better, to the point where the 99 OVR OL was allowing just 15 sacks per season compared to over 60 sacks for the 70 OVR.

Here's the full chart of passing stats I tracked:

Average Stats Per Season

OL Rtg QB Sacks Comp % Pass Yds
70 61 58% 2,892
80 42 62% 3,377
90 31 63% 3,622
99 15 61% 3,750

As OL gets better, QBs get sacked less and have more passing yards per season. Completion percentage stayed about the same.

So as with Practice Mode, a good Offensive Line obviously makes a difference in the passing game. 

How about the running game?

I'm glad you asked! The answer is much clearer here than in Practice Mode: It makes zero difference.

Average Stats Per Season

OL Rtg Att Yards YPC
70 557 2,020 3.63
80 503 1,799 3.57
90 547 2,005 3.67
99 553 2,039 3.68

Based on five simmed seasons at each level, the difference between a 70 OL and a 99 OL is only .05 yards per carry. That's four and a half extra centimeters per carry. Ryan Mathews and Darren Sproles had pretty much identical stats every season from every sim, regardless of how good or bad their offensive line was. This was pretty stunning to me, especially after seeing how stark the difference was for the passing game. How can Sam Bradford get thrown on his ass four times as often with a crappy line, but the running game isn't affected at all?

I decided to put it to the ultimate test: I created offensive linemen with 0 Awareness, 0 Strength, 0 Pass Block, 0 Run Block, and 0 Impact Block. This is essentially the same as starting your grandmother and her bridge partners on your o-line.

Here's what happened:

OL Rtg Att Yards YPC
0 567 2,117 3.73

A higher yards per carry than having OL with 99 in everything!

This is especially crazy considering how big of an effect having all zeroes on your line affected the passing game:

Behind the granny o-line, Sam Bradford completed just 50% of his passes for 1,790 yards, and was sacked 117 times (that's one sack every 2.5 attempts), en route to a 2-14 season. The passing game was completely useless, but the running game didn't even bat an eyelash.

In Conclusion

First off, it seems pretty clear there is some sort of bug in the simming engine as it pertains to the offensive line. Why would having godawful offensive linemen have such a drastic effect on the passing game but literally no effect at all on the rushing game? This makes no sense.

Second, I am now convinced that offensive line does matter for games that you play. It definitely matters for the passing game. The better OL you have, the more time you have to throw. It looks like it kind of matters for the running game, at least enough to help you not get blown up in the backfield. 

Third, I am an idiot and wish I had done this testing before completely neglecting offensive line in my fantasy draft league. Hopefully I can make some trades before the other guys in the league read this post.

Dan Chancellor

Orlando, Florida

I created Deep Dive Gaming as an excuse to post more charts about the games I love to play. I live in Florida with my wife, son, and a random assortment of cats.