How Cowboys’ Last 3 Drafts Are Hurting The Roster

One traditional draft mantra holds that you cannot really judge a draft until three or four years have gone by. But what if there was a way to do it much quicker?

What you’d probably need is to establish some kind of expectation for each draft pick and then measure that draft pick against that expectation. Sounds simple enough. And we have the tools to do it.

One of those tools is Approximate Value (AV) from Pro-Football-Reference.com, which is designed to assign a specific value to any player at any position for any given year. The algorithm behind AV weights position-specific metrics (i.e. yards or points scored/allowed) with an indicator for durability (total games played and seasons as their team’s primary starter) and quality (Pro Bowl and All-Pro nominations) and then normalizes all this at a team level. Follow the link for more details.

Because AV provides a value for every player in every season, we can use it to calculate the average AV for every single draft slot and then compare a player drafted in that slot against the average AV or as we’ll call it today, Expected AV (eAV).

For this exercise, I looked up the AV for each of a player’s first four years in the league from 2013-2025. I chose only the first four years because that’s usually the length of a rookie contract outside of the first round, and I chose 2013 as the cutoff because that was the last draft before Will McClay was put in charge of the draft in Dallas.

Here’s what the average AV looks like plotted against the draft slot for the rookie season of each player drafted since 2013.

What you can see is that there is some kind of curve inherent in the data, with early picks averaging more AV than later picks. This of course is not rocket science, it’s simply the nature of the draft where the better the player, the earlier he tends to be drafted.

But the chart is a bit “jumpy” – for lack of a better word – if you look at it more closely, with quite significant differences between adjacent draft slots. So I decided to smooth the data, by using a five-pick average to give me the average AV. So pick #1 is the average of picks #1-5, pick #2 is the average of picks #2-6, and so on. Here’s what the data looks like after the smoothing:

The average AV (the blue dots) now looks much smoother, and I added an exponential regression (red dotted line) that I’ll use to calculate the expected AV for each draft slot. For those interested, r² for the regression is a very strong 0.95.

I then repeated the exercise for the 2nd, 3rd, and 4th year of each player – understanding of course that the 2025 draft class only has the rookie season to look at, the 2024 draft class only has a rookie and 2nd season, and so on.

Here’s what the four curves look like:

The first thing that stands out is that we see actual evidence for the “second-year jump”. The difference between the rookie season curve and the 2nd season curve is 0.3 eAV points at the end of the draft, climbs to 0.5 eAV points in the bottom of the fourth round, increases to 0.8 eAV at the top of the second round, and maxes out at +1.7 eAV points for the first pick. That may not sound like much, but for the 104th pick (where the blue line hits 2.0 eAV in the chart above), the jump from 2.0 to 2.6 eAV points is an increase of +26%, for the 254th pick (0.6 to 0.9) that’s a jump of 47%, and even for the 12th pick (from 5.4 to 6.5), that’s still an improvement of 20%.

We can also see that the 3rd-season eAV is almost identical to the second-season eAV, while the 4th season dips slightly, most likely due to injuries and more players getting released/cut.

Also, r² decreases slightly by year, but remains strong. For those interested, here is r² and the equation for each of the exponential regression curves:

• Rookie season: r² = 0.95; y = -1.573 * ln(x) + 9.3499
• Second season: r² = 0.92; y = -1.883 * ln(x) + 11.085
• Third season: r² = 0.93; y = -1.79 * ln(x) + 10.835
• Fourth season: r² = 0.90; y = -1.718 * ln(x) + 10.386

With the eAV now established, we can begin to compare Cowboys draft classes against expectations. I’ll illustrate what that looks like with the 2022 Cowboys draft class.

body .sbnu-legacy-content-table td, body .sbnu-legacy-content-table th, body .sbnu-legacy-content-table { border: 1px solid #000 !important; border-collapse: collapse !important; }

eAV Surplus for 2022 Cowboys Draft Class
Player	Round	Pick	Total AV	Total eAV	eAV Surplus
Tyler Smith	1	24	35	19.7	+15.3
Sam Williams	2	56	6	13.8	-7.8
Jalen Tolbert	3	88	9	10.7	-1.7
Jake Ferguson	4	129	20	8.1	+11.9
Matt Waletzko	5	155	0	6.8	-6.8
DaRon Bland	5	167	22	6.3	+15.7
Damone Clark	5	176	14	5.9	+8.1
John Ridgeway	5	178	7	5.8	+1.2
Devin Harper	5	193	0	5.3	-5.3
Total			113	82.3	+30.7

Overall, this is quite a strong draft class. Despite picking “only” 24th, the Cowboys rank 7th overall in the NFL with their +30.7 eAV surplus points. Standouts DaRon Bland and Tyler Smith rank as the 18th and 21st best players in the 2022 draft class across the league in terms of eAV surplus. Jake Ferguson ranks 31st overall and is the second-ranked TE behind Trey McBride. Damone Clark may be a surprise for some, but he was a full-time starter in 2023, earning eight AV points in just that one season.

A disappointment relative to eAV is Sam Williams, but he had to sit out a complete season in 2024 and struggled to regain his form in 2025. Matt Waletzko saw spot duty over three seasons and didn’t get a single point of AV. Devin Harper was waived after just one season, but the eAV counter ticks mercilessly for four years, even if the player didn’t play four seasons.

Now that we’ve established what an individual draft class looks like, we can look at how all Cowboys draft classes since 2013 compare to the rest of the NFL.

body .sbnu-legacy-content-table td, body .sbnu-legacy-content-table th, body .sbnu-legacy-content-table { border: 1px solid #000 !important; border-collapse: collapse !important; }

NFL rank for each Cowboys draft class
Year	eAV Surplus	NFL Rank
2013	23.4	3
2014	33.2	3
2015	-8.4	18
2016	78.6	1
2017	-10.5	20
2018	25.4	7
2019	6.0	13
2020	45.9	2
2021	15.8	10
2022	30.7	7
2023	-25.5	29
2024	1.7	16
2025	-4.6	27
Total 2013-25	211.5	3

Up until 2023, all was well in Cowboys draft country. Out of the 10 drafts from 2013-22, seven delivered Top 10 results, and this is a big reason why many people believe the Cowboys draft well. Three more drafts were around average, and that’s to be expected, nobody hits on all the draft picks all the time.

Then over the last three years, the Cowboys draft results got much worse, and we’ll get to that in a little bit.

First though, and this is a key takeaway here even if it’s not obvious in the chart above, the Cowboys draft success (No. 3 overall since 2013!) is driven largely by their first-round picks. Those first-round picks account for 158.8 points of the entire 211.5 eAV surplus, the highest value in the league and 75% of the total surplus. Add in Dak Prescott with a 49.3 eAV surplus and we’re at 208.1, which means the entire Cowboys draft “success” is driven by their first-round picks plus Dak Prescott. Remove the underperformers among the first-rounders and you’re left with these players driving the entire draft success of the Cowboys:

body .sbnu-legacy-content-table td, body .sbnu-legacy-content-table th, body .sbnu-legacy-content-table { border: 1px solid #000 !important; border-collapse: collapse !important; }

Top Cowboys 1st-round picks + Prescott
Year	Player	POS	eAV Surplus
2022	Tyler Smith	OL	15.3
2021	Micah Parsons	LB	35.5
2020	CeeDee Lamb	WR	30.9
2018	Leighton Vander Esch	OLB	7.7
2016	Ezekiel Elliott	RB	18.9
2016	Dak Prescott	QB	49.3
2015	Byron Jones	CB	8.1
2014	Zack Martin	G	29.5
2013	Travis Frederick	C	27.1
Total			222.4

Now consider how many of these picks came down to pure luck.

I know that everyone’s mileage will vary on this, but these are the picks I consider “lucky” picks for the Cowboys:

• Dak Prescott was certainly a lucky pick. Nobody drafts their future franchise quarterback in the fourth round. Also consider that the Cowboys were desperately trying to trade up for Paxton Lynch (-17.1 eAV surplus) in that draft.
• Ceedee Lamb was also a lucky pick. Nobody had Lamb falling as far as he did in the draft, and to their credit, Cowboys had the wherewithal to grab him once he was available, but they were obviously lucky to get him.
• Drafting Zack Martin certainly wasn’t the plan going into the 2014 draft. At the time, the Cowboys had four players with the same grade on their board, DE Anthony Barr (+10.5 eAV surplus), DT Aaron Donald (+35.1), LB Ryan Shazier (+5.1), and OG Zack Martin. “All of those guys were right there with the same grade,” Jerry Jones said, referring to Martin, Barr, Donald and Shazier. The Cowboys were on the phone already with Shazier when Pittsburgh picked him from under their noses. Jon Machota of the Dallas Morning News at the time reported the Cowboys front office was forced to scramble while trying to figure out what to do with their pick, and were even looking to trade back before eventually selecting Zack Martin.
• The Travis Frederick pick was utter chaos. The Cowboys went into the draft with the 18th overall pick, and were looking hard at offensive guards Jonathan Cooper (-20.2) and Chance Warmack (-4.7), but both were taken in the top 10. And when safety Kenny Vaccaro (-3.9) went 15th to New Orleans – three ahead of Dallas’ original slot – things started to come apart. With Vaccaro gone, the focus shifted to Florida defensive lineman Sharrif Floyd (-7.0), who apparently had a top five grade on the Cowboys’ board, but the coaches didn’t like him, so the Cowboys traded down and ended up with Travis Frederick (+27.1) with the 31st pick. An argument could be made that they were hoping to get another offensive lineman, Justin Pugh (+2.7), with their 31st pick. Pugh was rated one spot above Frederick on the Cowboys’ draft board, but Pugh was picked 19th by the Giants, so the Cowboys took Frederick and ended up with the top-ranked player by eAV surplus taken in the first round that year. Lucky.
• Even the Micah Parsons pick was arguably a lucky pick. The Cowboys went into the draft with the 10th overall pick, and the understanding at the time was they would pick either of the two top CBs. But Jaycee Horn (-13.3) and Patrick Surtain (+17.5) were picked in consecutive picks at No. 8 and No. 9, after which the Cowboys traded down two spots and found Micah Parsons (+35.5) waiting for them, and once again got the player with the highest first-round eAV surplus that year.

So, if five of your 10 best picks were lucky picks, are you really good at drafting, or did you simply get lucky?

Consider also that the Cowboys have had a tough time getting two consecutive players with their top two picks in a draft with a positive eAV surplus.

body .sbnu-legacy-content-table td, body .sbnu-legacy-content-table th, body .sbnu-legacy-content-table { border: 1px solid #000 !important; border-collapse: collapse !important; }

Top Cowboys top 2 picks by year
Year	Player	AV	eAV	eAV Surplus
2013	Travis Frederick	45	17.9	27.1
	Gavin Escobar	2	15.0	-13.0
2014	Zack Martin	52	22.5	29.5
	DeMarcus Lawrence	24	17.3	6.7
2015	Byron Jones	27	18.9	8.1
	Randy Gregory	5	13.3	-8.3
2016	Ezekiel Elliott	51	32.1	18.9
	Jaylon Smith	25	17.3	7.7
2017	Taco Charlton	9	18.6	-9.6
	Chidobe Awuzie	16	13.3	2.7
2018	Leighton Vander Esch	29	21.3	7.7
	Connor Williams	23	14.6	8.4
2019	Trysten Hill	5	13.6	-8.6
	Connor McGovern	14	10.5	3.5
2020	CeeDee Lamb	53	22.1	30.9
	Trevon Diggs	29	14.5	14.5
2021	Micah Parsons	60	24.5	35.5
	Kelvin Joseph	2	15.5	-13.5
2022	Tyler Smith	35	19.7	15.3
	Sam Williams	6	13.8	-7.8
2023	Mazi Smith	9	14.3	-5.3
	Luke Schoonmaker	4	10.2	-6.2
2024	Tyler Guyton	9	9.0	0.0
	Marshawn Kneeland	2	6.7	-4.7
2025	Tyler Booker	6	5.4	0.6
	Donovan Ezeiruaku	3	3.4	-0.4

Even with two first-round picks this year, that chequered draft record does not make me optimistic about the Cowboys landing two standout players in this year’s draft.

More importantly though is that the Cowboys draft results have produced a top-heavy roster. That top-heavy roster was good enough to go on an eight-year run (2016-2023) that saw the Cowboys compile the third-best regular season W/L record (82-49) and make the playoffs five times – only to see the bottom fall out in the playoff games.

And when Prescott got injured in 2024, the Cowboys issues became plain for everyone to see. Mismanaged contract negotiations and two successive uninspiring drafts (2023 & 2024) left the Cowboys offense with few weapons and a shaky offensive line, and a defense full of holes that wasn’t held together by Dan Quinn anymore, and the bottom fell out on the 2024 season. Add another botched draft in 2025 along with another mismanaged contract negotiation that saw the Cowboys lose their best defender, and it’s no surprise the Cowboys strung together two losing seasons.

The Cowboys consider themselves a draft-and-develop team, except that they haven’t been drafting well for quite a while, which makes their stubborn refusal to participate in free agency even more baffling.

The front office is sitting on a decaying roster of their own making and is refusing to make any changes to the way they operate. Is anybody in the building looking at their abysmal draft record over the last three seasons and challenging the process? Is anybody challenging their personnel acquisition strategy? No way, Jose. The Cowboys are comfortable with the way things are, way too comfortable.

In the good ole days, Jerry used to take risks. Lots of risks, and many of them didn’t work out. And even though fans still love to crucify Jerry Jones for anything that goes wrong with the Cowboys, over the last 10 or so years, Stephen Jones has taken over more and more control of the day-to-day football operations in Dallas. And while 83-year-old Jerry retains all the fancy titles in the organization, Stephen is the guy running things behind the scenes.

Stephen has been involved with the Cowboys most of his adult life (he was 24 when his father bought the Cowboys in 1989), but he did not inherit Jerry’s penchant for taking risks. In fact, Stephen is running the Cowboys about as conservatively as possible, which is evident in the way the team handles player contracts and the salary cap, avoids free agency like the plague, was late to the game in analytics, and just now decided they would focus their draft efforts on Power 4 schools (eight years after the transfer portal was introduced) and avoid players with an injury history.

Because they’re comfortable with the way they are operating, they don’t want to change. They don’t want anybody questioning their personnel moves. They get to run the draft, they get to sign players, but they don’t do things many of the top teams in league do, people who are fighting for their jobs almost every year.

Nobody holds the Cowboys front office accountable. They are not held accountable if their boards are wrong, or if the players that they like can’t play, or if they don’t want to pay the players they like. Outside of the coaches, nobody’s job is in jeopardy. And if you occasionally get lucky in the draft, you are considered an expert talent evaluator.

The Cowboys have always been a stars & scrubs team, so that isn’t really new. But if you struggle to draft starter-level players outside of the first round, you’re hollowing out your roster. Fail to fill those gaps in free agency, and you will fail to beat teams with a better roster quality, and these are teams you regularly meet in the playoffs.

In this context, a look at the Cowboys’ draft performance by round is particularly instructive.

body .sbnu-legacy-content-table td, body .sbnu-legacy-content-table th, body .sbnu-legacy-content-table { border: 1px solid #000 !important; border-collapse: collapse !important; }

Cowboys performance by round
Round	NFL Rank	eAV Surplus
1	1	158.8
2	27	-22.6
3	6	54.4
4	2	80.3
5	26	-11.5
6	10	-1.6
7	32	-46.3

We already looked at the first-round performance and how much pure luck may have contributed to those results.

We looked at the second-round performance extensively in a recent post, and concluded that the root cause for the underperformance wasn’t really about drafting injured players, but more about reaching for a player in the second round was the primary cause of value collapse for the Cowboys.

When picking late in the first and all the way to the end of the second round, the Cowboys often seem to reach for specific traits (RAS) and end up with neutral or negative value. Mazi Smith (-9 reach versus consensus board), Sam Williams (-33 reach), Trysten Hill (-48 reach), and Luke Schoonmaker (-42 reach) were all reaches with elite athletic traits (RAS > 9.5) and all delivered negative surplus value. Internally, the Cowboys likely justified these picks with the players’ elite athletic traits, but the data shows this to be a high-risk and net negative drafting strategy when it ignores the consensus board rank, and the Cowboys have repeatedly run into this “Trait Trap”.

The single strongest driver of value in Round 2 is Reach/Steal, or in simpler terms: Players picked later than consensus rank (“Steal”) combined for a wAV Surplus of 29.6 points, while players picked higher than their consensus rank (“Reach”) combined for a wAV of -37.4 points. That’s quite a swing.

The third round is likely the round where the Cowboy scouting really shines, as the third round has delivered substantial value as measured by eAV. The same likely holds true for the fourth round, though the number here is inflated by Dak’s 49.3 eAV surplus. But even excluding Dak, the Cowboys would still rank sixth overall in the fourth round.

Rounds 5-7 all show a negative eAV surplus value, meaning the Cowboys have consistently failed to extract value from the later rounds. The seventh round is easily the worst: of the 22 players drafted since 2023, only two delivered a positive value, Terrance Mitchell (+2.6, though none of that was earned in Dallas) and Nathan Thomas (+1.1).

Realizing that you’re not consistently adding value to the roster with your picks in rounds 5-7, you’d be well-advised to follow three easy steps to fix this:

• Step 1: Trade your fifth-round pick for a veteran player.
• Step 2: Trade your sixth-round pick for another veteran player.
• Step 3: Trade your seventh-round pick for yet another veteran player.

And I’m only half-joking when I’m writing this. Consider these two trades:

• WR Brandin Cooks was acquired from Houston in exchange for a 2023 fifth-round pick and a 2024 sixth-round pick (the equivalent of a seventh-round pick in 2023)
• CB Stephon Gilmore was acquired from Indianapolis for a fifth-round draft pick.

Over the last 13 drafts, and Daron Bland (22 AV points) notwithstanding, the Cowboys averaged 4.7 AV over four years with their fifth-round picks and 1.0 AV points for their seventh rounders.

Brandin Cooks gave the Cowboys 11 AV points over two years, Stephon Gilmore delivered 4 AV points in just one year. That’s a win versus the four-year 4.7 eAV average any way you look at it. And even Logan Wilson, whom the Cowboy brought in for a seventh-rounder and then forgot to put on the field, gave the Cowboys 1 AV point in one year, which is what the average seventh-rounder delivers in four years.

So why aren’t the Cowboys doing this simple math and trading more of their late-round picks for veterans? Who knows, but probably because they don’t like to change things and are happy to sit there on the third day of the draft and strategize about players that will never work out in the NFL.

The end result of all this – three botched drafts in a row, reaching for players based on athletic traits, refusing to plug roster holes in free agency, and not using enough of their late-round picks as trade currency – is a hollowed-out roster that will not be able to compete against the better teams in the league. That may still be enough for a nine- or ten-win season if things go right, and maybe even a playoff berth. But it also means an early playoff exit.

Again.