What's new
Fantasy Football - Footballguys Forums

Welcome to Our Forums. Once you've registered and logged in, you're primed to talk football, among other topics, with the sharpest and most experienced fantasy players on the internet.

Instinctive's Most Divergent Projections - Upside players to take a second look at (1 Viewer)

Instinctive

Footballguy
Alright SP fam. It's been a while since I have done this exercise, what with finishing grad school and having a fairly time-intensive job, but I used a month of time off this year to get back to my old habits and create a full projection set for all 32 teams' QBs, 1-3 RBs, 2-3 WRs, and a TE. A couple teams I did with multiple QBs (SFO, CHI), some I only did one RB (e.g., TEN), some I did no TEs. For all teams, I project out total passing game stats and running game stats based on attempts/game and efficiency and then divvy it up amongst the fantasy players plus one player per team called "Other runners/receivers," so the totals make sense. 

The efficiency stats are created through a combination of historical data for an individual player as well as their coach + NFL averages, focused on the last 3 years unless I see a reason for exception (rare but it happens). 

I'm not great with college players and don't have a massive robust database or anything, so that's mostly what I think + the NFL averages for rookies in similar draft capital.

All that said: I tend to find my big outliers vs consensus and such with this exercise, and those guys become the people I revolve my drafts around. 14-4 years ago, it was the basis for about a decade of success in high stakes leagues that helped me pay for college. Some of these projections end up WAY OUT THERE - and rarely dead on. But they help really identify opportunities and I trust them in that sense. ALL of these are 16 game projections. They are NOT adjusted for the 17th game. All my historical data is in 16 game sets, so I stuck with that so I can compare them to old results. Also, NONE of this accounts for PPR. Because PPR is a stupid way to play. Come at me. 🔥 🔥 🔥 

Without further ado, here are the highlights (no TEs - they pretty much ended up mirroring most rankings):

Quarterbacks:

I have two players that appear to be outliers as measured against FBG: Stafford high at #3 and Josh Allen low at #7.

1. Matt Stafford: 656 attempts, 5445 yards, 33 TDs, 13 INTs, 112 rush yards

The major driver here is Stafford's attempts, combined with an 8.3 YPA. This would be below the 8.6 career high YPA for Stafford and within striking distance of his recent top years with Bevell in the high 7s, and the 3rd most attempts he's ever had. It's just barely more attempts for McVay than the 2019 Rams. Without Akers, I think they have to tilt pass heavy and use the short passing game as a de facto run game. You'll see this key factor come up again later on. Includes a little regression to a more normal TD rate of 4%.

2. Josh Allen: 560 attempts, 4200 yards, 25 TDs, 11 INTs, 572 rush yards, 7 rush TDs

A little regression to 7.5 YPA here, and a very slight downtick in pass attempts (1/game), and a regression to a more normal TD rate of 4.5%. I am giving him a bit better YPC on the ground game and maintaining his fairly historic rate of rushing for TDs. Story here is pretty simple, and I wouldn't be that surprised if this is one of the bigger misses I have and Josh Allen is just that talented. But I think they had excellent playcalling and a pretty weak passing defense schedule, and those things are combined with a tougher slate this year and me just not 100% buying it.

Running Backs:

Some interesting choices here, with my #1 as Dalvin Cook, Antonio Gibson up at RB5, and CEH at RB7. I don't really have many guys down lower than consensus by much, and Cook is in basically everyone's top three anyway, so we'll look at Gibson and Clyde.

1. Antonio Gibson: 256 for 1254 yards, 12 TDs, 77 targets, 61 recs, 445 yards, and 2 more TDs

I've got Gibson with a big ole workload coming his way, and he maintains a crisp 4.9 YPC. Less than double the receiving load, and in line with roughly his last 8 games. I'm also taking into account the turf toe and improved QB play this year to give him more leads (thus more carries) and more efficiency (uncommon, but certainly both numbers well within realistic range).

2. Clyde Edwards-Helaire: 240 for 1224 yards, 10 TDs, 74 targets, 48 recs, 481 yards, and 2 more TDs

Post-hype sleeper syndrome for the big time here. Even FBG has this guy at RB16?!?!? Here's what I see in Kansas City; I see a highly pedigreed back with build and measurables similar to previous monster Andy Reid fantasy backs, I see a new and significantly improved offensive line, I see a fairly complex offensive system that now has a non-COVID offseason of work, and I see a ton of TD opportunity...plus he did it for like the first 6 games last year anyway! I think I may be underestimating TDs and he could lead the league in total TDs. 

Wide Receivers:

Headline here is that my deal spit out Cooper Kupp as WR1 this year, then Mike Williams at WR7 (Keenan at WR6 too). I have Tyreek Hill all the way down at WR12 and Diggs at WR15 (both seem to be top 3 everywhere else). Diggs is low because of the same regression in Allen (minor) and a step back in targets from his 166 because Sanders is in town and Davis is good (low is still 140 targets for 98 recs, 1246 yards, and 8 TDs). Hill is purely TD regression, I have him at 9 instead of 15 (his career TD rate would be 10 TDs) as CEH gets to run a few more in. So we will focus on the two upside guys.

1. Cooper Kupp: 184 targets, 132/1690/13 line

Keeps his TD rate at 2018 and 2019 levels (so not a crazy outlier) and his YPT is jumping back up between 2018 (9.4) and 2019 (8.2) levels. This assumption is based on Kupp getting to play with Stafford instead of Goff. Most importantly, if the Rams throw the ball 650 times, Kupp's 28% target rate is the next biggest assumption. This is WAY out there, and I have Woods with a pretty big line on a 23% target share, but I'm seeing Kupp's opportunity to be huge here. I WANT pieces of the Rams this year.

2. Mike Williams: 150 targets, 90/1391/10 line

oooooo baby this guy. Everyone knows the issue is injuries. Guess what? They play football. Guys get hurt. And then they don't. If Herbert is as good as many think, or as decent as I think he'll be, there's a lot of room in this new offense with an improved line to boot. Parham seems to be well-liked, but he's not a target hog. I'm giving our boy Mike Williams 25% of the targets (less than Keenan, but admittedly still a lot, in part because they don't have much else outside of him, Allen, and Ekeler). 

Discussion time - suggested topics:

So, let's see what people think here. Too many targets for the big wideouts? Where do you think those shares are for those teams? Who's getting the targets instead? Are YPT or YPC or YPA numbers off for folks in a way you can point to? Do you think any of these offenses are going to have a different amount of plays on the run or pass side?

Please don't bring up health. We get it, some of you think some guys are injury prone. My greatest strength in this hobby for over a decade has been assuming that injuries are distributed equally and acting accordingly.

 
Last edited by a moderator:

Ilov80s

Footballguy
That is a lot of yards for Kupp. Is that the most yards WR is projected for or are there players higher? 

 

Instinctive

Footballguy
That is a lot of yards for Kupp. Is that the most yards WR is projected for or are there players higher? 
It's my highest. Ridley is close behind and Hopkins and Jefferson I have above 1500. 

I have found that my average projection is about 8% high across basically all positions, and it seems to be due to injury. But since I can't (and don't think anyone can) predict injuries, I go with everyone at 16 unless I see suspension or in the QB cases I try to predict rookies coming in based on byes and schedule.

 

Ilov80s

Footballguy
It's my highest. Ridley is close behind and Hopkins and Jefferson I have above 1500. 

I have found that my average projection is about 8% high across basically all positions, and it seems to be due to injury. But since I can't (and don't think anyone can) predict injuries, I go with everyone at 16 unless I see suspension or in the QB cases I try to predict rookies coming in based on byes and schedule.
16 or 17 now? 

 

Anarchy99

Footballguy
It's my highest. Ridley is close behind and Hopkins and Jefferson I have above 1500. 

I have found that my average projection is about 8% high across basically all positions, and it seems to be due to injury. But since I can't (and don't think anyone can) predict injuries, I go with everyone at 16 unless I see suspension or in the QB cases I try to predict rookies coming in based on byes and schedule.
Are you just assuming everyone will miss a game? Or are you projections not based on a 17 game schedule?

 

SeniorVBDStudent

Footballguy
As pass attempts go from 500 to 600 and beyond, they become less efficient.  See, for an extreme example, Big Ben last year.  So while I'm a fan of the Rams passing game this year, I think you need a modulating parameter in your model.  5445 yards (or 6000 in a 17 game season) would be very impressive.

Edited to add one of the interesting subtopics with Goff->Stafford is what are the proportionate gains between Kupp, Woods, Higbee, etc.  It doesn't take long in the FF innerwebs who make the case for each to receive the lions share of the gain.

 
Last edited by a moderator:

rschroeder1

Footballguy
Thank you for taking the time to put this together.

Even presuming some regression for the Rams' defense, I'm curious if they are going to need to be pass-happy enough to generate those gaudy numbers for Kupp.  I'm sure McVay wants to unleash the hounds, but he has always struck me as someone who understands practicality as well.  I have a lot of respect for McVay willfully shifting to grind-it-out offenses that the Rams have used to maximize whatever they could squeeze out of Goff and remain competitive.

I don't have any comment on the numbers for CEH, but I find the dichotomy on him to be fascinating.  I've read a number of people on this board and elsewhere claiming the exact opposite in regard to his build, measurables, pedigree, etc.  I don't think I have any skill in evaluating those types of things, but I've been struck with the almost brash confidence in declaring that the Chiefs front office and Andy Reid essentially got that pick wrong.  The pervasiveness of those attitudes (anecdotally) strikes me as a buy-low opportunity.

 

BobbyLayne

Footballguy
Thanks for sharing @Instinctive - I’ve noted you’ve been working on these awhile and I like your overall methodology - starting with pace, run/pass breakdown at the team level first.

Curious how you are handling GP for running backs? The only guys in the top.30 in PPR who played 16 last year were Henry and 4 satellite backs (McKissic, Hines, Edmonds, Gio.) Are you just projecting everyone at 17? Or presuming most will play 15-16?  

 

Instinctive

Footballguy
Are you just assuming everyone will miss a game? Or are you projections not based on a 17 game schedule?


16 or 17 now? 
I'm just doing everything on 16 games because that's what I had and I wanted to be able to check it with my past results. If you need 17 because you're doing something else, divide them all by 16 and multiply it by 17.

In the future I'll slowly adjust to more rate stuff to contextualize the 17 game season but I didn't want to go back and adjust everything else I had. c'est la vie. 

 

Instinctive

Footballguy
Thanks for sharing @Instinctive - I’ve noted you’ve been working on these awhile and I like your overall methodology - starting with pace, run/pass breakdown at the team level first.

Curious how you are handling GP for running backs? The only guys in the top.30 in PPR who played 16 last year were Henry and 4 satellite backs (McKissic, Hines, Edmonds, Gio.) Are you just projecting everyone at 17? Or presuming most will play 15-16?  
I projected everything at 16. Since I more or less use them just to rank in many respects, and I'm not willing to guess who will miss how many, I get the same effect as putting them all at "14.2 games" or whatever the average is. 

When I shift this to VBD/auction values, I'll take the total number of games played by the top 30 and divide it by 30 and use that as the average for them and for WRs/TEs, then just assume full seasons for QBs. 

 

Ilov80s

Footballguy
I agree with projecting everyone without injuries. It is impossible to know who will miss games, how many games they miss, etc. Project everyone equally and then when ranking, you can adjust players based on any gut feeling you have regarding health. 

 

Tanner9919

Footballguy
I'll sign up right now if you 're sure Stafford gets 5445 yards, 33td. that's fantastic! If Gibson gets what you're projecting I'd be thrilled.

for CEH, what is the history of teams in the year after losing the SB ?? Not good at all. KC has a revamped o-line, 5 new starters? going to take time to coalesce..Any Reid's teams typically implode after 3-4 seasons. I'm not saying it's going to happen in KC this season, but I'd be a bit worried that the team as a whole takes a giant step backwards. Even the mighty Seahawks teams of recent memory struggled after 3-4 seasons at the top. That KC thumping in the SB is still being felt. 

CEH was hurt last season, and when he played he was very ho-hum. You're projecting 421 more rushing yards, and many more TDs. your math doesn't add up, you're expecting 240 carries at 5.1 ypc average to get to 1224 yards, but CEH only made a paltry 4.4 ypc average last season. If he goes off at the same 4.4/carry, you'll need 278.18 carries to get to 1224 yards. there's no way he improves to 5.1 yards per carry, he doesn't look even remotely capable of that average in an NFL uniform.that would put him in Nick Chubb territory. he's nowhere near the RB Chubb and some of the other top 8 guys are. 

CEH is a decent player but let's not go crazy. you're more likely to see 900-1000 yards 8 TD than you are to see 1224 rushing yards, 5.1 per carry on your projected 240 carries. and doubtful he gets anything close to 240 carries. 200-225 maybe?

 

Biabreakable

Footballguy
Are you just assuming everyone will miss a game? Or are you projections not based on a 17 game schedule?
He said he projected for 16 games since that is the foundation that he is working from, based on the last 3 seasons which were 16 game seasons 

1/17 is 5.8% so his projections being 8% high across the board actually works out pretty well considering the extra game.

 

Biabreakable

Footballguy
I appreciate the effort and thanks for sharing some of the more surprising results you came up with instinctive.

I think your projection for Kupp is way too high. Have you noticed how many good WR the Rams have now?

I tend to agree with you about the Rams passing attempts going up this season, I think they are built for that and this is their plan, even without Akers being injured. I just think it will be more spread out than that and so Kupp wont get that many targets.

I also disagree with your projection for Mike William's, but that is more because I dont think Mike William's is good enough to command that many targets.

 

humpback

Footballguy
He said he projected for 16 games since that is the foundation that he is working from, based on the last 3 seasons which were 16 game seasons 

1/17 is 5.8% so his projections being 8% high across the board actually works out pretty well considering the extra game.
I'm reading this as the opposite- the numbers he gave are for 16 games, if you want to extrapolate that out for 17 you need to multiply by 1.0625 so his are more like 14-15% higher across the board than consensus.

 

Dr. Octopus

Footballguy
Any Reid's teams typically implode after 3-4 seasons.


Andy Reid coached 14 seasons in Philadelphia. Eagles lost in the Super Bowl in Year 6 and went to NFC Championship Game in Year 10.

Reid has coached KC for 8 seasons, winning a Super Bowl in Year 7 and losing in the Super Bowl in Year 8.

 
Last edited by a moderator:

zamboni

Footballguy
Any Reid's teams typically implode after 3-4 seasons.


Andy Reid coached 14 seasons in Philadelphia. Eagles lost in the Super Bowl in Year 6 and went to NFC Championship Game in Year 10.

Reid has coached KC for 8 seasons, winning a Super Bowl in Year 7 and losing in the Super Bowl in Year 8.
Yeah, that’s a bad take. Reid is a really good, proven coach and I bet most coaches wish they had a quarter of his success rate.

 

zamboni

Footballguy
1. Cooper Kupp: 184 targets, 132/1690/13 line

Keeps his TD rate at 2018 and 2019 levels (so not a crazy outlier) and his YPT is jumping back up between 2018 (9.4) and 2019 (8.2) levels. This assumption is based on Kupp getting to play with Stafford instead of Goff. Most importantly, if the Rams throw the ball 650 times, Kupp's 28% target rate is the next biggest assumption. This is WAY out there, and I have Woods with a pretty big line on a 23% target share, but I'm seeing Kupp's opportunity to be huge here. I WANT pieces of the Rams this year.
I admire the optimism, but I have a hard time he'll get anywhere close to that with Woods there, not to mention, D-Jax, Jefferson, D-Jax, Higbee and Henderson. Only 4 guys in the history of the league have had 132+ receptions in a season (Thomas, Harrison, A. Brown, Julio). If he got anywhere near that, he'd be more like a a DD-Kupp than a C-Kupp.

 

Instinctive

Footballguy
I admire the optimism, but I have a hard time he'll get anywhere close to that with Woods there, not to mention, D-Jax, Jefferson, D-Jax, Higbee and Henderson. Only 4 guys in the history of the league have had 132+ receptions in a season (Thomas, Harrison, A. Brown, Julio). If he got anywhere near that, he'd be more like a a DD-Kupp than a C-Kupp.
Lol :)  nice one

Yeah - I don't think he'll actually hit that. The model I run is just that - a model. It has outliers, and it is only so good at contextualizing things because I'm not a programmer or data scientist, so it's pretty rudimentary. 

I doubt Kupp actually finishes as #1 WR with 1600 yards and over 130 receptions...but I do trust that it means we're significantly under-ranking the guy. A regression to a typical TD rate, and a similar target share as the last 3 years would be monstrous if A) Stafford's better enough than Goff to increase YPT for receivers and B) Akers out + Stafford being better than Goff means they throw just a little bit more than they did in 2019 (when they launched it 620+ times). 

If I just look at those pieces and then add in the fact that McVay seems like the kind of guy who would use short slants and slots and bubble screens as run game stand-in plays...I think it makes sense that I'll pay above AV or draft a round or two above ADP to get value here and make sure he's on my team.

 

greenmountaingoat

Footballguy
I'm a relative newbie but here's my take:

Doing your own projections I'm sure historically gave you a big advantage but my hunch is that there's been a proliferation (in recent years) of data and scouting driven analysis that is catching up to the potential value in having good scouting and data analysis. 

Just a quick question, who's the last WR you'd take before him? 

 

zamboni

Footballguy
Lol :)  nice one

Yeah - I don't think he'll actually hit that. The model I run is just that - a model. It has outliers, and it is only so good at contextualizing things because I'm not a programmer or data scientist, so it's pretty rudimentary. 

I doubt Kupp actually finishes as #1 WR with 1600 yards and over 130 receptions...but I do trust that it means we're significantly under-ranking the guy. A regression to a typical TD rate, and a similar target share as the last 3 years would be monstrous if A) Stafford's better enough than Goff to increase YPT for receivers and B) Akers out + Stafford being better than Goff means they throw just a little bit more than they did in 2019 (when they launched it 620+ times). 

If I just look at those pieces and then add in the fact that McVay seems like the kind of guy who would use short slants and slots and bubble screens as run game stand-in plays...I think it makes sense that I'll pay above AV or draft a round or two above ADP to get value here and make sure he's on my team.
Fair enough. I guess I'm not sure about the target share with more mouths to feed versus a few years ago, when it had been basically all Cupp and Woods (and Gurley getting some as well). 

 

Instinctive

Footballguy
zamboni said:
Fair enough. I guess I'm not sure about the target share with more mouths to feed versus a few years ago, when it had been basically all Cupp and Woods (and Gurley getting some as well). 
It feels like there are fewer mouths to me. Everett gone. Reynolds gone. What am I missing?

 

Hot Sauce Guy

Footballguy
I’ve been targeting Stafford in mocks whenever I miss on the top 5 QB. 

This is the best OL & most complete weapons set he’s had to work with. While the perception seems to be that he’s on the downside of his career, IMO he’s been a byproduct of a fading offense in DET. 

And with the injury to Akers, I expect McVay to get even more pass happy. Bonus for Stafford. 

Even if he’s 85% of the projections above he’s likely to be a top 10 QB, and at a bargain ADP. 

 

Hot Sauce Guy

Footballguy
They signed free agent WR Desean Jackson and drafted WR Atwell. Second year WR Jefferson could also be a bigger factor.
Higbee should blossom as the TE1 and has sleeper appeal, and I agree that Jefferson is worth noting. He’s reportedly looked really good this preseason. He’s an end of draft target for me. 

 

greenmountaingoat

Footballguy
Higbee should blossom as the TE1 and has sleeper appeal, and I agree that Jefferson is worth noting. He’s reportedly looked really good this preseason. He’s an end of draft target for me. 
Agreed. I picked up Van in the puppy a month ago based (forget which pod it was) regarding his likely deep target role and the fragility of Jackson.  Still not sure he makes the cut for me in a 12 man league but he's definitely on my watch list. Guess I need some camp notes.

 

Biabreakable

Footballguy
humpback said:
I'm reading this as the opposite- the numbers he gave are for 16 games, if you want to extrapolate that out for 17 you need to multiply by 1.0625 so his are more like 14-15% higher across the board than consensus.
I dont understand what you are talking about.

 

Ilov80s

Footballguy
zamboni said:
Yeah, that’s a bad take. Reid is a really good, proven coach and I bet most coaches wish they had a quarter of his success rate.
I think most fan bases are kicking themselves for not having hired Andy Reid the moment he became available. 

 

Ilov80s

Footballguy
Since you posted guys who project to surprisingly good seasons, do you have any players projected to surprisingly low projections that you are avoiding?

 

Instinctive

Footballguy
Since you posted guys who project to surprisingly good seasons, do you have any players projected to surprisingly low projections that you are avoiding?
Josh Allen

Tyreek Hill

Stefon Diggs

All probably good, but I think towards the bottom of the top 12 at their positions. So I won't have any of them. I also notice myself fading Dallas somehow (other than Gallup). 

 

Ilov80s

Footballguy
Josh Allen

Tyreek Hill

Stefon Diggs

All probably good, but I think towards the bottom of the top 12 at their positions. So I won't have any of them. I also notice myself fading Dallas somehow (other than Gallup). 
Is Tyreek down because of a dip in his efficiency, the offenses efficiency or an increase in competition for targets? Is it Byron Pringle season? 

 

mcintyre1

Footballguy
Yeah I thought I'd sneak that in there. HAHA. 

I mean maybe PPFD but beyond that...
Ppfd is the way of the future


I feel like having points per first down is just weird... Say you give 1 point for earning a first down -- that means that a 1 yard run is potentially worth more than a 9 yard catch / 9 yard run just because it happened 1 yard away from the first down marker. Maybe I'm not understanding how it would work, though. Seems less fair than just giving a point per reception in that instance, though.

 

greenmountaingoat

Footballguy
You either believe Josh Allen's a guy that unlocked his potential and now can complete passes effectively or not. Everyone knows we probably just saw his best season but he was pretty consistent over a whole season, however the ceiling is still way high. The arm strength is insane and Diggs is a different receiver than he had before this past season. 

Saying you believe in some negative regression makes sense but expecting him to revert back to a low completion percentage guy seems like not a popular sharp analyst position. As a NE fan, we know that they already have two tougher games on the schedule than they did last year, Miami's D is good and the Jets might be better so baking in something of a setback makes sense. 

I mean I'm not taking a shot on him at ADP so I agree in some sense but Diggs? You're kind of taking the field by saying low end WR1 for him and Hill and in that way you're statement is hedged. I'm sorry if I'm coming across as harsh but I think we would have more fun picking on you if you posted your draft board :)  

 

Instinctive

Footballguy
I feel like having points per first down is just weird... Say you give 1 point for earning a first down -- that means that a 1 yard run is potentially worth more than a 9 yard catch / 9 yard run just because it happened 1 yard away from the first down marker. Maybe I'm not understanding how it would work, though. Seems less fair than just giving a point per reception in that instance, though.
I mean I'd do it based on success rate if I could but that seems impractical. 

A 9 yard catch is useless on your own 25 yard line with 13 seconds left in the half. A one yard run for a first down is incredibly useful with a stacked box and 4th and 1 with 1:20 left in the game.

A reception has no inherent value other than the yards gained. And we award points for the yards.

 

Instinctive

Footballguy
You either believe Josh Allen's a guy that unlocked his potential and now can complete passes effectively or not. Everyone knows we probably just saw his best season but he was pretty consistent over a whole season, however the ceiling is still way high. The arm strength is insane and Diggs is a different receiver than he had before this past season. 

Saying you believe in some negative regression makes sense but expecting him to revert back to a low completion percentage guy seems like not a popular sharp analyst position. As a NE fan, we know that they already have two tougher games on the schedule than they did last year, Miami's D is good and the Jets might be better so baking in something of a setback makes sense. 

I mean I'm not taking a shot on him at ADP so I agree in some sense but Diggs? You're kind of taking the field by saying low end WR1 for him and Hill and in that way you're statement is hedged. I'm sorry if I'm coming across as harsh but I think we would have more fun picking on you if you posted your draft board :)  
I don't think I said he'll become a low completion percentage guy? I expect some pretty major touchdown regression, and the team to run a little more. Some of the passes probably lower aDOT because the super high YPA last year doesn't seem sustainable to me. Maybe I'm wrong. 

I mean, I think Diggs will be close to WR 10 than to WR2. I don't really care if you think it sounds hedged or not. It's a good enough thought for me to know I won't have him on my teams. I only need information that's good enough to make a decision. Similarly, Kupp's not likely getting 1600 yards, but I think he's gonna have a great year and he'll be on all my teams as a result.

I don't really see any of this as harsh. Some posts I am sure will be uninformed or I'll find dumb (none yet), some will be unhelpful, but most of the SP folks who care about winning will engage and I'll learn some stuff and it will help me make decisions. I'm a selfish guy. Not trying to help you (or anyone) out, trying to drive some good discussion I'll learn from.

 

facook

Footballguy
Instinctive said:
Lol :)  nice one

Yeah - I don't think he'll actually hit that. The model I run is just that - a model. It has outliers, and it is only so good at contextualizing things because I'm not a programmer or data scientist, so it's pretty rudimentary. 

I doubt Kupp actually finishes as #1 WR with 1600 yards and over 130 receptions...but I do trust that it means we're significantly under-ranking the guy. A regression to a typical TD rate, and a similar target share as the last 3 years would be monstrous if A) Stafford's better enough than Goff to increase YPT for receivers and B) Akers out + Stafford being better than Goff means they throw just a little bit more than they did in 2019 (when they launched it 620+ times). 

If I just look at those pieces and then add in the fact that McVay seems like the kind of guy who would use short slants and slots and bubble screens as run game stand-in plays...I think it makes sense that I'll pay above AV or draft a round or two above ADP to get value here and make sure he's on my team.


I was curious on this one as well.  Are you willing to share where you've got Woods?  Most rankings have Woods/Kupp within a few spots of each other in the mid teens are so.  Do your projections have Woods much lower than that, or about the same with the Kupp explosion?

 

Instinctive

Footballguy
I was curious on this one as well.  Are you willing to share where you've got Woods?  Most rankings have Woods/Kupp within a few spots of each other in the mid teens are so.  Do your projections have Woods much lower than that, or about the same with the Kupp explosion?
151 targets, 103/1328/8 - my model's WR13

 

wgoldsph

Footballguy
I feel like having points per first down is just weird... Say you give 1 point for earning a first down -- that means that a 1 yard run is potentially worth more than a 9 yard catch / 9 yard run just because it happened 1 yard away from the first down marker. Maybe I'm not understanding how it would work, though. Seems less fair than just giving a point per reception in that instance, though.
A WR can catch a ball and fall 9 yards behind the los and end up with positive fantasy points, while a first down is not only always a positive play, it also has much greater impact on the offense as a whole no matter when it occurs.  As @Instinctive

said above, that one yard plunge against a stacked box to pick up the 4th and 1 is incredibly important, and should receive its due in fantasy.

I like ppr and ppfd, but if we're talking about which accurately measures the real impact of players on the field, I would be inclined to say first downs are more important than a handful of empty yards that lead to nowhere, let alone the ability to make an uncontested catch on a swing pass for 0 yards.

 

Biabreakable

Footballguy
His projections are ~8% higher on average despite being based on 16 games. If he extrapolated them out to 17 games they would be even more above consensus.
Ok I think I understand where you are coming from now.

I haven't seen any numbers, just going off of what instinctive said about them.

He hasn't accounted for the extra game because he is using 16 game seasons for his model. I see some merit in doing it that way so that calculations have some alignment.

He could divide all of the team numbers by 16 to find the per game numbers and then multiply those number by 17 which would be an even higher result as you mentioned.

However he didn't do that so my statement still stands here, that being 8% higher than historical averages seems fine to me since the extra game almost makes that even.

 

humpback

Footballguy
Ok I think I understand where you are coming from now.

I haven't seen any numbers, just going off of what instinctive said about them.

He hasn't accounted for the extra game because he is using 16 game seasons for his model. I see some merit in doing it that way so that calculations have some alignment.

He could divide all of the team numbers by 16 to find the per game numbers and then multiply those number by 17 which would be an even higher result as you mentioned.

However he didn't do that so my statement still stands here, that being 8% higher than historical averages seems fine to me since the extra game almost makes that even.
This isn't all that important but I'll try one more time since we're still not on the same page- I could be wrong but I took his statement that his projections are 8% high across the board to mean higher than consensus projections for this year, not historical. The only reason he kept his projections at 16 games is to make it comparable to previous years because that's what was used, but if you want to make it comparable with consensus for this year, you'd have to multiply by 17 and divide by 16, making his projections even higher than they are. The extra game does the opposite- it doesn't almost make his projections even with consensus, it makes them almost 15% above them.

For instance, take Kupp's numbers, multiply by 17 and divide by 16 and he'd project him at 140 catches on 195 targets for almost 1800 yards and 14 TDs over a full 17 game season.

 

LawFitz

Footballguy
Really appreciate threads like this one for the discussion they catalyze, but have never been a fan of heavy reliance on projections - they seem tedious and subject to too much input error, subjectivity and clouding of the big(ger) picture.

E.g. should we really be mean regressing a possible breakout star like Josh Allen? Maybe we should, but my eye test and gut are pretty high on what we saw from him last year, as a 3rd yr QB with a wunderkind offensive coach. He looked like Aaron Rodgers 2.0 to me. I need to hear more than just mean regression to sway that impression.

I keep it more basic with fantasy football analytics.

1. Who are likely to be the best teams?... Based on:

2. Who are the best coaches, and as importantly, which newer coaches are ascending?

3. Who are the best QBs/ascending QBs?

4. Who are the best/asc OLs?

5. Who are the best/asc Ds?

From there I plug and play the most talented RBs/WRs/TEs, especially those with discounted ADPs.

 

Biabreakable

Footballguy
This isn't all that important but I'll try one more time since we're still not on the same page- I could be wrong but I took his statement that his projections are 8% high across the board to mean higher than consensus projections for this year, not historical. The only reason he kept his projections at 16 games is to make it comparable to previous years because that's what was used, but if you want to make it comparable with consensus for this year, you'd have to multiply by 17 and divide by 16, making his projections even higher than they are. The extra game does the opposite- it doesn't almost make his projections even with consensus, it makes them almost 15% above them.

For instance, take Kupp's numbers, multiply by 17 and divide by 16 and he'd project him at 140 catches on 195 targets for almost 1800 yards and 14 TDs over a full 17 game season.
Ah yes I get your point now.

I guess it's not clear to me what the 8% is higher than.

If its higher than other projections based on 17 games then yes I agree with you.

If it's based on the historical average of the last 3 seasons (which is what I thought he meant) then an adjustment might not be needed much.

 

Instinctive

Footballguy
Really appreciate threads like this one for the discussion they catalyze, but have never been a fan of heavy reliance on projections - they seem tedious and subject to too much input error, subjectivity and clouding of the big(ger) picture.

E.g. should we really be mean regressing a possible breakout star like Josh Allen? Maybe we should, but my eye test and gut are pretty high on what we saw from him last year, as a 3rd yr QB with a wunderkind offensive coach. He looked like Aaron Rodgers 2.0 to me. I need to hear more than just mean regression to sway that impression.
Yeah I mean I wouldn't rely solely on this either and I made it. It's another input in my toolkit.

Aaron Rodgers' career TD % is below what Allen had last year. The point of regressing to something of a typical mean on TD% or YPA is that sometimes weird #### happens. Now, maybe he is great, and that 6.5% (which is not outrageous) from last year sticks. But TDs are notoriously fluky. So I think it falls back to earth some.

 

Biabreakable

Footballguy
@Biabreakable @humpback

I imagine that extra you're seeing is because I don't predict any injuries. Across the board. Whereas FBG, for example, has about a game to two games missing from every player.
Ok.

In this case then I agree with humpbacks point that the FBG projections are also for 17 games while yours were 16 correct?

So your projections are also high because of one less game.

 

Users who are viewing this thread

Top