This article means to break down the fallout of the DeAndre Hopkins trade specifically as it pertains to the Houston offense. Before it's over, the 2020 projections for Deshaun Watson, David Johnson, Duke Johnson, Will Fuller, and Kenny Stills will all be discussed.
(Ahem.)
Bill O'Brien is an innovator in the unthinkable, each of his feats more spellbinding than the last. Trading a third-round pick for Duke Johnson just to bench him for Carlos Hyde, trading Jadeveon Clowney for a special teams linebacker, coughing up two first-round picks for Laremy Tunsil right before he gets a record-setting contract – these accomplishments were once the sort of thing that would capture our awe, but O'Brien's work in Monday's opening of free agency has raised the bar to a galactic height.
Indeed, Bill-O set out Monday to make his past work look feeble in comparison, and by trading DeAndre Hopkins to the Cardinals he did just that. O'Brien flexed his inconceivably powerful brain muscles until Arizona general manager Steve Keim submitted under his might, accepting O'Brien's offer of Hopkins plus a fourth-round pick for a second-round pick, a 2021 fourth-round pick, and David Johnson's toxic, $14.16 million cap hit. You can believe Johnson still has something left – let's just assume he does – but Arizona was in no position to hold onto a backup running back with a $14 million salary, so when adjusting for leverage capital and the cap hit, it's basically as if
This article means to break down the fallout of the DeAndre Hopkins trade specifically as it pertains to the Houston offense. Before it's over, the 2020 projections for Deshaun Watson, David Johnson, Duke Johnson, Will Fuller, and Kenny Stills will all be discussed.
(Ahem.)
Bill O'Brien is an innovator in the unthinkable, each of his feats more spellbinding than the last. Trading a third-round pick for Duke Johnson just to bench him for Carlos Hyde, trading Jadeveon Clowney for a special teams linebacker, coughing up two first-round picks for Laremy Tunsil right before he gets a record-setting contract – these accomplishments were once the sort of thing that would capture our awe, but O'Brien's work in Monday's opening of free agency has raised the bar to a galactic height.
Indeed, Bill-O set out Monday to make his past work look feeble in comparison, and by trading DeAndre Hopkins to the Cardinals he did just that. O'Brien flexed his inconceivably powerful brain muscles until Arizona general manager Steve Keim submitted under his might, accepting O'Brien's offer of Hopkins plus a fourth-round pick for a second-round pick, a 2021 fourth-round pick, and David Johnson's toxic, $14.16 million cap hit. You can believe Johnson still has something left – let's just assume he does – but Arizona was in no position to hold onto a backup running back with a $14 million salary, so when adjusting for leverage capital and the cap hit, it's basically as if Houston traded Hopkins for something like a third-round pick.
That's good news for Arizona, and great news for Kyler Murray, but for Deshaun Watson this is a setback that could meaningfully alter the trajectory of his entire career. Not only is Hopkins one of the few truly elite receivers in the NFL, he's a beloved leader of the Texans whose exit could precipitate a legitimacy crisis for O'Brien, resulting in a talent exodus as star players develop tension toward a coach they don't respect. Watson needs a contract extension, and you have to wonder if he'll even listen to Houston's offers in light of this. Mad Dog Radio's/ESPN Radio's Mike Meltser reported that O'Brien was on strained terms with Hopkins, which would help explain how O'Brien could accept such a catastrophically bad trade offer, and it's hard to see how it could serve O'Brien's legitimacy now that these facts are in the open.
In the meantime, setting aside the questions of personal or political strain within the Texans, the Hopkins trade is a considerable hit to the projection of Watson, who we've generally assumed for a borderline elite fantasy quarterback the past two years. In BestBall10 drafts to this point Watson had gone, on average, 69.75 in the ADP. That put him about three rounds behind Patrick Mahomes, and one round ahead of Russell Wilson and Kyler Murray.
Over the last three years Watson completed 66.8 percent of his passes at 8.1 yards per attempt. In those same seasons Hopkins turned 487 targets into 315 receptions for 4,115 yards and 31 touchdowns (64.7 percent catch rate, 8.5 yards per target). Given the volume of Hopkins' production in that span, it's safe to say that he's almost singularly responsible for carrying Watson's per-attempt base line as a passer. That's not a slight on Watson, it just means that we should assume that Hopkins' replacement targets will provide lesser returns than what Hopkins did. Hopkins was the foundation of the Houston passing game in terms of volume and efficiency, opening up concerns for Watson on both fronts. After all, as much as Watson completed 66.8 percent of his passes at 8.1 YPA, the Texans completed only 63.5 percent of their passes at 6.2 YPA when throwing to someone other than Hopkins over the last three seasons. (Yes, I've triple checked this and every time I get the same numbers. Apologies if I'm missing something). Some of that was dragged down by Tom Savage's dreadful showing in 2017 (56.1 percent completed, 6.3 YPA), yet Hopkins still averaged 7.9 yards per target that year.
Hopkins' unique value is beyond questioning, and it's doubtful that Houston is in any position to replace his functions. Their first selection in the draft is at the 40th pick, so they won't get a shot at CeeDee Lamb or Jerry Jeudy. Realistically, Houston will have to go with more of a run-heavy offense, because an offensive model that presumes the presence of DeAndre Hopkins isn't going to function with Will Fuller and Kenny Stills as its top two receivers. And I like both of those guys! Hopkins is just a different breed.
Watson has played 2,583 regular season snaps, during which he's totaled 1,204 pass attempts and 217 rush attempts. That's one pass attempt every 2.15 snaps, and one rush attempt every 11.9 snaps. If O'Brien calls the same relative volume of pass attempts, I think we should project a considerable decline in Watson's per-attempt efficiency, at least in terms of YPA but potentially in terms of completion percentage too. If O'Brien calls a lower relative volume of passing attempts, Watson might have a shot of maintaining his per-attempt numbers since the average over a smaller sample can more easily spike upward than in a high-volume sample. Of course, Watson's price to this point wasn't reached with a decline in usage in mind, so I struggle to see how there's a means of treading water relative to prior fantasy output.
If Watson plays 1,050 snaps in a season and we presume a pass attempt every 2.2 snaps with the rush attempt frequency mostly maintained at once every 11.5 snaps, then it would project Watson for 477 pass attempts and 91 carries. If he averages 7.7 yards per pass attempt, Watson would project for 3,673 yards in that scenario. We should probably project a decline in passing touchdown percentage, too, given that Hopkins accounted for more than 38 percent of Houston's passing touchdowns in the last three years. Instead of Watson's customary 5.1-5.3 percent touchdown rate, let's say 4.85, which would project to 23 touchdowns. Watson's per-carry rushing average has fallen each year, so let's assume the 5.0 yards per carry from last year is the best-case scenario, projecting for 455 rushing yards. With a worse passing game should come fewer scoring opportunities, so after posting 12 rushing touchdowns over his last 181 attempts (6.6 percent), let's instead project something more like his rookie year rushing touchdown percentage of 5.5 percent (two on 36 attempts). On 91 attempts that would leave a projection of five rushing touchdowns.
Watson respectively averaged 24.8 and 25.8 standard scoring fantasy points per game in 2018 and 2019. If he were to throw for 3,673 yards and 23 touchdowns while posting 455 yards and five touchdowns on the ground, that would project for 21.95 fantasy points per game. Even in a largely dysfunctional offense as a rookie (and without Hopkins, crucially), Murray averaged 21.5 fantasy points per game last year. I think at the very least that this Hopkins trade means that Watson should fall behind Murray and Wilson in the quarterback draft order, which would entail a drop from 68th overall to around 90th overall, likely behind Dak Prescott in addition to Wilson and Murray. My projections could of course be wrong, but this is the sort of outlook I'm planning on in the meantime.
Of course, for as much as Hopkins is the star of this trade, David Johnson's inclusion also bears substantial repercussions for the fantasy football scene. Previously drafted around 79.6 in BestBall10s (80th overall), David's unambiguous starting status in Houston should boost his market, potentially by a lot. He at the very least should jump ahead of rookies Clyde Edwards-Helaire (74.82 ADP) and Cam Akers (77.16 ADP), while Kareem Hunt (74.79 ADP) and Derrius Guice (70.64 ADP) should drift past, too. I'd guess David breaks into at least the fifth round of fantasy drafts going forward, and quite possibly the third or fourth.
Guessing David's exact fit in the Houston offense isn't so easy, because it's difficult to tell what sort of utilization O'Brien has in mind. Johnson is nothing like Carlos Hyde and shouldn't be used the same way, but there's a real risk that an oaf like O'Brien is completely unaware of this. For that reason I'm not especially optimistic about David's ability to boost his per-carry rushing average, which never was a strength of his anyway. In fact, you could argue David Johnson is a below average pure runner – it was only as a route runner where he ever had any unique value. This particular value is conceivably substantial, and it's there for the taking if O'Brien can see it. But O'Brien didn't make a whole lot of use of fellow route-running back Duke Johnson last year, and if O'Brien simply envisions David as his Hyde replacement then David might not even be running that many routes, anyway. It's also disturbing for David's projection that both Kenyan Drake and Chase Edmonds enjoyed great success on the ground last year while David saw basically none.
Hyde and Duke combined for 328 carries and 78 targets on 1,069 snaps last year. Whereas Hyde was a running specialist and Duke primarily a passing-down specialist, we can assume the Johnsons will be more interchangeable if they're both on Houston's 2020 roster. That interchangeability should afford David to pick up a lot of Hopkins' vacated targets, so his target volume needn't fully occur at the expense of Duke. Let's assume 1,050 snaps up for grabs between the two over 16 games, then (generously) project David to be healthy enough to play 14 games. If we project David to take 55 percent of these available snaps, then that projects to roughly 505 snaps, leaving Duke with around 490 if he plays 45 percent of the snaps in games for which David is active and 58 percent in the two projected missed games for David.
Duke averaged one carry per 6.4 snaps and one target per 8.6 snaps last year. Let's lower the per-snap carry rate to once every 5.8 snaps with Hyde gone, but if you otherwise project that and last year's target rate against 490 snaps, he then projects for 85 carries and 57 targets. Because I can't be terribly optimistic about his per-touch efficiency increasing, especially with Hopkins gone and increased defensive pressure in light of that, this projection scenario implies a downgrade for Duke in light of David's arrival. David could easily miss more than two games, as the last couple seasons have shown, but I think David's greater pass-catching function over Hyde means that Duke will have to fight over a smaller pie in a worse offense than last year's. Of course, you could argue that Duke's ceiling is higher with David around since David is more likely to get hurt than Hyde was. It's also possible that, since David is much more of a pass catcher than Hyde, that Duke's per-snap carry rate could increase, pushing him toward 100 carries rather than the 85 in this exercise.
For David's projection I'll have to slightly lower the per-snap carry rate from what Hyde showed, if only to slightly subsidize the much greater per-snap target rate David figures to draw. If David secures a carry on 43 percent of his carries rather than the 45.5 percent from Hyde's case, then it'd project to 217 carries. If David were to average 3.75 yards per carry in this scenario, then it'd project for 814 rushing yards. If we project David for the same per-snap target rate as he showed in Arizona last year, then it'd come out to 53 targets, which at a 72 percent completion rate would result in 38 receptions. If he averages 9.5 yards per catch, this would leave David with 361 receiving yards – a total of 1,175 yards from scrimmage. After Duke and Hyde combined for just 11 touchdowns from scrimmage last year, I'm similarly inclined to project modest touchdown production for Duke and David in 2020. Let's say 12 total – five for Duke and seven for David. That would project David for roughly 14.1 PPR points per game in his 14 projected games.
As murky as the Texans backfield might be, figuring out its wide receiver usage could be even more difficult. You can't really anticipate what the exit of a high-volume, high-efficiency player like Hopkins might mean for the broader ecosystem of the offense. That's especially true when his top two replacement options, Fuller and Stills, are both injury-prone big-play specialists. Big plays are great, but neither Fuller nor Stills imply consistency, and with that we have to acknowledge a wide range of possible outcomes. It's worth noting, though, that Hopkins, Fuller, and Stills combined for 276 targets over 2,184 snaps. Hopkins' exit leaves 1,000 snaps and 150 targets unaccounted for. Even with all of their injury issues, and even with the chances that Houston adds another skilled wideout in the draft or free agency, this is clearly a breakout opportunity for both Fuller and Stills, who combined for 126 targets even with Hopkins present. If the two don't combine for at least 200 targets in 2020, then at least one of them likely got hurt.
For the sake of putting down something concrete, let's hazard a projection for Fuller and Stills despite their mammoth uncertainty in the meantime. Let's project 14 games active for both, with both playing 850 snaps. Fuller drew nearly twice the per-snap air yardage (1.74) last year than Stills did (0.91), so let's assume Fuller will get open more often and draw a higher per-snap target rate as a result. Let's assign Fuller a target for every 7.5 snaps, and Stills one target every nine snaps. That would leave Fuller with 113 targets and Stills with 94. A catch rate of 65 percent at 8.5 yards per target in this case would leave Fuller with roughly for 74 catches for 960 yards, and Fuller's per-target touchdown rate over the last two years (six percent) would project for roughly seven receiving touchdowns. If Stills catches 70 percent of his targets (74.1 last year) at 9.0 yards per target (10.2 last year) then he would end up with about 66 catches for 846 yards. With four touchdowns on 40 receptions last year, let's give Stills six touchdowns for this projection. These projections would leave Fuller with about 15.1 points per game in PPR, and Stills with 13.3. Perhaps these projections are too optimistic (I disagree!), but with Fuller often falling into the 10th round of BestBall10 drafts and Stills often times into the 20th round up to this point, I think it's safe to say that the prices of both can and should increase substantially. I'd expect Fuller to push for the sixth round now, and Stills probably something like the ninth. Then again, both Fuller and Stills have burned plenty of fantasy owners in the past, so maybe some lingering bitterness will keep their prices low despite the vastly improved projections.
I'm going to punt on the question of the tight ends (sorry, I'm tired), but I think the lumbering trio of Darren Fells, Jordan Thomas, and Kahale Warring can expect to log plenty of two-tight end snaps, while Jordan Akins will pitch in as more of a route-running specialist. Fells has the floor, but Thomas, Warring, and Akins are prospects who each have their own selling points. I'll probably stay away from them all unless we're offered more clarity in the upcoming days, but Fells is the safest one.
To recap, here are my adjusted projections for Watson, David Johnson, Duke Johnson, Fuller, and Stills:
Deshaun Watson (16 games): 3,673 yards and 23 touchdowns passing, 455 yards and five touchdowns rushing (22 PPG)
David Johnson (14 games): 814 yards rushing, 38 receptions for 361 yards, seven touchdowns from scrimmage (14.1 PPG PPR)
Duke Johnson (16 games): 412 yards rushing, 42 receptions for 386 yards receiving, five touchdowns from scrimmage (9.5 PPG PPR)
Will Fuller (14 games): 74 catches for 960 yards and seven touchdowns (15.1 PPG PPR)
Kenny Stills (14 games): 66 catches for 846 yards and six touchdowns (13.3 PPG PPR)
