WR stats since 1980 - 1,000 yard seasons (1 Viewer)

[Joe T]

One of my themes for this season is going to be, "What has Player X done to earn his ranking?"

IE - I will be using performance based measures to calculate whether or not a player earns his specific value. Here is a sample.

I looked at the number of times a player had a 1,000 yard receiving year and then followed it up with a 1,000 yard receiving year. Then I compared with the number of times a receiver did not have a 1,000 yard season, but was able to have a 1,000 yard season the following year.

Below are the results:

YardsRange # in Range  # w/1,000 next season  %w/1000 next season>1,000 370 170 46%900-999 129 38 29%800-899 166 38 22%700-799 209 34 16%600-699 217 24 11%500-599 232 14 6%<500 1988 58 3%Above you can see that if you have a 1,000 yard season, on average you have a 46% chance of repeating. This drops off dramatically as you scale this down as you would expect.As we go through the pre-season, one thing I am preparing to do is make people prove their rankings based on historical data. This is just a sample.

I'd be interested in other ideas of how we should apply statistics to our debates over rankings.

 

[John Mamula]

Statistics are not used enough as they have a pesky way of getting in the way of hyping up players.

I'M LOOKING AT YOU LEE EVANS!

 

[rzrback77]

One of my themes for this season is going to be, "What has Player X done to earn his ranking?"

IE - I will be using performance based measures to calculate whether or not a player earns his specific value.  Here is a sample.

I looked at the number of times a player had a 1,000 yard receiving year and then followed it up with a 1,000 yard receiving year.  Then I compared with the number of times a receiver did not have a 1,000 yard season, but was able to have a 1,000 yard season the following year.

Below are the results:

YardsRange # in Range  # w/1,000 next season  %w/1000 next season>1,000 370 170 46%900-999 129 38 29%800-899 166 38 22%700-799 209 34 16%600-699 217 24 11%500-599 232 14 6%<500 1988 58 3%Above you can see that if you have a 1,000 yard season, on average you have a 46% chance of repeating.  This drops off dramatically as you scale this down as you would expect.As we go through the pre-season, one thing I am preparing to do is make people prove their rankings based on historical data.  This is just a sample.

I'd be interested in other ideas of how we should apply statistics to our debates over rankings.

What time period were you looking at? 370 seasons with receivers >1000 yards? Only 19 last year, is this in the past twenty years? Seems more reasonable to only study the previous five years or so.

 

[Tiger Fan]

I like the historical aspect of it and the 5 year aspect. Either way, it's should make for some good analysis

 

[dgreen]

I'd be interested in other ideas of how we should apply statistics to our debates over rankings.

I think a better study would look at how often a player improves on prior production and by how much. Really, we only use 1000 yards because it's a nice round number. Someone going from 650 yds to 900 yds tells me more than someone going from 950 yds to 1050 yds.

 

[Tiger Fan]

I'd be interested in other ideas of how we should apply statistics to our debates over rankings.

I think a better study would look at how often a player improves on prior production and by how much. Really, we only use 1000 yards because it's a nice round number. Someone going from 650 yds to 900 yds tells me more than someone going from 950 yds to 1050 yds.

Maybe use 30% production increase as a benchmark

 

[Tick]

I'd be interested in other ideas of how we should apply statistics to our debates over rankings.

I think a better study would look at how often a player improves on prior production and by how much. Really, we only use 1000 yards because it's a nice round number. Someone going from 650 yds to 900 yds tells me more than someone going from 950 yds to 1050 yds.

I'd rather see fantasy points than any yardage numbers.

 

[BassNBrew]

I'd be interested in other ideas of how we should apply statistics to our debates over rankings.

I think a better study would look at how often a player improves on prior production and by how much. Really, we only use 1000 yards because it's a nice round number. Someone going from 650 yds to 900 yds tells me more than someone going from 950 yds to 1050 yds.

The answer to this and every other statistical question is pretty simple. 50% chance of improving, 50% chance declining, both adjusted for the average lifespan of a WR in the league. Apply the bell curve for overall production to any chosen starting point to determine the chance of those numbers improving/declining by x amount.

 

[The Jerk]

The concept of wanting people to prove their rankings is a good one. David Dodds had a comprehensive description of how he proves his projections vs. historical data.

And the fact that less than half of WRs over the past 25 years repeat as 1,000 yard receivers is interesting.

However, combining two valid points in this case isn't really very useful in preparing for the 2006 season, because it's not really a question of how many but of which ones:

Which of the 19 receivers over 1,000 yards in 2005 will repeat and not repeat?

Which of the other receivers under 1,000 yards in 2005 will exceed 1,000 yards in 2006?

In other words, you have to name names.

You need to predict who will be lost for the season in Week 1 like Steve Smith in 2004 and Javon Walker in 2005.

You need to predict who might be disciplined off the team (Keyshawn, TO).

For the record, a quick run to the Data Dominator and manual calculations showed the following data:

2002 22 WRs over 1,000: 9 repeats with 2 others short by less than 30 yards in '03

2003 14 WRs over 1,000: 7 repeats with 1 short by 50 yards in '04

2004 23 WRs over 1,000: 9 repeats with 1 short by 25 yards in '05

Here's another flaw in any "Boolean" statistical analysis (i.e., YES/NO at a cutoff):

Receivers 15-17 in 2003 finished between 973 and 981 yards then went over 1,000 the next season. So add one or two yards per game to these receivers and the result is very different.

Maybe the only "rule" this proves is the more WRs over 1,000 in any one season, the lower percentage of repeats. Fairly intuitive, and not helpful because it doesn't name names.

 

[Pskov1420]

1000 yards is not a very good indicator over such a long time period, as it is a lot easier to get 1000 yards recieveing now than it was 20 years ago. A better measure would be one that measured relative performance to the rest of the league and whether that held up over consecutive seasons.

 

[Sabertooth]

Wouldn't it be more useful to go on a PPG basis? Or perhaps ypg?

Two axoms in Fantasy football are the unpredictability of injuries and the volatility of touchdowns (especially for wideouts).

So is you were to this down as a points per game increase if might shed more light. For instance:

Steve Smith wasn't a 1000 yard receiver in 2004, but he sure as heck was in 2005. Same with Santana Moss. Now, both players went from not qualifying to qualifying for this study but their situations could not have been more dissimilar.

PPG or YPG might be the way to go.

 

[Sabertooth]

1000 yards is not a very good indicator over such a long time period, as it is a lot easier to get 1000 yards recieveing now than it was 20 years ago. A better measure would be one that measured relative performance to the rest of the league and whether that held up over consecutive seasons.

This is true as well. Good point. However I'd submit the early to middle nineties as a period where people seemed to break the reception record every season: Sterling Sharpe, Herman Moore, Jerry Rice, Marvin Harrison - all were over 100 catch mark within a few years of each other. The NFL is certainly trendy.

 

[KnowledgeReignsSupreme]

Age/NFL starts/years of experience has to factor into this somehow. Kennison didn't have his first 1,000 yard season until 31!

Who's a safer pick, Anquan Boldin in '04 or Joe Horn '01?

Repeating in year n+1 isn't necessarily as useful as repeats over his entire career.

Being able to say "a first round draft pick, who has his first 1,000 yard season in his 2nd year of experience will on average have 4 more 1,000 yard seasons" seems more useful to me than "any WR who gets 1,000 yards has a 46% chance of repeating in year 2"

 

[KnowledgeReignsSupreme]

1000 yards is not a very good indicator over such a long time period, as it is a lot easier to get 1000 yards recieveing now than it was 20 years ago. A better measure would be one that measured relative performance to the rest of the league and whether that held up over consecutive seasons.

This is true as well. Good point. However I'd submit the early to middle nineties as a period where people seemed to break the reception record every season: Sterling Sharpe, Herman Moore, Jerry Rice, Marvin Harrison - all were over 100 catch mark within a few years of each other. The NFL is certainly trendy.

I like to start my studies in 1994, since that's when the NFL adopted a salary cap (FA started in '92).

 

[videoguy505]

I'd be interested in other ideas of how we should apply statistics to our debates over rankings.

I think a better study would look at how often a player improves on prior production and by how much. Really, we only use 1000 yards because it's a nice round number. Someone going from 650 yds to 900 yds tells me more than someone going from 950 yds to 1050 yds.

I'd rather see fantasy points than any yardage numbers.

You have to account for stuff like Hines Ward, who went from 80/1004/4 over 16 games in 2004 (132FPs), to 69/975/11 in 15 games in 2005 (164FPs). He dropped below the 1000 yard mark, but went from 62.75 yds/gm to 65 yds/gm, and from 8.25 FPs/gm to 10.9, a 33% increase in fantasy production. If he was on my fantasy team, I'd be quite happy with that performance, even though he didn't repeat his 1000 yd season.

 

[Joe T]

The concept of wanting people to prove their rankings is a good one. David Dodds had a comprehensive description of how he proves his projections vs. historical data.

And the fact that less than half of WRs over the past 25 years repeat as 1,000 yard receivers is interesting.

However, combining two valid points in this case isn't really very useful in preparing for the 2006 season, because it's not really a question of how many but of which ones:

Which of the 19 receivers over 1,000 yards in 2005 will repeat and not repeat?

Which of the other receivers under 1,000 yards in 2005 will exceed 1,000 yards in 2006?

In other words, you have to name names.

You need to predict who will be lost for the season in Week 1 like Steve Smith in 2004 and Javon Walker in 2005.

You need to predict who might be disciplined off the team (Keyshawn, TO).

For the record, a quick run to the Data Dominator and manual calculations showed the following data:

2002 22 WRs over 1,000: 9 repeats with 2 others short by less than 30 yards in '03

2003 14 WRs over 1,000: 7 repeats with 1 short by 50 yards in '04

2004 23 WRs over 1,000: 9 repeats with 1 short by 25 yards in '05

Here's another flaw in any "Boolean" statistical analysis (i.e., YES/NO at a cutoff):

Receivers 15-17 in 2003 finished between 973 and 981 yards then went over 1,000 the next season. So add one or two yards per game to these receivers and the result is very different.

Maybe the only "rule" this proves is the more WRs over 1,000 in any one season, the lower percentage of repeats. Fairly intuitive, and not helpful because it doesn't name names.

Agree, the statistics above are only scratching the surface. Of course, when this gets serious, you will use it to name names. Lee Evans is a great example. If I have him ranked 31, it will be based on historical data not gut feel and bias which a lot of rankings are based on.

 

[Joe T]

I'd be interested in other ideas of how we should apply statistics to our debates over rankings.

I think a better study would look at how often a player improves on prior production and by how much. Really, we only use 1000 yards because it's a nice round number. Someone going from 650 yds to 900 yds tells me more than someone going from 950 yds to 1050 yds.

I'd rather see fantasy points than any yardage numbers.

Couldn't agree more.I think going forward I will use Fantasy Points produced and PPG more heavily. The only thing I worry about is that there are different scoring systems for different leagues. Point per reception makes a huge difference. That's why sometimes, it makes sense to use yardage.

But I agree, Fantasy Points will be more useful as a whole.

 

[Bri]

One of my themes for this season is going to be, "What has Player X done to earn his ranking?"

IE - I will be using performance based measures to calculate whether or not a player earns his specific value.  Here is a sample.

I looked at the number of times a player had a 1,000 yard receiving year and then followed it up with a 1,000 yard receiving year.  Then I compared with the number of times a receiver did not have a 1,000 yard season, but was able to have a 1,000 yard season the following year.

Below are the results:

YardsRange # in Range  # w/1,000 next season  %w/1000 next season>1,000 370 170 46%900-999 129 38 29%800-899 166 38 22%700-799 209 34 16%600-699 217 24 11%500-599 232 14 6%<500 1988 58 3%Above you can see that if you have a 1,000 yard season, on average you have a 46% chance of repeating.  This drops off dramatically as you scale this down as you would expect.As we go through the pre-season, one thing I am preparing to do is make people prove their rankings based on historical data.  This is just a sample.

I'd be interested in other ideas of how we should apply statistics to our debates over rankings.

This 1000 line doesn't seem true. I'm not questioning your work Joe. Rice, Owens(when not suspended), Bruce, Holt....can you name some WRs that didn't repeat 1000? I'd have thought that was a higher number. No coffee yet and not many are coming to mind.

 

[dgreen]

This 1000 line doesn't seem true. I'm not questioning your work Joe. Rice, Owens(when not suspended), Bruce, Holt....can you name some WRs that didn't repeat 1000? I'd have thought that was a higher number. No coffee yet and not many are coming to mind.

Since he goes back to 1980, I assume he includes the two strike seasons. In 1981, 19 receivers went over 1000 yards. Only one in the strike-shortened 1982 season. 1986 had 16 1000-yard receivers and only four in the scab season of 1987.

 

[Bri]

'Nother thought Joe.

NFL O's didn't pass as much til some point in the 90s when it just took off. Andre Reed's stats are a good indicator of this. He was clearly one of the best and most productive WRs in his day but yet his stats don't appear so when compared to today's top WRs.(I believe Aaron pointed this out in another thread a few months back)

Is there anyway you can make a cutoff date (1993?) and do this exercise again from that point til now?

 

[The Jerk]

This 1000 line doesn't seem true. I'm not questioning your work Joe. Rice, Owens(when not suspended), Bruce, Holt....can you name some WRs that didn't repeat 1000? I'd have thought that was a higher number. No coffee yet and not many are coming to mind.

Bri,From my earlier post improved with players named:

2002 22 WRs over 1,000: 9 repeats with 2 others short by less than 30 yards in '03

Repeaters: Harrison, R.Moss, Toomer, Ward, Holt, Owens, Coles, C.Johnson, Mason

Missed: Burress, Horn, Moulds, Price, K.Robinson, Rice, Booker, K.Johnson, Bruce, Driver, J.Smith, R.Smith, Gardner

2003 14 WRs over 1,000: 7 repeats with 1 short by 50 yards in '04

Repeaters: Holt, C.Johnson, Mason, Harrison, Ward, D.Jackson, Owens

Missed: R.Moss, Boldin, Coles, McCardell, S.Moss, S.Smith, Toomer

2004 23 WRs over 1,000: 9 repeats with 1 short by 25 yards in '05

Repeaters: Holt, C.Johnson, Wayne, Driver, J.Smith, Mason, R.Smith, Harrison, Kennison

Missed: Muhammad, Horn, Walker, Bennett, Owens, D.Jackson, Mi.Clayton, A.Johnson, Lelie, Stokley, Moulds, Burleson, Ward

 

[Bri]

This 1000 line doesn't seem true. I'm not questioning your work Joe. Rice, Owens(when not suspended), Bruce, Holt....can you name some WRs that didn't repeat 1000? I'd have thought that was a higher number. No coffee yet and not many are coming to mind.

Bri,From my earlier post improved with players named:

2002 22 WRs over 1,000: 9 repeats with 2 others short by less than 30 yards in '03

Repeaters: Harrison, R.Moss, Toomer, Ward, Holt, Owens, Coles, C.Johnson, Mason

Missed: Burress, Horn, Moulds, Price, K.Robinson, Rice, Booker, K.Johnson, Bruce, Driver, J.Smith, R.Smith, Gardner

2003 14 WRs over 1,000: 7 repeats with 1 short by 50 yards in '04

Repeaters: Holt, C.Johnson, Mason, Harrison, Ward, D.Jackson, Owens

Missed: R.Moss, Boldin, Coles, McCardell, S.Moss, S.Smith, Toomer

2004 23 WRs over 1,000: 9 repeats with 1 short by 25 yards in '05

Repeaters: Holt, C.Johnson, Wayne, Driver, J.Smith, Mason, R.Smith, Harrison, Kennison

Missed: Muhammad, Horn, Walker, Bennett, Owens, D.Jackson, Mi.Clayton, A.Johnson, Lelie, Stokley, Moulds, Burleson, Ward

see your post seems to make the 46% seem like 90% with just 1,1, and 2 being short of a repeat

 

[The Jerk]

This 1000 line doesn't seem true. I'm not questioning your work Joe. Rice, Owens(when not suspended), Bruce, Holt....can you name some WRs that didn't repeat 1000? I'd have thought that was a higher number. No coffee yet and not many are coming to mind.

Bri,From my earlier post improved with players named:

2002 22 WRs over 1,000: 9 repeats with 2 others short by less than 30 yards in '03

Repeaters: Harrison, R.Moss, Toomer, Ward, Holt, Owens, Coles, C.Johnson, Mason

Missed: Burress, Horn, Moulds, Price, K.Robinson, Rice, Booker, K.Johnson, Bruce, Driver, J.Smith, R.Smith, Gardner

2003 14 WRs over 1,000: 7 repeats with 1 short by 50 yards in '04

Repeaters: Holt, C.Johnson, Mason, Harrison, Ward, D.Jackson, Owens

Missed: R.Moss, Boldin, Coles, McCardell, S.Moss, S.Smith, Toomer

2004 23 WRs over 1,000: 9 repeats with 1 short by 25 yards in '05

Repeaters: Holt, C.Johnson, Wayne, Driver, J.Smith, Mason, R.Smith, Harrison, Kennison

Missed: Muhammad, Horn, Walker, Bennett, Owens, D.Jackson, Mi.Clayton, A.Johnson, Lelie, Stokley, Moulds, Burleson, Ward

see your post seems to make the 46% seem like 90% with just 1,1, and 2 being short of a repeat

I'm not sure what you're seeing here, Bri. The repeats are nowhere near 90%.Even with adding the players who were just short, that's 11 out of 22 in 2002-03, 8 out of 14 in 2003-04, and 10 out of 23 in 2004-05.

Added together, that's 25 out of 59 at the 1,000 cutoff or 29 out of 59 if you add the players who just missed.

So the 46% is actually pretty accurate. I just don't find it too useful due to the unpredictability of the reasons receivers are missing their repeats... (see my earlier post for details).

 

[Bri]

This 1000 line doesn't seem true. I'm not questioning your work Joe. Rice, Owens(when not suspended), Bruce, Holt....can you name some WRs that didn't repeat 1000? I'd have thought that was a higher number. No coffee yet and not many are coming to mind.

Bri,From my earlier post improved with players named:

2002 22 WRs over 1,000: 9 repeats with 2 others short by less than 30 yards in '03

Repeaters: Harrison, R.Moss, Toomer, Ward, Holt, Owens, Coles, C.Johnson, Mason

Missed: Burress, Horn, Moulds, Price, K.Robinson, Rice, Booker, K.Johnson, Bruce, Driver, J.Smith, R.Smith, Gardner

2003 14 WRs over 1,000: 7 repeats with 1 short by 50 yards in '04

Repeaters: Holt, C.Johnson, Mason, Harrison, Ward, D.Jackson, Owens

Missed: R.Moss, Boldin, Coles, McCardell, S.Moss, S.Smith, Toomer

2004 23 WRs over 1,000: 9 repeats with 1 short by 25 yards in '05

Repeaters: Holt, C.Johnson, Wayne, Driver, J.Smith, Mason, R.Smith, Harrison, Kennison

Missed: Muhammad, Horn, Walker, Bennett, Owens, D.Jackson, Mi.Clayton, A.Johnson, Lelie, Stokley, Moulds, Burleson, Ward

see your post seems to make the 46% seem like 90% with just 1,1, and 2 being short of a repeat

I'm not sure what you're seeing here, Bri. The repeats are nowhere near 90%.Even with adding the players who were just short, that's 11 out of 22 in 2002-03, 8 out of 14 in 2003-04, and 10 out of 23 in 2004-05.

Added together, that's 25 out of 59 at the 1,000 cutoff or 29 out of 59 if you add the players who just missed.

So the 46% is actually pretty accurate. I just don't find it too useful due to the unpredictability of the reasons receivers are missing their repeats... (see my earlier post for details).

Ahh I missed something22 WRs, 9 repeats, 1 bowed out....that leaves 12 newbies I missed this 22 part reading too quickly

 

[redman]

It's interesting data, and it may be useful to use as a reality check for one's own projections.

Unlike others, I have a different take on the time frame. I'd go back about 10 years, which is when teams' contracts fully reflected the salary cap and FA. This is a significant consideration because during that time, unlike the previous years, it was very possible for a 1000 yard receiver to sign with another team the following year, and we've seen how often that results in a successful receiver being used differently and his production falling off.

 

[Joe T]

This 1000 line doesn't seem true. I'm not questioning your work Joe. Rice, Owens(when not suspended), Bruce, Holt....can you name some WRs that didn't repeat 1000? I'd have thought that was a higher number. No coffee yet and not many are coming to mind.

Here's the full list of WR's who didn't repeat 1,000 yard seasons since 1980.

Alexander	Derrick	1997Alexander	Derrick	2000Anderson	Flipper	1990Bailey	Stacey	1984Barnett	Fred	1992Barnett	Fred	1994Bennett	Drew	2004Blades	Brian	1989Blades	Brian	1991Blades	Brian	1995Boldin	Anquan	2003Booker	Marty	2002Boston	David	2001Brooks	Bill	1986Brooks	Robert	1995Brooks	Robert	1997Brown	Charlie	1983Brown	Eddie	1988Brown	Tim	2001Brown	Troy	2001Bruce	Isaac	1996Bruce	Isaac	2002Bruce	Isaac	2004Burleson	Nate	2004Burress	Plaxico	2002Carmichael	Harold	1981Carrier	Mark	1989Carrier	Mark	1995Carson	Carlos	1984Carson	Carlos	1987Carter	Anthony	1990Carter	Cris	2000Chandler	Wes	1982Chandler	Wes	1985Chrebet	Wayne	1998Clark	Dwight	1981Clark	Gary	1987Clark	Gary	1991Clayton	Mark	1984Clayton	Mark	1986Clayton	Mark	1989Clayton	Mark	1991Clayton	Michael	2004Coles	Laveranues	2003Collinsworth	Cris	1981Collinsworth	Cris	1983Collinsworth	Cris	1986Connell	Albert	1999Conway	Curtis	1996Conway	Curtis	2001Crowell	Germane	1999Driver	Donald	2002Duper	Mark	1984Duper	Mark	1986Duper	Mark	1991Early	Quinn	1995Ellard	Henry	1991Ellard	Henry	1996Emanuel	Bert	1995Fernandez	Mervyn	1989Freeman	Antonio	1999Fryar	Irving	1991Fryar	Irving	1994Fryar	Irving	1997Galloway	Joey	1995Galloway	Joey	1998Gardner	Rod	2002Givins	Ernest	1986Glenn	Terry	1996Glenn	Terry	1999Graham	Jeff	1995Gray	Earnest	1983Green	Roy	1984Green	Roy	1988Haynes	Michael	1991Hill	Bruce	1988Hill	Drew	1986Hill	Drew	1988Hill	Drew	1991Hill	Tony	1980Hill	Tony	1985Horn	Joe	2002Horn	Joe	2004House	Kevin	1981House	Kevin	1984Irvin	Michael	1995Irvin	Michael	1998Ismail	Qadry	1999Ismail	Qadry	2001Ismail	Raghib	1999Jackson	Darrell	2001Jackson	Darrell	2004Jackson	Michael	1996Jackson	Willie	2001Jeffires	Haywood	1991Jefferson	John	1980Jeffers	Patrick	1999Jenkins	Alfred	1981Johnson	Andre	2004Johnson	Charles	1996Johnson	Keyshawn	1999Johnson	Keyshawn	2002Johnson	Kevin	2001Johnson	Richard	1989Johnson	Vance	1989Joiner	Charlie	1981Largent	Steve	1981Largent	Steve	1986Lelie	Ashley	2004Lewis	Frank	1981Lipps	Louis	1985Lofton	James	1981Lofton	James	1985Lofton	James	1991Manuel	Lionel	1988Martin	Eric	1989Martin	Eric	1992Martin	Tony	1996Martin	Tony	1999Mathis	Terance	1995Mathis	Terance	1999McCaffrey	Ed	2000McCardell	Keenan	1997McCardell	Keenan	2001McCardell	Keenan	2003McDuffie	O.J.	1998McGee	Tim	1989Metcalf	Eric	1995Miller	Anthony	1989Miller	Anthony	1995Monk	Art	1986Monk	Art	1989Monk	Art	1991Moore	Herman	1997Moore	Rob	1994Moore	Rob	1997Morgan	Stanley	1981Morgan	Stanley	1986Morton	Johnnie	1999Morton	Johnnie	2001Moss	Randy	2003Moss	Santana	2003Moulds	Eric	1998Moulds	Eric	2000Moulds	Eric	2002Moulds	Eric	2004Muhammad	Muhsin	2000Muhammad	Muhsin	2004Owens	Terrell	1998Owens	Terrell	2004Paige	Stephone	1990Perriman	Brett	1996Pickens	Carl	1996Pickens	Carl	1998Price	Peerless	2002Quick	Mike	1985Rashad	Ahmad	1980Reed	Andre	1989Reed	Andre	1991Reed	Andre	1994Reed	Andre	1996Reed	Jake	1997Rice	Jerry	1996Rice	Jerry	1998Rice	Jerry	2002Rison	Andre	1990Rison	Andre	1994Rison	Andre	1997Robinson	Koren	2002Robinson	Marcus	1999Sanders	Frank	1998Sanders	Ricky	1989Schroeder	Bill	1999Scott	Darnay	1999Scott	Freddie	1981Sharpe	Sterling	1990Slaughter	Webster	1989Smith	Jimmy	2002Smith	JT	1987Smith	Rod	2002Smith	Steve	2003Smith	Tim	1984Stallworth	John	1981Stallworth	John	1984Stokley	Brandon	2004Taylor	John	1989Taylor	John	1991Thigpen	Yancey	1995Thigpen	Yancey	1997Tilley	Pat	1981Toomer	Amani	2003Toon	Al	1986Toon	Al	1988Walker	Javon	2004Walker	Wesley	1986Ward	Hines	2004Watson	Steve	1981Watson	Steve	1984Westbrook	Michael	1999White	Sammy	1981

 

[Bri]

Thanks Joe T, interesting list.

I notice some were at their career's end or at the point where they realize they're old and are faced with that reality. Becaue of that, many seem they took a predictable fall.

Any thoughts on this?

 

[redman]

Moulds Eric 1998

Moulds Eric 2000

Moulds Eric 2002

Moulds Eric 2004

 

[The Jerk]

Thanks Joe T, interesting list.

I notice some were at their career's end or at the point where they realize they're old and are faced with that reality. Becaue of that, many seem they took a predictable fall.

Any thoughts on this?

I felt the same way, Bri.In addition, some of the others were avoidable due to known holdouts, preseason injuries, etc.

 

[Joe T]

'Nother thought Joe.

NFL O's didn't pass as much til some point in the 90s when it just took off. Andre Reed's stats are a good indicator of this. He was clearly one of the best and most productive WRs in his day but yet his stats don't appear so when compared to today's top WRs.(I believe Aaron pointed this out in another thread a few months back)

Is there anyway you can make a cutoff date (1993?) and do this exercise again from that point til now?

It makes a small difference. The % of 1,000 seasons the following season is slightly higher.

>1000     232   122   52.5%900-999    64    20   31%800-899    90    23   25%700-799   102    18   18%600-699    98    11   11%500-599   119     8    7%<500     1025    37    4%

 

[Joe T]

Thanks Joe T, interesting list.

I notice some were at their career's end or at the point where they realize they're old and are faced with that reality. Becaue of that, many seem they took a predictable fall.

Any thoughts on this?

I think its an interesting thought. And its the type of theory we can test (although I'm not going to be able to do it this morning).But we could test at what age does their % chance to repeat 1,000 yard season drop.

 

[dgreen]

'Nother thought Joe.

NFL O's didn't pass as much til some point in the 90s when it just took off.

Just thought I'd look into this common claim. Here are comp, att, yds, td, and int from 1977-2003 (I need to download the latest seasons file from Drinen to get up to 2005):
Year Comp Att Yds TD INT1977 4993 9715 63120 378 5561978 6246 11750 78815 465 6311979 6994 12925 88500 534 5931980 7669 13632 95413 602 6251981 7720 14126 99203 582 6031982 4462 7894 55286 314 3381983 7958 13971 100251 614 6101984 8052 14256 101739 610 5791985 7895 14379 101267 592 5991986 7992 14417 100739 579 5741987 6847 12364 86856 567 4831988 7655 14095 97359 555 5511989 7926 14186 101347 577 5471990 7559 13511 94472 569 4771991 7823 13604 93839 499 4741992 7680 13344 91603 513 5141993 8340 14384 96208 513 4651994 8734 15042 101722 579 4711995 9701 16657 112740 660 5081996 9185 15937 106412 622 5401997 8830 15706 104990 611 4761998 8751 15453 105843 651 5071999 9555 16727 112927 660 5622000 9483 16286 109856 632 5262001 9536 16164 109569 634 5442002 10317 17291 116183 692 5252003 9690 16477 109377 653 5371978 had the big rule changes in what DBs could do to WRs, so that's why I went ahead and wet back to '77; to show that jump. 1982 and 1987 were strike seasons. 1995 appears to have a jump, but that's the year Jax and Carolina joined. The Browns came back in 1999 and the Texans in 2002. Here are attempts per team for 1977-2003:

Year  Att/tm1977   3471978   4201979   4621980   4871981   5051982   501*1983   4991984   5091985   5141986   5151987   471*1988   5031989   5071990   4831991   4861992   5771993   5141994   5371995   5551996   5311997   5241998   5151999   5402000   5252001   5212002   5402003   515

*Inflated to 16 gamesSo, we do some jumps, but it's not like the 80s was an era that ignored the pass. In fact, 84-86 fit in nicely with the 90s. The bigger difference is in completion percentage.

Year Comp %1977 51.41978 53.21979 54.11980 56.31981 54.71982 56.51983 57.01984 56.51985 54.91986 55.41987 55.41988 54.31989 55.91990 55.91991 57.51992 57.61993 58.01994 58.11995 58.21996 57.61997 56.21998 56.61999 57.12000 58.22001 59.02002 59.72003 58.8A couple times in the 80s we see some bumps, but, for the most part, there appears to be a permanent jump up to the high 50s starting in 1991. Oddly enough, that's about when the 100 catch season become more common.What does all this mean? I don't know, but I had fun looking into it. Teams are passing more, but, more importantly, teams are completing more. Since this is a thread on WRs, it may be a good idea to go back no further than 1991 since that's where completions take a jump up.

Oh, and it goes to show that Art Monk played most of his career in a different era and deserves to be in the HOF.

 

[Bri]

Some like Carrier, Schroeder, and Frank Sanders, IMO, were playing above their head when they got a K and were more like 7-800 yard WRs normally.

Toon and Sharpe's careers cut short due to injury should maybe get an asterisk.

Koren and Owens last year would probably have qualified if they behaved.

I'm surprised Andre Reed is on their 3 times. Rice's 2002 isn't so surprising but the other is. I can't recall, did he get hurt?

Jake Reed IIRC tore his knee and never was the same player again.

OJ McDuffie wasn't a year in year out 1000 yard WR in my mind. He was a guy with great hands that wasn't the best athlete that just got it done like Chrebet in a way.

Joe T thanks for posting these stats and the follow up stats

 

[Bri]

nice post dgreen

 

[Velcro]

I guess I'll take the devil's advocate position and say that it's a waste of time. I'm in finance, and by nature I like to rely on hard data as opposed to hunches or "feel". But in FF, more indepth analysis tends to make things more hazy, not less so. Maybe there is some kind of chaos theory or Monte Carlo simulation that a quant could run that yields some statistical use. But for mere mortals like most of us, it will all come back to guesses (however educated they may be).

If we were to follow some data stated above and assume that 50% of 1000 yard receivers will repeat, we need to analyze why the 50% that didn't repeat failed. In that list you'll have retirements, injuries and simple fall offs due to changes in the offense or other. After retirements, nothing else is really predictable. So you're still left trying to GUESS at who will fall off. Drew Bennett recorded over 1000 yards in 2004 and didn't in 2005. How would statistical analysis support where you ranked him going into the 05 season? And how does the data help you rank where Andre Johnson should go, since he fell below 1000 yards last season?

Just meant as discussion, not intended to insult your efforts.

 

[Joe T]

I guess I'll take the devil's advocate position and say that it's a waste of time. I'm in finance, and by nature I like to rely on hard data as opposed to hunches or "feel". But in FF, more indepth analysis tends to make things more hazy, not less so. Maybe there is some kind of chaos theory or Monte Carlo simulation that a quant could run that yields some statistical use. But for mere mortals like most of us, it will all come back to guesses (however educated they may be).

If we were to follow some data stated above and assume that 50% of 1000 yard receivers will repeat, we need to analyze why the 50% that didn't repeat failed. In that list you'll have retirements, injuries and simple fall offs due to changes in the offense or other. After retirements, nothing else is really predictable. So you're still left trying to GUESS at who will fall off. Drew Bennett recorded over 1000 yards in 2004 and didn't in 2005. How would statistical analysis support where you ranked him going into the 05 season? And how does the data help you rank where Andre Johnson should go, since he fell below 1000 yards last season?

Just meant as discussion, not intended to insult your efforts.

You're exactly right. This is only meant to scratch the surface of what we can do. Were not making any predictions off this data.

In the future, as we debate players though, we can use statistics more than gut feel to prove points. If we are debating Drew Bennett, we will be able to use him as an example to frame our analysis and cut out players that 'aren't like him.' It will cut down on the 4200+ player seasons above to a more useful analysis. This was really just a sample. Again... only scratching the surface.

 

[Velcro]

In the future, as we debate players though, we can use statistics more than gut feel to prove points. If we are debating Drew Bennett, we will be able to use him as an example to frame our analysis and cut out players that 'aren't like him.' It will cut down on the 4200+ player seasons above to a more useful analysis. This was really just a sample. Again... only scratching the surface.

But even if you get to the point of saying "80% of 1000 yd receivers who fail to reach that benchmark in the following year will bounce back in the 3rd year", you're still left with the question of whether Bennett will be the 1 in 5 that remains below 1000 yards. And that question won't be answered by statistics because there are too many potential variables that impacted the previous performers such as coaching changes, strength of schedule, grass vs. turf, indoor vs. outdoor, changes to the offensive line, changes to other offensive players, cornerbacks or defenses faced, etc, etc.Ultimately, the decision on where to rank him will be based on your guess as to how factors affect him now (like if his QB will be Volek or Young) as opposed to how many previous WR reacted similarly when faced with factors that may or may not have been similar.

 

[Joe T]

In the future, as we debate players though, we can use statistics more than gut feel to prove points.  If we are debating Drew Bennett, we will be able to use him as an example to frame our analysis and cut out players that 'aren't like him.'  It will cut down on the 4200+ player seasons above to a more useful analysis.  This was really just a sample.  Again... only scratching the surface.

But even if you get to the point of saying "80% of 1000 yd receivers who fail to reach that benchmark in the following year will bounce back in the 3rd year", you're still left with the question of whether Bennett will be the 1 in 5 that remains below 1000 yards. And that question won't be answered by statistics because there are too many potential variables that impacted the previous performers such as coaching changes, strength of schedule, grass vs. turf, indoor vs. outdoor, changes to the offensive line, changes to other offensive players, cornerbacks or defenses faced, etc, etc.Ultimately, the decision on where to rank him will be based on your guess as to how factors affect him now (like if his QB will be Volek or Young) as opposed to how many previous WR reacted similarly when faced with factors that may or may not have been similar.

I think once you find a player to debate, you will then use his criteria to find players with similar criteria to cut your sample size. Comparing apples to apples basically.You would never compare him vs. a mass sample like the one above.

 

[Z-Dog]

Age/NFL starts/years of experience has to factor into this somehow. Kennison didn't have his first 1,000 yard season until 31!

Who's a safer pick, Anquan Boldin in '04 or Joe Horn '01?

Repeating in year n+1 isn't necessarily as useful as repeats over his entire career.

Being able to say "a first round draft pick, who has his first 1,000 yard season in his 2nd year of experience will on average have 4 more 1,000 yard seasons" seems more useful to me than "any WR who gets 1,000 yards has a 46% chance of repeating in year 2"

To be precise, the fact that 46% of WRs have repeated does NOT mean that a WR has a 46% chance of repeating, in the same way that if I flipped a coin 500 times and got 300 heads to 200 tails, it wouldn't change the probability of getting a tail to 60%, it would just mean that when I flipped the coin 500 times, 60% of my results were tails.

 

[Velcro]

But I'm saying that there really is no such thing as apples to apples. Last year, Joe Horn was (presumably) in the same situation as the year before - same coach, same QB, same RB. Yet his season was a disaster compared to prior years. Similar story with Andre Johnson - where was his improvement after his freshman success?

Jevon Walker is changing teams. With his new team and new offense, will he perform like Santana Moss (better) or Derrick Mason (worse)? What previous apple will you compare him to? For each factor that you think is similar, there will be 10 that are different.

 

[Joe T]

Age/NFL starts/years of experience has to factor into this somehow.  Kennison didn't have his first 1,000 yard season until 31!

Who's a safer pick, Anquan Boldin in '04 or Joe Horn '01? 

Repeating in year n+1 isn't necessarily as useful as repeats over his entire career.

Being able to say "a first round draft pick, who has his first 1,000 yard season in his 2nd year of experience will on average have 4 more 1,000 yard seasons" seems more useful to me than "any WR who gets 1,000 yards has a 46% chance of repeating in year 2"

To be precise, the fact that 46% of WRs have repeated does NOT mean that a WR has a 46% chance of repeating, in the same way that if I flipped a coin 500 times and got 300 heads to 200 tails, it wouldn't change the probability of getting a tail to 60%, it would just mean that when I flipped the coin 500 times, 60% of my results were tails.

This is fairly inaccruate.Player performance <> coin flip.

 

[Doug Drinen]

To be precise, the fact that 46% of WRs have repeated does NOT mean that a WR has a 46% chance of repeating.

My favorite example: roughly 40% of all people will get pregnant at some point in their lives. Does that mean that you have a 40% chance of getting pregnant?

If only the lurking variables in fantasy football were so easy to spot.

 

[BassNBrew]

To be precise, the fact that 46% of WRs have repeated does NOT mean that a WR has a 46% chance of repeating.

My favorite example: roughly 40% of all people will get pregnant at some point in their lives. Does that mean that you have a 40% chance of getting pregnant?

If only the lurking variables in fantasy football were so easy to spot.

That's why we started letting chicks in our league.

 

[Tau837]

But I'm saying that there really is no such thing as apples to apples. Last year, Joe Horn was (presumably) in the same situation as the year before - same coach, same QB, same RB. Yet his season was a disaster compared to prior years. Similar story with Andre Johnson - where was his improvement after his freshman success?

Jevon Walker is changing teams. With his new team and new offense, will he perform like Santana Moss (better) or Derrick Mason (worse)? What previous apple will you compare him to? For each factor that you think is similar, there will be 10 that are different.

, this and your others in this thread.The bottom line is that each situation must be assessed individually. I am in the camp that says injuries cannot be predicted, either to the player under evaluation or on other players that affect his performance (e.g., the QB). Then you have to weigh the effects of changes to team personnel (offense and defense) and coaching staff, age, schedule, and character. The statistics do not help with that.

 

[Z-Dog]

Age/NFL starts/years of experience has to factor into this somehow. Kennison didn't have his first 1,000 yard season until 31!

Who's a safer pick, Anquan Boldin in '04 or Joe Horn '01?

Repeating in year n+1 isn't necessarily as useful as repeats over his entire career.

Being able to say "a first round draft pick, who has his first 1,000 yard season in his 2nd year of experience will on average have 4 more 1,000 yard seasons" seems more useful to me than "any WR who gets 1,000 yards has a 46% chance of repeating in year 2"

To be precise, the fact that 46% of WRs have repeated does NOT mean that a WR has a 46% chance of repeating, in the same way that if I flipped a coin 500 times and got 300 heads to 200 tails, it wouldn't change the probability of getting a tail to 60%, it would just mean that when I flipped the coin 500 times, 60% of my results were tails.

This is fairly inaccruate.Player performance <> coin flip.

You're right. Coin flips are much easier to predict, and have no lurking variables or hidden causes. Player performance is so mcuh more unpredicatable, such that the fact that 46% of WRs who got 1,000 yards repeated the feat the following season is a relatively meaningless stat that should not be used for making predictions of any kind. This is particularly true b/c the 1,000 yard cutoff is arbitrary and meaningless. For all we know, if you extended the cutoff to 990 yards, you might find that 78% of WRs repeat. But hey, believe what you like. Some other smarter owner will be happy to take your money in December.

 

[Musesboy]

It's useful to have statistics such as 46% of 1000 yard receivers tend to follow it with another 1000 yard season. For fantasy purposes, the question is can we use what we know to ensure that we predict as many of the repeat performers as possible?

Looking at the 1000 yard receivers from last year (including one TE), how many are likely to miss that mark this year?

Antonio Bryant is one candidate as he changed teams and will be playing with an inexperienced QB as part of a poor offense.

Jimmy Smith has retired.

Gates has an inexperienced QB and may well miss 1000 yards this year.

Kennison has a more conservative coach and a new OC.

Rod Smith has to share with Javon Walker (although I still think Smith will get there).

Chambers has only reached 1000 once, and may have Harrington at QB for part of the year. We will know more as the season approaches, no doubt.

Terry Glenn will be opposite Owens and will likely see a lot less looks.

I think we know enough to predict which will repeat to a greater degree of accuracy than 46%.

I would also factor in the number of times each player has reached the 1000 yard mark. A guy like Harrison has a higher percentage chance of repeating than someone like Bryant who has only reached that mark once. I have added the number of 1000 yard seasons for each player in parentheses (not necessarily in consecutive seasons).

How many repeat with two, three, four or five consecutive 1000 yard seasons to their name?

1 Steve Smith CAR 1563 (2)

2 Santana Moss WAS 1483 (2)

3 Chad Johnson CIN 1432 (4)

4 Larry Fitzgerald ARI 1409 (1)

5 Anquan Boldin ARI 1402 (2)

6 Torry Holt STL 1331 (6)

7 Joey Galloway TB 1287 (4)

8 Donald Driver GB 1221 (3)

9 Plaxico Burress NYG 1214 (3)

10 Marvin Harrison IND 1146 (7)

11 Terry Glenn DAL 1136 (3)

12 Chris Chambers MIA 1118 (1)

13 Rod Smith DEN 1105 (8)

14 Eddie Kennison KC 1102 (2)

15 Antonio Gates SD 1101 (1)

16 Derrick Mason BAL 1073 (5)

17 Reggie Wayne IND 1055 (2)

18 Jimmy Smith JAC 1023 (9)

19 Antonio Bryant CLE 1009 (1)

20 Randy Moss OAK 1005 (7)

 

[Musesboy]

After checking the current FBGs projections, 15 of the 19 WRs that hit 1000 yards last year are expected to do it in 2006.

Galloway, Glenn and Bryant are all projected to fall below this year and Jimmy Smith has retired.

If correct, that would be a repeat rate of 78.9%.

History says that four or five of those other players will disappoint fantasy owners.

 

[Velcro]

I think we know enough to predict which will repeat to a greater degree of accuracy than 46%.

OK. You just went through and excluded players you don't think will repeat their performance. I don't disagree. But you didn't use any statistical data to support your reasons for downgrading them. You largely did it based on "feel", such as

Terry Glenn will be opposite Owens and will likely see a lot less looks

which just reiterates my point that there is no way to statistically validate current projections based on historical data.

 

[Musesboy]

I think we know enough to predict which will repeat to a greater degree of accuracy than 46%.

OK. You just went through and excluded players you don't think will repeat their performance. I don't disagree. But you didn't use any statistical data to support your reasons for downgrading them. You largely did it based on "feel", such as

Terry Glenn will be opposite Owens and will likely see a lot less looks

which just reiterates my point that there is no way to statistically validate current projections based on historical data.

I agree that I didn't back up any of my downgrades with hard statistics What I am really saying is that whatever formula you come up with, we as owners can look at players and automatically downgrade a few based on the things we know may have changed since last season. I think our instincts and feel are better than any mathematical formula. Our brains are making calculations based on information that a formula will not have. I guess it is possible to include every variable in a formula, but why not just use what we know?

I contend that we will do better than 46% based on our new information.

 

[KnowledgeReignsSupreme]

After checking the current FBGs projections, 15 of the 19 WRs that hit 1000 yards last year are expected to do it in 2006.

Galloway, Glenn and Bryant are all projected to fall below this year and Jimmy Smith has retired.

If correct, that would be a repeat rate of 78.9%.

History says that four or five of those other players will disappoint fantasy owners.

It would be useful to know how many failed for injury/retirement reasons. It may not be out of line to project a repeat rate of 79%.