Let's examine what we think about these topics as we head into Week 6.
- Aging Tight Ends with Starter Production
- Pre-emptive Pickups
- Revisiting Jordan Howard-Miles Sanders.
- Second-Half Surprises
Aging Tight Ends with STarter Production
Matt Waldman: Greg Olsen, Jason Witten, and Delanie Walker are in their mid-30s and ranked 8th, 9th, and 11th in PPR formats. Why do we write off older players at the position when there's a track record of quality performances from aging tight ends?
Dan Hindery: I wrote briefly about this topic in my dynasty trade value article last week. It seems like age-based regression does not hit tight ends nearly as hard as it does wide receiver and running backs. The aging guys ranking as low-end TE1s is both a testament to their staying power and veteran savvy and an indictment of many of the younger tight ends we were excited about. These three are averaging 9-10 PPG (PPR scoring), so it isn’t like they are providing a big advantage over the competition. But 10 PPGs probably doesn’t sound too bad for those who drafted O.J. Howard.
Daniel Simpkins: Maybe I’m going too deep with this answer, but speaking as a therapist who works primarily with geriatrics these days, I think our society is one that is ageist. Being old is undesirable and youth is paramount. That attitude filters all the way down into our fantasy football strategy. We’re afraid to fail because we picked a player that is too old to perform. We would rather take the younger player that is an unknown quantity because there is not as much of a stigma attached if we fall short in that way. This doesn’t just extend to the tight end position. One of the questions that we will cover later in this discussion will highlight that same dynamic.
Waldman: Maybe it's because I'm now nearly twice the age I was when I began playing fantasy football, but I find myself more trusting of experience and skill over youth and energy when skill shows up as questionable on tape. While I jokingly associate it with my age, the truth is that I long-gravitated to older players because they tend to process the game faster and they have a more robust library of experience and skill to draw upon that doesn't always require the athletic ability of a 25-year-old to execute.
Haseley: All three of those tight ends have two touchdowns in five games. If not for those, they would be outside of the Top 12, perhaps Top 15.
Waldman: Haseley, if not for the fact that you could spell, you might not be a writer. You're bringing me down, man...
Haseley: You can't spell, and they keep you around here, Waldman—I follow you on Twitter.
Waldman: I plead autocorrect...so was there more to your argument?
Haseley: Veteran tight ends generally know their offense well and are a key piece to the team’s success, especially in the red zone. It’s no big surprise that all three have multiple scores. That’s one of the benefits of having a veteran tight end on your team, the tendency to score. Tight ends who have been in the league awhile are also in tune with how the offense operates because they need to know blocking schemes, protection calls, play calls, etc.
Often, a veteran tight end is a vocal leader in the huddle, as well as the sidelines and practice sessions. Being so involved in the offense, their number is often called and the talented ones are the ones who benefit. We’ve seen that from Olsen, Witten, and Walker this season. Unfortunately, they are also the ones who are called upon when an extra blocker is needed in games where the pass rush is heavy and concerning. Injuries to offensive linemen can slow down their production as a result of this. We’ve seen it in each of the last two games with Olsen and Witten.
Waldman: If I was patient, I would have realized I was going to agree with much of this answer.
Haseley: Old and foolish. Great combination there...
Waldman: At least I'm still pretty.
Bob Henry: Delusional fantasies may be your saving grace, Wildman. As for this topic, the simplest answer is that we prefer younger players with unknown upside and untapped potential to players with a proven record of production, who in this case was coming or season-ending injuries or a year away from the game.
The other side of this, for the second year in a row, is the tight end position is top-heavy with very little value outside of those top performers. The lack of second-tier players erodes the quality depth at tight end creating a bubble not only for players like Olsen, Witten, and Walker but others who can produce inconsistently but enough to be in that same conversation. The value for this group of veterans has been their consistency and if you don’t have an elite player there is comfort in knowing that Witten consistently delivers an acceptable floor while generally lacking a more enticing, explosive week-to-week ceiling.
Maurile Tremblay: I don't think Olsen, Witten, and Walker are performing all that differently from expectations. All three came with risk. All three, if they stayed healthy, were going to be NFL starters and possibly fantasy starters—most likely at the low end. So far, they have stayed healthy and have settled in as low-end fantasy starters.
None of these three aging tight ends were likely to enter the top-tier with Kelce, Ertz, Kittle, Evans, et al. Players like Hunter Henry, O.J. Howard, David Njoku, Austin Hooper, Mark Andrews, and others all had better upside potential than Olsen, Witten, or Walker.
It's not that we wrote off the aging tight ends as potential fantasy contributors. We wrote them off—probably correctly—as potential top-tier fantasy starters.
Waldman: This point above may be the point of the thread thus far.
Tremblay: Thanks. Taking O.J. Howard over Greg Olsen didn't work out very well this year, but without the benefit of hindsight, I'd make the same decision again every time. Upside potential is worth shooting for.
Justin Howe: The issue is their general lack of upside. And I’m fine with that. I’m certainly not kicking myself for drafting Witten at TE9, after all. These guys are holding steady, but none of them will win you a week more than perhaps once a year. They’re just as likely to sink it, as when each and everyone has this year. Olsen lured a thousand fantasy players to sub him in after two big weeks but has laid eggs ever since. Witten caught two early-season touchdowns, but has yet to draw more than four targets - when would you start him? Walker hasn’t drawn a look from inside the 10 since Week 1.
First off, I believe that we tend to fundamentally misunderstand how aging works in the NFL. If you've been playing fantasy football for more than a minute you've undoubtedly seen countless examples of "age curves", graphs that average production by every player at every age to show when each position "peaks" and "declines". According to these curves, for example, running backs are in their prime at age 25 and 26 and then start declining quickly from there. Here's a phenomenal example from good friend and talented analyst Chase Stuart using one of the cleverer approaches I've seen.
Indeed, I used a similar approach of comparing running back performance from one year to the next, but I bucketed all backs into players who maintained their previous level of performance or even improved, players who declined by some reasonable amount, and players who essentially fell completely from relevance overnight. And older players were actually not any more likely to exhibit small declines than their younger peers (the yellow line below), but they were substantially more likely to suddenly implode entirely (red line).
An aging curve model suggests that players should typically be worse in their last year than they were in their second-to-last year. In other words, as they get older their play declines. This intuitively makes sense.
But if you look at the last two fantasy-relevant seasons from top 100 fantasy RBs and WRs from 1984-2014, you find that half of them were actually better in their last fantasy-relevant season than they were in their second-to-last season, which is exactly what you'd expect if improvement and decline were essentially random processes until the final and inevitable end.
In 2017, Chase noted that "Frank Gore [wasn't] aging", using this chart to make the point:
Because of this, I'd also be wary of Dan's takeaway that "age-based regression does not hit tight ends nearly as hard as it does wide receiver and running backs". It's possible that this is the case. It's also possible that the recent trend of tight ends lasting longer is just statistical noise.
Indeed, I wrote about this exact phenomenon last season, noting that contrary to popular belief, NFL careers at most positions were not getting longer. (To use one dramatic example: from 1997 to 2000 there were seven double-digit sack seasons by a player 36 years or older. In eighteen years since there has only been one more. Similar patterns hold at wide receiver, offensive line, and even among placekickers and punters.)
At the one position where careers seemingly were getting longer (quarterback), it was more a result of random fluctuations in incoming talent than anything else. It's not that great quarterbacks were playing longer than they had in the past, it's that we had an unusually large number of great quarterbacks enter the league at the same time, which fifteen years later left us with an unusually large number of old quarterbacks at the same time.
There's also an element of selective memory at play here. The list of the most valuable fantasy tight-ends of the last decade includes the likes of Tony Gonzalez (whose last TE1 season came at age 37), Antonio Gates (36), Jason Witten (35), and Delanie Walker (33). It also includes Jimmy Graham (31), Rob Gronkowski (29), and Jordan Reed (26).
And if you want to dismiss Gronkowski and Reed because of their injury histories, remember that Greg Olsen-- one of the players being held up as proof that old tight ends can still play-- hasn't had a TE1 season since age 31. He's missed 16 games over the last two years, and in the games he did play he averaged just 30 yards, down from 66 yards from ages 29-31.
Had Olsen failed to make it back, we could easily dismiss him, too, as someone who didn't age so much as he succumbed to injury. Since he did make it back, we hold him up as proof that tight ends can continue playing well as they age. This is how insidiously selection bias works.
The best way to combat selection bias is with large samples based on objective criteria. I happen to have a list of the most valuable fantasy tight ends since 1985, and the top 20 includes Keith Jackson (last top-12 season at age 31), Ben Coates (29), Brent Jones (32), Steve Jordan (32), Jay Novacek (33), Jeremy Shockey (27), Mark Bavaro (30), Eric Green (30), Dallas Clark (30), Frank Wycheck (30), Todd Heap (26), and Kellen Winslow Jr. (27). Indeed, historically tight ends have tended to fall off earlier than wide receivers.
Maybe you don't think Frank Wycheck is really comparable to someone like Greg Olsen, but again, this is selection bias; Wycheck made pro bowls at ages 27, 28, and 29, and was a top-5 fantasy tight end for six consecutive seasons from age 25 to age 30. The reason we don't think he's comparable is because he fell off so young.
Perhaps the league has trended more towards using tight ends like receivers and we should expect the positions to age more comparably going forward, though there's little reason to expect tight ends to last longer than their more-heralded receiving peers. And I also don't know if I believe today's tight ends are behaving more like wide receivers than, say, Kellen Winslow Sr. (30) or Todd Christensen (31), who led the NFL in receptions four times in seven years from 1980 through 1986.
In the past, I've attempted to create "mortality tables" for each fantasy position based on how frequently historical players have "survived" at any given age vs. how frequently they have faced the sudden and inevitable decline. To get around the problem of changing usage at the tight end position (and the astronomically small sample of true receiving weapons over the years), I used a blended approach that compared modern tight ends to a sample that included both tight ends and wide receivers.
Now, as this pertains to tight ends Olsen, Witten, and Walker: because they are older, we should assume heading into this year that they were at heightened risk of a dramatic decline in their quality of play. But absent that sudden and dramatic decline, there's little reason to believe that they should have been noticeably less effective than they were the last time we saw them. And indeed, it looks like all three players have managed to outrun Father Time for at least one more season. Given their strong starts I don't think any of the three are at any heightened risk of disappointing the rest of the way than any other comparably productive tight end.
At the same time, I feel hesitant to suggest (as the question implies) that it's unwise to discount aging players at tight end. I think they genuinely are riskier assets than their younger peers. We know with certainty that the end comes for everyone, and the older a player becomes the more likely it is that this year is the year they are no longer who they once were. The fact that this year wasn't the year for Olsen, Witten, or Walker doesn't change that fact.
I'd compare player aging not to a smooth curve, but to a series of weighted coin flips. Imagine that every player has made a deal with the devil. Before the season, they flip a coin. If the coin comes up heads they survive another year unscathed. If the coin comes up tails they have reached the end of the road. Perhaps they continue on as a shell of their former selves, perhaps they suffer a career-ending injury or suddenly retire, but one way or another they are no longer who they once were.
Every year that coin becomes weighted more and more toward tails, but just like anything involving coin flips, there's nothing fundamentally preventing someone like Jerry Rice from flipping heads every year into his 40s. Nor is there anything preventing Andre Johnson from flipping tails at 33, or Herman Moore at 30, or Dez Bryant at 27, or David Boston at 25. Each of these events differs in terms of relative likelihood, but again, assuming probabilistic statements that are true at the group level will apply to individual members of that group is how we got into this mess in the first place.
So to sum up: age is a serious concern for tight ends (and for all players at every position), not because it tends to make players a little bit worse, but because it tends to make players a little bit more likely to become a lot worse all at once. But if we can see that a player hasn't become a lot worse all of a sudden, then I don't know that knowing their age adds any further explanatory power.
Haseley: Wow, and that was only one of the questions.