Steve’s RB Study, Part Zero: Introducing and explaining numbers behind the study

Transparency is pretty dang important when using any process to make strong claims, especially when taking an unconventional approach. Thus, I’m attempting to explain how all of my activities behind the curtain work. In the future I might have to release a second installment of Part Zero (0.5? 0.0? 0.2.0? The possibilities are endless.) to answer frequently-asked questions that I miss this time around, but that’s something to get into in the future. As for now, let me run down how each of my stats works.

As for my first caveat, be aware that these stats are all, by nature, quite subjective. I realize that and hope you do too. However, my process is consistent, which is what matters most when recording such observations. I’m getting out in front of this now because otherwise, you’d read those last few sentences at least five times throughout this article. Let’s begin!

OL-Generated Yards

Simply put, this stat measures the amount of yards the team blocks for a back. The practical approach I take to recording this contribution comes through asking myself a question: “How many yards does a replacement-level back pick up, here?”

This turns out to open a small additional can of worms. This hypothetical back (let’s call him “Adam Lerch”) might actually be below replacement-level because I assume all of his attributes to be replacement-level (speed, quickness, strength, etc.), while most replacement-level backs are at least okay at something. The point here is to reward backs for any particular skill (or, lack of deficiency) they possess. Lerch is not fast. He doesn’t break any direct tackles and will probably get taken down by most arm tackles. His vision is ordinary and he takes the guaranteed yards unless there’s an obviously-better path. He acts rationally and never makes bone-headed decisions.

The yards that Lerch would gain on a carry count as the OL-generated yards. The name of the stat is somewhat misleading, however. Receivers and tight ends also play a part. OL-GY increase the further Lerch goes down the field, and depending on his blocking from skill positions, he can get even further before a defender gets to him. Therefore, it might be more accurate to call the stat Blocker-Generated Yards, but because most of it comes down to line play, we’ll leave the name for simplicity’s sake.

The only other part of the yard-generating equation now is RB-Generated Yards.

RB-Generated Yards

The natural flipside to the OL-GY definition for RB-GY is “how many more yards the back being discussed gets than a replacement-level back would.” So, if the new back makes it through a hole that Adam Lerch wouldn’t for five additional yards, that would count as 5 RB-GY due to quickness, and so on. I imagine I’ve described the theory behind this well enough, so let’s now turn to some specific examples of Generated Yards in practice. We’ll start by looking at some tape from Kamryn Pettway. It’s easier to dip our toes with him because of a lack of top-end speed (which can make things tricky). With his embedded game against Vanderbilt following, I’ll detail the RB-GY breakdown for each play–I’d do OL-GY too, but it’s somewhat redundant because you can derive those numbers quite easily.

0:00- A defender meets Pettway at the line of scrimmage head-on while Pettway uses his strength and low pad level to fall forward for a yard. 1 yard generated.

0:05- Again, he’s solidly engaged at the LOS. He pulls through this tackle, and everything past the original LOS goes to his credit. 12 yards generated (Could be 11, but when we expand the sample size, ball marking with a one-yard difference doesn’t impact our recording).

0:14- A DL wraps him up 1 yard down the field. The driving he does thereafter is his gain. 2 yards generated.

0:21- Hits the hole and can be brought down after a 2 yard gain. He pulls forward for a couple more. 2 yards generated.

0:29- If he doesn’t plant and lower his pad level (and were a smaller, weaker back), he could be brought down at the LOS. 2 yards generated.

0:52- He’s met in the backfield and goes nowhere. Well, he goes backward, but forward progress means he didn’t actually lose any additional yardage. 0 yards generated.

0:59- A smaller back has much less advantage on #5 at the LOS. I’d say Adam Lerch (small, of course) gets brought down there for no gain. 7 yards generated.

1:10- Bizarre play. Pettway was out of control, and an in-control player at least gets past the outside corner. Would Lerch (a bad player, but rational actor) be in control? I’d say no, because he’d be hurrying to beat 86’s man to the spot. Thus, I’ll give KP the benefit of the doubt. 0 yards generated (if you say Lerch would make it, there’d be -2 yards generated). Subjectivity, my friends.

Intermission! Time for a penalty discussion.

1:17- Holding called on the play. Let’s assume the penalty was accepted. How do we treat such a play?

There’s a couple ways to do so. You could discount the play entirely. You could give the OL -10 yards, because that’s how much the penalty is, after all. You could give the OL the opposite of however many RB-GY there were because they were the reason the back’s carrying stats couldn’t reflect his last contribution. Honestly, I’m not sure how to go about this exactly, and I’m happy to hear your guys’ opinions.

My answer, though? Zero OL-GY and keep the same RB-GY regardless of the penalty. Why? First, we can’t disregard what the back just did. If a back breaks a 40 yard gain only for it to be called back for an illegal procedure (that does NOT spring the play—that’d mean those yards would belong to the OL), that doesn’t mean the back isn’t creative. Cancelling out the RB-GY in that play is effectively punishing him for teammates’ actions—which is exactly the opposite of what we set out to do. As for the OL-GY, negating such a play and lumping -40 or so OL-GY is just too strong of a reaction to a penalty. Such plays would cancel out entire games’ worth of contribution for a blocking unit. Simply put, a 0 OL-GY penalty catch-all ensures a penalty from a single play would not have too strong of an impact on the OL’s overall production while still giving the OL a slight bump down.

As a result, box score carries and yardage often don’t exactly match with my recorded combined generated yard records.

1:22- The lead in to the tackle and KP’s weird failed spin cause an effective wash from Lerch getting tackled in the open field. 0 yards generated.

1:31- There were two spaces that Pettway could’ve run through. I think his choice was fine, so I won’t detract from that. He then steams through the man who meets him after lowering his pads. 2 yards generated.

1:38- He doesn’t really do anything after engaging the tackler. If he pushed a little more, he could’ve gotten a yard. 0 yards generated.

1:44- Defender has a chance to engage at the 42 but Pettway’s stiff arm keeps him off until the back has stepped out of bounds, creating a couple yards in turn. 3 yards generated. (Refer to above section on penalties)

2:02- A routine 2 yards generated.

2:09- Dude’s a bowling ball, and his momentum lets him fall forward through his tackler. 2 yards generated.

2:17- 14 is the first defender with a real chance to bring him down. He manages to struggle ahead for a bit. 1 yard generated.

2:29- I think it’s safe to say we know the pattern to how Pettway creates. 2 yards generated.

2:47- Tempted to say he gets credit for patience here, but I think he might’ve found the hole earlier if he started scanning sooner. Things get even more difficult after he’s tripped up getting through the hole. Looking at that hole, I think Lerch gets through without being tripped up. The safety flies in for a tackle that I’d say comes at about the 47. After some deliberation, -3 yards generated.

2:54- This is getting a little repetitive. 4 yards generated.

3:00- *gets beer from fridge* 1 yard generated.

3:15- Missed the hole. If he runs underneath where his H-Back blocks, there’s a lot of daylight. Detective time! At the bottom of the screen, it appears 31 never was engaged with his receiver. With a lot of subjective calculation, I say he and Lerch would meet in the middle of the field at the 28. Turns out Pettway’s poor choice worked out after 70 ended up hooking his man! 6 yards generated.

3:24- A couple different paths would’ve led to the same result. 2 yards generated.

3:29- Given the second defender flying in from the edge, I don’t think Lerch gets anywhere even if he’s past the guy who grabbed KP’s leg. 0 yards generated.

3:36- After 14 runs through the hole, Pettway made the right choice to bounce. 4 yards generated.

3:52- I think most guys hit this hole. After he’s through, there’s nobody left to pursue. He doesn’t make it to the end zone but… I can’t fault him for pulling a hammy. This is where I make an executive decision and grant him 0 yards generated (which he was bound to get after the easy hole + sprint) and give the OL all 72 yards generated. Sometimes I have to make weird subjective decisions about this stuff. Again, GY sums won’t always match box scores.

So that’s how an entire game’s worth of carries would score for GY.

Briefly, to conclude our GY section, I’ll discuss skill sets. We just examined GY for a fairly textbook power back. The process is the same for any kind of back, but how they go about getting those yards changes:

  • A speed back will squeeze out yards on the perimeter and the second level. Let’s say a pursuing defender is 5 lateral yards away from Jamaal Charles. Charles then hits another gear and beats the defender’s pursuit angle. Any yards Charles gains past the point the pursuit angle pointed to count as his own, if they hadn’t started counting already. These sorts of plays are hardest to judge, but after 60+ hours of RB scouting in the last few months, it’s gotten pretty routine to judge.
  • LeVeon Bell is essentially one-of-a-kind with his patience. That trait appears more with Success Rate (to be discussed later), but can occasionally arise here. If he waits so long that a hole opens after Lerch would have already ran into the muck, any yards past the muck belong to Bell.
  • Saquon Barkley is a similarly unique runner in regards to his elusiveness. He makes exceptional use of space, bouncing away from would-be tacklers once he’s in space and then running past them (5:03, where he gets past the safety To a lazy, untrained eye, those could be either speed yards or more yards to the OL’s credit, but in reality Barkley created those yards with his bubbling. It doesn’t matter, of course, if we’re choosing between speed and bubbling yards, but it does when we’re choosing between bubbling yards and OL-generated yards.

Now, onto some simpler stats.

Holes, Holes Identified, and Holes Hit

These are pretty simple. You can get technical with holes if you want, but that isn’t really necessary. I’ll describe a hole as “an opening in the defensive front that provides a clear path for gaining yards.”

It doesn’t have to be a textbook crease where the back splits blockers who have their defenders turned opposite ways, however. This measure originated as a mark to measure vision, so clear paths to more yardage are included, even if there isn’t some opening to pure daylight in front of it. For example: If the defense overpursues and allows an opening for the back to bounce (which wasn’t actually created by the line), that counts as a hole.

Holes Identified are pretty easy to record once you find a hole. Do they make their way to it? If so, then they’ve identified it. If not, then they haven’t. That’s about it. The stat effectively rates a back’s vision and patience, as the former allows a player to quickly recognize the hole while the latter allows more time to recognize that hole.

Subjectivity creeps back into the process when I break down Holes Hit. Assuming a player has identified the hole, we are judging whether or not he made it to and/or through the hole. The and/or that I mention is decided on a case-by-case basis. NFL football is fast, so most of the time when holes open, they close quite quickly. Essentially, if a player gets to the hole but gets brought down once he’s there, I judge the hit based on if I think a reasonably good back (this is a different stat than RB-GY, so we’re ignoring Adam Lerch) could get through the hole without going down. If not, the back hit the hole. If so, he didn’t. It’s a judgment call, but it’s also a consistent one.

At this point, my measurements here are pretty routine, so I don’t think I need to dive deeper into this, but if folks don’t understand this as clearly as I do, I can add in a similar set of examples like I did in the RB-GY section. Onto our final section!

Broken Tackles, Powers, and Eludes

These are the most self-explanatory. A player gets a broken tackle for each tackle he breaks. Each broken tackle fits into one of two categories: Powers and Eludes. Powers rely more on strength, balance, and pad level while eludes depend on quickness and agility. Eludes do not necessarily have to be clean. If a back sidesteps a tackler, and with the tackler unbalanced, gets past him despite the tackler making contact, that counts as an elude. Basically, it’s a power if he doesn’t use any sort of move to set his defender up, and instead just plows through. Again, I can bring examples for these if needed. Nonetheless, there’s not much of a need at this point as I’ve yet to differentiate powers and eludes in my greater analysis.


Let me lead into the summary of my explanatory article with some disclaimers. I have many, so I’m bound to forget a few. Nonetheless, take note of all of the following:

  • Once more, this is very subjective work. I will probably not outsource these tasks to anyone anytime soon, partially because I have no volunteers, but more importantly because it’d be quite difficult to account for bias in observations between recorders.
  • I will present some statistics for individual players. Unless I note otherwise, I would not put much into these individual stats. At this point, I’ve watched five backs enough to feel comfortable enough to assess the abilities of both them and their lines. For now, I’m sticking to greater trends from my 120 (and counting) games sampled.
  • I’m steadily improving my process. I started charting these games with totals and would like to eventually move into play-by-play stats. That format would add a lot to my ability to break production down, but that’s simply way too time-consuming at this point.
  • I’m really bad at explaining things and will need to be told to explain stuff more at times. You can help me with this simply by tweeting me!
  • Like my play charting process, I have a lot of things I aim to assess in the future. Defenses, box defenders, and game situations are all things I want to analyze in the future.


In this foundational part of my production series, I explained how each of my observational statistics works in the hopes of helping readers understand the statistics I’ll be diving into in the future. This has to be one of the more dull fantasy football articles over 2,500 words that you can read, but it’s important for everyone to be on the same page with this big of an undertaking. Once more, I’ll ask anyone confused with my methodology to bring up the lack of clarity to me (over Twitter). Otherwise, with all the boring stuff out of the way, we’re ready to dive in!