Compare these RB statistics (1 Viewer)

[Chase Stuart]

Good discussion.

Once again, it's important to remember to ignore supporting casts here. The statistics don't know the supporting casts, so we can't use supporting casts to say one formula is better. Sure, Thomas Jones was in a much worse situation last year than Marion Barber, but Barber's numbers were way better (IMO) than Jones. Sure, if Jones was a Cowboy and Barber a Jet the numbers would have looked different, but that's irrelevant here. I also think numbers wise, ADP was at least equal to LT last year, even though clearly he's not ahead of him as a player. He ranks pretty low in your formula.

 

[DawnBTVS]

Here's a revised formula for ya, Chase... I also agree that there should be something in relation to yard per carry/catch for performance. I took out fumbles because it skewed the numbers way too much and IMO is too fluctuating.

(GP * 1) + (Rushing Yards/3.5)/(Rushing Attempts) + (Rushing TD * 14.5) + (Receiving Yards/8)/(Receptions) + (Receiving TD * 14.5)

Still struggling with how to "weigh" the YPC/YPR factor. This would essentially give you a player's "yards" over the benchmark while still leaving in the ability to judge the YPC by seeing how beneficial their YPC was compared to the benchmark of 3.5 (same for receiving). The higher over 1, the better.

Quickie Comparison

LaDainian Tomlinson: 279.33

Adrian Peterson: 205.87

Marion Barber: 192.17

Thomas Jones: 47.00

 

[Chase Stuart]

Here's a revised formula for ya, Chase... I also agree that there should be something in relation to yard per carry/catch for performance. I took out fumbles because it skewed the numbers way too much and IMO is too fluctuating.

(GP * 1) + (Rushing Yards/3.5)/(Rushing Attempts) + (Rushing TD * 14.5) + (Receiving Yards/8)/(Receptions) + (Receiving TD * 14.5)

Still struggling with how to "weigh" the YPC/YPR factor. This would essentially give you a player's "yards" over the benchmark while still leaving in the ability to judge the YPC by seeing how beneficial their YPC was compared to the benchmark of 3.5 (same for receiving). The higher over 1, the better.

Quickie Comparison

LaDainian Tomlinson: 279.33

Adrian Peterson: 205.87

Marion Barber: 192.17

Thomas Jones: 47.00

That formula is essentially games played plus total TDs * 14.5.I don't think it's exactly what yuo mean. Rushing yards/3.5 is going to be very similar to rushes; then you divide that number by rush attempts. That will give you, always, a number that's less than two, which is obviously essentially meaningless. For receiving yards, same problem.

Do you mean YPC divided by 3.5, times rush attempts?

 

[Chase Stuart]

Here's the top 25 with my system:

707	181	888	Adrian Peterson519	361	880	Brian Westbrook679	198	877	LaDainian Tomlinson409	184	593	Joseph Addai533	 12	545	Fred Taylor420	115	535	Jamal Lewis307	203	510	Maurice Jones-Drew428	 77	505	Marion Barber285	200	485	Ronnie Brown292	175	467	Frank Gore467	- 8	459	Ryan Grant318	133	451	Chester Taylor314	134	448	Jerious Norwood340	 96	436	Laurence Maroney267	149	416	Clinton Portis318	 76	394	Brandon Jacobs284	108	392	Kenny Watson295	 92	387	Marshawn Lynch307	 74	381	Earnest Graham300	 67	367	DeAngelo Williams256	 87	343	Steven Jackson289	 52	341	Selvin Young293	 46	339	Willie Parker315	 21	336	Willis McGahee223	112	335	Najeh Davenport223	105	328	Maurice Morris215	 81	296	Edgerrin James189	 84	273	Thomas Jones246	 25	271	Ron Dayne

Oops. That included fumble data. Here's the list without that -- top 30:

569	361	930	Brian Westbrook747	181	928	Adrian L. Peterson679	198	877	LaDainian Tomlinson500	115	615	Jamal Lewis583	 12	595	Fred Taylor409	184	593	Joseph Addai443	133	576	Chester Taylor357	203	560	Maurice Jones-Drew372	175	547	Frank Gore397	149	546	Clinton Portis463	 77	540	Marion Barber443	 76	519	Brandon Jacobs285	200	485	Ronnie Brown472	- 8	464	Ryan Grant314	134	448	Jerious Norwood345	 92	437	Marshawn Lynch340	 96	436	Laurence Maroney341	 87	428	Steven Jackson373	 46	419	Willie Parker395	 21	416	Willis McGahee299	108	407	Kenny Watson332	 74	406	Earnest Graham320	 81	401	Edgerrin James325	 67	392	DeAngelo Williams319	 52	371	Selvin Young248	105	353	Maurice Morris228	112	340	Najeh Davenport257	 56	313	Derrick Ward199	 84	283	Thomas Jones271	 12	283	LenDale White

Here's the biggest problem as I see it -- Chestor Taylor ranks too high;Jerrious Norwood ranks too high;Here's what I did. I added one yard for every carry over 200, and one yard for every reception over 30.

647	421   1068	Brian Westbrook794	228   1022	LaDainian Tomlinson785	181	966	Adrian L. Peterson598	115	713	Jamal Lewis522	166	688	Clinton Portis470	195	665	Joseph Addai432	198	630	Frank Gore606	 12	618	Fred Taylor443	133	576	Chester Taylor357	213	570	Maurice Jones-Drew467	 91	558	Marion Barber494	 46	540	Willie Parker444	 81	525	Edgerrin James489	 34	523	Willis McGahee445	 76	521	Brandon Jacobs425	 92	517	Marshawn Lynch285	209	494	Ronnie Brown378	 95	473	Steven Jackson472	- 8	464	Ryan Grant314	134	448	Jerious Norwood354	 93	447	Earnest Graham340	 96	436	Laurence Maroney299	130	429	Kenny Watson309	 84	393	Thomas Jones325	 67	392	DeAngelo Williams374	 12	386	LenDale White319	 57	376	Selvin Young248	105	353	Maurice Morris228	112	340	Najeh Davenport257	 56	313	Derrick Ward

Norwood drops from 15 to 20. Chestor Taylor drops from 7 to 9. Jamal Lewis jumps ahead of Taylor and Addai.

 

[DawnBTVS]

No on the formula but maybe add/detract for every hundredth that a RB is over/under a benchmark for YPC and YPR (Think Napoleon Kaufman vs. Eddie George)? Like say (100) * (YPC-Benchmark) and same with YPR.

Assume the benchmarks are 3.50 and 8.00.

RB1: 238 carries for 1341 yards (5.63) and 19 catches for 268 yards (14.11) - Total Yards = 1,609 on 257 touches (6.26)

RB2: 315 carries for 1474 yards (4.68) and 60 catches for 475 yards (7.92) - Total Yards = 1,949 on 375 touches (5.20)

RB3: 325 carries for 1262 yards (3.88) and 47 catches for 389 yards (8.28) - Total Yards = 1,651 on 372 touches (4.44)

RB1 would score a combined 213 points + 611 points (824)

RB2 would score a combined 118 points - 8 points (110)

RB3 would score a combined 38 points + 28 points (66)

I think most fantasy football players would agree that if RB1 were assured of getting the same touches as RB2/RB3, that they would snatch him up without a second thought. They would also agree that between the latter two RB, that RB2 is clearly the better player because he gained almost 300 more yards on only 3 more touches.

I just think there should be a way to quantify that within the above formula. This would also weigh a guy who's better at both running and receiving over a player who is more one dimensional (either great at running but a poor receiver or vice versa).

 

[Chase Stuart]

No on the formula but maybe add/detract for every hundredth that a RB is over/under a benchmark for YPC and YPR (Think Napoleon Kaufman vs. Eddie George)? Like say (100) * (YPC-Benchmark) and same with YPR.Assume the benchmarks are 3.50 and 8.00.RB1: 238 carries for 1341 yards (5.63) and 19 catches for 268 yards (14.11) - Total Yards = 1,609 on 257 touches (6.26)RB2: 315 carries for 1474 yards (4.68) and 60 catches for 475 yards (7.92) - Total Yards = 1,949 on 375 touches (5.20)RB3: 325 carries for 1262 yards (3.88) and 47 catches for 389 yards (8.28) - Total Yards = 1,651 on 372 touches (4.44)RB1 would score a combined 213 points + 611 points (824)RB2 would score a combined 118 points - 8 points (110)RB3 would score a combined 38 points + 28 points (66)I think most fantasy football players would agree that if RB1 were assured of getting the same touches as RB2/RB3, that they would snatch him up without a second thought. They would also agree that between the latter two RB, that RB2 is clearly the better player because he gained almost 300 more yards on only 3 more touches.I just think there should be a way to quantify that within the above formula. This would also weigh a guy who's better at both running and receiving over a player who is more one dimensional (either great at running but a poor receiver or vice versa).

Doesn't your formula totally ignore carries and catches, which is what you were advocating before? Or do you mean carries * YPC - benchmark, in which case:RB1 - 507 + 138 = 645RB2 - 372 + 0 = 372RB3 - 124 + 13 = 137My system would give them, for the rushing part:RB1 - 665 + RB2 - 644 + RB3 - 412 + That's not too bad. The carry bonuses there definitely help RB3.

 

[DawnBTVS]

No on the formula but maybe add/detract for every hundredth that a RB is over/under a benchmark for YPC and YPR (Think Napoleon Kaufman vs. Eddie George)? Like say (100) * (YPC-Benchmark) and same with YPR.Assume the benchmarks are 3.50 and 8.00.RB1: 238 carries for 1341 yards (5.63) and 19 catches for 268 yards (14.11) - Total Yards = 1,609 on 257 touches (6.26)RB2: 315 carries for 1474 yards (4.68) and 60 catches for 475 yards (7.92) - Total Yards = 1,949 on 375 touches (5.20)RB3: 325 carries for 1262 yards (3.88) and 47 catches for 389 yards (8.28) - Total Yards = 1,651 on 372 touches (4.44)RB1 would score a combined 213 points + 611 points (824)RB2 would score a combined 118 points - 8 points (110)RB3 would score a combined 38 points + 28 points (66)I think most fantasy football players would agree that if RB1 were assured of getting the same touches as RB2/RB3, that they would snatch him up without a second thought. They would also agree that between the latter two RB, that RB2 is clearly the better player because he gained almost 300 more yards on only 3 more touches.I just think there should be a way to quantify that within the above formula. This would also weigh a guy who's better at both running and receiving over a player who is more one dimensional (either great at running but a poor receiver or vice versa).

Doesn't your formula totally ignore carries and catches, which is what you were advocating before? Or do you mean carries * YPC - benchmark, in which case:RB1 - 507 + 138 = 645RB2 - 372 + 0 = 372RB3 - 124 + 13 = 137My system would give them, for the rushing part:RB1 - 665 + RB2 - 644 + RB3 - 412 + That's not too bad. The carry bonuses there definitely help RB3.

I didn't mean for it to be it's own separate formula but as another "sub" formula within the overall (Yours or Switz).Something like: (Rushing Yards + Rushing TDs *10) - Rushes *3 + (10)* (YPC-Benchmark) + ((Receiving Yards + 10 * Receiving TDs)/ Receptions) - 5.11) * Receptions + (10)* (YPR-Benchmark)Quick ComparisonAdrian Peterson: 1010.31Brian Westbrook: 948.78LaDainian Tomlinson: 888.96Frank Gore: 556.82Clinton Portis: 552.43Marion Barber: 537.05Thomas Jones: 281.52It really shows how much better Brian Westbrook was at getting yards compared to LT (4.79 vs. 4.68 and 8.57 vs. 7.92) and how MB compares far more with the Gores/Portis' than he does Thomas Jones (and that's with a poor -15.91 YPR-Benchmark score).

 

[Tau837]

Rushing value - I give 10 yards for each rushing TD. This is controversial, although some think that's too much, most think it's too little. A good explanation can be found here.

It is too little. There is a good discussion about it here: Ranking the QBs — Methodology Discussion. At the end of the posts, Doug finally shows some analysis that reveals that the value of a TD is roughly 14.5 yards, although that is based on some assumptions he describes, and thus could be more in a range of 12 to 15 yards.The post you linked ends with this: "I don’t know if ten is exactly the right answer, but it feels about right." We can do better than that, as shown in the post I linked above.

I assume changing TD worth from 10 yards to 14.5 yards or thereabouts would have a significant effect on your rankings.

Chase, I don't believe you have responded to this yet. Have you changed the value of TDs from 10 to 14.5 (or other) in your formula? If not, why not?

 

[Tau837]

How about receiving? The NFL average adjusted net yards per pass play this past season was 5.11. That means on every pass play, if you hit just 5.11 yards, that's average. I don't want to give receiving RBs too much credit, so while I used a really low baseline for RB ypc (3.0), I'll use average here. So RBs only get points for their yards above that number.

I don't understand your logic in choosing 3.0 ypc due to it being a "very easy barometer" but then going with the average of 5.11 adjusted net yards per pass play. I have little faith in this formula if you are applying different criteria in choosing your thresholds.Furthermore, if you're not going to go with the "easy barometer" route, why would you choose an NFL wide average of adjusted net yards per pass play? That includes sacks and pass plays to WRs and TEs, correct? Why not keep the focus on RBs, and determine what the average ypr is for RBs, then determine a threshold based on that?

The average YPR for RBs is something like 8. Therefore, about half of the RBs will have no receiving value. I think a RB with 30 catches for 180 yards brings more value to the table than a RB with 0 catches for 0 yards. Using the RB-wide YPR average would be too high.I think RB rushing and RB receiving do need different thresholds. People don't view them as the same (i.e., a RB with 1500 rushing yards and 100 receiving yards is generally viewed as a better RB -- and perhaps rightly so -- than one with 800/800).

I didn't say to use the average ypr, nor did I say to use the same thresholds for running and receiving. I suggested to (a) be consistent in how you determine the two thresholds and (b) determine the receiving threshold based on the average RB ypr rather than NFL average adjusted net yards per pass play.

Chase, I don't believe you responded to this. Are you sticking with NFL average adjusted net yards per pass play? If so, can you elaborate on why, relative to my two points above?

 

[Chase Stuart]

Rushing value - I give 10 yards for each rushing TD. This is controversial, although some think that's too much, most think it's too little. A good explanation can be found here.

It is too little. There is a good discussion about it here: Ranking the QBs — Methodology Discussion. At the end of the posts, Doug finally shows some analysis that reveals that the value of a TD is roughly 14.5 yards, although that is based on some assumptions he describes, and thus could be more in a range of 12 to 15 yards.The post you linked ends with this: "I don’t know if ten is exactly the right answer, but it feels about right." We can do better than that, as shown in the post I linked above.

I assume changing TD worth from 10 yards to 14.5 yards or thereabouts would have a significant effect on your rankings.

Chase, I don't believe you have responded to this yet. Have you changed the value of TDs from 10 to 14.5 (or other) in your formula? If not, why not?

I'm in the process of updating this now. I'm not sure 14.5 will be the final answer, but will let you know when I come up with it.

 

[Chase Stuart]

How about receiving? The NFL average adjusted net yards per pass play this past season was 5.11. That means on every pass play, if you hit just 5.11 yards, that's average. I don't want to give receiving RBs too much credit, so while I used a really low baseline for RB ypc (3.0), I'll use average here. So RBs only get points for their yards above that number.

I don't understand your logic in choosing 3.0 ypc due to it being a "very easy barometer" but then going with the average of 5.11 adjusted net yards per pass play. I have little faith in this formula if you are applying different criteria in choosing your thresholds.Furthermore, if you're not going to go with the "easy barometer" route, why would you choose an NFL wide average of adjusted net yards per pass play? That includes sacks and pass plays to WRs and TEs, correct? Why not keep the focus on RBs, and determine what the average ypr is for RBs, then determine a threshold based on that?

The average YPR for RBs is something like 8. Therefore, about half of the RBs will have no receiving value. I think a RB with 30 catches for 180 yards brings more value to the table than a RB with 0 catches for 0 yards. Using the RB-wide YPR average would be too high.I think RB rushing and RB receiving do need different thresholds. People don't view them as the same (i.e., a RB with 1500 rushing yards and 100 receiving yards is generally viewed as a better RB -- and perhaps rightly so -- than one with 800/800).

I didn't say to use the average ypr, nor did I say to use the same thresholds for running and receiving. I suggested to (a) be consistent in how you determine the two thresholds and (b) determine the receiving threshold based on the average RB ypr rather than NFL average adjusted net yards per pass play.

Chase, I don't believe you responded to this. Are you sticking with NFL average adjusted net yards per pass play? If so, can you elaborate on why, relative to my two points above?

I think I explained before why average RB ypr won't work. I'm going to stick with average adjusted net yards per pass play for now. I'm not sure if I can fully explain why, but it seems right to me. Every pass play is worth whatever the average ANY/A is. So if a RB has a reception where he gets more than that, he's going to get credit. Remember, for QB rushing, I used adjusted rushing yards above 4.0. That was totally and completely arbitrary, yet I think it worked as well or better than any other study I've seen. When a player gets credit for doing something another position generally does (QB rushing, RB receiving), I want to set the standard a bit higher. We think of a great QB as Marino or Montana, and a great RB as Jim Brown or Barry Sanders. It's not Randall Cunningham and Brian Westbrook that pops into our head when we hear the term RB. I like the results given to me from RB yards above ANY/A, but am willing to check out another formula. I currently think ANY/A is a high baseline, yet the results (before my latest tweak to reward ballcarriers) gave me too many receiving RBs on the all time list. That's a sign to me that I need to devalue, or at least not increase, the worth of RB receiving.

 

[Tau837]

How about receiving? The NFL average adjusted net yards per pass play this past season was 5.11. That means on every pass play, if you hit just 5.11 yards, that's average. I don't want to give receiving RBs too much credit, so while I used a really low baseline for RB ypc (3.0), I'll use average here. So RBs only get points for their yards above that number.

I don't understand your logic in choosing 3.0 ypc due to it being a "very easy barometer" but then going with the average of 5.11 adjusted net yards per pass play. I have little faith in this formula if you are applying different criteria in choosing your thresholds.Furthermore, if you're not going to go with the "easy barometer" route, why would you choose an NFL wide average of adjusted net yards per pass play? That includes sacks and pass plays to WRs and TEs, correct? Why not keep the focus on RBs, and determine what the average ypr is for RBs, then determine a threshold based on that?

The average YPR for RBs is something like 8. Therefore, about half of the RBs will have no receiving value. I think a RB with 30 catches for 180 yards brings more value to the table than a RB with 0 catches for 0 yards. Using the RB-wide YPR average would be too high.I think RB rushing and RB receiving do need different thresholds. People don't view them as the same (i.e., a RB with 1500 rushing yards and 100 receiving yards is generally viewed as a better RB -- and perhaps rightly so -- than one with 800/800).

I didn't say to use the average ypr, nor did I say to use the same thresholds for running and receiving. I suggested to (a) be consistent in how you determine the two thresholds and (b) determine the receiving threshold based on the average RB ypr rather than NFL average adjusted net yards per pass play.

Chase, I don't believe you responded to this. Are you sticking with NFL average adjusted net yards per pass play? If so, can you elaborate on why, relative to my two points above?

I think I explained before why average RB ypr won't work. I'm going to stick with average adjusted net yards per pass play for now. I'm not sure if I can fully explain why, but it seems right to me. Every pass play is worth whatever the average ANY/A is. So if a RB has a reception where he gets more than that, he's going to get credit. Remember, for QB rushing, I used adjusted rushing yards above 4.0. That was totally and completely arbitrary, yet I think it worked as well or better than any other study I've seen. When a player gets credit for doing something another position generally does (QB rushing, RB receiving), I want to set the standard a bit higher. We think of a great QB as Marino or Montana, and a great RB as Jim Brown or Barry Sanders. It's not Randall Cunningham and Brian Westbrook that pops into our head when we hear the term RB. I like the results given to me from RB yards above ANY/A, but am willing to check out another formula. I currently think ANY/A is a high baseline, yet the results (before my latest tweak to reward ballcarriers) gave me too many receiving RBs on the all time list. That's a sign to me that I need to devalue, or at least not increase, the worth of RB receiving.

Actually, I think it's a sign that you were not properly valuing the rushing side, not that you were overvaluing receiving.ETA: And I still don't understand why you can't apply the same logic you used with ypc for ypr. For example, if the average RB ypr is 8.0 (which actually sounds high to me), perhaps you could establish 4.0 ypr as your "easy barometer" for receiving yards. At least then you'd be using a consistent approach with rushing and receiving yards.

 

[Tau837]

Rushing value - I give 10 yards for each rushing TD. This is controversial, although some think that's too much, most think it's too little. A good explanation can be found here.

It is too little. There is a good discussion about it here: Ranking the QBs — Methodology Discussion. At the end of the posts, Doug finally shows some analysis that reveals that the value of a TD is roughly 14.5 yards, although that is based on some assumptions he describes, and thus could be more in a range of 12 to 15 yards.The post you linked ends with this: "I don’t know if ten is exactly the right answer, but it feels about right." We can do better than that, as shown in the post I linked above.

I assume changing TD worth from 10 yards to 14.5 yards or thereabouts would have a significant effect on your rankings.

Chase, I don't believe you have responded to this yet. Have you changed the value of TDs from 10 to 14.5 (or other) in your formula? If not, why not?

I'm in the process of updating this now. I'm not sure 14.5 will be the final answer, but will let you know when I come up with it.

 

[switz]

Quick ComparisonAdrian Peterson: 1010.31Brian Westbrook: 948.78LaDainian Tomlinson: 888.96Frank Gore: 556.82Clinton Portis: 552.43Marion Barber: 537.05Thomas Jones: 281.52It really shows how much better Brian Westbrook was at getting yards compared to LT (4.79 vs. 4.68 and 8.57 vs. 7.92) and how MB compares far more with the Gores/Portis' than he does Thomas Jones (and that's with a poor -15.91 YPR-Benchmark score).

Hmmm.... I think there is an error then.If you were told you could pick between LT and Westbrook for your NFL team, which would you choose? I'd guess 95% would say LT, without even blinking an eye.And I think Barber is much closer to Thomas Jones than Gore or Portis. Gore and Portis are stud RBs. Jones and Barber are serviceable starters. Barber isn't even a proven starter yet. I think the flaw in this is trying to get it to measure up to what you "think" is right. Because some players are overvalued or undervalued by our own biases. Reality is, some players just aren't as good as we'd like to believe, and some are better than we think. If it lines up perfectly with our "view" then something is wrong. In fact, an accurate system would probably have some real outliers.

 

[Chase Stuart]

How about receiving? The NFL average adjusted net yards per pass play this past season was 5.11. That means on every pass play, if you hit just 5.11 yards, that's average. I don't want to give receiving RBs too much credit, so while I used a really low baseline for RB ypc (3.0), I'll use average here. So RBs only get points for their yards above that number.

I don't understand your logic in choosing 3.0 ypc due to it being a "very easy barometer" but then going with the average of 5.11 adjusted net yards per pass play. I have little faith in this formula if you are applying different criteria in choosing your thresholds.Furthermore, if you're not going to go with the "easy barometer" route, why would you choose an NFL wide average of adjusted net yards per pass play? That includes sacks and pass plays to WRs and TEs, correct? Why not keep the focus on RBs, and determine what the average ypr is for RBs, then determine a threshold based on that?

The average YPR for RBs is something like 8. Therefore, about half of the RBs will have no receiving value. I think a RB with 30 catches for 180 yards brings more value to the table than a RB with 0 catches for 0 yards. Using the RB-wide YPR average would be too high.I think RB rushing and RB receiving do need different thresholds. People don't view them as the same (i.e., a RB with 1500 rushing yards and 100 receiving yards is generally viewed as a better RB -- and perhaps rightly so -- than one with 800/800).

I didn't say to use the average ypr, nor did I say to use the same thresholds for running and receiving. I suggested to (a) be consistent in how you determine the two thresholds and (b) determine the receiving threshold based on the average RB ypr rather than NFL average adjusted net yards per pass play.

Chase, I don't believe you responded to this. Are you sticking with NFL average adjusted net yards per pass play? If so, can you elaborate on why, relative to my two points above?

I think I explained before why average RB ypr won't work. I'm going to stick with average adjusted net yards per pass play for now. I'm not sure if I can fully explain why, but it seems right to me. Every pass play is worth whatever the average ANY/A is. So if a RB has a reception where he gets more than that, he's going to get credit. Remember, for QB rushing, I used adjusted rushing yards above 4.0. That was totally and completely arbitrary, yet I think it worked as well or better than any other study I've seen. When a player gets credit for doing something another position generally does (QB rushing, RB receiving), I want to set the standard a bit higher. We think of a great QB as Marino or Montana, and a great RB as Jim Brown or Barry Sanders. It's not Randall Cunningham and Brian Westbrook that pops into our head when we hear the term RB. I like the results given to me from RB yards above ANY/A, but am willing to check out another formula. I currently think ANY/A is a high baseline, yet the results (before my latest tweak to reward ballcarriers) gave me too many receiving RBs on the all time list. That's a sign to me that I need to devalue, or at least not increase, the worth of RB receiving.

Actually, I think it's a sign that you were not properly valuing the rushing side, not that you were overvaluing receiving.ETA: And I still don't understand why you can't apply the same logic you used with ypc for ypr. For example, if the average RB ypr is 8.0 (which actually sounds high to me), perhaps you could establish 4.0 ypr as your "easy barometer" for receiving yards. At least then you'd be using a consistent approach with rushing and receiving yards.

There's not much of a difference between 4.0 and 5.11. And I think the number should change with era, as the passing game has improved. As such, a RB needs to do more to be impressive now. I understand your complaints about consistency, but I'd think the 3.0 number is a lot more arbitrary than league average.

 

[Tau837]

How about receiving? The NFL average adjusted net yards per pass play this past season was 5.11. That means on every pass play, if you hit just 5.11 yards, that's average. I don't want to give receiving RBs too much credit, so while I used a really low baseline for RB ypc (3.0), I'll use average here. So RBs only get points for their yards above that number.

I don't understand your logic in choosing 3.0 ypc due to it being a "very easy barometer" but then going with the average of 5.11 adjusted net yards per pass play. I have little faith in this formula if you are applying different criteria in choosing your thresholds.Furthermore, if you're not going to go with the "easy barometer" route, why would you choose an NFL wide average of adjusted net yards per pass play? That includes sacks and pass plays to WRs and TEs, correct? Why not keep the focus on RBs, and determine what the average ypr is for RBs, then determine a threshold based on that?

The average YPR for RBs is something like 8. Therefore, about half of the RBs will have no receiving value. I think a RB with 30 catches for 180 yards brings more value to the table than a RB with 0 catches for 0 yards. Using the RB-wide YPR average would be too high.I think RB rushing and RB receiving do need different thresholds. People don't view them as the same (i.e., a RB with 1500 rushing yards and 100 receiving yards is generally viewed as a better RB -- and perhaps rightly so -- than one with 800/800).

I didn't say to use the average ypr, nor did I say to use the same thresholds for running and receiving. I suggested to (a) be consistent in how you determine the two thresholds and (b) determine the receiving threshold based on the average RB ypr rather than NFL average adjusted net yards per pass play.

Chase, I don't believe you responded to this. Are you sticking with NFL average adjusted net yards per pass play? If so, can you elaborate on why, relative to my two points above?

I think I explained before why average RB ypr won't work. I'm going to stick with average adjusted net yards per pass play for now. I'm not sure if I can fully explain why, but it seems right to me. Every pass play is worth whatever the average ANY/A is. So if a RB has a reception where he gets more than that, he's going to get credit. Remember, for QB rushing, I used adjusted rushing yards above 4.0. That was totally and completely arbitrary, yet I think it worked as well or better than any other study I've seen. When a player gets credit for doing something another position generally does (QB rushing, RB receiving), I want to set the standard a bit higher. We think of a great QB as Marino or Montana, and a great RB as Jim Brown or Barry Sanders. It's not Randall Cunningham and Brian Westbrook that pops into our head when we hear the term RB. I like the results given to me from RB yards above ANY/A, but am willing to check out another formula. I currently think ANY/A is a high baseline, yet the results (before my latest tweak to reward ballcarriers) gave me too many receiving RBs on the all time list. That's a sign to me that I need to devalue, or at least not increase, the worth of RB receiving.

Actually, I think it's a sign that you were not properly valuing the rushing side, not that you were overvaluing receiving.ETA: And I still don't understand why you can't apply the same logic you used with ypc for ypr. For example, if the average RB ypr is 8.0 (which actually sounds high to me), perhaps you could establish 4.0 ypr as your "easy barometer" for receiving yards. At least then you'd be using a consistent approach with rushing and receiving yards.

There's not much of a difference between 4.0 and 5.11. And I think the number should change with era, as the passing game has improved. As such, a RB needs to do more to be impressive now. I understand your complaints about consistency, but I'd think the 3.0 number is a lot more arbitrary than league average.

Well, one could argue that if there isn't much difference between 4.0 and 5.11, there isn't a compelling reason to prefer 5.11. And that was merely an example... I wasn't saying 4.0 should be your "easy barometer." My point was that you should consider the average RB ypr and determine the right "easy barometer." But maybe that isn't so "easy."