2021 Rookie Tiers
- Dynasty DeLorean
- Degenerate
- Posts: 8920
- Joined: Tue Feb 24, 2015 8:45 am
Re: 2021 Rookie Tiers
“Twisting myself in knots” aka actually doing some research and using sound logic and data. This is a joke right?
“Unwilling to investigate” says the person clinging to an opinion they did zero investigating on, lmao.
“Unwilling to investigate” says the person clinging to an opinion they did zero investigating on, lmao.
Re: 2021 Rookie Tiers
Nice chart. Looks like under 92 (around 30th percentile) and lower are worth steering clear from or at least be cautious aboutOhCruelestRanter wrote: ↑Sun Jan 31, 2021 8:11 am Here. I’ll do it.
As you can see on this chart, there is a correlation, but that trend line is strongly influenced by the terrible outcomes of RBs with speed scores under 92. Remove them, and there’s basically no trend.
Like I said earlier, we know low speed scores are really bad prognostic indicators, but we can’t really say that there’s a meaningful difference between 114 and 118. This is demonstrated by the data we have.
The problem with speed score is that it’s way more predictive for rushing success than it is for overall fantasy/NFL success, likely because it doesn’t correlate at all with success in the receiving game. It’s still a useful tool, and if everything is truly equal I’ll take the guy with the better score, but it’s most useful as a cutoff.
Slower than 4.6 and under 200 lbs would fit that criteria. However someone at 225 could run 4.7 and be at 92 score and I don't think we've ever had some one run over 4.7 and be relevant at RB?? Edit: Blount looks like the only example I can find recently
Of course, I don't think 40 times or speed scores matter too much for RB either but it's an interesting discussion.
JT 121, CEH 92.5
-
- MVP
- Posts: 2732
- Joined: Sun Jul 06, 2014 5:33 pm
Re: 2021 Rookie Tiers
Yeah- like anything of these formulaic scores, it’s likely to fall apart at the extremes. I doubt a 300 lb guy who ran a 4.9 could play RB, even if that would be a speed score above 104.Jigga94 wrote: ↑Sun Jan 31, 2021 10:00 amNice chart. Looks like under 92 (around 30th percentile) and lower are worth steering clear from or at least be cautious aboutOhCruelestRanter wrote: ↑Sun Jan 31, 2021 8:11 am Here. I’ll do it.
As you can see on this chart, there is a correlation, but that trend line is strongly influenced by the terrible outcomes of RBs with speed scores under 92. Remove them, and there’s basically no trend.
Like I said earlier, we know low speed scores are really bad prognostic indicators, but we can’t really say that there’s a meaningful difference between 114 and 118. This is demonstrated by the data we have.
The problem with speed score is that it’s way more predictive for rushing success than it is for overall fantasy/NFL success, likely because it doesn’t correlate at all with success in the receiving game. It’s still a useful tool, and if everything is truly equal I’ll take the guy with the better score, but it’s most useful as a cutoff.
Slower than 4.6 and under 200 lbs would fit that criteria. However someone at 225 could run 4.7 and be at 92 score and I don't think we've ever had some one run over 4.7 and be relevant at RB?? Edit: Blount looks like the only example I can find recently
Of course, I don't think 40 times or speed scores matter too much for RB either but it's an interesting discussion.
JT 121, CEH 92.5
If you’re curious, the highest speed score by a RB was Keith Marshall’s 126.9. Vernon Davis’ score would be 136.8. Montez Sweat would be 137.5.
COOGAN IS A CHEATER AND A THIEF
- Dynasty DeLorean
- Degenerate
- Posts: 8920
- Joined: Tue Feb 24, 2015 8:45 am
Re: 2021 Rookie Tiers
Is there any way to filter by draft capital. For example, use only players that went top 3 rounds or something.
edit: never mind, thought you had done it yourself
edit: never mind, thought you had done it yourself
Last edited by Dynasty DeLorean on Sun Jan 31, 2021 6:06 pm, edited 1 time in total.
-
- Role Player
- Posts: 369
- Joined: Tue May 15, 2018 4:53 am
Re: 2021 Rookie Tiers
I usually read the Football Outsiders speed score analysis each year (link for those interested https://www.footballoutsiders.com/stat- ... score-2020).
They mention 2 lines, being the “Arian Foster line” and “Ahmad Bradshaw line” whereby Foster is the only player with a speed score under 100 to average 1000 yards rushing over his first 5 seasons. Whilst that’s a pretty arbitrary number they include a table showing speed scores correlation to average yards rushing in a running backs first 5 seasons, which is pretty interesting.
Like the above chart there looks to be a pretty clear cut off at around 90, which is close to the “Ahmad Bradshaw line” whereby he is the only player with a speed score below 87.7 to manage 1000 yards rushing in their first 5 seasons combined.
As pointed out above, what these tables don’t do is account for passing yardage and a backs involvement in the passing game, which is obviously a big part of the modern game. Based on the above tables it shows that, for example, Clyde and James Robinson statistically are unlikely to average 1000 yards rushing over their first 5 seasons - but they both showed the ability to be impactful in the passing game which bodes well for their future success.
I think where the tables are most useful is highlighting how someone with a sub 100 speed score and no involvement in the passing game is an easy sell if they produce early in the rushing game - players like Alfred Morris, for example.
They mention 2 lines, being the “Arian Foster line” and “Ahmad Bradshaw line” whereby Foster is the only player with a speed score under 100 to average 1000 yards rushing over his first 5 seasons. Whilst that’s a pretty arbitrary number they include a table showing speed scores correlation to average yards rushing in a running backs first 5 seasons, which is pretty interesting.
Like the above chart there looks to be a pretty clear cut off at around 90, which is close to the “Ahmad Bradshaw line” whereby he is the only player with a speed score below 87.7 to manage 1000 yards rushing in their first 5 seasons combined.
As pointed out above, what these tables don’t do is account for passing yardage and a backs involvement in the passing game, which is obviously a big part of the modern game. Based on the above tables it shows that, for example, Clyde and James Robinson statistically are unlikely to average 1000 yards rushing over their first 5 seasons - but they both showed the ability to be impactful in the passing game which bodes well for their future success.
I think where the tables are most useful is highlighting how someone with a sub 100 speed score and no involvement in the passing game is an easy sell if they produce early in the rushing game - players like Alfred Morris, for example.
- Attachments
-
- 6C1EEB60-91D0-45B9-ABA1-290975A42F1E.png (19.44 KiB) Viewed 686 times
Last edited by cantguardjake on Sun Jan 31, 2021 12:07 pm, edited 1 time in total.
-
- Degenerate
- Posts: 7557
- Joined: Sat May 28, 2011 11:53 pm
Re: 2021 Rookie Tiers
Interesting graph. Some questions-OhCruelestRanter wrote: ↑Sun Jan 31, 2021 8:11 am Here. I’ll do it.
As you can see on this chart, there is a correlation, but that trend line is strongly influenced by the terrible outcomes of RBs with speed scores under 92. Remove them, and there’s basically no trend.
Like I said earlier, we know low speed scores are really bad prognostic indicators, but we can’t really say that there’s a meaningful difference between 114 and 118. This is demonstrated by the data we have.
The problem with speed score is that it’s way more predictive for rushing success than it is for overall fantasy/NFL success, likely because it doesn’t correlate at all with success in the receiving game. It’s still a useful tool, and if everything is truly equal I’ll take the guy with the better score, but it’s most useful as a cutoff.
The X axis says combine speed score. Does that mean players with only pro day numbers are eliminated?
The Y axis is unclear- NFL AVG/G measures what?
The legend indicates the line refers to a number. What is it? Seconds, ie 40 time? Years, ie career length? Something else?
League #1- 14 tm ppr, 1Q, 2R, 3W, 1T, 1 R/W/T, 1K
1 DT, 2 DE, 2 LB, 1 CB, 1 S, 1 flex
League #2- 12 team PPR, 1Q, 1R, 2W, 1T, 1 R/W/T, 1 W/R/T, 1 Def
League #3- 12 tm PPR, 1Q, 0R (yes, ZERO RB) 3W, 1T, 2 R/W/T flex, 1 Def
1 DT, 2 DE, 2 LB, 1 CB, 1 S, 1 flex
League #2- 12 team PPR, 1Q, 1R, 2W, 1T, 1 R/W/T, 1 W/R/T, 1 Def
League #3- 12 tm PPR, 1Q, 0R (yes, ZERO RB) 3W, 1T, 2 R/W/T flex, 1 Def
-
- Ring of Fame
- Posts: 3835
- Joined: Tue Jul 31, 2018 7:12 am
Re: 2021 Rookie Tiers
That is really well done. Great jobOhCruelestRanter wrote: ↑Sun Jan 31, 2021 8:11 am Here. I’ll do it.
As you can see on this chart, there is a correlation, but that trend line is strongly influenced by the terrible outcomes of RBs with speed scores under 92. Remove them, and there’s basically no trend.
Like I said earlier, we know low speed scores are really bad prognostic indicators, but we can’t really say that there’s a meaningful difference between 114 and 118. This is demonstrated by the data we have.
The problem with speed score is that it’s way more predictive for rushing success than it is for overall fantasy/NFL success, likely because it doesn’t correlate at all with success in the receiving game. It’s still a useful tool, and if everything is truly equal I’ll take the guy with the better score, but it’s most useful as a cutoff.
- Dynasty DeLorean
- Degenerate
- Posts: 8920
- Joined: Tue Feb 24, 2015 8:45 am
Re: 2021 Rookie Tiers
Thanks for posting. Appears to confirm what I’ve been saying.cantguardjake wrote: ↑Sun Jan 31, 2021 12:02 pm I usually read the Football Outsiders speed score analysis each year (link for those interested https://www.footballoutsiders.com/stat- ... score-2020).
They mention 2 lines, being the “Arian Foster line” and “Ahmad Bradshaw line” whereby Foster is the only player with a speed score under 100 to average 1000 yards rushing over his first 5 seasons. Whilst that’s a pretty arbitrary number they include a table showing speed scores correlation to average yards rushing in a running backs first 5 seasons, which is pretty interesting.
Like the above chart there looks to be a pretty clear cut off at around 90, which is close to the “Ahmad Bradshaw line” whereby he is the only player with a speed score below 87.7 to manage 1000 yards rushing in their first 5 seasons combined.
As pointed out above, what these tables don’t do is account for passing yardage and a backs involvement in the passing game, which is obviously a big part of the modern game. Based on the above tables it shows that, for example, Clyde and James Robinson statistically are unlikely to average 1000 yards rushing over their first 5 seasons - but they both showed the ability to be impactful in the passing game which bodes well for their future success.
I think where the tables are most useful is highlighting how someone with a sub 100 speed score and no involvement in the passing game is an easy sell if they produce early in the rushing game - players like Alfred Morris, for example.
I have Bradshaws speed score as 92.4 by the way. Combine times and weights are often listed incorrectly. There is no single resource with all accurate information. I did my best to verify all the info for my own database.
edit: I have a 4.55 listed for Bradshaw and they have a 4.61. I'll have to check this.
ESPN Article from 2008 has it listed as 4.55 https://www.espn.com/nfl/news/story?id=3337822
The combineresults page has 4.61 but they don't always have accurate info. https://nflcombineresults.com/playerpage.php?i=7613The thing is, not all 40 times are created equal. Brandon Jacobs' 4.56 40 is incredible when you consider his 267-pound frame. On the other hand, Ahmad Bradshaw's 4.55 40 was disappointing for a player who weighed in nearly 70 pounds lighter. Adjusting 40 times for weight helps translate a raw metric into something that's more indicative of a player's NFL potential.
*shrug*
Arian Foster was injured and therefor his was a pro day time in which he may not have even been fully healthy. So just FYI to everyone reading that both speed scores for foster and Bradshaw listed in the article are potentially misleading.
-
- MVP
- Posts: 2732
- Joined: Sun Jul 06, 2014 5:33 pm
Re: 2021 Rookie Tiers
Great questions. It’s not my data or graph, so I’m not sure if proday time were eliminated or adjusted.Pullo Vision wrote: ↑Sun Jan 31, 2021 12:04 pmInteresting graph. Some questions-OhCruelestRanter wrote: ↑Sun Jan 31, 2021 8:11 am Here. I’ll do it.
As you can see on this chart, there is a correlation, but that trend line is strongly influenced by the terrible outcomes of RBs with speed scores under 92. Remove them, and there’s basically no trend.
Like I said earlier, we know low speed scores are really bad prognostic indicators, but we can’t really say that there’s a meaningful difference between 114 and 118. This is demonstrated by the data we have.
The problem with speed score is that it’s way more predictive for rushing success than it is for overall fantasy/NFL success, likely because it doesn’t correlate at all with success in the receiving game. It’s still a useful tool, and if everything is truly equal I’ll take the guy with the better score, but it’s most useful as a cutoff.
The X axis says combine speed score. Does that mean players with only pro day numbers are eliminated?
The Y axis is unclear- NFL AVG/G measures what?
The legend indicates the line refers to a number. What is it? Seconds, ie 40 time? Years, ie career length? Something else?
AV/G is a pro-football-reference metric. https://www.pro-football-reference.com/ ... x37a8.html
The line is the R-squared between the Speed Score and AV/G.
COOGAN IS A CHEATER AND A THIEF
-
- Role Player
- Posts: 369
- Joined: Tue May 15, 2018 4:53 am
Re: 2021 Rookie Tiers
I’ve also found it difficult to get accurate measurements. Claypool is usually listed at 228 on both Wikipedia and nflcombineresults but that was actually his senior bowl weigh in, not the weight he actually tested at.Dynasty DeLorean wrote: ↑Sun Jan 31, 2021 12:43 pmThanks for posting. Appears to confirm what I’ve been saying.cantguardjake wrote: ↑Sun Jan 31, 2021 12:02 pm I usually read the Football Outsiders speed score analysis each year (link for those interested https://www.footballoutsiders.com/stat- ... score-2020).
They mention 2 lines, being the “Arian Foster line” and “Ahmad Bradshaw line” whereby Foster is the only player with a speed score under 100 to average 1000 yards rushing over his first 5 seasons. Whilst that’s a pretty arbitrary number they include a table showing speed scores correlation to average yards rushing in a running backs first 5 seasons, which is pretty interesting.
Like the above chart there looks to be a pretty clear cut off at around 90, which is close to the “Ahmad Bradshaw line” whereby he is the only player with a speed score below 87.7 to manage 1000 yards rushing in their first 5 seasons combined.
As pointed out above, what these tables don’t do is account for passing yardage and a backs involvement in the passing game, which is obviously a big part of the modern game. Based on the above tables it shows that, for example, Clyde and James Robinson statistically are unlikely to average 1000 yards rushing over their first 5 seasons - but they both showed the ability to be impactful in the passing game which bodes well for their future success.
I think where the tables are most useful is highlighting how someone with a sub 100 speed score and no involvement in the passing game is an easy sell if they produce early in the rushing game - players like Alfred Morris, for example.
I have Bradshaws speed score as 92.4 by the way. Combine times and weights are often listed incorrectly. There is no single resource with all accurate information. I did my best to verify all the info for my own database.
edit: I have a 4.55 listed for Bradshaw and they have a 4.61. I'll have to check this.
ESPN Article from 2008 has it listed as 4.55 https://www.espn.com/nfl/news/story?id=3337822
The combineresults page has 4.61 but they don't always have accurate info. https://nflcombineresults.com/playerpage.php?i=7613The thing is, not all 40 times are created equal. Brandon Jacobs' 4.56 40 is incredible when you consider his 267-pound frame. On the other hand, Ahmad Bradshaw's 4.55 40 was disappointing for a player who weighed in nearly 70 pounds lighter. Adjusting 40 times for weight helps translate a raw metric into something that's more indicative of a player's NFL potential.
*shrug*
Arian Foster was injured and therefor his was a pro day time in which he may not have even been fully healthy. So just FYI to everyone reading that both speed scores for foster and Bradshaw listed in the article are potentially misleading.
It would be nice if there was a uniform data base.
-
- Ring of Fame
- Posts: 3835
- Joined: Tue Jul 31, 2018 7:12 am
Re: 2021 Rookie Tiers
Dynasty DeLorean wrote: ↑Sun Jan 31, 2021 8:12 am “Twisting myself in knots” aka actually doing some research and using sound logic and data. This is a joke right?
“Unwilling to investigate” says the person clinging to an opinion they did zero investigating on, lmao.
I guess this is how you get a 90% accuracy rating. Revisionist history:
Dynasty DeLorean wrote: ↑Sat Jan 30, 2021 11:26 am If a poor speed score is an indicator of failure then a high speed score is likely an indicator of success.
Hope you have a great day today.Dynasty DeLorean wrote: ↑Sun Jan 31, 2021 12:43 pm
Thanks for posting. Appears to confirm what I’ve been saying.
-
- Role Player
- Posts: 369
- Joined: Tue May 15, 2018 4:53 am
Re: 2021 Rookie Tiers
Did you see the table I posted?Bronco Billy wrote: ↑Mon Feb 01, 2021 5:40 amDynasty DeLorean wrote: ↑Sun Jan 31, 2021 8:12 am “Twisting myself in knots” aka actually doing some research and using sound logic and data. This is a joke right?
“Unwilling to investigate” says the person clinging to an opinion they did zero investigating on, lmao.
I guess this is how you get a 90% accuracy rating. Revisionist history:
Dynasty DeLorean wrote: ↑Sat Jan 30, 2021 11:26 am If a poor speed score is an indicator of failure then a high speed score is likely an indicator of success.Hope you have a great day today.Dynasty DeLorean wrote: ↑Sun Jan 31, 2021 12:43 pm
Thanks for posting. Appears to confirm what I’ve been saying.
-
- Role Player
- Posts: 369
- Joined: Tue May 15, 2018 4:53 am
Re: 2021 Rookie Tiers
I feel like it’s going to be another lean year like 2019 for your model DD, there is essentially zero productive size / speed freaks this year.
Hopefully Harris turns in a low 4.5 to make it interesting.
Hopefully Harris turns in a low 4.5 to make it interesting.
-
- MVP
- Posts: 2732
- Joined: Sun Jul 06, 2014 5:33 pm
Re: 2021 Rookie Tiers
If Etienne weighs in at like 208 and blazes a 4.40 (after adjusting for the pro-day) then that could also make things interesting.cantguardjake wrote: ↑Mon Feb 01, 2021 3:32 pm I feel like it’s going to be another lean year like 2019 for your model DD, there is essentially zero productive size / speed freaks this year.
Hopefully Harris turns in a low 4.5 to make it interesting.
COOGAN IS A CHEATER AND A THIEF
- Dynasty DeLorean
- Degenerate
- Posts: 8920
- Joined: Tue Feb 24, 2015 8:45 am
Re: 2021 Rookie Tiers
Yeah, looking similar to 2019 so far. With regards to 40 times, etienne likely needs a high 4.3x or low 4.4x and najee needs a low 4.5x to make things interesting. There are people out there who do think both could happen, so we’ll see. Of course it’s going to be a mess with adjusting for pro day times. No combine really hurts.cantguardjake wrote: ↑Mon Feb 01, 2021 3:32 pm I feel like it’s going to be another lean year like 2019 for your model DD, there is essentially zero productive size / speed freaks this year.
Hopefully Harris turns in a low 4.5 to make it interesting.