TT drafting vs the other GM's (Career approx Value)

[Norm]

1 minute ago, TheOnlyThing said:

Lots of hard work Squire, thanks for the effort.

While I have repeatedly challenged the claim that Ted's drafting since 2011 has continued to be elite, not even I expected to see the Packers "ranked" of 21st out of 32 teams. That is not even mediocre drafting.

Still not sure CaV is the be all and end all of measuring drafting success for all the reasons stated above and elsewhere, but your data certainly lends support to the notion that Green Bay's drafting has declined, and declined significantly, during the second half of Ted's reign as GM of the Packers.

Nothing is the be all end all of measuring draft success.

[incognito_man]

3 minutes ago, skibrett15 said:

No, it's not that simple.  Some teams have more draft capital because of compensatory picks.  If they get the same value from the draft as a similar slotted team with no comp picks that's a tangibly worse outcome, yes?

I get the point that they got the comp pick because the player left after their contract, but what does a player signing here like Cook, leaving, and earning us a comp pick have anything to do with drafting? 

As soon as you open this up to salary cap and that kind of discussion we lose focus on the topic at hand which is ability to hit on picks.  We're trying to find a batting percentage for these teams, right, not create some sort of GM god score.

Similarly, if BB trades away 2nd round picks for years and turns into 1st round picks in the following year, we're meant to evaluate the CAV as if he had used all his picks in a given year even though he hasn't and he's deferred the investment until later?

why not just evaluate the drafting by looking at the picks that were actually made in the NFL draft?

A big part of being a good drafter is working the draft system to one's advantage, especially if you lack the inherited draft capital. If you gain an 'unearned' compensatory pick like the Cook example, it doesn't hurt the evaluation of a GM because that pick isn't factored into the capital. It will only help a GM who 'gamed the system' by doing so. This reflects reality.

There is no need to open anything to the salary cap - I don't think I suggested or implied it is a factor for this. And my working assumption was more of the GM god score - but really only on how they build a team using their given draft capital. i.e. sustaining success (which is impossibly difficult).

I don't get the BB example. If he trades a 2nd in 2017 for a first in 2018 he trades a year of AV for a better draft slot in 2018. His hope is that he is around long enough to witness the moment the 2018 1st rounder eclipses the 2017 2nd rounder in career AV. Then repeat the year after, if you'd like. I'm not sure what your comment about "used all his picks in a given year" is referencing?

If you only look at actual players drafted, people start to credit a team like Seattle with Sheldon Richardson and Jimmy Graham's contributions while ignoring the 1st round pick it took to get them. It artificially deflates their draft capital. You can't give free credit for picks traded away for players.

[incognito_man]

14 minutes ago, TheOnlyThing said:

Lots of hard work Squire, thanks for the effort.

While I have repeatedly challenged the claim that Ted's drafting since 2011 has continued to be elite, not even I expected to see the Packers "ranked" of 21st out of 32 teams. That is not even mediocre drafting.

Still not sure CaV is the be all and end all of measuring drafting success for all the reasons stated above and elsewhere, but your data certainly lends support to the notion that Green Bay's drafting has declined, and declined significantly, during the second half of Ted's reign as GM of the Packers.

you're not great at science are you

 

[skibrett15]

1 minute ago, incognito_man said:

If you only look at actual players drafted, people start to credit a team like Seattle with Sheldon Richardson and Jimmy Graham's contributions while ignoring the 1st round pick it took to get them. It artificially deflates their draft capital. You can't give free credit for picks traded away for players.

This isn't reflected in the scores i posted.

 

2 minutes ago, incognito_man said:

I don't get the BB example. If he trades a 2nd in 2017 for a first in 2018 he trades a year of AV for a better draft slot in 2018. His hope is that he is around long enough to witness the moment the 2018 1st rounder eclipses the 2017 2nd rounder in career AV. Then repeat the year after, if you'd like. I'm not sure what your comment about "used all his picks in a given year" is referencing?

Hypothetical:

So, BB trades 2010's 20 chase stuart points for 2011's 26 chase stuart points by trading his 2nd for a 1st.  But you are saying let's treat 2010 as if he invested the full amount of draft capital of the 29th ranked team.  But he didn't.  He invested 140% of the 32nd ranked team in 2011 and 70% of the 29th ranked team in 2010.  The metric should reflect that.

5 minutes ago, incognito_man said:

If you gain an 'unearned' compensatory pick like the Cook example, it doesn't hurt the evaluation of a GM because that pick isn't factored into the capital. It will only help a GM who 'gamed the system' by doing so. This reflects reality.

If you game the system and whiff on the pick vs gaming the system and hit on the pick.... many times... to a statistically significant level... one is a much better GM and drafter than the other.  So i think you want to factor in the capital of the comp pick. 

Even if the comp pick is "earned" in the traditional sense, the GM should be penalized for whiffing on a 3rd round pick that he got for a good player leaving.  Fact is, he had extra chips that year and he should have extra CAV that year.

Your point below is good.  I think you want to adjust the CAV/investment by including a factor which would adjust for "actual capital" earned vs "capital assigned" due to team rank.

12 minutes ago, incognito_man said:

There is no need to open anything to the salary cap - I don't think I suggested or implied it is a factor for this. And my working assumption was more of the GM god score - but really only on how they build a team using their given draft capital. i.e. sustaining success (which is impossibly difficult).

I stand by CAV/invested point as a really good metric of drafting success which is akin to batting average.  To me, that's an actionable and digestible (barely) statistic which would be worth further exploration. 

I think that we could also take a look at projecting the CAV of 2nd year players by multiplying by 2x or something, 3rd year players by 1.5x, and 1st year players by 2.5x.

Also, I think replacing CAV with AV over the rookie deal with the drafted team would be a great change/exploration as well, but that doesn't really work for QBs.  Still, you don't give one team a ton of draft credit for a guy who goes on to a great career on another team after being cut in camp, do you?

[ChaRisMa]

Only value from the draft you need to know when judging a GM. We’re top 5 the last 5 years. And it’s pretty much all via draft, so...

[Norm]

4 minutes ago, ChaRisMa said:

Only value from the draft you need to know when judging a GM. We’re top 5 the last 5 years. And it’s pretty much all via draft, so...

But only because Rodgers

[AlexGreen#20]

2011-2017 is tough to nail down using career AV. It's not going to be like the previous bigger dataset where you have a lot of sample size to filter out the "noise".

If draft picks are judged based on their production through out their first contract, (and once that has completed, they're judge against their extension) half of that sample size isn't finished being judged. The 2014-2017 draft classes are weighted only make up 10 "years of data" where as 2011-2013 make up "18 years of data". 

The one player who should theoretically be our biggest gainer in this study, having been our first round pick in 2011, is Derrick Sherrod. 

Let's say we make Sherrod the average CareerAV of the other tackles in his round (Tyron Smith, Nate Solder, Anthony Costanzo, James Carpenter, Gabe Carimi. That's a fair mix of hit and bust IMO.) that's a CareerAV of 37.8 as opposed to 3. That pushes you all the way up to 16th in Career AV average for that time period just based on one draft pick. 

[squire12]

58 minutes ago, AlexGreen#20 said:

 

The one player who should theoretically be our biggest gainer in this study, having been our first round pick in 2011, is Derrick Sherrod. 

Let's say we make Sherrod the average CareerAV of the other tackles in his round (Tyron Smith, Nate Solder, Anthony Costanzo, James Carpenter, Gabe Carimi. That's a fair mix of hit and bust IMO.) that's a CareerAV of 37.8 as opposed to 3. That pushes you all the way up to 16th in Career AV average for that time period just based on one draft pick. 

That is a great point, but you also need to do that for the other 31 teams and factor in players lost to injury/suspension/etc.  That gets really messy.

[spilltray]

Suspensions usually have the warning signs in college. Thing is with only using 2011-2017 rookie contracts as your evaluation, losing your 2011 1st round pick completely to unforeseeable circumstances really skews the scale

[skibrett15]

8 hours ago, squire12 said:

That is a great point, but you also need to do that for the other 31 teams and factor in players lost to injury/suspension/etc.  That gets really messy.

yes yes yes.

Tired of people suggesting we make adjustments to packers stats when the stats are run the same way across the league for every team.  Sherrod is an example, not necessarily an outlier/exception.

The small sample size argument is valid.  My suggestion is to juice up the CaV for players still under rookie contract.

[AlexGreen#20]

2 minutes ago, skibrett15 said:

yes yes yes.

Tired of people suggesting we make adjustments to packers stats when the stats are run the same way across the league for every team.  Sherrod is an example, not necessarily an outlier/exception.

The small sample size argument is valid.  My suggestion is to juice up the CaV for players still under rookie contract.

Don't get me wrong, what you've done here is great work and extremely interesting.

My reason for picking out Sherrod is because it fits in so concisely with the small sample size issue. Though I disagree that it's something insignificant or common enough to be explained away by the volume of the study. How many teams have lost a first round pick to a career ending injury over that time period? Kevin White for the Bears is the only one I can think of off the top of my head and that isn't a real career ending injury, more like an awful string of season enders. 

[skibrett15]

1 minute ago, AlexGreen#20 said:

Don't get me wrong, what you've done here is great work and extremely interesting.

My reason for picking out Sherrod is because it fits in so concisely with the small sample size issue. Though I disagree that it's something insignificant or common enough to be explained away by the volume of the study. How many teams have lost a first round pick to a career ending injury over that time period? Kevin White for the Bears is the only one I can think of off the top of my head and that isn't a real career ending injury, more like an awful string of season enders. 

David wilson, Manziel?  It hasn't happened often, but sherrod was pick 32, so has it happened to high 2nds? High 3rds?  It's a slippery slope type of thing and the draft value chart for low 1sts and 2nd rounders isn't all that different.

 

@squire12 any interest in putting a correction factor on the rookie deals, and limiting CAV to rookie contract AV (i.e. first 5 years AV for 1st rounders, first 4 years AV for 2nd-7th)?  Perhaps a better way to do that would be to stick with CaV, but count the rookie contract seasons as double.  So CAV+RookAV = DraftPickAV(DPAV)

In terms of a correction, I'd say that 2nd year guys get a 2x bump, 1st year guys get a 2.5x bump, and 3rd year guys get 1.5x bump to their AVs.

Any interest in opening up the data so that I might copy it?

[AlexGreen#20]

1 hour ago, skibrett15 said:

David wilson, Manziel?  It hasn't happened often, but sherrod was pick 32, so has it happened to high 2nds? High 3rds?  It's a slippery slope type of thing and the draft value chart for low 1sts and 2nd rounders isn't all that different.

 

@squire12 any interest in putting a correction factor on the rookie deals, and limiting CAV to rookie contract AV (i.e. first 5 years AV for 1st rounders, first 4 years AV for 2nd-7th)?  Perhaps a better way to do that would be to stick with CaV, but count the rookie contract seasons as double.  So CAV+RookAV = DraftPickAV(DPAV)

In terms of a correction, I'd say that 2nd year guys get a 2x bump, 1st year guys get a 2.5x bump, and 3rd year guys get 1.5x bump to their AVs.

Any interest in opening up the data so that I might copy it?

David Wilson is a good call.

I disagree with Manziel. The kid had problems with drugs and alcohol that everybody knew about. The Browns choosing to ignore those issues is a classic example of ****ty GMing.

[squire12]

6 minutes ago, AlexGreen#20 said:

David Wilson is a good call.

I disagree with Manziel. The kid had problems with drugs and alcohol that everybody knew about. The Browns choosing to ignore those issues is a classic example of ****ty GMing.

RG3, Ryan Tannehill in 2017, Justin Blackmon, Jahvid Best, Andrew Luck, Jake Locker.

All these players have had injuries and suspensions that have affected their development.  Whether the injury and situation is on the same level as Sherrod's is debatable, but to find a easy criteria to say 1 player should be handled differently than another is dicey.

[AlexGreen#20]

Just now, squire12 said:

RG3, Ryan Tannehill in 2017, Justin Blackmon, Jahvid Best, Andrew Luck, Jake Locker.

All these players have had injuries and suspensions that have affected their development.  Whether the injury and situation is on the same level as Sherrod's is debatable, but to find a easy criteria to say 1 player should be handled differently than another is dicey.

I'm not doubting that it's dicey. I'm just pointing out the issue I take with the method. I'm not even saying that there's a better way to breakdown this data. What you've done is extremely impressive. I just don't know how you break down a draft history with a small sample size.

Just looking at it on an individual basis.

RG3 is arguable. The Redskins got 3 seasons out of him and truthfully the Redskins are responsible for breaking him. Letting him keep playing on his obviously busted knee was retarded.

Ryan Tannehill missed one year.

Justin Blackmon's drug history was known by everybody. 

Andrew Luck missed one season.

I don't know what you do with Locker, probably give the team the benefit of the doubt, but he played 4 seasons and never once looked the part of an average starter.