Last offseason, the Dallas Cowboys' roster construction plans became very clear, very early on. 

The Cowboys, coming off a surprising 13-3 campaign, had two obvious areas where they needed to improve. Dallas had the 19th-ranked pass defense in 2016, per Football Outsiders' DVOA; but the unit struggled as the year wore on (especially after an injury to a surprisingly effective Morris Claiborne), and they were torn up by Aaron Rodgers in the playoffs. There were problems on the back end, and for the second straight time during a divisional round playoff game, they found it nearly impossible to get pressure. 

When free agency started, they saw several of their players targeted by opposing teams early in the process. For the most part, the Cowboys let them walk. Dallas had 17 players reach free agency in 2017, and re-signed only four -- three of them on one-year deals. Everyone else was let go. That included four players from the secondary: safeties Barry Church and J.J. Wilcox, and corners Brandon Carr and Morris Claiborne. Combined, they accounted for 49.7 percent of the snaps played by Cowboys defensive backs in 2016. 

The Cowboys signed just two players to replace that departed foursome (Nolan Carroll and Robert Blanton), but they went heavy on defensive backs in the draft, selecting Chidobe Awuzie in the second round and Jourdan Lewis in the third, trading up to get Xavier Woods in the sixth, and nabbing Marquez White in the sixth as well. The Cowboys used three of their additional five picks on the defensive line, adding edge rusher Taco Charlton in the first round and defensive tackles Joey Ivey and Jordan Carrell in the seventh. Of the six free agents the Cowboys signed away from other teams, two (the aforementioned Carroll and Blanton) were defensive backs and two (Stephen Paea and Damontre Moore) were defensive linemen. The Cowboys came into the offseason clearly wanting to overhaul the front and back of their defense, and that's what they did. 

Looking for a hot new NFL podcast that's your home for NFL coverage? Look no further. The Pick Six Podcast with Will Brinson has you covered each day with new episodes around 30 minutes each. Subscribe: via iTunes | via Stitcher | via TuneIn | via Google Play

The team's plan in the 2018 season was much more difficult to suss out. First of all, the Cowboys did not have that many free agents targeted by other teams, so it didn't become immediately clear that there was a position at which they were planning a dramatic overhaul. The one significant move they made early in free agency was letting Anthony Hitchens leave for a huge contract with the Chiefs. Every other team in the NFL signed at least one free agent away from another team before the Cowboys did so, and when they finally did, it was special teamer (not that) Joe Thomas. There was not much to glean from them cutting Orlando Scandrick or letting ineffective or sparsely-used players like Jonathan Cooper, Benson Mayowa, Kyle Wilber, and Bene Benwikere leave. 

The plan did not become clear until much later, and arguably, not until after their three days of wheeling and dealing at the draft. 

The departure of Hitchens left Dallas with a hole at the MIKE linebacker spot, and while many expected that they'd use their first-round pick on a wide receiver, they instead found a replacement here with Leighton Vander Esch. A monster-sized (6-4, 256 pound) athletic marvel out of Boise State, Vander Esch helps the Cowboys kill two (or even three) birds with one stone. He seems likely to be an immediate starter at MIKE, slotting in between Sean Lee at the WILL and Jaylon Smith at the SAM. His athleticism and versatility should help Dallas make up for the fact that the defense has absolutely fallen apart whenever Lee has been injured over the last several seasons and is coming up on the end of his contract, and that Smith struggled badly in coverage in his first game action in over a year. 

The Cowboys came into the draft with 10 picks in all, and aside from Vander Esch, only two were used on defensive players. The first was Kansas pass-rusher Dorance Armstrong, who followed up an excellent sophomore season (10 sacks, 20 tackles for loss) with a meeker 2017 campaign, thanks to the team's switch to a 3-4 scheme. Armstrong figures to be a rotational rushman at best as a rookie, but if he flashes, he could end up allowing the Cowboys to save a bunch of money next season by cutting ties with Tyrone Crawford. The last defender the team picked was Indiana linebacker Chris Covington, who seems likely ticketed for the Wilber role of deep-sub-package linebacker and special teams contributor, if he makes the roster.

Those positions were obvious adds for the Cowboys from the time free agency got into full swing, and it wasn't surprising to see them addressed last weekend. But the Cowboys' most significant signings early in free agency were tackle Cameron Fleming and guard Marcus Martin. Heading into the draft, it seemed completely reasonable that Fleming might start at right tackle, kicking La'el Collins back inside to left guard, where he played in 2015 and 2016 before shifting across the line after the retirement of Doug Free. Then again, the Cowboys were at their best in 2014 and 2016 when their offensive line was dominating the league; so it maybe should not have been surprising when they tried to stack on top of their strength by selecting Texas offensive lineman Connor Williams in the second round. Williams was considered a potential first-round selection, but his up-and-down 2017 season and poor measurables for a tackle knocked him down the board to No. 50. The Cowboys made clear that they see him as an interior lineman, which means we should likely expect him to start at left guard and Collins to stay at tackle, while Fleming and Martin provide depth. It's fairly clear there was a concerted effort to upgrade and add depth to this unit, even though it was considered the strongest on the team and one of the strongest in the league. That effort, at least on paper, looks like a success. 

The team's much more ambitious, much more controversial, and much less obvious plan this offseason was apparently to dramatically overhaul the offensive skill position groups. 

It's no secret that the Cowboys struggled to create splash plays offensively last season, with a lack of speed anywhere on the offense being pegged as a definitive culprit. Dallas added two bodies to the wide receiver room in March with Allen Hurns and Deonte Thompson and executed a trade with the Raiders that landed versatile fullback Jamize Olawale; but until they surprisingly cut Dez Bryant two weeks before the draft, you might not have known this overhaul was coming. After a weekend full of picks and trades, the team has a much different look than it did before the draft: 

That's Morris, McFadden, (Keith) Smith, Bryant, Butler, Switzer, Witten, and Hanna out; and Olawale, Austin, Scarborough, Hurns, Thompson, Gallup, Wilson, and Schultz in. The departed players account for 54.6 percent of all snaps played by Dallas skill position players last season; 28.9 percent of the snaps in the backfield (this figure is obviously affected by Ezekiel Elliott's six-game suspension), 46.7 percent of the snaps at receiver, and 88.4 percent of the snaps at tight end. In all, there are 2,415 snaps that need to be replaced by eight incoming players. Some of those players obviously figure to see more than others, but that is a whole lot of change -- and if you're reading the tea leaves, it's obvious the Cowboys might not even be done. The operative question, of course, is how all these pieces fit together, and how the Cowboys plan to utilize them.

To start with, the roster spots of the three holdover receivers seem extremely tenuous. The Cowboys can save $3.25 million against the salary cap by cutting Beasley, and not only did they add a receiver in free agency who has extensive experience in the slot (Hurns), they also traded for a gadget weapon who can move to different spots on the field and run jet-sweeps, quick screens, and short-breaking routes (Austin). Beasley was arguably the team's best receiver in 2016 but he has not been the same since suffering a hamstring injury late in that season. He had a career-low catch rate last season and also posted his lowest figures since his rookie year in both receptions per game and yards per reception. 

The Cowboys re-signed Williams last offseason to a deal they touted as a discount (four years, $17 million), but he came out and had his worst professional season (career-low 10.7 yards per reception, zero touchdowns) and the best thing you can say about him at this point is that he's a willing and above-average blocker. By signing Hurns and drafting Gallup, it seems pretty clear the Cowboys are planning for a post-Williams world. I'd be shocked if he was on the roster in 2019. Brown was a sixth-rounder in 2017 who is basically a hybrid tight end most known for his blocking, but considering the Cowboys have preseason star Gathers coming back from injury and that they just drafted both a tight end that is known as an excellent blocker and a big-bodied wide receiver on Day 3, I would not be feeling too safe if I was Brown, either.

The remainder of the pass-catching corps is, well, interesting, I guess. With Witten and Hanna retiring, the Cowboys have a group of tight ends who have combined for NINE career catches. Witten had that many or more in 20 of his 239 career games. This seems like a tight-end-by-committee kind of group, with Swaim getting more work than expected merely because he is by far the most experienced player of the bunch. Gathers has a ton of upside and can be a mismatch nightmare if he puts it all together. Schultz was a great blocker at Stanford and runs excellent short-yardage routes. Jarwin is largely an unknown. (He played three offensive snaps last season.) Two-tight end sets with Schultz and Gathers on the field seem like they'd stretch the defense the most, but given their inexperience, it's fair to wonder how often that combination will see the field. 

Hurns is versatile and big-bodied, and seems like he might be the team's de facto No. 1 to start the year. (The Joneses and Jason Garrett said the team knows it's not going to have a traditional No. 1 receiver and will have to spread the wealth, and that definitely sounds like the kind of thing a team says when it's not too confident in its wide receiver corps.) Thompson is a pure speed threat. They clearly have high hopes for Gallup, who some of the front office wanted to select in the second round but was still there in Round 3. He's well-rounded and versatile as well, if not necessarily super speedy. Wilson showed that he could make big plays at Boise State, but as a sixth-round pick, he's not guaranteed to make the team. At the post-draft presser, the team brass all talked of their big plans for Austin. Like the Rams always did, the Cowboys brass said they want to get him 8-12 touches a game. It seems far more likely that he gets 3-5 incredibly obvious looks and their effectiveness is neutered by the obviousness that he's going to get the ball; why else would he be on the field? 

That said, the Austin acquisition might be emblematic of the kind of offense the Cowboys plan to run. It's clear they want to get more versatile offensively, and more "Dak-friendly," as they've been saying all offseason. Austin's experience at West Virginia lends itself to employing Chiefs-stlye trickeration with jet sweeps and misdirection and mesh concepts like the ones the Eagles utilized to win the Super Bowl. As we've written previously, the Cowboys should take advantage of Prescott's ability to make the correct reads on those type of plays, as well as the pressure his mobility puts on the defense. Using Austin as a ball-carrier and decoy in the style of, say, Tyreek Hill or Tarik Cohen makes some degree of sense. If he can make defensive ends hesitate for even a moment before Prescott hands the ball to Elliott, that can mean the difference between a short gain or a first down. If he can take advantage of the fact that defenses will key on Elliott and speed around the edge, that can help the Cowboys create the kind of big plays they were missing last season. 

Similarly, players like Hurns and Gallup make sense as versatile weapons who can line up at several positions (Hurns can line up anywhere and Gallup got extensive experience at both X and Z while at Colorado State) and run all kinds of routes. A big issue with the Cowboys' receiving trio of Bryant, Beasley, and Williams was their limited individual capabilities. Dez was always running hitches, slants, or digs. Williams was always going deep. Beasley was always staying short. (And Witten was always running eight-to-10 yards down the middle of the field.) Adding players that can do a little bit of everything so the defense doesn't necessarily know what to expect should help create wider throwing lanes for Prescott, who had to throw into tight coverage more often than almost any other passer in the league last season. It helps that Gallup excelled at making contested catches and at beating press coverage, considering how big an issue that was for the Cowboys' primary receivers last season. 

Working through all of this brings up a lot of uncertainty, and so it seems logical to expect that the Cowboys will once again lean on the running game as the foundation of the offense. That makes sense, given the strength of the offensive line and the explosive and efficient capabilities of Elliott. But they can expect to see a lot of stacked boxes until Garrett, offensive coordinator Scott Linehan, Prescott, and the pass-catchers prove that opposing defenses need to pay attention to what's going on outside. Using deception and versatility to attack defenses in more varied ways than they have in the past makes sense as a tactic for doing that. Whether the Cowboys acquired the correct players to put that plan into action remains to be seen.