RB Touches (1 Viewer)

[mggoilers]

Adrian Peterson and Michael Turner both had over 380 touches in 2008(usually the mark is 400+ touches). What does history tell us about RB's who get this much work?

In some cases, the results are realized pretty quick...LJ and Jamal Anderson.

Sometimes, it takes a few years for the problems to arise...Tomlinson.

In the age of the RBBC, the bell cow is a pretty rare thing these days. Just looking to get a fresh perspective. Peterson and Turner both have a history of getting nicked up, and both have very capable backups. I don't expect an immediate falloff or injury but curious about the history.

 

[CalBear]

Adrian Peterson and Michael Turner both had over 380 touches in 2008(usually the mark is 400+ touches). What does history tell us about RB's who get this much work?In some cases, the results are realized pretty quick...LJ and Jamal Anderson.Sometimes, it takes a few years for the problems to arise...Tomlinson.In the age of the RBBC, the bell cow is a pretty rare thing these days. Just looking to get a fresh perspective. Peterson and Turner both have a history of getting nicked up, and both have very capable backups. I don't expect an immediate falloff or injury but curious about the history.

The history is:1) Exemplary populations regress to the mean.2) People fallaciously believe that point 1) means that an individual will regress to the mean.3) People cherry-pick data to prove what they already want to believe. (Check out why it's the 370-carry rule and not the 365-carry rule).4) Since the merger, the last 10 years is the decade with the lowest amount of RBBC.

 

[Maroney=Speed]

Adrian Peterson and Michael Turner both had over 380 touches in 2008(usually the mark is 400+ touches). What does history tell us about RB's who get this much work?In some cases, the results are realized pretty quick...LJ and Jamal Anderson.Sometimes, it takes a few years for the problems to arise...Tomlinson.In the age of the RBBC, the bell cow is a pretty rare thing these days. Just looking to get a fresh perspective. Peterson and Turner both have a history of getting nicked up, and both have very capable backups. I don't expect an immediate falloff or injury but curious about the history.

The history is:1) Exemplary populations regress to the mean.2) People fallaciously believe that point 1) means that an individual will regress to the mean.3) People cherry-pick data to prove what they already want to believe. (Check out why it's the 370-carry rule and not the 365-carry rule).4) Since the merger, the last 10 years is the decade with the lowest amount of RBBC.

Good points. Why is the age 30 for RB's, and not 28 1/2? Those cherry-pickers.

 

[JamesTheScot]

Adrian Peterson and Michael Turner both had over 380 touches in 2008(usually the mark is 400+ touches). What does history tell us about RB's who get this much work?In some cases, the results are realized pretty quick...LJ and Jamal Anderson.Sometimes, it takes a few years for the problems to arise...Tomlinson.In the age of the RBBC, the bell cow is a pretty rare thing these days. Just looking to get a fresh perspective. Peterson and Turner both have a history of getting nicked up, and both have very capable backups. I don't expect an immediate falloff or injury but curious about the history.

The history is:1) Exemplary populations regress to the mean.2) People fallaciously believe that point 1) means that an individual will regress to the mean.3) People cherry-pick data to prove what they already want to believe. (Check out why it's the 370-carry rule and not the 365-carry rule).4) Since the merger, the last 10 years is the decade with the lowest amount of RBBC.

Doesn't #4 kind of run afoul of #3? I'm not disagreeing with you on #4, but lumping what was going on 10 years ago in the NFL with today is rather thin to me. What's the trend over the last 3 or 4 years and is it explanable (and predictable) by virtue of pedigree. In other words, are RBBC coaches taking over for bell-cow coaches the same way the WCO spread?

 

[CalBear]

I'm not disagreeing with you on #4, but lumping what was going on 10 years ago in the NFL with today is rather thin to me. What's the trend over the last 3 or 4 years and is it explanable (and predictable) by virtue of pedigree. In other words, are RBBC coaches taking over for bell-cow coaches the same way the WCO spread?

Since 2002 (7 seasons), there have been 15 RB seasons of 350+ carries. That's an average of just over two RBs per year. Seven of those are within the past four years, including two this year. From 1960-2001 (42 seasons), there were a grand total of 38 RB seasons with 350+ carries, an average of less than one per season.By any statistical measure, there are more "bell cow" RBs now than ever before.

 

[Synthesizer]

Tomlinson had 400 touches in each of his first 3 seasons (2001, 2002, 2003). But that didn't seem to slow him down much in 2004, 2005, 2006, and 2007. People are always looking for negatives in a RB, but I guess I consider it a good thing when he's getting 400 touches.

 

[rzrback77]

I'm not disagreeing with you on #4, but lumping what was going on 10 years ago in the NFL with today is rather thin to me. What's the trend over the last 3 or 4 years and is it explanable (and predictable) by virtue of pedigree. In other words, are RBBC coaches taking over for bell-cow coaches the same way the WCO spread?

Since 2002 (7 seasons), there have been 15 RB seasons of 350+ carries. That's an average of just over two RBs per year. Seven of those are within the past four years, including two this year. From 1960-2001 (42 seasons), there were a grand total of 38 RB seasons with 350+ carries, an average of less than one per season.

By any statistical measure, there are more "bell cow" RBs now than ever before.

Not a good comparison, as the season switched from 14 games to 16 games in 1978. I don't believe that there were many 350 carry seasons when there were only 14 games. That would be averaging 25 carries per game for the entire season.

 

[rzrback77]

Just looking at over 300 carries between 1960 and 1977, when there were 14 games

1 Walter Payton rb 1977 23 3 14 339 1852 5.46 14 308.10

2 Mark VanEeghen rb 1977 25 4 14 324 1273 3.93 7 182.80

3 Lydell Mitchell rb 1977 28 6 14 301 1159 3.85 3 219.90

4 Franco Harris rb 1977 27 6 14 300 1162 3.87 11 188.40

5 Walter Payton rb 1976 22 2 14 311 1390 4.47 13 231.90

6 O.J. Simpson rb 1975 28 7 14 329 1817 5.52 16 362.30

7 O.J. Simpson rb 1973 26 5 14 332 2003 6.03 12 279.30

8 Jim Brown rb 1961 25 5 14 305 1408 4.62 8 246.70

One guy in 61 Jim Brown 305 for 1408 4.6 ypc

Next one in 73 OJ Simpson 332 2003 6.0 ypc

Simpson again in 75 329 1817 5.52 ypc

Payton in 76 and 77

Three others in 77

 

[CalBear]

I'm not disagreeing with you on #4, but lumping what was going on 10 years ago in the NFL with today is rather thin to me. What's the trend over the last 3 or 4 years and is it explanable (and predictable) by virtue of pedigree. In other words, are RBBC coaches taking over for bell-cow coaches the same way the WCO spread?

Since 2002 (7 seasons), there have been 15 RB seasons of 350+ carries. That's an average of just over two RBs per year. Seven of those are within the past four years, including two this year. From 1960-2001 (42 seasons), there were a grand total of 38 RB seasons with 350+ carries, an average of less than one per season.

By any statistical measure, there are more "bell cow" RBs now than ever before.

Not a good comparison, as the season switched from 14 games to 16 games in 1978. I don't believe that there were many 350 carry seasons when there were only 14 games. That would be averaging 25 carries per game for the entire season.

Fine. From 1978-2001 (24 seasons) there were 38 350-carry seasons, or just over 1.5 per year, and far less than from 2002-2008.Though really, I would put the dividing line at 1997; I used 2002 just because that's what's in the Data Dominator. 1997 saw Bettis, TD, and Eddie George all get 350+ rushes. Since then (12 seasons), there have been 28 RB seasons of 350+ carries (2.3/year). From 1978-1996 (19 seasons), there were just 26 RB seasons of 350+ carries (1.4/year).

And before anyone quibbles with the number of carries I chose, it doesn't matter. If you use 300 carries, you find 85 RB seasons from 1978-1996, and 104 RB seasons from 1997-2008.

What's more, those single-back carry totals occur despite the fact that per-game carries are down over the same time frame. Teams have averaged 446 carries per seasons from 1997-2008; from 1978-1996 they averaged 470 carries per season (even including the shortened strike year).

We are in the era of the Bell Cow RB, which started in 1997 and appears to be continuing.

 

[mggoilers]

Wow, was not expecting the thread to go this way...contradicts what we are led to believe when it comes to the RB situation.

The hype tells you that RBBC is in vogue, but this thread says otherwise.

Saying that, I don't trust a whole lot of the guys going into 2009. The list of RB's which have a track record greater than one year is pretty short, and even those guys have their issues.

 

[puckalicious]

Adrian Peterson and Michael Turner both had over 380 touches in 2008(usually the mark is 400+ touches). What does history tell us about RB's who get this much work?

In some cases, the results are realized pretty quick...LJ and Jamal Anderson.

Sometimes, it takes a few years for the problems to arise...Tomlinson.

In the age of the RBBC, the bell cow is a pretty rare thing these days. Just looking to get a fresh perspective. Peterson and Turner both have a history of getting nicked up, and both have very capable backups. I don't expect an immediate falloff or injury but curious about the history.

The history is:1) Exemplary populations regress to the mean.

2) People fallaciously believe that point 1) means that an individual will regress to the mean.

3) People cherry-pick data to prove what they already want to believe. (Check out why it's the 370-carry rule and not the 365-carry rule).

4) Since the merger, the last 10 years is the decade with the lowest amount of RBBC.

This article is the end of the argument over the Myth of 370./thread

 

[JamesTheScot]

I'm not disagreeing with you on #4, but lumping what was going on 10 years ago in the NFL with today is rather thin to me. What's the trend over the last 3 or 4 years and is it explanable (and predictable) by virtue of pedigree. In other words, are RBBC coaches taking over for bell-cow coaches the same way the WCO spread?

Since 2002 (7 seasons), there have been 15 RB seasons of 350+ carries. That's an average of just over two RBs per year. Seven of those are within the past four years, including two this year. From 1960-2001 (42 seasons), there were a grand total of 38 RB seasons with 350+ carries, an average of less than one per season.

By any statistical measure, there are more "bell cow" RBs now than ever before.

What's the significance of 350+ carries to RBBC?I'm more curious about the distribution of touches between all RB's on the roster? That seems more to me to reflect RBBC. Has the league average for the RB1 (in touches) for each team gone up, down or remained stable in terms of % of total offensive plays and in terms of total RB touches.

 

[CalBear]

Adrian Peterson and Michael Turner both had over 380 touches in 2008(usually the mark is 400+ touches). What does history tell us about RB's who get this much work?

In some cases, the results are realized pretty quick...LJ and Jamal Anderson.

Sometimes, it takes a few years for the problems to arise...Tomlinson.

In the age of the RBBC, the bell cow is a pretty rare thing these days. Just looking to get a fresh perspective. Peterson and Turner both have a history of getting nicked up, and both have very capable backups. I don't expect an immediate falloff or injury but curious about the history.

The history is:1) Exemplary populations regress to the mean.

2) People fallaciously believe that point 1) means that an individual will regress to the mean.

3) People cherry-pick data to prove what they already want to believe. (Check out why it's the 370-carry rule and not the 365-carry rule).

4) Since the merger, the last 10 years is the decade with the lowest amount of RBBC.

This article is the end of the argument over the Myth of 370./thread

Dang, I'd been meaning to write that article.The multiple endpoints bit is the fun bit. One of the most exemplary populations in year N+1 is RBs who had 368-369 carries.

1 E.James 1999 16 369 1553 4.21 13 62 586 9.45 4 315.902 C.Martin 1998 15 369 1287 3.49 8 43 365 8.49 1 219.203 T.Davis 1997 15 369 1750 4.74 15 42 287 6.83 0 293.704 C.Martin 1995 16 368 1487 4.04 14 30 261 8.70 1 264.805 E.Smith 1994 15 368 1484 4.03 21 50 341 6.82 1 314.506 E.Campbell 1979 16 368 1697 4.61 19 16 94 5.88 0 293.107 W.Payton 1979 16 369 1610 4.36 14 31 313 10.10 2 288.30Year N+1 for this group includes Edge's 2300-total-yard season, Terrell Davis' 2000-yard rushing season, Emmitt's 2148-total-yard season, and Earl Campbell's 1900-yard rushing season. That means that in a sample size of 7, all HOFers or possible HOFers, FOUR of them had the best year of their career in year N+1 after having 368 or 369 carries in year N. (Payton and Martin had decent N+1 years as well). What does it mean? Absolutely nothing. It's a funny coincidence, the kind which always show up when you have enough data and you slice it in a million different ways. It has absolutely no predictive value--if it did, you would take a RB at 1.01 just because he had 369 carries last year.This is why the FO study is so intellectually dishonest. They just kept discarding data points until their study said what they wanted it to say. They knew that these seasons existed just below where they drew the line--that's why they drew the line where they did.

 

[JamesTheScot]

I'm not disagreeing with you on #4, but lumping what was going on 10 years ago in the NFL with today is rather thin to me. What's the trend over the last 3 or 4 years and is it explanable (and predictable) by virtue of pedigree. In other words, are RBBC coaches taking over for bell-cow coaches the same way the WCO spread?

Since 2002 (7 seasons), there have been 15 RB seasons of 350+ carries. That's an average of just over two RBs per year. Seven of those are within the past four years, including two this year. From 1960-2001 (42 seasons), there were a grand total of 38 RB seasons with 350+ carries, an average of less than one per season.

By any statistical measure, there are more "bell cow" RBs now than ever before.

Not a good comparison, as the season switched from 14 games to 16 games in 1978. I don't believe that there were many 350 carry seasons when there were only 14 games. That would be averaging 25 carries per game for the entire season.

Fine. From 1978-2001 (24 seasons) there were 38 350-carry seasons, or just over 1.5 per year, and far less than from 2002-2008.Though really, I would put the dividing line at 1997; I used 2002 just because that's what's in the Data Dominator. 1997 saw Bettis, TD, and Eddie George all get 350+ rushes. Since then (12 seasons), there have been 28 RB seasons of 350+ carries (2.3/year). From 1978-1996 (19 seasons), there were just 26 RB seasons of 350+ carries (1.4/year).

And before anyone quibbles with the number of carries I chose, it doesn't matter. If you use 300 carries, you find 85 RB seasons from 1978-1996, and 104 RB seasons from 1997-2008.

What's more, those single-back carry totals occur despite the fact that per-game carries are down over the same time frame. Teams have averaged 446 carries per seasons from 1997-2008; from 1978-1996 they averaged 470 carries per season (even including the shortened strike year).

We are in the era of the Bell Cow RB, which started in 1997 and appears to be continuing.

Do we know how the number of players per year (that meet whatever milestone you select) relates to the number of teams in the NFL? Also, are we selectively using rushes and ignoring receptions here? With the advent and expansion of the WCO and it's hybrids, I'm wondering if we aren't getting an accurate idea of how much RB's are used?

 

[CalBear]

Do we know how the number of players per year (that meet whatever milestone you select) relates to the number of teams in the NFL?

A larger population will produce a larger number of exemplary individuals, so the number of teams in the league has some effect. But the difference between 28 and 32 shouldn't really have that much effect; we're looking at 50-75% increases in the number of backs above the threshholds.I'm wondering if it has something to do with roster limits. (With smaller rosters you'll have more reps per individual). When did the NFL go to the current 53/45 man roster limits?

Also, are we selectively using rushes and ignoring receptions here? With the advent and expansion of the WCO and it's hybrids, I'm wondering if we aren't getting an accurate idea of how much RB's are used?

Reasonable question, but it looks even more pronounced that way. Since 1997 there have been 87 RBs with 350 rushes+receptions; from 1978-1996 there were only 70.

 

[thatguy]

Adrian Peterson and Michael Turner both had over 380 touches in 2008(usually the mark is 400+ touches). What does history tell us about RB's who get this much work?

In some cases, the results are realized pretty quick...LJ and Jamal Anderson.

Sometimes, it takes a few years for the problems to arise...Tomlinson.

In the age of the RBBC, the bell cow is a pretty rare thing these days. Just looking to get a fresh perspective. Peterson and Turner both have a history of getting nicked up, and both have very capable backups. I don't expect an immediate falloff or injury but curious about the history.

The history is:1) Exemplary populations regress to the mean.

2) People fallaciously believe that point 1) means that an individual will regress to the mean.

3) People cherry-pick data to prove what they already want to believe. (Check out why it's the 370-carry rule and not the 365-carry rule).

4) Since the merger, the last 10 years is the decade with the lowest amount of RBBC.

This article is the end of the argument over the Myth of 370./thread

Dang, I'd been meaning to write that article.The multiple endpoints bit is the fun bit. One of the most exemplary populations in year N+1 is RBs who had 368-369 carries.

1 E.James 1999 16 369 1553 4.21 13 62 586 9.45 4 315.902 C.Martin 1998 15 369 1287 3.49 8 43 365 8.49 1 219.203 T.Davis 1997 15 369 1750 4.74 15 42 287 6.83 0 293.704 C.Martin 1995 16 368 1487 4.04 14 30 261 8.70 1 264.805 E.Smith 1994 15 368 1484 4.03 21 50 341 6.82 1 314.506 E.Campbell 1979 16 368 1697 4.61 19 16 94 5.88 0 293.107 W.Payton 1979 16 369 1610 4.36 14 31 313 10.10 2 288.30Year N+1 for this group includes Edge's 2300-total-yard season, Terrell Davis' 2000-yard rushing season, Emmitt's 2148-total-yard season, and Earl Campbell's 1900-yard rushing season. That means that in a sample size of 7, all HOFers or possible HOFers, FOUR of them had the best year of their career in year N+1 after having 368 or 369 carries in year N. (Payton and Martin had decent N+1 years as well). What does it mean? Absolutely nothing. It's a funny coincidence, the kind which always show up when you have enough data and you slice it in a million different ways. It has absolutely no predictive value--if it did, you would take a RB at 1.01 just because he had 369 carries last year.This is why the FO study is so intellectually dishonest. They just kept discarding data points until their study said what they wanted it to say. They knew that these seasons existed just below where they drew the line--that's why they drew the line where they did.

 

[Doug Drinen]

We are in the era of the Bell Cow RB, which started in 1997 and appears to be continuing.

I think this is a little misleading, as it pertains to fantasy football players in the year 2009.Yes, it is absolutely true that, in the broad sweep of history, we are in a low-RBBC high-BellCow era for RBs. We fret over how Leon Washington is stealing touches from Thomas Jones, which Walter Payton's fantasy owners would have thought absurd. Walter Payton, Jim Brown, Franco Harris, Tony Dorsett, none of those guys ever --- not in one single season --- got as high a percentage of their team's rushes as Thomas Jones got in 2008.It's also true that people are hypersensitive about RBBC and have been complaining about increasing RBBC for as long as this message board has been around, even when it was clearly not true.But the OP said:

In the age of the RBBC, the bell cow is a pretty rare thing these days.

Compared to 1979, that's a not at all true. But, since our FF instincts were more likely to have been developed in 200X and not 1979, 1979 is probably not a very useful starting point for this discussion. Take a look at the micro-trend:

+------+-------------+-------------+-------------+-------------+| year | 275+ rushes | 300+ rushes | 325+ rushes | 350+ rushes |+------+-------------+-------------+-------------+-------------+| 2000 | 16		  | 9		   | 5		   | 3		   || 2001 | 12		  | 10		  | 5		   | 1		   || 2002 | 13		  | 9		   | 5		   | 2		   || 2003 | 16		  | 13		  | 7		   | 4		   || 2004 | 11		  | 9		   | 7		   | 3		   || 2005 | 12		  | 10		  | 8		   | 4		   || 2006 | 12		  | 10		  | 7		   | 1		   || 2007 | 10		  | 6		   | 1		   | 0		   || 2008 | 8		   | 5		   | 3		   | 2		   |+------+-------------+-------------+-------------+-------------+

+------+--------------+--------------+--------------+-------------+--------------+| year | 300+ touches | 325+ touches | 350+ touches | 375+ rushes | 400+ touches |+------+--------------+--------------+--------------+-------------+--------------+| 2000 | 19		   | 16		   | 6			| 3		   | 2			|| 2001 | 13		   | 11		   | 9			| 4		   | 0			|| 2002 | 16		   | 13		   | 10		   | 4		   | 2			|| 2003 | 15		   | 15		   | 11		   | 7		   | 5			|| 2004 | 12		   | 9			| 9			| 6		   | 1			|| 2005 | 14		   | 11		   | 8			| 5		   | 2			|| 2006 | 14		   | 11		   | 8			| 5		   | 3			|| 2007 | 11		   | 8			| 3			| 1		   | 0			|| 2008 | 9			| 7			| 4			| 3		   | 0			|+------+--------------+--------------+--------------+-------------+--------------+

I don't claim these numbers are definitive. But they are suggestive. Just because the '08 numbers look more like the '04 numbers than the '79 numbers, doesn't necessarily mean that there isn't a real trend here that's could have an impact on FF strategy.

 

[rzrback77]

We are in the era of the Bell Cow RB, which started in 1997 and appears to be continuing.

I think this is a little misleading, as it pertains to fantasy football players in the year 2009.Yes, it is absolutely true that, in the broad sweep of history, we are in a low-RBBC high-BellCow era for RBs. We fret over how Leon Washington is stealing touches from Thomas Jones, which Walter Payton's fantasy owners would have thought absurd. Walter Payton, Jim Brown, Franco Harris, Tony Dorsett, none of those guys ever --- not in one single season --- got as high a percentage of their team's rushes as Thomas Jones got in 2008.It's also true that people are hypersensitive about RBBC and have been complaining about increasing RBBC for as long as this message board has been around, even when it was clearly not true.But the OP said:

In the age of the RBBC, the bell cow is a pretty rare thing these days.

Compared to 1979, that's a not at all true. But, since our FF instincts were more likely to have been developed in 200X and not 1979, 1979 is probably not a very useful starting point for this discussion. Take a look at the micro-trend:

+------+-------------+-------------+-------------+-------------+| year | 275+ rushes | 300+ rushes | 325+ rushes | 350+ rushes |+------+-------------+-------------+-------------+-------------+| 2000 | 16		  | 9		   | 5		   | 3		   || 2001 | 12		  | 10		  | 5		   | 1		   || 2002 | 13		  | 9		   | 5		   | 2		   || 2003 | 16		  | 13		  | 7		   | 4		   || 2004 | 11		  | 9		   | 7		   | 3		   || 2005 | 12		  | 10		  | 8		   | 4		   || 2006 | 12		  | 10		  | 7		   | 1		   || 2007 | 10		  | 6		   | 1		   | 0		   || 2008 | 8		   | 5		   | 3		   | 2		   |+------+-------------+-------------+-------------+-------------+

+------+--------------+--------------+--------------+-------------+--------------+| year | 300+ touches | 325+ touches | 350+ touches | 375+ rushes | 400+ touches |+------+--------------+--------------+--------------+-------------+--------------+| 2000 | 19		   | 16		   | 6			| 3		   | 2			|| 2001 | 13		   | 11		   | 9			| 4		   | 0			|| 2002 | 16		   | 13		   | 10		   | 4		   | 2			|| 2003 | 15		   | 15		   | 11		   | 7		   | 5			|| 2004 | 12		   | 9			| 9			| 6		   | 1			|| 2005 | 14		   | 11		   | 8			| 5		   | 2			|| 2006 | 14		   | 11		   | 8			| 5		   | 3			|| 2007 | 11		   | 8			| 3			| 1		   | 0			|| 2008 | 9			| 7			| 4			| 3		   | 0			|+------+--------------+--------------+--------------+-------------+--------------+

I don't claim these numbers are definitive. But they are suggestive. Just because the '08 numbers look more like the '04 numbers than the '79 numbers, doesn't necessarily mean that there isn't a real trend here that's could have an impact on FF strategy.

To me these last two tables say a lot, especially the touches. The column for >325 touches (just over 20 per game for the entire season) indicated that there were only eight RBs in 07 and eight in 08 that met that level. Thaks for the solid summary data DD.

 

[Matt Waldman]

Kind of drifting back to the first point of this post with this comment:

I had a colleague do a bit of research on touches and he developed a way of viewing backs according to what we called "f/carries"

1 rushing attempt = 1 f/carry

1 reception = .5 f/carry

If the RB had over 370 f carries, we found a high correlation to that player's drop off in subsequent seasons. We're not just talking about the high probability that the top fantasy back in year x is not likely to repeat production in year y. We're talking about a significant drop off the continues in the years after.

In fact, only a handful of backs in a 16-game season actually came close to repeating a baseline level of production after that 370 f/carry season. I believe their names were LT, Edgerrin James, and Emmit Smith (I might be missing someone else, but I'm sure of these three).

I'm sure it's not as statistically sound as some perspectives, but I've come to rely upon it as a decent guide marker when considering RBs.

Although I haven't looked at the final f/carries for 2008 backs, I know Turner, Forte, Peterson, and Portis were close and I think Turner might have gone over. He's a player I'm probably going to be wary of taking too early. He was a bargain last year - I got him as my 2nd RB in 3-4 drafts at the top of the third round - but as much as last year's stats make him a good candidate as a top 5 back - I'm not sure I'd want to pursue him that high again.

 

[rzrback77]

In response to previous post, only two over 370 f carries for 08:

1) Turner 377 carries and 6 receptions for 380 f carries

2) Peterson 364 carries and 21 receptions for 374.5 f carries

all others < 370 f carries and showing only those > 300 f carries

3) Portis 342 carries and 28 receptions for 356 f carries

4) Forte 315 carries and 64 receptions for 347 f carries

5) Grant 312 carries and 18 receptions for 321 f carries

6) Tomlinson 292 carries and 52 receptions for 318 f carries

7) T Jones 290 carries and 36 receptions for 308 f carries

 

[puckalicious]

Kind of drifting back to the first point of this post with this comment: I had a colleague do a bit of research on touches and he developed a way of viewing backs according to what we called "f/carries"1 rushing attempt = 1 f/carry1 reception = .5 f/carryIf the RB had over 370 f carries, we found a high correlation to that player's drop off in subsequent seasons. We're not just talking about the high probability that the top fantasy back in year x is not likely to repeat production in year y. We're talking about a significant drop off the continues in the years after. In fact, only a handful of backs in a 16-game season actually came close to repeating a baseline level of production after that 370 f/carry season. I believe their names were LT, Edgerrin James, and Emmit Smith (I might be missing someone else, but I'm sure of these three). I'm sure it's not as statistically sound as some perspectives, but I've come to rely upon it as a decent guide marker when considering RBs. Although I haven't looked at the final f/carries for 2008 backs, I know Turner, Forte, Peterson, and Portis were close and I think Turner might have gone over. He's a player I'm probably going to be wary of taking too early. He was a bargain last year - I got him as my 2nd RB in 3-4 drafts at the top of the third round - but as much as last year's stats make him a good candidate as a top 5 back - I'm not sure I'd want to pursue him that high again.

Maybe you didn't read the article from advancednflstats I linked above? It seems to state what you claim isn't supported by the stats.

 

[puckalicious]

In response to previous post, only two over 370 f carries for 08:1) Turner 377 carries and 6 receptions for 380 f carries2) Peterson 364 carries and 21 receptions for 374.5 f carriesall others < 370 f carries and showing only those > 300 f carries3) Portis 342 carries and 28 receptions for 356 f carries4) Forte 315 carries and 64 receptions for 347 f carries5) Grant 312 carries and 18 receptions for 321 f carries6) Tomlinson 292 carries and 52 receptions for 318 f carries7) T Jones 290 carries and 36 receptions for 308 f carries

I think this puts into perspective how few RB's ever eclipse this mark, and it is natural for a RB to get less carries the next season but it has nothing to do with the fact they had a high carry amount.The same can be said for any position that has a career year, don't expect them to repeat! That being said, I would expect Turner to have less carries but he seems a lock to go 300-350, while I think Peterson could easily repeat his carries but I wouldn't expect it to happen.

 

[Bri]

I'm not disagreeing with you on #4, but lumping what was going on 10 years ago in the NFL with today is rather thin to me. What's the trend over the last 3 or 4 years and is it explanable (and predictable) by virtue of pedigree. In other words, are RBBC coaches taking over for bell-cow coaches the same way the WCO spread?

Since 2002 (7 seasons), there have been 15 RB seasons of 350+ carries. That's an average of just over two RBs per year. Seven of those are within the past four years, including two this year. From 1960-2001 (42 seasons), there were a grand total of 38 RB seasons with 350+ carries, an average of less than one per season.By any statistical measure, there are more "bell cow" RBs now than ever before.

There are more useable backups in FF now than before. There's a predictability to it that we didn't have. The NFL may be changing toward more RBBC and people jumped the gun saying it was more prevalent before it was. I think it's simply that the quality of RBs is especially high now. Just consider how many people like Bradshaw on the Giants and how their 4th stringer is GB's starting RB. Ravens were 3 deep at RB this year. Cowboys had 3 RBs do significantly well, so did the Bucs, Saints, and Raiders. There's plenty of teams where people feel that both RBs would make good starters, not just one guy. I think in the past we didn't get to discussing the 3rd guy unless he was a 3rd down back and even still some teams with good 3rd down backs didn't necessarily have a very good primary backup. Keeping RBs fresh and healthy has been a challenge for coaches probably since the game began. It's something they can accomplish fairly easily now yet for much of the history of the NFL it seems it was more challenging. As I said last year, I think we'll see more NFL teams(as much as the college talent can feed it) use a big back to beat up/wear down a D and then another back to help clean up. Baltimore and NYG already do it with Jacobs and McClain. Raiders can do it with Michael Bush. Ward is a big back himself and could be in a Jacobs role for another team in 09.Even a guy like Lendale White that isn't very much liked around here, he was invaluable to setting up Chris Johnson and he isn't that big when compared to Jacobs. Not dissimilar to the teams above though, it seems the Ds were not prepared for CJ's speed. Rarely but when Leon did get time, you'd see his speed and moves prove to be too much after Thomas Jones wore down a D. Jerrious Norwood's speed was an awesome compliment to Turner's bruising style.By not prepared I mean worn down or tired-NFL teams will have to get better depth on DL and at LB to stop these teams and the ones that did have the depth were successful.IMO more often than not, it's the backup that is the straw that broke the camel's back not the starter. That brings attention to it and while we may not be at more of a RBBC now, it's that backup being useful in FF that is new.

 

[CalBear]

IMO more often than not, it's the backup that is the straw that broke the camel's back not the starter. That brings attention to it and while we may not be at more of a RBBC now, it's that backup being useful in FF that is new.

Is that real? Let's look at it. I don't have a systematic way to look at this easily, so I'll do some spot checks. In 2008, the #33 RB (first backup) in fantasy scoring is Warrick Dunn, who managed 124 fantasy points, 160 less than #1 Deangelo Williams. The #49 RB (median backup) is Julius Jones, who had 88 (-194), and the #64 RB (last backup) was Cadillac Williams, who had 52 (-232).In 1980, the #29 RB was Jim Jodat, who had 118 (-170), the #43 was Bruce Harper. who had 94 (-194), and the #56 was Lenvil Elliott, who had 81 (-207).In 1990, the #29 RB was Tom Rathman, who had 107 (-167), the #43 was Steve Broussard, who had 85 (-189), and the #56 was Dave Meggett, who had 63 (-210).So a cursory glance at the numbers suggests that backups were at least as useful, and possibly more useful from a fantasy perspective in 1980 and 1990 as they were this year. That's the result I would expect from looking at stats where the top RBs are getting more carries while the total carries are going down.

 

[Matt Waldman]

Kind of drifting back to the first point of this post with this comment: I had a colleague do a bit of research on touches and he developed a way of viewing backs according to what we called "f/carries"1 rushing attempt = 1 f/carry1 reception = .5 f/carryIf the RB had over 370 f carries, we found a high correlation to that player's drop off in subsequent seasons. We're not just talking about the high probability that the top fantasy back in year x is not likely to repeat production in year y. We're talking about a significant drop off the continues in the years after. In fact, only a handful of backs in a 16-game season actually came close to repeating a baseline level of production after that 370 f/carry season. I believe their names were LT, Edgerrin James, and Emmit Smith (I might be missing someone else, but I'm sure of these three). I'm sure it's not as statistically sound as some perspectives, but I've come to rely upon it as a decent guide marker when considering RBs. Although I haven't looked at the final f/carries for 2008 backs, I know Turner, Forte, Peterson, and Portis were close and I think Turner might have gone over. He's a player I'm probably going to be wary of taking too early. He was a bargain last year - I got him as my 2nd RB in 3-4 drafts at the top of the third round - but as much as last year's stats make him a good candidate as a top 5 back - I'm not sure I'd want to pursue him that high again.

Maybe you didn't read the article from advancednflstats I linked above? It seems to state what you claim isn't supported by the stats.

I didn't see the link, but I did see the article this summer. He makes a good point not to hypothesize over use as the reason. His knowledge of stats is great and he points out really substantive points. Still, he doesn't bust the myth that a high workload one year usually results in a decline in production the next. This is something he even stated in the piece: "Note that there is in fact a very strong tendency for high YPC RBs to decline the following year, regardless of whether a RB exceeded 370 carries.Very-high-carry RBs tend to have very high YPC stats, and they naturally suffer bigger declines the following season. 370+ carry RBs decline so much the following year simply because they peaked so high. This phenomenon is purely expected and not caused by overuse."As a fantasy owner, I'm not as concerned why this happens as much as I'm concerned that it does happen. So when I see a back with a combination of 370 touches based on his combo of attempts and receptions, the numbers appear to show that these players decline and decline significantly without returning to the form that people expect them to meet. It would be nice to know why, but it's exactly why I didn't come near Tom Brady in 2008 - his value was way too high based on a once in a career season.Maybe this is an ignorant approach and I apologize if it comes across this way but sometimes doing everything purely based on the numbers can cause a person to lose the overall point. It's like nutrition. We have all these scientific studies showing links to cancer, diabetes, or heat disease (or their prevention) and a single chemical that's found in foods. Year after year we hear "x causes cancer" or "z is no longer good for you", when 5 years before it was a viewed as healthful. Next think you know we're recommended to take supplements of one chemical rather than eat the food that contains dozens of them that work in harmony with your body to make your body healthy. I may not know why RBs have a higher rate of decline in subsequent seasons when they reach a certain threshold of work. If someone really learns it, I'm sure NFL coaches and GMs will want to see it so they extend the careers of superstars. But for right now, the info is good enough for me to consider until there is a definitive reason.

 

[JamesTheScot]

I didn't see the link, but I did see the article this summer. He makes a good point not to hypothesize over use as the reason. His knowledge of stats is great and he points out really substantive points. Still, he doesn't bust the myth that a high workload one year usually results in a decline in production the next. This is something he even stated in the piece:

"Note that there is in fact a very strong tendency for high YPC RBs to decline the following year, regardless of whether a RB exceeded 370 carries.

Very-high-carry RBs tend to have very high YPC stats, and they naturally suffer bigger declines the following season. 370+ carry RBs decline so much the following year simply because they peaked so high. This phenomenon is purely expected and not caused by overuse."

As a fantasy owner, I'm not as concerned why this happens as much as I'm concerned that it does happen. So when I see a back with a combination of 370 touches based on his combo of attempts and receptions, the numbers appear to show that these players decline and decline significantly without returning to the form that people expect them to meet.

It would be nice to know why, but it's exactly why I didn't come near Tom Brady in 2008 - his value was way too high based on a once in a career season.

Maybe this is an ignorant approach and I apologize if it comes across this way but sometimes doing everything purely based on the numbers can cause a person to lose the overall point.

It's like nutrition. We have all these scientific studies showing links to cancer, diabetes, or heat disease (or their prevention) and a single chemical that's found in foods. Year after year we hear "x causes cancer" or "z is no longer good for you", when 5 years before it was a viewed as healthful. Next think you know we're recommended to take supplements of one chemical rather than eat the food that contains dozens of them that work in harmony with your body to make your body healthy.

I may not know why RBs have a higher rate of decline in subsequent seasons when they reach a certain threshold of work. If someone really learns it, I'm sure NFL coaches and GMs will want to see it so they extend the careers of superstars. But for right now, the info is good enough for me to consider until there is a definitive reason.

I kinda feel like you read the article this summer but lost the point since then. The article alleges that the 370 mark was intentionally chosen to prove the hypothesis. But if you look at the sample size and choose your threshold as lower or HIGHER than 370, the hypothesis isn't supported.All you can say is that a RB who has a peak year isn't likely to repeat it and the higher the peak, the larger the regression. But that really isn't insightful for draft strategy in my mind.

How FAR the RB regresses is the question pertinent to draft strategy. It's still all about comparative value to the field of other RB's. Unfortunately, the "curse of 370" doesn't do **** to tell me how to value him comparatively to a RB who didn't hit 370. Why would you AVOID a RB who had a peak year? You shouldn't. Is it possible that someone thinks the RB will repeat his spectacular season and will over value him? Sure. But to say you think his value should otherwise be LOWERED simply because he's coming off a career year and is thus more likely to have something bad happen to him is simply not supported by the data.

Your Brady example is illustrative. So you steered away from Brady in 2008 because of his career year in 2007? You saw how he got hurt. Are you suggesting that there is some kind of "fate" that is keeping score and increases the liklihood that a defensive player will hit his knee just right in 2008 than in 2007? Arguments for bounties notwithstanding.

Was I expecting him to duplicate his 2007 numbers in 2008? No. I expected a regression.

Was I expecting him to finish top 5 among QB's absent injury? You bet. He was a top 5 QB before his 2007 season. Why would I expect him to regress beyond that?

Was I expecting him to finish the top scoring QB absent injury? I thought it was very possible bordering on probable. I had him ranked as the #1 QB on my board.

So where did you have Brady ranked?

 

[encaitar]

Kind of drifting back to the first point of this post with this comment: I had a colleague do a bit of research on touches and he developed a way of viewing backs according to what we called "f/carries"1 rushing attempt = 1 f/carry1 reception = .5 f/carryIf the RB had over 370 f carries, we found a high correlation to that player's drop off in subsequent seasons. We're not just talking about the high probability that the top fantasy back in year x is not likely to repeat production in year y. We're talking about a significant drop off the continues in the years after. In fact, only a handful of backs in a 16-game season actually came close to repeating a baseline level of production after that 370 f/carry season. I believe their names were LT, Edgerrin James, and Emmit Smith (I might be missing someone else, but I'm sure of these three). I'm sure it's not as statistically sound as some perspectives, but I've come to rely upon it as a decent guide marker when considering RBs. Although I haven't looked at the final f/carries for 2008 backs, I know Turner, Forte, Peterson, and Portis were close and I think Turner might have gone over. He's a player I'm probably going to be wary of taking too early. He was a bargain last year - I got him as my 2nd RB in 3-4 drafts at the top of the third round - but as much as last year's stats make him a good candidate as a top 5 back - I'm not sure I'd want to pursue him that high again.

Maybe you didn't read the article from advancednflstats I linked above? It seems to state what you claim isn't supported by the stats.

I didn't see the link, but I did see the article this summer. He makes a good point not to hypothesize over use as the reason. His knowledge of stats is great and he points out really substantive points. Still, he doesn't bust the myth that a high workload one year usually results in a decline in production the next. This is something he even stated in the piece: "Note that there is in fact a very strong tendency for high YPC RBs to decline the following year, regardless of whether a RB exceeded 370 carries.Very-high-carry RBs tend to have very high YPC stats, and they naturally suffer bigger declines the following season. 370+ carry RBs decline so much the following year simply because they peaked so high. This phenomenon is purely expected and not caused by overuse."As a fantasy owner, I'm not as concerned why this happens as much as I'm concerned that it does happen. So when I see a back with a combination of 370 touches based on his combo of attempts and receptions, the numbers appear to show that these players decline and decline significantly without returning to the form that people expect them to meet. It would be nice to know why, but it's exactly why I didn't come near Tom Brady in 2008 - his value was way too high based on a once in a career season.Maybe this is an ignorant approach and I apologize if it comes across this way but sometimes doing everything purely based on the numbers can cause a person to lose the overall point. It's like nutrition. We have all these scientific studies showing links to cancer, diabetes, or heat disease (or their prevention) and a single chemical that's found in foods. Year after year we hear "x causes cancer" or "z is no longer good for you", when 5 years before it was a viewed as healthful. Next think you know we're recommended to take supplements of one chemical rather than eat the food that contains dozens of them that work in harmony with your body to make your body healthy. I may not know why RBs have a higher rate of decline in subsequent seasons when they reach a certain threshold of work. If someone really learns it, I'm sure NFL coaches and GMs will want to see it so they extend the careers of superstars. But for right now, the info is good enough for me to consider until there is a definitive reason.

I understand what you're saying here in terms of a player's value. But the thing to consider is that nobody really knew what Turner's role would be on the Falcons team this year. Now, I think we can accurately say that he's gonna get the bulk of the carries. Will he get 370? Probably not, but I would bet he gets over 300 and that he's still a top 10, if not top 5 fantasy back next year. Same goes for ADP. He went in the top 3 in every draft I saw this past year, yet I doubt anyone expected 370+ touches from him. So, even if his carries do drop a bit next year, I don't think hisvalue does.People bring up the carry thing because they say avoid this running back at all cost. That is a huge mistake imo. If I have the third pick in my draft and the first two avoid ADP just because he has a high carry total, I'm laughing as I make my pick.

 

[JamesTheScot]

IMO more often than not, it's the backup that is the straw that broke the camel's back not the starter. That brings attention to it and while we may not be at more of a RBBC now, it's that backup being useful in FF that is new.

Is that real? Let's look at it. I don't have a systematic way to look at this easily, so I'll do some spot checks. In 2008, the #33 RB (first backup) in fantasy scoring is Warrick Dunn, who managed 124 fantasy points, 160 less than #1 Deangelo Williams. The #49 RB (median backup) is Julius Jones, who had 88 (-194), and the #64 RB (last backup) was Cadillac Williams, who had 52 (-232).In 1980, the #29 RB was Jim Jodat, who had 118 (-170), the #43 was Bruce Harper. who had 94 (-194), and the #56 was Lenvil Elliott, who had 81 (-207).In 1990, the #29 RB was Tom Rathman, who had 107 (-167), the #43 was Steve Broussard, who had 85 (-189), and the #56 was Dave Meggett, who had 63 (-210).So a cursory glance at the numbers suggests that backups were at least as useful, and possibly more useful from a fantasy perspective in 1980 and 1990 as they were this year. That's the result I would expect from looking at stats where the top RBs are getting more carries while the total carries are going down.

Not picking a fight but you keep going back to these numbers from the 80's and 90's. As someone posted, it does appear that there is a declining trend over the past two or three seasons in the number of backs getting a lion's share of the RB touches.Given how copycat this league is, I think you have to pay attention to the micro-trends. I don't have a draft strategy based upon the 80's or 90's game. This isn't the stock market where I can get ahead by patiently waiting for the rebound to come along in a couple of years. We have to be day traders in fantasy football.In other words, what's been happening since 2005 is more important and more impacting than what's happened since 1985 or even 1995.I'm not accusing you of cherry-picking your years or being intellectually dishonest, BTW! I want to be clear on that. I'm just saying that putting 2008 into perspective against the history of the NFL since the merger isn't insightful to my mind for drafting in 2009. Putting 2008 into perspective based upon what's happened in the last couple of years in the league however is!

 

[Matt Waldman]

I kinda feel like you read the article this summer but lost the point since then. The article alleges that the 370 mark was intentionally chosen to prove the hypothesis. But if you look at the sample size and choose your threshold as lower or HIGHER than 370, the hypothesis isn't supported.

All you can say is that a RB who has a peak year isn't likely to repeat it and the higher the peak, the larger the regression. But that really isn't insightful for draft strategy in my mind.

How FAR the RB regresses is the question pertinent to draft strategy. It's still all about comparative value to the field of other RB's. Unfortunately, the "curse of 370" doesn't do **** to tell me how to value him comparatively to a RB who didn't hit 370. Why would you AVOID a RB who had a peak year? You shouldn't. Is it possible that someone thinks the RB will repeat his spectacular season and will over value him? Sure. But to say you think his value should otherwise be LOWERED simply because he's coming off a career year and is thus more likely to have something bad happen to him is simply not supported by the data.

Your Brady example is illustrative. So you steered away from Brady in 2008 because of his career year in 2007? You saw how he got hurt. Are you suggesting that there is some kind of "fate" that is keeping score and increases the liklihood that a defensive player will hit his knee just right in 2008 than in 2007? Arguments for bounties notwithstanding.

Was I expecting him to duplicate his 2007 numbers in 2008? No. I expected a regression.

Was I expecting him to finish top 5 among QB's absent injury? You bet. He was a top 5 QB before his 2007 season. Why would I expect him to regress beyond that?

Was I expecting him to finish the top scoring QB absent injury? I thought it was very possible bordering on probable. I had him ranked as the #1 QB on my board.

So where did you have Brady ranked?

Thanks for your explanation - the "fate" quip was funny. While I had Brady ranked as my top QB based on projections, one of the things I took into consideration when discussing his draft status was the environment in which people regarded Brady's value was sky high. We all know of numerous leagues where Brady was taken as a top five or top ten pick in scoring systems where a typically good year for a top 3 QB wouldn't warrant near that value.

Although clever, I don't view this as some sort of karmic debt coming a player's way. My advice last year was he was overrated based on expectations people were setting in 2008 based on his 2007 number and you would have to dramatically over pay to get a player who wasn't going to come close to those number again - the odds were against it based on history.

While I appreciate that the higher the peak the higher the regression point wasn't that insightful to you, it's a point worth making because there are a lot of people who choose to ignore a simple point like this year after year.

That said, I'll have to look over this piece and the data so I have a clearer understanding of the author's point and compare it with the work I mentioned.

Thanks again.

 

[Couch Potato]

To CalBear, JamesTheScot and Doug Drinen... OK, I've read and understood the points made, and so to me the question then becomes whether the recent micro-trend is truly a trend to be relied upon going forward or an aberration.

My background in statistics encompasses mostly college courses nearly three decades ago, so which statistics do we rely on? The longer term as CalBear suggests because we will trend back in that direction, or the post-2005 period because it's a trend and not a glitch?

How much statistical evidence is needed to determine this? Do we have enough to know, or in the end is throwing darts just as good if we want to guess what to expect in 2009, 2010, etc.?

 

[JamesTheScot]

I kinda feel like you read the article this summer but lost the point since then. The article alleges that the 370 mark was intentionally chosen to prove the hypothesis. But if you look at the sample size and choose your threshold as lower or HIGHER than 370, the hypothesis isn't supported.

All you can say is that a RB who has a peak year isn't likely to repeat it and the higher the peak, the larger the regression. But that really isn't insightful for draft strategy in my mind.

How FAR the RB regresses is the question pertinent to draft strategy. It's still all about comparative value to the field of other RB's. Unfortunately, the "curse of 370" doesn't do **** to tell me how to value him comparatively to a RB who didn't hit 370. Why would you AVOID a RB who had a peak year? You shouldn't. Is it possible that someone thinks the RB will repeat his spectacular season and will over value him? Sure. But to say you think his value should otherwise be LOWERED simply because he's coming off a career year and is thus more likely to have something bad happen to him is simply not supported by the data.

Your Brady example is illustrative. So you steered away from Brady in 2008 because of his career year in 2007? You saw how he got hurt. Are you suggesting that there is some kind of "fate" that is keeping score and increases the liklihood that a defensive player will hit his knee just right in 2008 than in 2007? Arguments for bounties notwithstanding.

Was I expecting him to duplicate his 2007 numbers in 2008? No. I expected a regression.

Was I expecting him to finish top 5 among QB's absent injury? You bet. He was a top 5 QB before his 2007 season. Why would I expect him to regress beyond that?

Was I expecting him to finish the top scoring QB absent injury? I thought it was very possible bordering on probable. I had him ranked as the #1 QB on my board.

So where did you have Brady ranked?

Thanks for your explanation - the "fate" quip was funny. While I had Brady ranked as my top QB based on projections, one of the things I took into consideration when discussing his draft status was the environment in which people regarded Brady's value was sky high. We all know of numerous leagues where Brady was taken as a top five or top ten pick in scoring systems where a typically good year for a top 3 QB wouldn't warrant near that value.

Although clever, I don't view this as some sort of karmic debt coming a player's way. My advice last year was he was overrated based on expectations people were setting in 2008 based on his 2007 number and you would have to dramatically over pay to get a player who wasn't going to come close to those number again - the odds were against it based on history.

While I appreciate that the higher the peak the higher the regression point wasn't that insightful to you, it's a point worth making because there are a lot of people who choose to ignore a simple point like this year after year.

That said, I'll have to look over this piece and the data so I have a clearer understanding of the author's point and compare it with the work I mentioned.

Thanks again.

I looks as though you weren't avoiding Brady, then, you were just cognizant of the liklihood that others would expect another 2007 performance and thus would overpay relative to a value which takes into account a regression?If that's the case, I think you are spot on in your analysis.

What I wasn't sure about was if you devalue a player BELOW a value which anticipates a regression because he hits a high carry threshold - like predicting an increased chance of injury, so to speak.

But I don't think you are saying that as I originally suspected. I think we are actually in agreement, then, that we don't automatically expect a career year to be repeated.

 

[JamesTheScot]

To CalBear, JamesTheScot and Doug Drinen... OK, I've read and understood the points made, and so to me the question then becomes whether the recent micro-trend is truly a trend to be relied upon going forward or an aberration.My background in statistics encompasses mostly college courses nearly three decades ago, so which statistics do we rely on? The longer term as CalBear suggests because we will trend back in that direction, or the post-2005 period because it's a trend and not a glitch? How much statistical evidence is needed to determine this? Do we have enough to know, or in the end is throwing darts just as good if we want to guess what to expect in 2009, 2010, etc.?

This is where my lack of understanding and lack of background in statistics puts me in the art camp instead of the science camp. I think the individual mindset of the coaches involved is the predominant influence. To some extent I think that is shaped by their pedigree and their perception of their competitors' success or failure.As teams like the Giants, Titans, Ravens, Panthers, etc. do well, that puts pressure on the non-comforming clubs to conform. But assistant coaches on those succesful teams also aid the trend by getting hired to HC positions...thus the doctrine spreads. So if I see a team with a lackluster running game, I believe the RBBC/specialist approach looks appealing as a remedy to that team's running game woes and that team may shift in that direction for the 2009 season. Teams that run the ball well with a bellcow will continue to do so. Teams that run the ball well with a RBBC will continue to do so. But the hearts and minds of the teams that don't run the ball so well are up for grabs...as are often the HC jobs for those teams.In that sense, I think the statistics will prove the trend, but they lag behind the trend so to speak. By the time the trend appears clear and convincing statistically, the RBBC bias may have run it's course and the league may already be primed for another paradigm shift which the stats don't yet reflect and may only hint at.The danger is trying to catch up to a trend rather than anticpating and capitalizing on it.

 

[GroveDiesel]

Two of the best articles on this subject from the FootballOutsiders:

From 2004

From 2007

I think the point is not to predict an injury or something like that, but to simply say that your most likely looking at a 30-35% decline in that player's numbers the next season. If the carries and overall stats are high enough, then that player should still be among the top performers in the league. But the numbers most likely will take a serious hit compared to the year before.

If a guy has 1900 yards rushing, I'll still gladly take 65% of that. But I'm not sure that's worth the #1 overall pick that a 1900 yard rusher would probably garner.

 

[CalBear]

Two of the best articles on this subject from the FootballOutsiders:

From 2004

From 2007

I think the point is not to predict an injury or something like that, but to simply say that your most likely looking at a 30-35% decline in that player's numbers the next season. If the carries and overall stats are high enough, then that player should still be among the top performers in the league. But the numbers most likely will take a serious hit compared to the year before.

These articles have already been thoroughly debunked, in this thread.

If a guy has 1900 yards rushing, I'll still gladly take 65% of that. But I'm not sure that's worth the #1 overall pick that a 1900 yard rusher would probably garner.

This is a common fallacy (that FO also falls into). You suggest that the 1900-yard rusher in year N may not be worth the #1 overall pick in year N+1. So, would you rather have the guy who was a 1200-yard rusher in year N?Exemplary populations will always regress to the mean. That does nothing to predict what a specific individual will do.

 

[CalBear]

This is where my lack of understanding and lack of background in statistics puts me in the art camp instead of the science camp. I think the individual mindset of the coaches involved is the predominant influence. To some extent I think that is shaped by their pedigree and their perception of their competitors' success or failure.As teams like the Giants, Titans, Ravens, Panthers, etc. do well, that puts pressure on the non-comforming clubs to conform. But assistant coaches on those succesful teams also aid the trend by getting hired to HC positions...thus the doctrine spreads.

You might want to look up the term "confirmation bias." We all have a very strong cognitive force which makes us see what we expect to see. You have to use statistics to check whether your assumptions are correct; if you don't, you'll always believe that your assumptions are correct, because our brains are built to discard any data which disagree with our assumptions.(This is why 30% of the country still has a positive view of George W. Bush.)I think the sample sizes are way too small to consider two years of marginal decline in the numbers a real trend. Consider what happened in the league in 2008: LT, Jacobs, Lynch, LJ, Gore, Parker, Addai, Westbrook all got hurt. RBs will always get hurt, but that's a list of eight RBs who all could go for 300+ carries, who didn't because of their situations, not because of changes in coaching philosophy. Three of the top 12, 4 of the top 16 in carries were rookies; all of them will likely get more carries next year.

 

[puckalicious]

Wow. Not enough signs to go around, great discussion in this thread.

 

[Bri]

IMO more often than not, it's the backup that is the straw that broke the camel's back not the starter. That brings attention to it and while we may not be at more of a RBBC now, it's that backup being useful in FF that is new.

Is that real? Let's look at it. I don't have a systematic way to look at this easily, so I'll do some spot checks. In 2008, the #33 RB (first backup) in fantasy scoring is Warrick Dunn, who managed 124 fantasy points, 160 less than #1 Deangelo Williams. The #49 RB (median backup) is Julius Jones, who had 88 (-194), and the #64 RB (last backup) was Cadillac Williams, who had 52 (-232).In 1980, the #29 RB was Jim Jodat, who had 118 (-170), the #43 was Bruce Harper. who had 94 (-194), and the #56 was Lenvil Elliott, who had 81 (-207).In 1990, the #29 RB was Tom Rathman, who had 107 (-167), the #43 was Steve Broussard, who had 85 (-189), and the #56 was Dave Meggett, who had 63 (-210).So a cursory glance at the numbers suggests that backups were at least as useful, and possibly more useful from a fantasy perspective in 1980 and 1990 as they were this year. That's the result I would expect from looking at stats where the top RBs are getting more carries while the total carries are going down.

I think you missed my point in that we find the backup useful for FF now. I'd take a guess that all the 08 backs started and none of the others did. Even if it's wrong I still think I'm on the right track because it also "occured to me" that leagues were different then as FF was new,(to alot of us) they weren't so deep. I'd have to dig some but I would guess we only had 3 RBs per team when I started in a league, because of that the backup RBs also had less value.I'm not saying I'm dead on right here but I really think there's something to each of these causing the confusion about RBBC wherein so many are dead certain(and wrong) that it's more prevalent now. There's gotta be a good reason so many are wrong and well.....it's been a long day, hope this makes some sense

 

[CalBear]

I'm not saying I'm dead on right here but I really think there's something to each of these causing the confusion about RBBC wherein so many are dead certain(and wrong) that it's more prevalent now. There's gotta be a good reason so many are wrong and well.....it's been a long day, hope this makes some sense

People are wrong about lots of things, and because of confirmation bias, once they're wrong they tend to stay wrong. They believe things like "Vinateri is a clutch kicker" or "Elway was the best QB of all time" no matter how much evidence to the contrary is presented; they accept the data that agree with their beliefs, and reject the data that don't.My theory on how the RBBC myth got started is this: Most people start out with an unrealistic expectation of how many carries their RBs are going to get. You draft someone like Marshawn Lynch or Brandon Jacobs, you want 300 carries and 15 TDs. But the reality is that 300 carries and 15 TDs are very rare, even in the Bell Cow Age. Game situations, injuries, and fatigue all play a role in how many carries your guy winds up getting, and when you lose a game because Fred Jackson got two TDs instead of Lynch, you exclaim "freakin' RBBC!", when what was really wrong was your expectations going in.

 

[Bri]

I spent a bunch of time in a debate with SSOG on the Vinatieri thing 2-3 years ago. He was a clutch kicker then. I can't find it but we did pass about alot of stats and all.

 

[Tau837]

I spent a bunch of time in a debate with SSOG on the Vinatieri thing 2-3 years ago. He was a clutch kicker then. I can't find it but we did pass about alot of stats and all.

His "clutchness" is vastly overrated. I have posted on it before, but I don't want to take this thread on a tangent.

 

[AB in DC]

+------+--------------+--------------+--------------+-------------+--------------+| year | 300+ touches | 325+ touches | 350+ touches | 375+ rushes | 400+ touches |+------+--------------+--------------+--------------+-------------+--------------+| 2000 | 19		   | 16		   | 6			| 3		   | 2			|| 2001 | 13		   | 11		   | 9			| 4		   | 0			|| 2002 | 16		   | 13		   | 10		   | 4		   | 2			|| 2003 | 15		   | 15		   | 11		   | 7		   | 5			|| 2004 | 12		   | 9			| 9			| 6		   | 1			|| 2005 | 14		   | 11		   | 8			| 5		   | 2			|| 2006 | 14		   | 11		   | 8			| 5		   | 3			|| 2007 | 11		   | 8			| 3			| 1		   | 0			|| 2008 | 9			| 7			| 4			| 3		   | 0			|+------+--------------+--------------+--------------+-------------+--------------+

I don't claim these numbers are definitive. But they are suggestive.

The problem with this sort of analysis is that it picks up injury trends more than usage trends. In 2007, almost every top RB missed games due to injury -- that was unusual for that year, but ultimately meaningless for predicting workload.

 

[CalBear]

I spent a bunch of time in a debate with SSOG on the Vinatieri thing 2-3 years ago. He was a clutch kicker then. I can't find it but we did pass about alot of stats and all.

His "clutchness" is vastly overrated. I have posted on it before, but I don't want to take this thread on a tangent.

The relevant point is that clutch kicking as a personal characteristic probably doesn't exist. The fact that Vinatieri made two very difficult clutch kicks in 2001 created the idea that he is a clutch kicker, and not even the fact that he's missed more Super Bowl kicks than Scott Norwood shakes the idea that there's something unusually talented about him. Vinatieri is a very good kicker--but that means he should make 80% of his kicks, anyway. So when he misses two Super Bowl kicks and then makes the third, people who've already decided he's a clutch kicker will ignore the first two and accept the last one as confirmation of what they already want to believe.

 

[Doug Drinen]

The problem with this sort of analysis is that it picks up injury trends more than usage trends. In 2007, almost every top RB missed games due to injury -- that was unusual for that year, but ultimately meaningless for predicting workload.

That's a reasonable point. Here's another look that may or may not clarify anything.I looked at only players who led their team in rushes AND played 16 games. Here are the number of such people per season, along with their average number of rushes:

+------+--------+----------------+| year | number | average rushes |+------+--------+----------------+| 2000 |	 22 | 272.1		  || 2001 |	 18 | 276.1		  || 2002 |	 21 | 272.8		  || 2003 |	 19 | 282.5		  || 2004 |	  9 | 309.1		  || 2005 |	 11 | 302.4		  || 2006 |	 17 | 289.8		  || 2007 |	  9 | 270.3		  || 2008 |	 15 | 275.7		  |+------+--------+----------------+

Same thing with touches instead of rushes:

+------+--------+-----------------+| year | number | average touches |+------+--------+-----------------+| 2000 |	 23 | 309.9		   || 2001 |	 16 | 329.7		   || 2002 |	 21 | 318.2		   || 2003 |	 19 | 329.5		   || 2004 |	  9 | 343.6		   || 2005 |	 11 | 334.3		   || 2006 |	 17 | 326.8		   || 2007 |	  9 | 307.8		   || 2008 |	 15 | 309.4		   |+------+--------+-----------------+

If you want to increase the sample a smidge by allowing in players who played 15 games, you get this:

+------+--------+----------------+| year | number | average rushes |+------+--------+----------------+| 2000 |	 24 | 276.1		  || 2001 |	 20 | 272.2		  || 2002 |	 24 | 269.7		  || 2003 |	 22 | 272.0		  || 2004 |	 19 | 288.6		  || 2005 |	 21 | 276.7		  || 2006 |	 21 | 283.0		  || 2007 |	 18 | 265.4		  || 2008 |	 20 | 259.5		  |+------+--------+----------------+

+------+--------+-----------------+| year | number | average touches |+------+--------+-----------------+| 2000 |	 25 | 313.3		   || 2001 |	 18 | 323.6		   || 2002 |	 24 | 315.0		   || 2003 |	 23 | 311.7		   || 2004 |	 19 | 322.8		   || 2005 |	 21 | 308.0		   || 2006 |	 21 | 321.2		   || 2007 |	 19 | 297.8		   || 2008 |	 22 | 282.6		   |+------+--------+-----------------+

You can make of this what you will. But I can assure CalBear that I'm not suffering from confirmation bias. I've been railing against RBBC-chicken-little-ism for years. And again, I'm not saying any of this is definitive. I am, however, much more open to the possibility of a trend toward RBBC than I was two or three years ago. Just because people have been wrongly crying RBBC for a long time doesn't mean that people crying RBBC right now are necessarily wrong.

 

[CalBear]

You can make of this what you will. But I can assure CalBear that I'm not suffering from confirmation bias.

Present company excepted, of course.

 

[puckalicious]

The problem with this sort of analysis is that it picks up injury trends more than usage trends. In 2007, almost every top RB missed games due to injury -- that was unusual for that year, but ultimately meaningless for predicting workload.

That's a reasonable point. Here's another look that may or may not clarify anything.I looked at only players who led their team in rushes AND played 16 games. Here are the number of such people per season, along with their average number of rushes:...You can make of this what you will. But I can assure CalBear that I'm not suffering from confirmation bias. I've been railing against RBBC-chicken-little-ism for years. And again, I'm not saying any of this is definitive. I am, however, much more open to the possibility of a trend toward RBBC than I was two or three years ago. Just because people have been wrongly crying RBBC for a long time doesn't mean that people crying RBBC right now are necessarily wrong.

What about compensating for the # of rushes/touches available? What if the NFL as a whole is just using the RB position less frequently? Can you normalize on a % of team rushes/touches per year? I think this will be more clear in determining if RBBC is increasing or not. If it is, then I would expect the leading rusher's average rushes/touches in terms of the team average rushes/touches (for that year) to be in decline.