2020 NFL Draft Discussion

[GWH87]

Has this thread discussed Isaiah Wilson OT from Georgia yet?

I think he’s a good bet to be there at 30 and would be a bookend with Bahk for awhile.

Reminds me some of Becton, but actually think he’s a better pick... Would love him Rnd 1 and Davon Hamilton Rnd 2 to start winning in the trenches.

[PossibleCabbage]

The Packers place a priority on mobility skills from OL.  The Zone blocking run game that LaFleur places a priority on moblity skills from OL.   The Packers place a priority on OL who can play multiple positions.

Isaiah Wilson is a man scheme right tackle only who doesn't move very well.  If he's the pick, it basically breaks every single tendency we've seen.

If you put a gun to Gute's head and said "you have to pick Isaiah Wilson, Ezra Cleveland, or Ben Bartch"  Wilson would probably be his third choice.

Edited April 8, 2020 by PossibleCabbage

[MrBobGray]

2 hours ago, AlexGreen#20 said:

I'm having a hard time buying it though. We accounted for all the run plays and like 20 percent of the passing plays and there isn't a significant observable difference. The 80% of the passing plays we aren't seeing has to be an absolutely wild difference.

After reading the breakdown of EPA, I think I get it, and let me say that calculating EPA for individual players is massively, massively stupid in my opinion.

First thing to note is that EPA doesn't consider clock or game situation, so if you're in clock killing mode (ie, if you're resting your starter to let your power back run it down) you still have the same EPA adjustment as running in the first quarter:

Quote

To mitigate that problem, the baseline EP values for each down-distance-field position situation must be created based on real game situations when points are equally valuable and time is not yet a factor. The baseline EP values are therefore based only on game situations when the score was within 10 points and in the first and third quarters. This eliminates situations like ‘trash time,’ and other distortions.

If you run the ball for 2 yards on 1st and 10, you're getting dinged for it, even if it's at the end of a blowout.  This is obviously going to hurt Williams more than Jones.

Second is that total EPA is hugely influenced by big plays.  The average EPA adjustment of any given play, especially run play, is fairly small regardless of direction.  Big plays represent big swings, but this can be misleading in its own way.  If the offense with Jones went -1 yard run, -1 yard run, then Rodgers hits a receiver for 30, the EPA of the offense on those three plays is going to be very positive even though Jones himself had nothing to do with it.  If Jamaal goes 2, 4, and then they convert the 3rd and he comes out, his EPA change is minimal, even though he had more positive impact on the offense than Jones.  This is why I really don't understand the value of calculating the EPA change for an individual player, and especially a RB.  The vast majority of every play is going to be entirely out of his hands.  

Third, EPA is weighted more heavily as you get closer to the goal line:

Quote

Field position is also an important consideration. Yards gained near the goal line are tougher to come by and are more valuable than yards gained at midfield. Yards lost near one’s own goal line can be more costly as well.

Jones being a red zone monster is another huge boost, as running a TD in from the five is worth more than a full point if I'm understanding a lot of these breakdowns. This isn't necessarily a criticism, as Jones's ability to score in the red zone was vital given Aaron hasn't been the killer there he once was in several years.  However, his proportions of red zone carries and touches in general mean that he's also by design going to have a higher EPA.

So Jones is a big play threat who doesn't play often in clock-killing situations or in catch-up mode but does see a lot of goal line carries, which means he's perfectly suited to consistently add huge EPA gains.  Williams is a pass blocker and change of pace guy who's going to see all the garbage time carries and plays that aren't expected to succeed.  The difference in their EPA is exactly what I would expect.

[Shanedorf]

2 minutes ago, MrBobGray said:

This is why I really don't understand the value of calculating the EPA change for an individual player, and especially a RB.

So, based on your extensive research into this ( thx) what are you going to advise Gute about Jones for 2021 and beyond ?

A) let him walk and take the comp pick
B) see if he'd agree to a shorter, more palatable deal
C) there is no C because the Packers aren't going to offer him a market rate 4 or 5 year deal
D) draft his replacement and trade him at the deadline (lol)

[MrBobGray]

13 minutes ago, Shanedorf said:

So, based on your extensive research into this ( thx) what are you going to advise Gute about Jones for 2021 and beyond ?

A) let him walk and take the comp pick
B) see if he'd agree to a shorter, more palatable deal
C) there is no C because the Packers aren't going to offer him a market rate 4 or 5 year deal
D) draft his replacement and trade him at the deadline (lol)

Depends on if you're asking the fan or the pretend GM.

As a fan, I'd love to see Jones come back even at market rate.  Seems like a good guy who works his *** off, extremely productive and talented, prime of his career and a ton of fun to watch.  How can you not want that guy on your team?

As a pretend GM: let him walk and take the comp pick 200%.  The team won't replace his production, but they don't need to; they need Aaron to replace his production.  Part of the reason Jones had as many TDs as he did is because Rodgers was ineffective in the redzone.  Keeping Jones obviously helps mitigate that, but only as long as he stays healthy and really, that's a band-aid solution.  You win in the NFL because you can throw the ball; the 49ers dominated that Super Bowl for 3 quarters and lost by 11 because that's what the NFL is these days.  You either pass or you dream about lifting the Lombardi next year.  As good as Jones is, he's just masking team issues that need to be addressed if the team is going to win a Super Bowl; he's like a nice coat of paint on a building with water damage.  

The capital goes to the pass rush, the secondary, the OL and QB, and it doesn't matter how much talent you already have there.  I'd rather draft pass rusher in the first for the next five years than a RB once, and as much as that sort of sucks as a guy who loves the run game, it's reality.

Edited April 8, 2020 by MrBobGray

[incognito_man]

45 minutes ago, MrBobGray said:

So Jones is a big play threat who doesn't play often in clock-killing situations or in catch-up mode but does see a lot of goal line carries, which means he's perfectly suited to consistently add huge EPA gains.  Williams is a pass blocker and change of pace guy who's going to see all the garbage time carries and plays that aren't expected to succeed.  The difference in their EPA is exactly what I would expect.

It seems that since Jones is used more in critical situations by our coaching staff, that is pretty telling. If Williams is seeing the garbage time and plays that aren't expected to succeed, it seems his value to the team is appropriately represented by EPA

This also is assuming a lot of things about their usage. I haven't seen any snap/situation breakdowns that demonstrates these assumptions.

[MrBobGray]

11 minutes ago, incognito_man said:

It seems that since Jones is used more in critical situations by our coaching staff, that is pretty telling. If Williams is seeing the garbage time and plays that aren't expected to succeed, it seems his value to the team is appropriately represented by EPA

This also is assuming a lot of things about their usage. I haven't seen any snap/situation breakdowns that demonstrates these assumptions.

Not necessarily.  It just means that the staff views Williams as less likely to produce that Jones.  Which is fair and born out by all the evidence.  However, that doesn't tell you anything about the magnitude of the gap between the two.  If Jones is a 9 and Williams a 2, you want Jones to have all the carries.  If Jones is an 8 and Williams is a 7, you want Jones to get all the carries.  The better player is going to see the lion's share of the prime snaps because you play the best you have.  That's not a knock on Williams, it's a credit to Jones.

As far as their usage, there's no good way I can find to look at the breakdown of their non-touch snaps, so unfortunately the pass blocking assumption is difficult to prove out.  But we can look at PFR's advanced splits and get a pretty good idea of the differences in how they were used:

Aaron Jones:

Split	Value	Att	Yds	Y/A	TD	1D	Tgt	Rec	Yds	Y/R	TD	1D	Ctch%	Y/Tgt
Quarter	1st Qtr	57	229	4.0	5	12	24	20	178	8.9	2	5	83.3%	7.4
	2nd Qtr	54	265	4.9	3	14	13	7	48	6.9	0	3	53.8%	3.7
	3rd Qtr	62	303	4.9	7	17	13	8	74	9.3	0	4	61.5%	5.7
	4th Qtr	63	287	4.6	1	12	18	14	174	12.4	1	6	77.8%	9.7
	1st Half	111	494	4.5	8	26	37	27	226	8.4	2	8	73.0%	6.1
	2nd Half	125	590	4.7	8	29	31	22	248	11.3	1	10	71.0%	8.0

Jamaal Williams:

Split	Value	Att	Yds	Y/A	TD	1D	Tgt	Rec	Yds	Y/R	TD	1D	Ctch%	Y/Tgt
Quarter	1st Qtr	17	48	2.8	1	3	7	6	45	7.5	1	2	85.7%	6.4
	2nd Qtr	34	180	5.3	0	7	18	15	75	5.0	2	6	83.3%	4.2
	3rd Qtr	21	105	5.0	0	9	9	8	63	7.9	0	3	88.9%	7.0
	4th Qtr	35	127	3.6	0	4	11	10	70	7.0	2	5	90.9%	6.4
	1st Half	51	228	4.5	1	10	25	21	120	5.7	3	8	84.0%	4.8
	2nd Half	56	232	4.1	0	13	20	18	133	7.4	2	8	90.0%	6.7

Jones carries are very consistent.  He varies from 54 in the 2nd to 63 in the fourth, but overall you can see that he generally sees pretty consistent usage.  Not Williams; he clearly gets more carries in the 2nd and 4th quarter by a fair margin.  Now, there's two reasons a player would see a significant variance tilting towards those two quarters; either they're so elite but fragile you only want to play them in critical situations (nope), or you're using them to run the clock down.  Williams' high YPC in the 2nd quarter and low YPC in the 4th quarter is consistent with this; teams will give you the run at the end of the half, so it's easier for the clock-killer to pick up some yards.  In the 4th, the other team desperately needs the ball back and commits to the run.

Jones:

Split	Value	Att	Yds	Y/A	TD	1D	Tgt	Rec	Yds	Y/R	TD	1D	Ctch%	Y/Tgt
Field Position	Own 1-10	8	38	4.8	0	1	5	4	49	12.3	0	2	80.0%	9.8
	Own 1-20	30	194	6.5	0	6	11	7	71	10.1	0	2	63.6%	6.5
	Own 21-50	110	511	4.6	1	20	28	22	292	13.3	1	11	78.6%	10.4
	Opp 49-20	67	297	4.4	1	15	18	11	75	6.8	1	2	61.1%	4.2
	Red Zone	33	116	3.5	14	15	12	10	42	4.2	1	3	83.3%	3.5
	Opp 1-10	19	37	1.9	11	12	4	4	17	4.3	1	2	100.0%	4.3

Williams:

Split	Value	Att	Yds	Y/A	TD	1D	Tgt	Rec	Yds	Y/R	TD	1D	Ctch%	Y/Tgt
Field Position	Own 1-10	5	16	3.2	0	0	3	3	8	2.7	0	0	100.0%	2.7
	Own 1-20	12	48	4.0	0	2	7	6	36	6.0	0	0	85.7%	5.1
	Own 21-50	43	243	5.7	0	11	19	15	78	5.2	0	4	78.9%	4.1
	Opp 49-20	40	141	3.5	0	7	11	11	89	8.1	0	5	100.0%	8.1
	Red Zone	12	28	2.3	1	3	8	7	50	7.1	5	7	87.5%	6.3
	Opp 1-10	5	7	1.4	1	2	5	5	28	5.6	4	5	100.0%	5.6

Similar thing here.  We can see that Williams sees a nearly perfectly even split of carries, with very few in high leverage situations (near his own goal line or the opponents), while Jones's carries are very heavily split toward what is traditionally the starting point for an offense (21-50) and the red zone.  

Finally we can look at the down and distance splits:

Jones: 

Split	Value	Att	Yds	Y/A	TD	1D	Tgt	Rec	Yds	Y/R	TD	1D	Ctch%	Y/Tgt
Down & Yards to Go	1st & 10	129	675	5.2	4	16	27	19	167	8.8	1	6	70.4%	6.2
	1st & <10	10	11	1.1	3	3	2	2	6	3.0	0	0	100.0%	3.0
	1st & >10	2	5	2.5	0	0	2	2	15	7.5	0	0	100.0%	7.5
	2nd & 4-6	30	131	4.4	2	12	8	5	23	4.6	1	2	62.5%	2.9
	2nd & 7-9	20	109	5.5	1	6	1	1	4	4.0	0	0	100.0%	4.0
	2nd & 10+	21	82	3.9	1	1	12	9	85	9.4	0	2	75.0%	7.1
	3rd & 4-6	3	11	3.7	0	1	7	4	35	8.8	0	3	57.1%	5.0
	3rd & 7-9	0	0		0	0	1	0	0		0	0	.0%	0
	3rd & 10+	0	0		0	0	3	3	41	13.7	0	1	100.0%	13.7
	3rd/4th & short	9	17	1.9	1	5	2	0	0		0	0	.0%

Williams:

Split	Value	Att	Yds	Y/A	TD	1D	Tgt	Rec	Yds	Y/R	TD	1D	Ctch%	Y/Tgt
Down & Yards to Go	1st & 10	57	227	4.0	0	4	15	13	103	7.9	2	5	86.7%	6.9
	1st & <10	2	0	.0	0	0	0	0	0		0	0
	1st & >10	1	3	3.0	0	0	2	2	1	.5	0	0	100.0%	.5
	2nd & 4-6	10	73	7.3	0	4	5	4	26	6.5	1	4	80.0%	5.2
	2nd & 7-9	8	21	2.6	0	0	4	2	15	7.5	0	1	50.0%	3.8
	2nd & 10+	8	53	6.6	0	3	8	8	32	4.0	0	0	100.0%	4.0
	3rd & 4-6	1	3	3.0	0	1	2	2	17	8.5	0	2	100.0%	8.5
	3rd & 7-9	2	15	7.5	0	1	2	1	10	10.0	0	1	50.0%	5.0
	3rd & 10+	4	23	5.8	0	0	4	4	29	7.3	0	0	100.0%	7.3
	4th & 4-6	0	0		0	0	1	1	15	15.0	0	1	100.0%	15.0
	3rd/4th & short	7	21	3.0	0	5	2	2	10	5.0	1	2	100.0%	5.0

So not surprisingly, both saw about half their carries on first and 10.  That would make sense, as you're less likely to run the worse the down and distance gets.  But there's two interesting things to note here: Williams was significantly better and more reliable on 3rd/4th and short and 3rd & 4-6 (he converted 83.3% of his rushes + targets on those downs into 1st downs vs 42.8% for Jones), and Jones really saw the lion's share of the high leverage downs.  Williams saw 2 1st and <10's vs 10 for Jones, and 10 2nd & 4-6 compared to 30 for Jones.  Given that Jones only had about 2x the carries, you wouldn't expect to see him get 5x and 3x the carries of those downs unless he was getting the majority of the run game snaps in those situations.  Conversely, Williams saw 4 runs on 3rd and 10 (clear give up downs) vs 0 for Jones. 

It seems pretty clear to me from my own eye test as well as the stats I can gather that Williams simply didn't see the kind of usage that Jones did.

[AlexGreen#20]

3 minutes ago, MrBobGray said:

Not necessarily.  It just means that the staff views Williams as less likely to produce that Jones.  Which is fair and born out by all the evidence.  However, that doesn't tell you anything about the magnitude of the gap between the two.  If Jones is a 9 and Williams a 2, you want Jones to have all the carries.  If Jones is an 8 and Williams is a 7, you want Jones to get all the carries.  The better player is going to see the lion's share of the prime snaps because you play the best you have.  That's not a knock on Williams, it's a credit to Jones.

As far as their usage, there's no good way I can find to look at the breakdown of their non-touch snaps, so unfortunately the pass blocking assumption is difficult to prove out.  But we can look at PFR's advanced splits and get a pretty good idea of the differences in how they were used:

Aaron Jones:

Split	Value	Att	Yds	Y/A	TD	1D	Tgt	Rec	Yds	Y/R	TD	1D	Ctch%	Y/Tgt
Quarter	1st Qtr	57	229	4.0	5	12	24	20	178	8.9	2	5	83.3%	7.4
	2nd Qtr	54	265	4.9	3	14	13	7	48	6.9	0	3	53.8%	3.7
	3rd Qtr	62	303	4.9	7	17	13	8	74	9.3	0	4	61.5%	5.7
	4th Qtr	63	287	4.6	1	12	18	14	174	12.4	1	6	77.8%	9.7
	1st Half	111	494	4.5	8	26	37	27	226	8.4	2	8	73.0%	6.1
	2nd Half	125	590	4.7	8	29	31	22	248	11.3	1	10	71.0%	8.0

Jamaal Williams:

Split	Value	Att	Yds	Y/A	TD	1D	Tgt	Rec	Yds	Y/R	TD	1D	Ctch%	Y/Tgt
Quarter	1st Qtr	17	48	2.8	1	3	7	6	45	7.5	1	2	85.7%	6.4
	2nd Qtr	34	180	5.3	0	7	18	15	75	5.0	2	6	83.3%	4.2
	3rd Qtr	21	105	5.0	0	9	9	8	63	7.9	0	3	88.9%	7.0
	4th Qtr	35	127	3.6	0	4	11	10	70	7.0	2	5	90.9%	6.4
	1st Half	51	228	4.5	1	10	25	21	120	5.7	3	8	84.0%	4.8
	2nd Half	56	232	4.1	0	13	20	18	133	7.4	2	8	90.0%	6.7

Jones carries are very consistent.  He varies from 54 in the 2nd to 63 in the fourth, but overall you can see that he generally sees pretty consistent usage.  Not Williams; he clearly gets more carries in the 2nd and 4th quarter by a fair margin.  Now, there's two reasons a player would see a significant variance tilting towards those two quarters; either they're so elite but fragile you only want to play them in critical situations (nope), or you're using them to run the clock down.  Williams' high YPC in the 2nd quarter and low YPC in the 4th quarter is consistent with this; teams will give you the run at the end of the half, so it's easier for the clock-killer to pick up some yards.  In the 4th, the other team desperately needs the ball back and commits to the run.

Jones:

Split	Value	Att	Yds	Y/A	TD	1D	Tgt	Rec	Yds	Y/R	TD	1D	Ctch%	Y/Tgt
Field Position	Own 1-10	8	38	4.8	0	1	5	4	49	12.3	0	2	80.0%	9.8
	Own 1-20	30	194	6.5	0	6	11	7	71	10.1	0	2	63.6%	6.5
	Own 21-50	110	511	4.6	1	20	28	22	292	13.3	1	11	78.6%	10.4
	Opp 49-20	67	297	4.4	1	15	18	11	75	6.8	1	2	61.1%	4.2
	Red Zone	33	116	3.5	14	15	12	10	42	4.2	1	3	83.3%	3.5
	Opp 1-10	19	37	1.9	11	12	4	4	17	4.3	1	2	100.0%	4.3

Williams:

Split	Value	Att	Yds	Y/A	TD	1D	Tgt	Rec	Yds	Y/R	TD	1D	Ctch%	Y/Tgt
Field Position	Own 1-10	5	16	3.2	0	0	3	3	8	2.7	0	0	100.0%	2.7
	Own 1-20	12	48	4.0	0	2	7	6	36	6.0	0	0	85.7%	5.1
	Own 21-50	43	243	5.7	0	11	19	15	78	5.2	0	4	78.9%	4.1
	Opp 49-20	40	141	3.5	0	7	11	11	89	8.1	0	5	100.0%	8.1
	Red Zone	12	28	2.3	1	3	8	7	50	7.1	5	7	87.5%	6.3
	Opp 1-10	5	7	1.4	1	2	5	5	28	5.6	4	5	100.0%	5.6

Similar thing here.  We can see that Williams sees a nearly perfectly even split of carries, with very few in high leverage situations (near his own goal line or the opponents), while Jones's carries are very heavily split toward what is traditionally the starting point for an offense (21-50) and the red zone.  

Finally we can look at the down and distance splits:

Jones: 

Split	Value	Att	Yds	Y/A	TD	1D	Tgt	Rec	Yds	Y/R	TD	1D	Ctch%	Y/Tgt
Down & Yards to Go	1st & 10	129	675	5.2	4	16	27	19	167	8.8	1	6	70.4%	6.2
	1st & <10	10	11	1.1	3	3	2	2	6	3.0	0	0	100.0%	3.0
	1st & >10	2	5	2.5	0	0	2	2	15	7.5	0	0	100.0%	7.5
	2nd & 4-6	30	131	4.4	2	12	8	5	23	4.6	1	2	62.5%	2.9
	2nd & 7-9	20	109	5.5	1	6	1	1	4	4.0	0	0	100.0%	4.0
	2nd & 10+	21	82	3.9	1	1	12	9	85	9.4	0	2	75.0%	7.1
	3rd & 4-6	3	11	3.7	0	1	7	4	35	8.8	0	3	57.1%	5.0
	3rd & 7-9	0	0		0	0	1	0	0		0	0	.0%	0
	3rd & 10+	0	0		0	0	3	3	41	13.7	0	1	100.0%	13.7
	3rd/4th & short	9	17	1.9	1	5	2	0	0		0	0	.0%

Williams:

Split	Value	Att	Yds	Y/A	TD	1D	Tgt	Rec	Yds	Y/R	TD	1D	Ctch%	Y/Tgt
Down & Yards to Go	1st & 10	57	227	4.0	0	4	15	13	103	7.9	2	5	86.7%	6.9
	1st & <10	2	0	.0	0	0	0	0	0		0	0
	1st & >10	1	3	3.0	0	0	2	2	1	.5	0	0	100.0%	.5
	2nd & 4-6	10	73	7.3	0	4	5	4	26	6.5	1	4	80.0%	5.2
	2nd & 7-9	8	21	2.6	0	0	4	2	15	7.5	0	1	50.0%	3.8
	2nd & 10+	8	53	6.6	0	3	8	8	32	4.0	0	0	100.0%	4.0
	3rd & 4-6	1	3	3.0	0	1	2	2	17	8.5	0	2	100.0%	8.5
	3rd & 7-9	2	15	7.5	0	1	2	1	10	10.0	0	1	50.0%	5.0
	3rd & 10+	4	23	5.8	0	0	4	4	29	7.3	0	0	100.0%	7.3
	4th & 4-6	0	0		0	0	1	1	15	15.0	0	1	100.0%	15.0
	3rd/4th & short	7	21	3.0	0	5	2	2	10	5.0	1	2	100.0%	5.0

So not surprisingly, both saw about half their carries on first and 10.  That would make sense, as you're less likely to run the worse the down and distance gets.  But there's two interesting things to note here: Williams was significantly better and more reliable on 3rd/4th and short and 3rd & 4-6 (he converted 83.3% of his rushes + targets on those downs into 1st downs vs 42.8% for Jones), and Jones really saw the lion's share of the high leverage downs.  Williams saw 2 1st and <10's vs 10 for Jones, and 10 2nd & 4-6 compared to 30 for Jones.  Given that Jones only had about 2x the carries, you wouldn't expect to see him get 5x and 3x the carries of those downs unless he was getting the majority of the run game snaps in those situations.  Conversely, Williams saw 4 runs on 3rd and 10 (clear give up downs) vs 0 for Jones. 

It seems pretty clear to me from my own eye test as well as the stats I can gather that Williams simply didn't see the kind of usage that Jones did.

Great post

[Shanedorf]

5 minutes ago, MrBobGray said:

It seems pretty clear to me from my own eye test as well as the stats I can gather that Williams simply didn't see the kind of usage that Jones did.

Does MLF sub by series like MM did, or is he subbing within a series ?
I know up tempo plays a role, just curious about the coaching philosophy on subbing RBs and I'm way too lazy to look into it
I'm gonna watch some Packer football tonight and see what I can see

[incognito_man]

2 minutes ago, MrBobGray said:

Jones carries are very consistent.  He varies from 54 in the 2nd to 63 in the fourth, but overall you can see that he generally sees pretty consistent usage.  Not Williams; he clearly gets more carries in the 2nd and 4th quarter by a fair margin.  Now, there's two reasons a player would see a significant variance tilting towards those two quarters; either they're so elite but fragile you only want to play them in critical situations (nope), or you're using them to run the clock down.  Williams' high YPC in the 2nd quarter and low YPC in the 4th quarter is consistent with this; teams will give you the run at the end of the half, so it's easier for the clock-killer to pick up some yards.  In the 4th, the other team desperately needs the ball back and commits to the run.

Another possibility is that Jones is on a pitch count and gets rest after 2-3 series in each half. That would explain higher Williams usage in the 2nd quarter (and lower-leverage 4th qtr). I don't think it's fair to summarize the entire 2nd and 4th quarters as "clock-killing" plays.

5 minutes ago, MrBobGray said:

Jones:

Split	Value	Att	Yds	Y/A	TD	1D	Tgt	Rec	Yds	Y/R	TD	1D	Ctch%	Y/Tgt
Field Position	Own 1-10	8	38	4.8	0	1	5	4	49	12.3	0	2	80.0%	9.8
	Own 1-20	30	194	6.5	0	6	11	7	71	10.1	0	2	63.6%	6.5
	Own 21-50	110	511	4.6	1	20	28	22	292	13.3	1	11	78.6%	10.4
	Opp 49-20	67	297	4.4	1	15	18	11	75	6.8	1	2	61.1%	4.2
	Red Zone	33	116	3.5	14	15	12	10	42	4.2	1	3	83.3%	3.5
	Opp 1-10	19	37	1.9	11	12	4	4	17	4.3	1	2	100.0%	4.3

Williams:

Split	Value	Att	Yds	Y/A	TD	1D	Tgt	Rec	Yds	Y/R	TD	1D	Ctch%	Y/Tgt
Field Position	Own 1-10	5	16	3.2	0	0	3	3	8	2.7	0	0	100.0%	2.7
	Own 1-20	12	48	4.0	0	2	7	6	36	6.0	0	0	85.7%	5.1
	Own 21-50	43	243	5.7	0	11	19	15	78	5.2	0	4	78.9%	4.1
	Opp 49-20	40	141	3.5	0	7	11	11	89	8.1	0	5	100.0%	8.1
	Red Zone	12	28	2.3	1	3	8	7	50	7.1	5	7	87.5%	6.3
	Opp 1-10	5	7	1.4	1	2	5	5	28	5.6	4	5	100.0%	5.6

Similar thing here.  We can see that Williams sees a nearly perfectly even split of carries, with very few in high leverage situations (near his own goal line or the opponents), while Jones's carries are very heavily split toward what is traditionally the starting point for an offense (21-50) and the red zone. 

I don't see this split you are claiming? As a percentage of total carries the breakdown is:

Jones // Williams

3% // 4%

11% // 10%

41% // 37%

25% // 34%

12% // 10%

7% // 4%

These are nearly identical, with the exception of Williams getting more carries in opponent territory before the red zone, and then Jones getting slightly more RZ touches.

11 minutes ago, MrBobGray said:

Conversely, Williams saw 4 runs on 3rd and 10 (clear give up downs) vs 0 for Jones. 

It seems pretty clear to me from my own eye test as well as the stats I can gather that Williams simply didn't see the kind of usage that Jones did.

4 carries is noise amongst nearly 1000 total plays going into these calcs, and the EPA for those 4 plays (since they weren't converted) are effective zero anyway. That's more or less the point here: IF (and I'm still not convinced this is true) Williams was used more as a percent of his snaps on low-percentage plays, why was he used this way? The answer, to me, is obvious: because there's no point is using up Jones' snaps in low% plays because he's far more valuable. If he wasn't significantly more valuable, this assumed discrepancy wouldn't exist.

 

 

[MrBobGray]

9 minutes ago, Shanedorf said:

Does MLF sub by series like MM did, or is he subbing within a series ?
I know up tempo plays a role, just curious about the coaching philosophy on subbing RBs and I'm way too lazy to look into it
I'm gonna watch some Packer football tonight and see what I can see

My memory off rip is that he subbed largely by series, but there was some subbing in the longer series if they took the whole field, which is to be expected.  But I'll see if I can find a way to dig that info out.

[MrBobGray]

14 minutes ago, incognito_man said:

Another possibility is that Jones is on a pitch count and gets rest after 2-3 series in each half. That would explain higher Williams usage in the 2nd quarter (and lower-leverage 4th qtr). I don't think it's fair to summarize the entire 2nd and 4th quarters as "clock-killing" plays.

I doubt it's either/or, but that's sort of immaterial; the point is that overall Williams saw a higher percentage of his carries in those quarters.  A player that sees his usage tilt towards specific scenarios is going to have his average value affected, because he's not playing the same spread that the starter does.

14 minutes ago, incognito_man said:

I don't see this split you are claiming? As a percentage of total carries the breakdown is:

Jones // Williams

3% // 4%

11% // 10%

41% // 37%

25% // 34%

12% // 10%

7% // 4%

These are nearly identical, with the exception of Williams getting more carries in opponent territory before the red zone, and then Jones getting slightly more RZ touches.

 

Your numbers don't seem right here.  Jones saw 33 red zone carries of his 236 total, which is 14%, no 12%.  He also saw 8% of his snaps inside the 10, not 7%.  The breakdown should be:

Jones:  
3.4%
12.7%
46.6%
28.4%
13.9%
8.1%

Williams:

4.6%
11.2%
40.1%
37.3%
11.2%
4.6%

Jones getting significantly more carries inside the opponents 10 and less inside his own 10 alone will acount for a massive difference in EPA production.  Like, I literally can't stress this enough: goal line carries are worth enormous waiting in EPA.  That alone gives you context.  

17 minutes ago, incognito_man said:

4 carries is noise amongst nearly 1000 total plays going into these calcs, and the EPA for those 4 plays (since they weren't converted) are effective zero anyway. That's more or less the point here: IF (and I'm still not convinced this is true) Williams was used more as a percent of his snaps on low-percentage plays, why was he used this way? The answer, to me, is obvious: because there's no point is using up Jones' snaps in low% plays because he's far more valuable. If he wasn't significantly more valuable, this assumed discrepancy wouldn't exist.

That's now how this works.  EPA is a function of the change in EP for a given snap.  A 3rd and 10 snap that isn't converted is worth negative EP, although yes, admittedly not much.  And that's the thing with EPA; it's almost entirely noise if you try to use it for anything other than broad generalizations.  An individual player only gets so many chances for a significant contribution.  Your back-up running back?  He's getting hardly any.  That's the nature of the position.  This is especially true now, as the passing game has changed EP:

Quote

That doesn’t mean EPA is perfect, however. Trey Causey observed that EP has increased over time, which tracks with offenses becoming more efficient. But that means calculations of EP either have to use a small data set from the current year, increasing the statistical noise, or go back to prior years that might undersell EP. Josh Hermsmeyer illustrates many of the challenges calculating EP in a recent post, summarizing EPA as “difficult to calculate, hard to explain, inherently noisy,” and not particularly stable.

As the passing game gets more prevalent, EP becomes heavier weighted toward efficient passing offenses that can strike with big plays.  Jones is absolutely a big play guy, while Williams is not, and yes you're not wrong that that makes him more valuable.  But that doesn't mean Williams is used to his full potential either.

Edited April 8, 2020 by MrBobGray

[craig]

2nd quarter is not "clock-killing" at all.  4th quarter could be. 

Both quarters also have some 2-minute drill, and the 4th sometimes catchup-mode.  Both of those involve an increased attempt to throw, which might suggest increased usage of your blocking back rather than your running back.  As Alex discussed yesterday regarding 3rd-down situations, 2-minute drills and catchup-mode might be analogous in being ill-suited to optimal success.  

 

[incognito_man]

13 minutes ago, MrBobGray said:

Your numbers don't seem right here.  Jones saw 33 red zone carries of his 236 total, which is 14%, no 12%.  He also saw 8% of his snaps inside the 10, not 7%.  The breakdown should be:

he had 267 total carries, not 236. I think you arrived erroneously at 236 somehow

[MrBobGray]

3 minutes ago, incognito_man said:

he had 267 total carries, not 236. I think you arrived erroneously at 236 somehow

Uh....where are you seeing this?  PFR and NFL.com both have him at 236.