RBBC is here (1 Viewer)

[Jeff Pasquino]

Several people have questioned the validity of the alleged trend of the NFL going to a Running Back By Committee (RBBC) approach to team backfields. They've argued the total carries or touches for a season - which I believe is intrinsically flawed.

Here's the best way to prove my point.

Fact 1: Teams put in a game plan for a given week / opponent. They will call plays and personnel based on that game plan. That means substitutions are scripted / planned for that contest.

Why is that significant? Well, if you believe that Fact 1 is true (I stand on that point) then there should be a reflection - positive or negative - on the RBBC trend in the NFL. Looking at the data on a per game basis eliminates the issues of RB injury (a player sidelined in the middle of the year) which often happens and skews annual data.

Hypothesis 1 - If there is indeed a trend towards RBBC, RB touches per player per game will be trending downward.

To prove (or disprove) the above Hypothesis (Hypothesis 1), I just went through the Game Logs for 2004-2010 (thanks to the Game Log Dominator) and broke down the number of games per season where a RB had 15, 20 and 25 touches in each contest. If there is no trend towards RBBC, then those numbers should be pretty steady. If there is indeed a trend to RBBC, there will be a definitive rolloff / dropoff at either 20 or 25 touches per game.



Please note that I consider touches a better measure than carries, given that some RBs are used more as receivers than rushers. If there is a feature back trend, however, then the usage of touches vs. carries shouldn't matter.

Hypothesis 2 - The NFL went to a copycat system of Chicago and Indianapolis after each went to the Super Bowl in 2006 (Feb. 2007). Starting in 2007, RBBC became much more prevalent.

Again by looking at the same set of data, we should see the dropoff / rolloff starting in 2007 and continuing through 2010.

Now let's look at some numbers:

YEAR / # of games for a RB with 15+ touches / 20+ touches / 25+ touches2004 / 457 / 304 / 1712005 / 445 / 290 / 1452006 / 466 / 305 / 1522007 / 446 / 290 / 1492008 / 442 / 259 / 1102009 / 426 / 241 / 1012010 / 425 / 238 / 101Analysis: There is indeed a rolloff between 2007 and 2008 at both the 15+ touches/game and 20+ touches/game level. That rolloff is significant. The dropoff is bigger in 2009 and nearly the same in 2010. The difference at 15+ touches/game is over 20% at 15+ touches and almost 33% at the 20+ touches level for all three seasons.Conclusion: Hypothesis 1 is correct. The NFL is in full-blown RBBC mode, with only about 100 games a year with "feature back" 20+ touch usage, a huge decrease from 2004-2007. That dropoff is almost the same rate at the 15+ touch range over the same period.

Hypothesis 2 is nearly correct, except the trend really kicked in two seasons after the Bears-Colts Super Bowl, not one. I can understand that as it might have taken some teams two seasons to develop their RBBC approach and acquire their second (or both) backs.

 

[Go deep]

Glad someone finally did this. The numbers confirm what i think most of us thought. What im left wondering now is how does this affect RB's numbers. In other words, even though RB's are getting less touches per game, is it really having a negative affect on their production. It seems the top 10 or 20 RB's are putting up just as many, if not more fantasty points as they always have.

 

[moderated]

Very interesting. Not many games with a primary bell cow anymore.

 

[ZWK]

There is a trend in those numbers, but if it's from more teams going to a full-fledged committee shouldn't we expect to see an increase in the number of games where a RB has 10+ touches? In a committee backfield there should be lots of games where both backs cross that threshold. The 2006 Colts and Bears each had 27 (15 for Addai, 12 for Rhodes, 16 for Thomas Jones, and 11 for Benson).

Other trends that could explain your data:

- more passing, less running

- more concern about the effects of a heavy workload, leading to a kind of "pitch count"

If there was a key event that sparked the change, I'd point to Shaun Alexander's collapse.

 

[Raiderfan32904]

Good stuff. But will there still be the handful of bell cows remaining after they decide to go to an 18 game season? Is it possible that their touches are game managed much like baseball pitcher's pitch count?

 

[Hoosier16]

YEAR    / # of games for a RB with 10+ touches / 15+ touches / 20+ touches2004    /                  457                 /    304      /    1712005    /                  445                 /    290      /    1452006    /                  466                 /    305      /    1522007    /                  446                 /    290      /    1492008    /                  442                 /    259      /    1102009    /                  426                 /    241      /    1012010    /                  425                 /    238      /    101

Your columns should read 15+, 20+, and 25+.

 

[gianmarco]

There is a trend in those numbers, but if it's from more teams going to a full-fledged committee shouldn't we expect to see an increase in the number of games where a RB has 10+ touches? In a committee backfield there should be lots of games where both backs cross that threshold. The 2006 Colts and Bears each had 27 (15 for Addai, 12 for Rhodes, 16 for Thomas Jones, and 11 for Benson).Other trends that could explain your data:- more passing, less running- more concern about the effects of a heavy workload, leading to a kind of "pitch count"If there was a key event that sparked the change, I'd point to Shaun Alexander's collapse.

There's something else that may be skewing the numbers. In the past, it seemed, there was more of a clear backup guy. What I mean is, if the primary ball carrier went down, someone else stepped in getting the 20+ carries until the main guy returned. Now, it seems, there's a RBBC approach for RB's that miss games. Given the number of injuries at the position, this could partly explain the trend as well. Lots of factors potentially not accounted for here.

 

[Tanner9919]

Glad someone finally did this. The numbers confirm what i think most of us thought. What im left wondering now is how does this affect RB's numbers. In other words, even though RB's are getting less touches per game, is it really having a negative affect on their production. It seems the top 10 or 20 RB's are putting up just as many, if not more fantasty points as they always have.

great point! seems plausible..RB's appear to be doing more with less these days..

 

[Caveman_Nick]

Glad someone finally did this. The numbers confirm what i think most of us thought. What im left wondering now is how does this affect RB's numbers. In other words, even though RB's are getting less touches per game, is it really having a negative affect on their production. It seems the top 10 or 20 RB's are putting up just as many, if not more fantasty points as they always have.

So maybe it's not the point production at the position which is changing, but rather the number of available scorers. RB dominance early in drafts in the past was driven as much by points as it was the opportunity to grab one of the few bell cows. Draft philosophies will adapt and allow people to build teams differently. For several years some folks have been double tapping WR in rounds 1 and 2 and had it work for them, grabbing under rated guys like Joe Addai or Matt Forte later in drafts.The tough part is finding the right guys (of course)

 

[We Tigers]

There is a trend in those numbers, but if it's from more teams going to a full-fledged committee shouldn't we expect to see an increase in the number of games where a RB has 10+ touches? In a committee backfield there should be lots of games where both backs cross that threshold. The 2006 Colts and Bears each had 27 (15 for Addai, 12 for Rhodes, 16 for Thomas Jones, and 11 for Benson).Other trends that could explain your data:- more passing, less running

This was my first thought as well. I'd be interested in seeing the data for the number of distinct RBs hitting each mark--i.e., were 50 players responsible for all of the 15+ touch games? 60? 3 dudes with a time machine?I think the "pitch count" interpretation could make a lot of sense, given that 10-touch games haven't increased to compensate for the big dropoff in 20-touch games, and have in fact decreased by a small amount. Are we seeing RBBC, or are we seeing a league that, since 2004, has just turned 5 or 6 touches once assigned to the lead back into pass plays?

 

[Jeff Pasquino]

YEAR    / # of games for a RB with 10+ touches / 15+ touches / 20+ touches2004    /                  457                 /    304      /    1712005    /                  445                 /    290      /    1452006    /                  466                 /    305      /    1522007    /                  446                 /    290      /    1492008    /                  442                 /    259      /    1102009    /                  426                 /    241      /    1012010    /                  425                 /    238      /    101

Your columns should read 15+, 20+, and 25+.

Fixed in Post 1, thanks.

 

[Dinsy Ejotuz]

I'm not sure whether I agree with Jeff's OP or not, but ultimately it doesn't seem to matter much.

VBD points for the top 23 RBs from 2004-2010, followed by the VBD points total for RBs 13-23:

2010 1422 3022009 1194 1672008 1346 2972007 1256 2402006 1375 1662005 1572 2112004 1357 265
If there's a pattern there I don't see it. So regardless of whether teams are moving to RBBC or not, it appears that RBs should be valued very similarly to what they have been in the past.

 

[Jeff Pasquino]

There is a trend in those numbers, but if it's from more teams going to a full-fledged committee shouldn't we expect to see an increase in the number of games where a RB has 10+ touches? In a committee backfield there should be lots of games where both backs cross that threshold. The 2006 Colts and Bears each had 27 (15 for Addai, 12 for Rhodes, 16 for Thomas Jones, and 11 for Benson).Other trends that could explain your data:- more passing, less running- more concern about the effects of a heavy workload, leading to a kind of "pitch count"If there was a key event that sparked the change, I'd point to Shaun Alexander's collapse.

I looked at 10+ touches and the numbers weren't nearly as telling - they were all slammed together. There may be a sweet spot between 10 and 15, but I think the trend presents itself just fine at 20 and 25 not being there.Most teams have 60-70 offensive plays, so if there was a feature back then he'd get at least a third of the touches at the 20-touch level.

 

[Jeff Pasquino]

There is a trend in those numbers, but if it's from more teams going to a full-fledged committee shouldn't we expect to see an increase in the number of games where a RB has 10+ touches? In a committee backfield there should be lots of games where both backs cross that threshold. The 2006 Colts and Bears each had 27 (15 for Addai, 12 for Rhodes, 16 for Thomas Jones, and 11 for Benson).Other trends that could explain your data:- more passing, less running- more concern about the effects of a heavy workload, leading to a kind of "pitch count"If there was a key event that sparked the change, I'd point to Shaun Alexander's collapse.

There's something else that may be skewing the numbers. In the past, it seemed, there was more of a clear backup guy. What I mean is, if the primary ball carrier went down, someone else stepped in getting the 20+ carries until the main guy returned. Now, it seems, there's a RBBC approach for RB's that miss games. Given the number of injuries at the position, this could partly explain the trend as well. Lots of factors potentially not accounted for here.

You're missing the point of looking at each particular game. There are 256 NFL games during a season, and there are not even 50 injuries that sideline a RB for a full contest. By looking at each particular game, the "noise" generated from a RB going down mid-season and being replaced by another back or backs for the next several games (with each back getting 100+ touches for the year, for example) doesn't muck up the results.To put it another way - there are far more games without RB injuries than there are with RB injuries. By looking at each game's results for touches, we do not have injuries skewing the results nearly as much as when we look at seasonal totals.

 

[Jeff Pasquino]

I'm not sure whether I agree with Jeff's OP or not, but ultimately it doesn't seem to matter much. VBD points for the top 23 RBs from 2004-2010, followed by the VBD points total for RBs 13-23:

2010	1422	3022009	1194	1672008	1346	2972007	1256	2402006	1375	1662005	1572	2112004	1357	265

If there's a pattern there I don't see it. So regardless of whether teams are moving to RBBC or not, it appears that RBs should be valued very similarly to what they have been in the past.

RBBC makes the RBs beyond the Top 20 more valuable. Perhaps you need to look further down the list of RBs. Not sure what your RB baseline is for these numbers - 24?

 

[Sweet Love]

There is a trend in those numbers, but if it's from more teams going to a full-fledged committee shouldn't we expect to see an increase in the number of games where a RB has 10+ touches? In a committee backfield there should be lots of games where both backs cross that threshold. The 2006 Colts and Bears each had 27 (15 for Addai, 12 for Rhodes, 16 for Thomas Jones, and 11 for Benson).Other trends that could explain your data:- more passing, less running- more concern about the effects of a heavy workload, leading to a kind of "pitch count"If there was a key event that sparked the change, I'd point to Shaun Alexander's collapse.

Yeah, if we are not going to look at ratio of touches from RB1 to RB2 during this time frame (and just take into account total touches by RB1), we HAVE TO evaluate whether the league has gone heavier in passing. You could have a 20% increase in passing and 15% decrease in RB touches (assuming RBs get some of those passes) and it would do nothing to explain a RBBC...it would just prove a decrease in the overall RB workload.There's something else that may be skewing the numbers. In the past, it seemed, there was more of a clear backup guy. What I mean is, if the primary ball carrier went down, someone else stepped in getting the 20+ carries until the main guy returned. Now, it seems, there's a RBBC approach for RB's that miss games. Given the number of injuries at the position, this could partly explain the trend as well. Lots of factors potentially not accounted for here.

 

[Dinsy Ejotuz]

Not sure what your RB baseline is for these numbers - 24?

Yeah... just grabbed the historical stats from FBGs, which sets it at 24. I'll let someone else go back and look at 25+ - but just eyeballing things, it didn't appear there were any glaring differences. I also did a bit more with the data I had... For each RB position (RB1, RB2, RB3, etc) I ranked them by VBD points for each year, 2004-2010. So for RB1 you see that the most VBD points were scored as follows: 2006, 2005, 2009, 2010, 2007, 2004, 2008 (ranked #1 to #7). And for RB 13 you find that most to least VBD points were scored: 2010, 2008, 2004, 2007, 2009, 2005, 2006 (again ranked from #1 to #7)Averaging the 23 ranks associated with each year leaves you with:

2010 3.13 2005 3.22 2008 3.30 2004 3.57 2007 4.22 2006 4.78 2009 5.78

And again there doesn't seem to be a pattern. (2009 was a terrible year for RBs!)I haven't thought this all the way through, but wouldn't you expect that the top RBs would be losing value if the RB3 types were gaining touches at their expense? That doesn't seem to be indicated here (or the later years would be associated with higher averages).

 

[Sweet Love]

There is a trend in those numbers, but if it's from more teams going to a full-fledged committee shouldn't we expect to see an increase in the number of games where a RB has 10+ touches? In a committee backfield there should be lots of games where both backs cross that threshold. The 2006 Colts and Bears each had 27 (15 for Addai, 12 for Rhodes, 16 for Thomas Jones, and 11 for Benson).Other trends that could explain your data:- more passing, less running- more concern about the effects of a heavy workload, leading to a kind of "pitch count"If there was a key event that sparked the change, I'd point to Shaun Alexander's collapse.

There's something else that may be skewing the numbers. In the past, it seemed, there was more of a clear backup guy. What I mean is, if the primary ball carrier went down, someone else stepped in getting the 20+ carries until the main guy returned. Now, it seems, there's a RBBC approach for RB's that miss games. Given the number of injuries at the position, this could partly explain the trend as well. Lots of factors potentially not accounted for here.

You're missing the point of looking at each particular game. There are 256 NFL games during a season, and there are not even 50 injuries that sideline a RB for a full contest. By looking at each particular game, the "noise" generated from a RB going down mid-season and being replaced by another back or backs for the next several games (with each back getting 100+ touches for the year, for example) doesn't muck up the results.To put it another way - there are far more games without RB injuries than there are with RB injuries. By looking at each game's results for touches, we do not have injuries skewing the results nearly as much as when we look at seasonal totals.

This could definitely be the case, but the study does not address the possibility of passing increases. If ADP and CJ average 15 touches a game, and RB2 on those teams average one, your data (based on touches of RB1) suggest a RBBC. It is not taking into account the fact that maybe teams are just passing more often. Now, your data could reveal that RB touches by RB1 is down, but i don't see how it proves a RBBC trend (not confirming or denying).

 

[Raiderfan32904]

There is a trend in those numbers, but if it's from more teams going to a full-fledged committee shouldn't we expect to see an increase in the number of games where a RB has 10+ touches? In a committee backfield there should be lots of games where both backs cross that threshold. The 2006 Colts and Bears each had 27 (15 for Addai, 12 for Rhodes, 16 for Thomas Jones, and 11 for Benson).Other trends that could explain your data:- more passing, less running- more concern about the effects of a heavy workload, leading to a kind of "pitch count"If there was a key event that sparked the change, I'd point to Shaun Alexander's collapse.

There's something else that may be skewing the numbers. In the past, it seemed, there was more of a clear backup guy. What I mean is, if the primary ball carrier went down, someone else stepped in getting the 20+ carries until the main guy returned. Now, it seems, there's a RBBC approach for RB's that miss games. Given the number of injuries at the position, this could partly explain the trend as well. Lots of factors potentially not accounted for here.

You're missing the point of looking at each particular game. There are 256 NFL games during a season, and there are not even 50 injuries that sideline a RB for a full contest. By looking at each particular game, the "noise" generated from a RB going down mid-season and being replaced by another back or backs for the next several games (with each back getting 100+ touches for the year, for example) doesn't muck up the results.To put it another way - there are far more games without RB injuries than there are with RB injuries. By looking at each game's results for touches, we do not have injuries skewing the results nearly as much as when we look at seasonal totals.

This could definitely be the case, but the study does not address the possibility of passing increases. If ADP and CJ average 15 touches a game, and RB2 on those teams average one, your data (based on touches of RB1) suggest a RBBC. It is not taking into account the fact that maybe teams are just passing more often. Now, your data could reveal that RB touches by RB1 is down, but i don't see how it proves a RBBC trend (not confirming or denying).

I can see where the passing game may skew the numbers over the last decade. The NFL has definitely become a passing league. The NFL rules have changed over that time to expand the passing at the expense of the running game. I think that the argument that Chris Johnson and ADP having any statistical correlation that puts them in a RBBC would have to be examined. I dont think that is what the data will show.

 

[Jeff Pasquino]

Yeah, if we are not going to look at ratio of touches from RB1 to RB2 during this time frame (and just take into account total touches by RB1), we HAVE TO evaluate whether the league has gone heavier in passing. You could have a 20% increase in passing and 15% decrease in RB touches (assuming RBs get some of those passes) and it would do nothing to explain a RBBC...it would just prove a decrease in the overall RB workload.

More numbers:

Year	Total Carries	Team Rushes/Game	Total Rushing Yards	Team Rushing/Game2004		14428	     28.2	                   59709	116.62005		14375	     28.1	                   57583	112.52006		14448	     28.2	                   60063	117.32007		13986	     27.3	                   56790	110.92008		14119	     27.6	                   59376	116.02009		14088	     27.5	                   59739	116.72010		13920	     27.2	                   58607	114.5

 

[Clifford]

There is a trend in those numbers, but if it's from more teams going to a full-fledged committee shouldn't we expect to see an increase in the number of games where a RB has 10+ touches? In a committee backfield there should be lots of games where both backs cross that threshold. The 2006 Colts and Bears each had 27 (15 for Addai, 12 for Rhodes, 16 for Thomas Jones, and 11 for Benson).Other trends that could explain your data:- more passing, less running- more concern about the effects of a heavy workload, leading to a kind of "pitch count"If there was a key event that sparked the change, I'd point to Shaun Alexander's collapse.

More passing less running. More specifically emergence of te and 3wr sets. Run two more sets of data for the same years: catches by Te and wrs. I bet you see increases there that go along with the dropoff for rbs.

 

[Jeff Pasquino]

There is a trend in those numbers, but if it's from more teams going to a full-fledged committee shouldn't we expect to see an increase in the number of games where a RB has 10+ touches? In a committee backfield there should be lots of games where both backs cross that threshold. The 2006 Colts and Bears each had 27 (15 for Addai, 12 for Rhodes, 16 for Thomas Jones, and 11 for Benson).Other trends that could explain your data:- more passing, less running- more concern about the effects of a heavy workload, leading to a kind of "pitch count"If there was a key event that sparked the change, I'd point to Shaun Alexander's collapse.

More passing less running. More specifically emergence of te and 3wr sets. Run two more sets of data for the same years: catches by Te and wrs. I bet you see increases there that go along with the dropoff for rbs.

Post #20 shows that running numbers haven't changed league-wide from 2004-2010.

 

[Sweet Love]

Yeah, if we are not going to look at ratio of touches from RB1 to RB2 during this time frame (and just take into account total touches by RB1), we HAVE TO evaluate whether the league has gone heavier in passing. You could have a 20% increase in passing and 15% decrease in RB touches (assuming RBs get some of those passes) and it would do nothing to explain a RBBC...it would just prove a decrease in the overall RB workload.

More numbers:

Year	Total Carries	Team Rushes/Game	Total Rushing Yards	Team Rushing/Game2004		14428	     28.2	                   59709	116.62005		14375	     28.1	                   57583	112.52006		14448	     28.2	                   60063	117.32007		13986	     27.3	                   56790	110.92008		14119	     27.6	                   59376	116.02009		14088	     27.5	                   59739	116.72010		13920	     27.2	                   58607	114.5

That is what I needed to see...while a tiny decrease, it would not suggest a major dip (at all) in rushing.

 

[Aabye]

Why not add in 2002 and 2003? Prior to 2002 there were fewer teams, so the data will be screwed up if we go back any further.

YEAR / # of games for a RB with 15+ touches / 20+ touches / 25+ touches2002 / 443 / 294 / 1552003 / 447 / 298 / 1642004 / 457 / 304 / 1712005 / 445 / 290 / 1452006 / 466 / 305 / 1522007 / 446 / 290 / 1492008 / 442 / 259 / 1102009 / 426 / 241 / 1012010 / 425 / 238 / 101
ZWK offered a couple of alternative explanations:

(1) There are just fewer RB touches overall.

So look at total rushes and receptions by RBs from 2002-10:

YEAR / Total RB carries / Total RB receptions / Total RB touches2002 / 12016 / 2838 / 148542003 / 12698 / 2635 / 153332004 / 12654 / 2430 / 150842005 / 12726 / 2323 / 150492006 / 12734 / 2525 / 152592007 / 12414 / 2473 / 148872008 / 12498 / 2405 / 149032009 / 12384 / 2439 / 148232010 / 12068 / 2448 / 14516
Fewer overall touches may explain a little bit, but it seems to me that it's not enough to account for the sharp decline, especially in 25+ touch games.

(2) Pitch count?

This strikes me as a plausible explanation. Maybe some coaches are doing this.

The best way I can think of to do this is to use the Game Log Dominator again. I can't sort by total touches so I sort by team, then cap the number of rush attempts by a single player at 21+ and look at the game logs from 2002-10. I'm not going to do it for every team, but here are the first 5 teams:

These are the total games in which one RB recorded 21 or more carries:

YEAR / Arizona / Atlanta / Baltimore / Buffalo / Carolina2002 / 5 / 4 / 7 / 7 / 82003 / 5 / 4 / 12 / 11 / 122004 / 5 / 4 / 8 / 9 / 82005 / 0 / 4 / 5 / 9 / 62006 / 9 / 5 / 7 / 6 / 42007 / 10 / 3 / 8 / 7 / 12008 / 3 / 11 / 9 / 4 / 32009 / 0 / 5 / 2 / 2 / 62010 / 0 / 9 / 6 / 4 / 4
In Arizona, almost all of the high-volume games were compliments of Edgerrin James. In Atlanta, the high-volume games pretty much line up with when Michael Turner can play. In Baltimore, it's when Jamal Lewis is running well. In Buffalo, it's mostly Travis Henry and Willis McGahee, with a reasonable dose of early-career Lynch.

In other words, when teams have talented, durable backs, it looks to me like they use them as much as they can. The only exception for these 5 teams is Carolina, where John Fox has had better backs in recent years but seems to be committed to limiting their per-game carries somewhat.

So another possible explanation is this:

(3a) A few coaches want to limit per-game carries. This could skew the 20+ and 25+ touch games down a little.

and

(3b) There is a slight decrease in the number of really good NFL RBs right now. This might account for the rest of it.

 

[Aabye]

Yeah, if we are not going to look at ratio of touches from RB1 to RB2 during this time frame (and just take into account total touches by RB1), we HAVE TO evaluate whether the league has gone heavier in passing. You could have a 20% increase in passing and 15% decrease in RB touches (assuming RBs get some of those passes) and it would do nothing to explain a RBBC...it would just prove a decrease in the overall RB workload.

More numbers:

Year	Total Carries	Team Rushes/Game	Total Rushing Yards	Team Rushing/Game2004		14428	     28.2	                   59709	116.62005		14375	     28.1	                   57583	112.52006		14448	     28.2	                   60063	117.32007		13986	     27.3	                   56790	110.92008		14119	     27.6	                   59376	116.02009		14088	     27.5	                   59739	116.72010		13920	     27.2	                   58607	114.5

I got numbers that were way off of this using Data Dominator.

 

[Jeff Pasquino]

Yeah, if we are not going to look at ratio of touches from RB1 to RB2 during this time frame (and just take into account total touches by RB1), we HAVE TO evaluate whether the league has gone heavier in passing. You could have a 20% increase in passing and 15% decrease in RB touches (assuming RBs get some of those passes) and it would do nothing to explain a RBBC...it would just prove a decrease in the overall RB workload.

More numbers:

Year	Total Carries	Team Rushes/Game	Total Rushing Yards	Team Rushing/Game2004		14428	     28.2	                   59709	116.62005		14375	     28.1	                   57583	112.52006		14448	     28.2	                   60063	117.32007		13986	     27.3	                   56790	110.92008		14119	     27.6	                   59376	116.02009		14088	     27.5	                   59739	116.72010		13920	     27.2	                   58607	114.5

I got numbers that were way off of this using Data Dominator.

Sometimes the Data Dominator isn't 100% accurate.I used NFL.com team stats.

 

[CalBear]

I think three years of data are just not enough to draw significant conclusions from--especially since, even at 2008-2010 levels, RBs are being used more than they were at any point from 1960-1999. The anomaly is 2000-2007, when there were a few really durable, high-carry backs like Tomlinson and Curtis Martin. If another Tomlinson comes along, you can count on him to get 300+ touches.

 

[Cato]

Why not add in 2002 and 2003? Prior to 2002 there were fewer teams, so the data will be screwed up if we go back any further.

YEAR    / # of games for a RB with 15+ touches / 20+ touches / 25+ touches2002    /                  443                 /    294      /    1552003    /                  447                 /    298      /    1642004    /                  457                 /    304      /    1712005    /                  445                 /    290      /    1452006    /                  466                 /    305      /    1522007    /                  446                 /    290      /    1492008    /                  442                 /    259      /    1102009    /                  426                 /    241      /    1012010    /                  425                 /    238      /    101

ZWK offered a couple of alternative explanations:(1) There are just fewer RB touches overall.So look at total rushes and receptions by RBs from 2002-10:

YEAR    / Total RB carries / Total RB receptions / Total RB touches2002    /      12016       /      2838           /      148542003    /      12698       /      2635           /      153332004    /      12654       /      2430           /      150842005    /      12726       /      2323           /      150492006    /      12734       /      2525           /      152592007    /      12414       /      2473           /      148872008    /      12498       /      2405           /      149032009    /      12384       /      2439           /      148232010    /      12068       /      2448           /      14516

Fewer overall touches may explain a little bit, but it seems to me that it's not enough to account for the sharp decline, especially in 25+ touch games.(2) Pitch count?This strikes me as a plausible explanation. Maybe some coaches are doing this.The best way I can think of to do this is to use the Game Log Dominator again. I can't sort by total touches so I sort by team, then cap the number of rush attempts by a single player at 21+ and look at the game logs from 2002-10. I'm not going to do it for every team, but here are the first 5 teams:These are the total games in which one RB recorded 21 or more carries:

YEAR    / Arizona / Atlanta / Baltimore / Buffalo / Carolina2002    /    5    /    4    /     7     /    7    /    82003    /    5    /    4    /     12    /    11   /    122004    /    5    /    4    /     8     /    9    /    82005    /    0    /    4    /     5     /    9    /    62006    /    9    /    5    /     7     /    6    /    42007    /    10   /    3    /     8     /    7    /    12008    /    3    /    11   /     9     /    4    /    32009    /    0    /    5    /     2     /    2    /    62010    /    0    /    9    /     6     /    4    /    4

In Arizona, almost all of the high-volume games were compliments of Edgerrin James. In Atlanta, the high-volume games pretty much line up with when Michael Turner can play. In Baltimore, it's when Jamal Lewis is running well. In Buffalo, it's mostly Travis Henry and Willis McGahee, with a reasonable dose of early-career Lynch.In other words, when teams have talented, durable backs, it looks to me like they use them as much as they can. The only exception for these 5 teams is Carolina, where John Fox has had better backs in recent years but seems to be committed to limiting their per-game carries somewhat.So another possible explanation is this:(3a) A few coaches want to limit per-game carries. This could skew the 20+ and 25+ touch games down a little.and(3b) There is a slight decrease in the number of really good NFL RBs right now. This might account for the rest of it.

I like this a lot, I would like to suggest it's a cumulation of all the factors mentioned here, some of them correlated/inter-twined to others. It will be interesting to see if RBBC RBs do last longer if the pitch count concept does hold true. DeAngelo Williams should last longer than the typical RB in theory.

 

[gianmarco]

I think one of the important things we need to do to determine if this is the case is to actually DEFINE RBBC. What constitutes a true RBBC? Main guy getting only 55%? 60%?

I mean, if (and I'm making these #'s completely up) the average % of carries for all teams by the main ball carrier averaged 62% of the carries back in 2000 and now that number is down to 58%, does that mean we're trending more to RBBC?

Or are we trying to say that the # of 250+ carry RBs is down?

The 2nd question is a lot more fantasy relevant than the 1st. I don't care if there's overall a decrease in 25+ carry games as long as my main guy is getting 300+ touches. I'd be much more concerned if there were, say, on average 10 RBs that were getting 300+ carries/year and that number is now down to 5. THAT would be an indication of a shift toward RBBC.

 

[gianmarco]

I just used the historical data dominator for the following:

RBs with 300+ rushes/receptions per year:

2002 -- 16

2003 -- 15

2004 -- 12

2005 -- 14

2006 -- 14

2007 -- 11

2008 -- 9

2009 -- 10

2010 -- 11

RBs with 250+ rushes/receptions per year:

2002 -- 26

2003 -- 22

2004 -- 23

2005 -- 19

2006 -- 26

2007 -- 17

2008 -- 19

2009 -- 17

2010 -- 19

There's definitely a trend downward over the last 4 years. Statistically, I don't know the significance of it. I also don't know how much talent level plays into it as well.

 

[fsufan]

most NFL teams has been going to a RBBC for about 10 years now. the league is almost full blown RBBC

 

[Jeff Pasquino]

I just used the historical data dominator for the following:RBs with 300+ rushes/receptions per year:2002 -- 162003 -- 152004 -- 122005 -- 142006 -- 142007 -- 112008 -- 92009 -- 102010 -- 11RBs with 250+ rushes/receptions per year:2002 -- 262003 -- 222004 -- 232005 -- 192006 -- 262007 -- 172008 -- 192009 -- 172010 -- 19There's definitely a trend downward over the last 4 years. Statistically, I don't know the significance of it. I also don't know how much talent level plays into it as well.

Do you agree or disagree that the more important data is on a weekly / game basis?Annual numbers don't matter for most fantasy purposes. Weekly production does.

 

[gianmarco]

I just used the historical data dominator for the following:RBs with 300+ rushes/receptions per year:2002 -- 162003 -- 152004 -- 122005 -- 142006 -- 142007 -- 112008 -- 92009 -- 102010 -- 11RBs with 250+ rushes/receptions per year:2002 -- 262003 -- 222004 -- 232005 -- 192006 -- 262007 -- 172008 -- 192009 -- 172010 -- 19There's definitely a trend downward over the last 4 years. Statistically, I don't know the significance of it. I also don't know how much talent level plays into it as well.

Do you agree or disagree that the more important data is on a weekly / game basis?Annual numbers don't matter for most fantasy purposes. Weekly production does.

It depends on what you're looking for. In this case, I think end of year numbers are just as helpful as per game numbers. Per game numbers can vary due to any number of reasons: team deciding to pass more in a specific game, minor injury, matchup, ineffectiveness, winning vs. losing. If a team gets out to a huge lead from some quick passing and/or ST TDs, then the running game just may not be very involved. That doesn't mean that they went to RBBC because the main guy didn't get 20+ touches that game.Just as a quick example, let's look at Adrian Peterson. I think we can all agree he is the main guy and is a bell-cow RB. The guy finished with 320 touches and missed one week.From weeks 9-14, his carries were as follows:15, 17, 14, 6, 16, 14 (and his receptions were 4, 4, 1, 1, 1, 0).In those same games, Gerhart had 2, 4, 1, 22, 14, and 1 carries.So in all of those games, if you go by the number of 20+ touch games, only Toby's game of 22 carries hits that mark. Yet, that most definitely is NOT a RBBC situation over those 6 weeks. This is the kind of situation where I think your data can be flawed. For whatever reasons in those games, Minnesota simply didn't run the ball much.But, at the end of the year, AP had 283 carries and 36 receptions. Gerhart finished the year with 81 carries and got 22 of them in 1 game alone when AP was limited. There is ZERO RBBC in Minnesota yet AP failed to hit your criteria in 6 of 16 weeks (he only had 14 carries in week 17 too). I think we can all agree that there is absolutely NO RBBC in Minnesota yet your criteria suggests that they did for almost half the season.

 

[CalBear]

most NFL teams has been going to a RBBC for about 10 years now. the league is almost full blown RBBC

This is just not true. Take any 10 year period in the history of the league, and any metric you like to determine what is RBBC, and the past 10 years are at the absolute bottom in terms of RBBC situations.

 

[Jeff Pasquino]

I just used the historical data dominator for the following:RBs with 300+ rushes/receptions per year:2002 -- 162003 -- 152004 -- 122005 -- 142006 -- 142007 -- 112008 -- 92009 -- 102010 -- 11RBs with 250+ rushes/receptions per year:2002 -- 262003 -- 222004 -- 232005 -- 192006 -- 262007 -- 172008 -- 192009 -- 172010 -- 19There's definitely a trend downward over the last 4 years. Statistically, I don't know the significance of it. I also don't know how much talent level plays into it as well.

Do you agree or disagree that the more important data is on a weekly / game basis?Annual numbers don't matter for most fantasy purposes. Weekly production does.

It depends on what you're looking for. In this case, I think end of year numbers are just as helpful as per game numbers. Per game numbers can vary due to any number of reasons: team deciding to pass more in a specific game, minor injury, matchup, ineffectiveness, winning vs. losing. If a team gets out to a huge lead from some quick passing and/or ST TDs, then the running game just may not be very involved. That doesn't mean that they went to RBBC because the main guy didn't get 20+ touches that game.Just as a quick example, let's look at Adrian Peterson. I think we can all agree he is the main guy and is a bell-cow RB. The guy finished with 320 touches and missed one week.From weeks 9-14, his carries were as follows:15, 17, 14, 6, 16, 14 (and his receptions were 4, 4, 1, 1, 1, 0).In those same games, Gerhart had 2, 4, 1, 22, 14, and 1 carries.So in all of those games, if you go by the number of 20+ touch games, only Toby's game of 22 carries hits that mark. Yet, that most definitely is NOT a RBBC situation over those 6 weeks. This is the kind of situation where I think your data can be flawed. For whatever reasons in those games, Minnesota simply didn't run the ball much.But, at the end of the year, AP had 283 carries and 36 receptions. Gerhart finished the year with 81 carries and got 22 of them in 1 game alone when AP was limited. There is ZERO RBBC in Minnesota yet AP failed to hit your criteria in 6 of 16 weeks (he only had 14 carries in week 17 too). I think we can all agree that there is absolutely NO RBBC in Minnesota yet your criteria suggests that they did for almost half the season.

So you're going to pull six games out of the season as your basis for this argument instead of going with 512 data points?I'll play along - even though this is not a good argument. Doing just some basic research, ADP left one of those games in the second quarter with an ankle injury (the game where Gerhart had 24 carries). Seems to me that the next week the workload was split nearly 50-50 to make sure ADP was healthy. Looking over the ENTIRE Minnesota 2010 season, ADP played 15 games, Gerhart all 16. Their touches broke down as:

Player   / Under 15 Touches / 15-19 / 20-24 / 25+Peterson /       2          /   5   /   3   /  5Gerhart  /      14          /   1   /   1   /  0

So ADP is much more of a feature back as evidenced by 5 games with 25+ touches (out of 101 for the entire league). Gerhart had under 15 touches in every game but the one when ADP was injured (Week 12) and when ADP missed the entire week (Week 15).So if you define a feature back as 20+ touches, ADP hit that mark more than 50% of the time - with one game limited by an early injury. If you moved the line to 19, ADP hit that mark twice, making it even more dramatic (10 of 15 games with 19+ touches).

 

[gianmarco]

I just used the historical data dominator for the following:

RBs with 300+ rushes/receptions per year:

2002 -- 16

2003 -- 15

2004 -- 12

2005 -- 14

2006 -- 14

2007 -- 11

2008 -- 9

2009 -- 10

2010 -- 11

RBs with 250+ rushes/receptions per year:

2002 -- 26

2003 -- 22

2004 -- 23

2005 -- 19

2006 -- 26

2007 -- 17

2008 -- 19

2009 -- 17

2010 -- 19

There's definitely a trend downward over the last 4 years. Statistically, I don't know the significance of it. I also don't know how much talent level plays into it as well.

Do you agree or disagree that the more important data is on a weekly / game basis?

Annual numbers don't matter for most fantasy purposes. Weekly production does.

It depends on what you're looking for. In this case, I think end of year numbers are just as helpful as per game numbers. Per game numbers can vary due to any number of reasons: team deciding to pass more in a specific game, minor injury, matchup, ineffectiveness, winning vs. losing. If a team gets out to a huge lead from some quick passing and/or ST TDs, then the running game just may not be very involved. That doesn't mean that they went to RBBC because the main guy didn't get 20+ touches that game.Just as a quick example, let's look at Adrian Peterson. I think we can all agree he is the main guy and is a bell-cow RB. The guy finished with 320 touches and missed one week.

From weeks 9-14, his carries were as follows:

15, 17, 14, 6, 16, 14 (and his receptions were 4, 4, 1, 1, 1, 0).

In those same games, Gerhart had 2, 4, 1, 22, 14, and 1 carries.

So in all of those games, if you go by the number of 20+ touch games, only Toby's game of 22 carries hits that mark. Yet, that most definitely is NOT a RBBC situation over those 6 weeks. This is the kind of situation where I think your data can be flawed. For whatever reasons in those games, Minnesota simply didn't run the ball much.

But, at the end of the year, AP had 283 carries and 36 receptions. Gerhart finished the year with 81 carries and got 22 of them in 1 game alone when AP was limited. There is ZERO RBBC in Minnesota yet AP failed to hit your criteria in 6 of 16 weeks (he only had 14 carries in week 17 too). I think we can all agree that there is absolutely NO RBBC in Minnesota yet your criteria suggests that they did for almost half the season.

So you're going to pull six games out of the season as your basis for this argument instead of going with 512 data points?I'll play along - even though this is not a good argument. Doing just some basic research, ADP left one of those games in the second quarter with an ankle injury (the game where Gerhart had 24 carries). Seems to me that the next week the workload was split nearly 50-50 to make sure ADP was healthy.

Looking over the ENTIRE Minnesota 2010 season, ADP played 15 games, Gerhart all 16. Their touches broke down as:

Player / Under 15 Touches / 15-19 / 20-24 / 25+Peterson / 2 / 5 / 3 / 5Gerhart / 14 / 1 / 1 / 0So ADP is much more of a feature back as evidenced by 5 games with 25+ touches (out of 101 for the entire league). Gerhart had under 15 touches in every game but the one when ADP was injured (Week 12) and when ADP missed the entire week (Week 15).

Huh? I picked ONE player and used it as an example. I never stated it completely refuted all of your data. I simply wasn't going to take the time to look through player after player in each individual situation. So, I picked one of the most obvious bell-cows and decided to look. Now look at what you just posted at the very bottom. ADP is a feature back based on his 25+ touches. Not only that, we KNOW that ADP is the feature back because he is obviously the guy and gets the majority of the carries. Yet, he had SEVEN games under 20 touches. SEVEN.

So if you tell me an unnamed RB has less than 20 touches in 7 of the 15 games he played in, I would conclude that the guy is likely in a RBBC. I mean, that's why you're looking at # of 20+ touch games, right? Yet we KNOW that isn't the case. There is no RBBC whatsoever in Minnesota save the one game where AP came back from injury and split the load with Gerhart.

This is why your data CAN BE flawed. It's not going to be in every instance, but it clearly is in this one and it's the first one I looked at. It's why I'm suggesting you can't ONLY look at the data you did to conclude the RBBC is trending upward. If you look at the year end numbers, it's crystal clear that AP is the main ball carrier (320 touches to 100 touches or about 75%). Yet based on per game numbers, AP isn't even hitting 20 touches in ~half the games he played and looks to be a RBBC guy. If we're not going to draw that conclusion on your data, then why are we looking at it and saying "well, it's trending downward so there's more RBBC"?

 

[gianmarco]

So if you define a feature back as 20+ touches, ADP hit that mark more than 50% of the time - with one game limited by an early injury. If you moved the line to 19, ADP hit that mark twice, making it even more dramatic (10 of 15 games with 19+ touches).

You just added this part so I'll address it:I'll flip it the other way: ADP failed to hit 20+ touches 47% of the time (7 of 15 games), thus it's unlikely he's a feature back. Same data, different conclusion.

 

[Aabye]

Huh? I picked ONE player and used it as an example. I never stated it completely refuted all of your data. I simply wasn't going to take the time to look through player after player in each individual situation. So, I picked one of the most obvious bell-cows and decided to look. Now look at what you just posted at the very bottom. ADP is a feature back based on his 25+ touches. Not only that, we KNOW that ADP is the feature back because he is obviously the guy and gets the majority of the carries. Yet, he had SEVEN games under 20 touches. SEVEN. So if you tell me an unnamed RB has less than 20 touches in 7 of the 15 games he played in, I would conclude that the guy is likely in a RBBC. I mean, that's why you're looking at # of 20+ touch games, right? Yet we KNOW that isn't the case. There is no RBBC whatsoever in Minnesota save the one game where AP came back from injury and split the load with Gerhart.This is why your data CAN BE flawed. It's not going to be in every instance, but it clearly is in this one and it's the first one I looked at. It's why I'm suggesting you can't ONLY look at the data you did to conclude the RBBC is trending upward. If you look at the year end numbers, it's crystal clear that AP is the main ball carrier (320 touches to 100 touches or about 75%). Yet based on per game numbers, AP isn't even hitting 20 touches in ~half the games he played and looks to be a RBBC guy. If we're not going to draw that conclusion on your data, then why are we looking at it and saying "well, it's trending downward so there's more RBBC"?

I agree that there is an issue here. You mentioned in post 29 that we should actually get clear about stipulating just what a RBBC is.The most natural definition (to me) seems to be per-game % of a team's carries. If a team gives 2/3 or more of the carries to the starter when he's healthy, we can pretty safely conclude that he is not in a RBBC.Jeff's numbers show just what they show: there are fewer games in which RBs touch the ball 20+ and 25+ times. And as you've pointed out, that doesn't necessarily mean that there are more RBBCs. There are alternative explanations that we can propose and test to see if they might account for this reduction.It seems a little strange to me to worry about whether these statistics support the claim that there are more RBBCs when we can just:(1) Stipulate what a RBBC actually isand then(2) directly look up whether or not there are more of them in 2009 than in 2006 (or whenever).In other words, this is a factual question, not a matter of interpretation.

 

[gianmarco]

Here is another example:

Player X had 8 games last year with 20+ touches. This is 1 more 20+ touch game than Adrian Peterson. This includes 5 games with 20+ carries alone.

Based on your conclusion above with ADP, you'd say that this guy is not in a RBBC.

Guess what: This guy was 2nd on his team in total touches to Jamaal Charles. Player X = Thomas Jones and is what you would call the very definition of RBBC as they split it right down the middle.

So, using your data, ADP and Thomas Jones look identical. They had 7 and 8 games of 20+ touches respectively. Yet they couldn't be more different in terms of RBBC as one is the CLEAR lead ball carrier with no RBBC whatsoever and the other guy is in a completely RBBC and was actually 2nd in touches on his own team. And no, it didn't trend differently as the year went on as T. Jones still had 22 carries and 23 carries in weeks 15 and 16.

You're using the above criteria to define RBBC and I've pretty clearly showed how the data doesn't fit that definition based only on the numbers.

 

[Alex P Keaton]

I just used the historical data dominator for the following:

RBs with 300+ rushes/receptions per year:

2002 -- 16

2003 -- 15

2004 -- 12

2005 -- 14

2006 -- 14

2007 -- 11

2008 -- 9

2009 -- 10

2010 -- 11

RBs with 250+ rushes/receptions per year:

2002 -- 26

2003 -- 22

2004 -- 23

2005 -- 19

2006 -- 26

2007 -- 17

2008 -- 19

2009 -- 17

2010 -- 19

There's definitely a trend downward over the last 4 years. Statistically, I don't know the significance of it. I also don't know how much talent level plays into it as well.

Do you agree or disagree that the more important data is on a weekly / game basis?

Annual numbers don't matter for most fantasy purposes. Weekly production does.

It depends on what you're looking for. In this case, I think end of year numbers are just as helpful as per game numbers. Per game numbers can vary due to any number of reasons: team deciding to pass more in a specific game, minor injury, matchup, ineffectiveness, winning vs. losing. If a team gets out to a huge lead from some quick passing and/or ST TDs, then the running game just may not be very involved. That doesn't mean that they went to RBBC because the main guy didn't get 20+ touches that game.Just as a quick example, let's look at Adrian Peterson. I think we can all agree he is the main guy and is a bell-cow RB. The guy finished with 320 touches and missed one week.

From weeks 9-14, his carries were as follows:

15, 17, 14, 6, 16, 14 (and his receptions were 4, 4, 1, 1, 1, 0).

In those same games, Gerhart had 2, 4, 1, 22, 14, and 1 carries.

So in all of those games, if you go by the number of 20+ touch games, only Toby's game of 22 carries hits that mark. Yet, that most definitely is NOT a RBBC situation over those 6 weeks. This is the kind of situation where I think your data can be flawed. For whatever reasons in those games, Minnesota simply didn't run the ball much.

But, at the end of the year, AP had 283 carries and 36 receptions. Gerhart finished the year with 81 carries and got 22 of them in 1 game alone when AP was limited. There is ZERO RBBC in Minnesota yet AP failed to hit your criteria in 6 of 16 weeks (he only had 14 carries in week 17 too). I think we can all agree that there is absolutely NO RBBC in Minnesota yet your criteria suggests that they did for almost half the season.

So you're going to pull six games out of the season as your basis for this argument instead of going with 512 data points?I'll play along - even though this is not a good argument. Doing just some basic research, ADP left one of those games in the second quarter with an ankle injury (the game where Gerhart had 24 carries). Seems to me that the next week the workload was split nearly 50-50 to make sure ADP was healthy.

Looking over the ENTIRE Minnesota 2010 season, ADP played 15 games, Gerhart all 16. Their touches broke down as:

Player / Under 15 Touches / 15-19 / 20-24 / 25+Peterson / 2 / 5 / 3 / 5Gerhart / 14 / 1 / 1 / 0So ADP is much more of a feature back as evidenced by 5 games with 25+ touches (out of 101 for the entire league). Gerhart had under 15 touches in every game but the one when ADP was injured (Week 12) and when ADP missed the entire week (Week 15).

Huh? I picked ONE player and used it as an example. I never stated it completely refuted all of your data. I simply wasn't going to take the time to look through player after player in each individual situation. So, I picked one of the most obvious bell-cows and decided to look. Now look at what you just posted at the very bottom. ADP is a feature back based on his 25+ touches. Not only that, we KNOW that ADP is the feature back because he is obviously the guy and gets the majority of the carries. Yet, he had SEVEN games under 20 touches. SEVEN.

So if you tell me an unnamed RB has less than 20 touches in 7 of the 15 games he played in, I would conclude that the guy is likely in a RBBC. I mean, that's why you're looking at # of 20+ touch games, right? Yet we KNOW that isn't the case. There is no RBBC whatsoever in Minnesota save the one game where AP came back from injury and split the load with Gerhart.

This is why your data CAN BE flawed. It's not going to be in every instance, but it clearly is in this one and it's the first one I looked at. It's why I'm suggesting you can't ONLY look at the data you did to conclude the RBBC is trending upward. If you look at the year end numbers, it's crystal clear that AP is the main ball carrier (320 touches to 100 touches or about 75%). Yet based on per game numbers, AP isn't even hitting 20 touches in ~half the games he played and looks to be a RBBC guy. If we're not going to draw that conclusion on your data, then why are we looking at it and saying "well, it's trending downward so there's more RBBC"?

On the flipside, Jeff's analysis will have far more data points than yours, therefore it is FAR less likely to be skewed by outliers. Your data can be (at a minimum) equally flawed, as Jeff has already explained.Either way, both of you have provided great analysis to support the idea that RBBC is increasingly common.

 

[gianmarco]

Huh? I picked ONE player and used it as an example. I never stated it completely refuted all of your data. I simply wasn't going to take the time to look through player after player in each individual situation. So, I picked one of the most obvious bell-cows and decided to look. Now look at what you just posted at the very bottom. ADP is a feature back based on his 25+ touches. Not only that, we KNOW that ADP is the feature back because he is obviously the guy and gets the majority of the carries. Yet, he had SEVEN games under 20 touches. SEVEN. So if you tell me an unnamed RB has less than 20 touches in 7 of the 15 games he played in, I would conclude that the guy is likely in a RBBC. I mean, that's why you're looking at # of 20+ touch games, right? Yet we KNOW that isn't the case. There is no RBBC whatsoever in Minnesota save the one game where AP came back from injury and split the load with Gerhart.This is why your data CAN BE flawed. It's not going to be in every instance, but it clearly is in this one and it's the first one I looked at. It's why I'm suggesting you can't ONLY look at the data you did to conclude the RBBC is trending upward. If you look at the year end numbers, it's crystal clear that AP is the main ball carrier (320 touches to 100 touches or about 75%). Yet based on per game numbers, AP isn't even hitting 20 touches in ~half the games he played and looks to be a RBBC guy. If we're not going to draw that conclusion on your data, then why are we looking at it and saying "well, it's trending downward so there's more RBBC"?

I agree that there is an issue here. You mentioned in post 29 that we should actually get clear about stipulating just what a RBBC is.The most natural definition (to me) seems to be per-game % of a team's carries. If a team gives 2/3 or more of the carries to the starter when he's healthy, we can pretty safely conclude that he is not in a RBBC.Jeff's numbers show just what they show: there are fewer games in which RBs touch the ball 20+ and 25+ times. And as you've pointed out, that doesn't necessarily mean that there are more RBBCs. There are alternative explanations that we can propose and test to see if they might account for this reduction.It seems a little strange to me to worry about whether these statistics support the claim that there are more RBBCs when we can just:(1) Stipulate what a RBBC actually isand then(2) directly look up whether or not there are more of them in 2009 than in 2006 (or whenever).In other words, this is a factual question, not a matter of interpretation.

It's very possible that those stating RBBC is increasing are just as right as those saying RBBC is decreasing. It depends on how you define it. And unfortunately, we will never all agree on a definition. To some, anything less than 65/35 is a RBBC. Others will go with 60/40. Others will say that the main guy needs to have 250 touches. Another may say 275 touches. And they would all be right. Maybe some care if ALL the RBs are getting less touches compared to 10 yrs ago. Others may care that a higher % of RBs aren't getting enough carries to score well. In the end, all I care about is a guy that's getting enough to consistently score. I think the above information Jeff posted is very useful but it does not definitively answer the question of RBBC increasing or decreasing. Factor in differences in talent level and it makes it very difficult to accurately answer to everyone's satisfaction.

 

[Alex P Keaton]

Here is another examplelayer X had 8 games last year with 20+ touches. This is 1 more 20+ touch game than Adrian Peterson. This includes 5 games with 20+ carries alone.Based on your conclusion above with ADP, you'd say that this guy is not in a RBBC.Guess what: This guy was 2nd on his team in total touches to Jamaal Charles. Player X = Thomas Jones and is what you would call the very definition of RBBC as they split it right down the middle.So, using your data, ADP and Thomas Jones look identical. They had 7 and 8 games of 20+ touches respectively. Yet they couldn't be more different in terms of RBBC as one is the CLEAR lead ball carrier with no RBBC whatsoever and the other guy is in a completely RBBC and was actually 2nd in touches on his own team. And no, it didn't trend differently as the year went on as T. Jones still had 22 carries and 23 carries in weeks 15 and 16.You're using the above criteria to define RBBC and I've pretty clearly showed how the data doesn't fit that definition based only on the numbers.

You are fixating too much on individual player examples, rather than a summation across all players AND a comparison year-by-year.

 

[gianmarco]

Here is another examplelayer X had 8 games last year with 20+ touches. This is 1 more 20+ touch game than Adrian Peterson. This includes 5 games with 20+ carries alone.Based on your conclusion above with ADP, you'd say that this guy is not in a RBBC.Guess what: This guy was 2nd on his team in total touches to Jamaal Charles. Player X = Thomas Jones and is what you would call the very definition of RBBC as they split it right down the middle.So, using your data, ADP and Thomas Jones look identical. They had 7 and 8 games of 20+ touches respectively. Yet they couldn't be more different in terms of RBBC as one is the CLEAR lead ball carrier with no RBBC whatsoever and the other guy is in a completely RBBC and was actually 2nd in touches on his own team. And no, it didn't trend differently as the year went on as T. Jones still had 22 carries and 23 carries in weeks 15 and 16.You're using the above criteria to define RBBC and I've pretty clearly showed how the data doesn't fit that definition based only on the numbers.

You are fixating too much on individual player examples, rather than a summation across all players AND a comparison year-by-year.

I'm not fixating on anything. I'm pointing out that the conclusion being reached based on the data is flawed. Using # of 20+ touch games as an indicator of being or not being in a RBBC is not the best indicator. I showed an example of 2 players with the same data set that are in completely different situations. Thus, how can you use that data to define and conclude what's going on with RBBC when it simply doesn't answer that question?Let's pick another example:LeSean McCoy had 285 touches last year and over 200 carries. The next closest guy was Jerome Harrison with 40 carries and 8 receptions. Clearly NOT a RBBC, correct?McCoy had NINE games under 20 touches and he only played in 15 games. He didn't hit 20+ touches in 60% of the games he played in. Yet, he had almost 300 touches for the year and was absolutely the lead ball carrier. But, using 20+ touches as a measure of being the main ball carrier, he fails pretty miserably. How many more examples do you need to see that the data isn't answering the question proposed?

 

[gianmarco]

I'm going to propose another hypothesis:

While the number of clear lead ball carrier RBs hasn't changed, the number of games where they are used as such by defining it with 20+ touches in a game has decreased.

What do I mean? Again, I didn't look at every single guy, but I looked back at guys like LT, Edge, Curtis Martin, SA.....

These guys had very few games each year where they didn't get 20+ touches. From 2002-2007, for example, LT never had more than 3 games where he failed to get 20+ touches. Similar #s for Edge, SA, etc. Nowadays, however, it seems it's more common for the main guy to have games where he doesn't do so. Looking just at least year with the guys that had 300+ carries, only SJax and MJD failed to hit 20+ touches in 3 games. Here's the others

AP -- 7 games under 20+ touches (320 touches for the year)

Mendenhall -- 8 games under 20+ touches (almost 350 touches for the year)

Ray Rice -- 6 games under 20+ touches (370+ touches for the year)

Cedric Benson -- 5 games under 20+ touches (350 touches for the year)

Michael Turner -- 8 games under 20+ touches ( 350 touches for the year)

CJ3 -- 4 games under 20+ touches

Foster -- 4 games under 20+ touches

When you start to add those up, this can start to account for some of the decrease in 20+ touch games. It doesn't mean that there's more RBBC for these guys in particular, it just means their carries aren't as evenly distributed game by game. Maybe coaches aren't using their guys as heavily if they come off a high workload game. Maybe they are trying to rest little nagging injuries. But, it does seem that the main guys seem to have a lot more variability and a lot more games where they simply aren't used as much when compare to their counterparts from earlier in the decade. This may be wrong if you look at each guy for every year (which I'm not about to do), but based on preliminary looks, this seems to be the case. In fact, I'm surprised how often these main guys are not being used for complete games.

 

[Aabye]

I'm going to propose another hypothesis:While the number of clear lead ball carrier RBs hasn't changed, the number of games where they are used as such by defining it with 20+ touches in a game has decreased. What do I mean? Again, I didn't look at every single guy, but I looked back at guys like LT, Edge, Curtis Martin, SA.....These guys had very few games each year where they didn't get 20+ touches. From 2002-2007, for example, LT never had more than 3 games where he failed to get 20+ touches. Similar #s for Edge, SA, etc. Nowadays, however, it seems it's more common for the main guy to have games where he doesn't do so. Looking just at least year with the guys that had 300+ carries, only SJax and MJD failed to hit 20+ touches in 3 games. Here's the othersAP -- 7 games under 20+ touches (320 touches for the year)Mendenhall -- 8 games under 20+ touches (almost 350 touches for the year)Ray Rice -- 6 games under 20+ touches (370+ touches for the year)Cedric Benson -- 5 games under 20+ touches (350 touches for the year)Michael Turner -- 8 games under 20+ touches ( 350 touches for the year)CJ3 -- 4 games under 20+ touchesFoster -- 4 games under 20+ touchesWhen you start to add those up, this can start to account for some of the decrease in 20+ touch games. It doesn't mean that there's more RBBC for these guys in particular, it just means their carries aren't as evenly distributed game by game. Maybe coaches aren't using their guys as heavily if they come off a high workload game. Maybe they are trying to rest little nagging injuries. But, it does seem that the main guys seem to have a lot more variability and a lot more games where they simply aren't used as much when compare to their counterparts from earlier in the decade. This may be wrong if you look at each guy for every year (which I'm not about to do), but based on preliminary looks, this seems to be the case. In fact, I'm surprised how often these main guys are not being used for complete games.

I never would have considered this hypothesis. Really, really insightful and interesting idea, gianmarco.This is a fantastic discussion thread.

 

[CalBear]

It is clear that "games with at least X% of the team's RB carries" is a much better metric for whether something is RBBC than "games with at least X carries." Unfortunately it's a lot harder to gather data for.

I also think we have to look at the multiple endpoints problem. 2008-2010 might have fewer bell cow RBs than 2004-2007, but 2004-2007 is part of a seriously anomalous blip (1998-2007) which includes the only two seasons with 410+ carries, three of the five seasons with 400+ carries, five of the ten seasons with 390+ carries, eight of the 16 seaons with 380+ carries, and so on. Any way you slice it, 1998-2007 produced half or more of all the serious bell cow seasons in NFL history. It's like looking at WR numbers from 1996 to 1998 and concluding that WR numbers are way down because we're not seeing four players with 1500+ yards in those years like we did in 1995.

I also think the small sample size skews the analysis significantly. High-carry RB seasons come when the right RB gets the right situation, and right now we haven't seen a direct replacement of backs like Martin, Tomlinson, Alexander. That doesn't mean we won't see one tomorrow.

 

[Jeff Pasquino]

To be clear - the debates with gianmarco are healthy. He knows that I'm not picking on him, nor is he on me (at least I think not).

I think what everyone is stating here is that there is a trend towards RBBC - but finding the data is a big pain in the ### to prove it. I would agree that yes, a back who gets 30-35% of the touches per game as the RB2 is in a RBBC. Now what that means to a particular offense is completely different on a per team viewpoint. Carolina and the Jets run more than Philly. I know that, and most NFL fans know that. But if McCoy gets 10 touches in an infurating "Andy Reid is calling 60 pass play" week, that doesn't mean that Jerome Harrison's 5 touches would make him a committee back.

All the lines are arbitrary, but the data in Post 1 does prove that feature backs (which I believe would be 20-25 touch/game for one player) is diminishing. Proving each game as RBBC is a bit harder to do due to other contributing factors (injuries, game plans, low number of runs, falling behind 21-0 in the first quarter, etc.).

Looking at annual data isn't a good idea, IMHO, especially in fantasy football since most teams score on a weekly - not annual - basis. Weekly lineup decisions and add/drop decisions are made every week. When it comes to Week 16, we don't care if RB X has 280 carries that year or 20 - what is he projected to do for the FF Championship Week?

I'll contend that weekly data matters more, but I will admit that more study to really REALLY get at a true RBBC may be required. I don't have that data and I don't necessarily want to delve into that exercise. I am OK with my initial analysis being better than an annual viewpoint in telling me the general league-wide trend that fewer backs are featured backs and then the data later in this thread that the league isn't running the ball any less in the last three years than in 2004-2007.

 

[Alex P Keaton]

Here is another example:

Player X had 8 games last year with 20+ touches. This is 1 more 20+ touch game than Adrian Peterson. This includes 5 games with 20+ carries alone.

Based on your conclusion above with ADP, you'd say that this guy is not in a RBBC.

Guess what: This guy was 2nd on his team in total touches to Jamaal Charles. Player X = Thomas Jones and is what you would call the very definition of RBBC as they split it right down the middle.

So, using your data, ADP and Thomas Jones look identical. They had 7 and 8 games of 20+ touches respectively. Yet they couldn't be more different in terms of RBBC as one is the CLEAR lead ball carrier with no RBBC whatsoever and the other guy is in a completely RBBC and was actually 2nd in touches on his own team. And no, it didn't trend differently as the year went on as T. Jones still had 22 carries and 23 carries in weeks 15 and 16.

You're using the above criteria to define RBBC and I've pretty clearly showed how the data doesn't fit that definition based only on the numbers.

You are fixating too much on individual player examples, rather than a summation across all players AND a comparison year-by-year.

I'm not fixating on anything. I'm pointing out that the conclusion being reached based on the data is flawed. Using # of 20+ touch games as an indicator of being or not being in a RBBC is not the best indicator. I showed an example of 2 players with the same data set that are in completely different situations. Thus, how can you use that data to define and conclude what's going on with RBBC when it simply doesn't answer that question?

Let's pick another example:

LeSean McCoy had 285 touches last year and over 200 carries. The next closest guy was Jerome Harrison with 40 carries and 8 receptions. Clearly NOT a RBBC, correct?

McCoy had NINE games under 20 touches and he only played in 15 games. He didn't hit 20+ touches in 60% of the games he played in. Yet, he had almost 300 touches for the year and was absolutely the lead ball carrier. But, using 20+ touches as a measure of being the main ball carrier, he fails pretty miserably.

How many more examples do you need to see that the data isn't answering the question proposed?

Actually, I do think you're either a) fixating, or b) being pedantic.If he rephrased this to say "the day of the bellcow RB is coming to an end" then the data clearly demonstrates this to be true. No?

So rather than fixating on "RBBC", I'm far more interested in whether or not workload is decreasing for RB1 on teams. And again, the data does very clearly demonstrate this. Almost to the point where I would say "who cares whether we call it RBBC - all that matters is that is becoming much more rare for a RB to get a ton of touches - whether in individual games or across a 16-game season."

 

[Alex P Keaton]

To be clear - the debates with gianmarco are healthy. He knows that I'm not picking on him, nor is he on me (at least I think not).

I think what everyone is stating here is that there is a trend towards RBBC - but finding the data is a big pain in the ### to prove it. I would agree that yes, a back who gets 30-35% of the touches per game as the RB2 is in a RBBC. Now what that means to a particular offense is completely different on a per team viewpoint. Carolina and the Jets run more than Philly. I know that, and most NFL fans know that. But if McCoy gets 10 touches in an infurating "Andy Reid is calling 60 pass play" week, that doesn't mean that Jerome Harrison's 5 touches would make him a committee back.

All the lines are arbitrary, but the data in Post 1 does prove that feature backs (which I believe would be 20-25 touch/game for one player) is diminishing. Proving each game as RBBC is a bit harder to do due to other contributing factors (injuries, game plans, low number of runs, falling behind 21-0 in the first quarter, etc.).

Looking at annual data isn't a good idea, IMHO, especially in fantasy football since most teams score on a weekly - not annual - basis. Weekly lineup decisions and add/drop decisions are made every week. When it comes to Week 16, we don't care if RB X has 280 carries that year or 20 - what is he projected to do for the FF Championship Week?

I'll contend that weekly data matters more, but I will admit that more study to really REALLY get at a true RBBC may be required. I don't have that data and I don't necessarily want to delve into that exercise. I am OK with my initial analysis being better than an annual viewpoint in telling me the general league-wide trend that fewer backs are featured backs and then the data later in this thread that the league isn't running the ball any less in the last three years than in 2004-2007.

I don't think gianmarco is saying this at all.

 

[Jeff Pasquino]

It is clear that "games with at least X% of the team's RB carries" is a much better metric for whether something is RBBC than "games with at least X carries." Unfortunately it's a lot harder to gather data for.

Exactly.

I also think we have to look at the multiple endpoints problem. 2008-2010 might have fewer bell cow RBs than 2004-2007, but 2004-2007 is part of a seriously anomalous blip (1998-2007) which includes the only two seasons with 410+ carries, three of the five seasons with 400+ carries, five of the ten seasons with 390+ carries, eight of the 16 seaons with 380+ carries, and so on. Any way you slice it, 1998-2007 produced half or more of all the serious bell cow seasons in NFL history. It's like looking at WR numbers from 1996 to 1998 and concluding that WR numbers are way down because we're not seeing four players with 1500+ yards in those years like we did in 1995. I also think the small sample size skews the analysis significantly. High-carry RB seasons come when the right RB gets the right situation, and right now we haven't seen a direct replacement of backs like Martin, Tomlinson, Alexander. That doesn't mean we won't see one tomorrow.

I disagree on the sample size, simply because the RBBC phenomenon really was spurred on after the Colts and Bears showed how successful it can be. The cat seems to be out of the bag as far as using RBBC, and I don't know that even bell cow caliber RBs will be rode as hard as they once were. Add in that teams want to preserve their assets (not overwork them) along with the data that the FF value / production isn't greatly diminished with fewer touches and involving a second back and the trend may become the norm - at least until a feature back team wins a championship.As for passing / receiving yardage and stats changes, that's a different issue. A rule change (no contact after five yards) significantly opened up the passing game. I don't know of any watershed rules that impacted RBBC.