In 1923, British mountaineer George Mallory was asked why he would attempt to climb Mount Everest. “Because it’s there,” Mallory responded, a quote that would go on to be called “the most famous three words in mountaineering.” In 1924, Mallory died while climbing Mount Everest. His words are a testament to human bravery, or at least to the human compulsion to take on unnecessary challenges with little hope of success for no particularly good reason.
I think of Mallory’s words often when I see people make online sports predictions. We don’t have to pretend that we know what will happen in future sporting events, nor do we have to read predictions made by other people. And yet, if there is a sporting event, somebody will inevitably predict what’ll happen—because it’s there.
The Mount Everest of sports predictions is the seven-round NFL mock draft. Nobody has ever produced an accurate seven-round mock. This is not a reflection of the intelligence or expertise of the people who make them, but rather of the inherent impossibility of the task. So why do so many people attempt an exercise that is doomed from the start?
Predicting the first round of the NFL draft is much like offering a seven-day forecast. There is ample information available pertaining to what might happen—although meteorologists reading weather data never have to consider if a low-pressure system is really a smokescreen designed to mask the weather’s true intention to select a quarterback. Since both weather and the NFL draft are extremely complicated, it is helpful to have an expert provide a reasonable approximation of what could happen. Besides, these predictions give us something to talk about at the water cooler. It’s going to be beautiful on Sunday—we’re having a cookout! The Browns might take the kid from Wyoming who doesn’t know how to throw a football—we’re having a cookout!
At a certain point, it’s like, ‘Oh, I’ve spent 20 minutes thinking about where Trent Taylor is going to land in the fifth round, and even though I really like Louisiana Tech slot receivers, maybe I should just pound this through.’Dieter Kurtenbach, Bay Area News Group columnist
Predicting all seven rounds of the NFL draft, however, is akin to giving a 365-day forecast. There are simply too many variables in play to make anything resembling an accurate guess. It is harder to predict who the Panthers will take with the 234th pick in this year’s draft than it is to predict the exact temperature on October 16. After all, October can’t trade its 16th day to June as part of a package deal to pick up an extra weekend.
“They’re literally impossible to get correct,” says Dieter Kurtenbach, who put together multiple seven-round mock drafts last year for Fox Sports. “You’re trying to put out a static product predicting a very fluid event.”
Despite the Sisyphean nature of the task, Kurtenbach, who now works for the Bay Area News Group, tried his best. He says that in the run-up to releasing his final 2017 seven-round mock, he would start working when his girlfriend went to bed at 10 p.m. and still be working when she woke up at 7 a.m. Kurtenbach’s job called for him to “spend all day watching college football [on Saturdays] and then all Sunday watching the NFL,” and a previous gig covering mid-major college football left him feeling “like I knew the lower rounds of the draft pretty well.” He accepted the assignment as “a fun challenge.”
But it was more than he bargained for. “I tried to spend most of my time finding interesting, cool fits for the later rounds. But at a certain point, it’s like, ‘Oh, I’ve spent 20 minutes thinking about where Trent Taylor is going to land in the fifth round, and even though I really like Louisiana Tech slot receivers, maybe I should just pound this through.’”
For all his expertise and effort, Kurtenbach got one and a half picks right in the second through seventh rounds. He accurately projected that the Buccaneers would select Boise State running back Jeremy McNichols in the fifth round—McNichols was later cut during training camp—and that the Rams would take South Alabama tight end Gerald Everett—correct in theory, although Los Angeles took Everett in the second round, whereas Kurtenbach had him going in the third.
One and a half accurate picks in a six-round stretch might sound like a catastrophic failure, but it was par for the course. Bleacher Report’s Matt Miller, who also writes a seven-round mock draft, acknowledges that the exercise doesn’t lead to a lot of accurate projections. “I’m trying to hit on positions more than players after pick 100,” Miller says via email. “If the Patriots have three picks on Day 3, I’m more worried about getting close to what positions they draft and not as focused on nailing the player to the team. Any of those you get are either luck or you’ve been given the information directly from a source, so it’s already penciled in.”
Miller’s projections from last year reflect that mind-set. His fifth-round mock was wildly inaccurate; he didn’t project any players to a team that ultimately drafted them, and of his 40 projected fifth-rounders, only one (wide receiver Shelton Gibson) was even selected in that round. But he correctly assessed that the Bengals would use their fifth-round pick on a kicker—they took Jake Elliott, not, as he projected, Zane Gonzalez—and that the 49ers would use their fifth-rounder on a tight end. (Miller projected Jonnu Smith, who was already off the board in the real draft, and the 49ers took George Kittle.)
To get a sense for how futile the average seven-round mock is, I tallied up the results of seven such mocks from last year: Kurtenbach’s at Fox Sports, Miller’s at Bleacher Report, Peter Schrager’s at Fox Sports, Walter Cherepinsky’s at Walter Football, Chad Reuter’s at NFL.com, Cole Thompson’s at USA Today, and R.J. White’s at CBS Sports, all published during the week of the draft. Here’s how they did:
- The mockers started out all right: All seven correctly predicted that the Browns would select Myles Garrett with the first overall pick and that the Panthers would take Christian McCaffrey at no. 8. Four of the seven also had the Jaguars grabbing Leonard Fournette at no. 4. As it turns out, a first-round mock draft can be useful and effective!
- After the end of the first round, the seven mocks got just four combined picks correct. I’m not saying the mocks had an average of four accurate picks, or that all seven agreed on four correct picks. I’m saying that in the 221 picks following the first round, only four individual picks were correctly predicted. Miller forecast that the Jets would take safety Marcus Maye in the second round; Thompson suggested that the Chargers would take safety Rayshawn Jenkins in the fourth round; Kurtenbach projected that the Bucs would take McNichols in the fifth round; and Schrager predicted that Washington would take cornerback Joshua Holsey in the seventh round. Among these mocks, the average hit rate for picks after the first round was 0.26 percent.
- If we include near-hits—for example, Reuter projecting tight end Adam Shaheen to the Bears in the third round, when he was actually taken by the Bears in the second round—the mockers combined to get 45 picks right beyond the first round. That’s a 2.9 percent success rate.
- Hypothetically, mock drafts should be useful for understanding which players will be selected in a given round, but they fall well short even at that. From the fourth round onward, these mocks projected just 13.7 percent of players into the accurate round. In the 34-pick sixth round, each of the seven mocks failed to project five players actually selected in the sixth round.
- Five of the seven mock drafts featured a player who eventually went undrafted as high as the third round. All seven included eventual undrafted players in the fourth round.
Schrager had the most successful mock by a significant margin. He explained his process in a video on NFL.com: “I don’t watch the tape ... but what I have is this!”—he triumphantly held up his phone—“and I’ve been texting and calling and trying to get as much information as I can from teams.” It worked better than anybody else’s plan: He made 11 right or almost-right picks after the first round, with at least one in six of the seven rounds. Most mockers had Cam Robinson going in the first round, but Schrager accurately slotted him to the Jaguars in the second. He was the only mocker who put Offensive Rookie of the Year Alvin Kamara on the Saints, albeit a round early; he was the lone mocker to get a pick completely correct after the fifth round. But what qualifies as a massive success by seven-round-mock standards would be terrible in almost any other field: Getting 11 picks almost right after the first round means getting 211 picks completely wrong.
The people who assemble seven-round mock drafts do not fare poorly because they don’t understand the NFL draft. Most of these writers know the draft process extensively and are better informed about late-round prospects than basically everyone who isn’t employed by the league. These people fared poorly because putting together an accurate seven-round mock draft, by its very nature, is impossible.
“I really don’t like thinking about people actually using my [seven-round mock draft] as a valuable service,” Kurtenbach says. “That kinda fucks up the whole charade.”
To Miller, putting together a seven-round mock is a helpful exercise, even if it doesn’t lead to many accurate projections. “People who look at mock drafts as predictions will see a seven-rounder as pointless,” he says. “A seven-rounder makes me think about team needs, player values and forces me to look at drafting trends for each team.”
People who look at mock drafts as predictions will see a seven-rounder as pointless.Matt Miller, Bleacher Report NFL draft writer
But while draft aficionados like Miller might see a seven-round mock as an intriguing thought experiment, the people who widely consume them online often take the content literally. “I know exactly how people read it,” Kurtenbach says. “They would click on it, they would hit control-F, they would look for their team, and they would see who their team picked. And then they’d get happy or mad.”
Seven-round mocks serve as the foundation for a miniature internet economy: National NFL sites post them. NFL team blogs write posts rounding them up—or create their own seven-round mocks, featuring exclusively their own team’s selections. NFL fans click.
It makes sense why this is so appealing. Fans want to have an idea for which players could end up on their teams. Yet for those who don’t know much about non-first-round prospects—that is to say, almost every NFL fan in existence—it’s tough to determine where to start. Thankfully, somebody else knows a lot about these players, has a feel for each team’s needs, and spent hours compiling a seemingly authoritative list of players their team might take. Only being authoritative is hardly the goal.
“If the end goal is to better educate football fans about the players that are available in the draft and for them to better understand the prospects at hand, the mock draft is almost the worst possible way to do that. It doesn’t bring context or deep analysis,” Kurtenbach says. “If fans really wanted to know, they would go read [NFL.com analyst] Lance Zierlein’s scouting reports. There are so many different $20 e-books you can read.”
For true draft aficionados, seven-round mocks can have value. Kurtenbach’s could have served as an introduction to otherwise uncelebrated players from small schools. Miller’s offered insight into the thought processes of various front offices on draft day. The key is understanding that the mockers themselves realize their predictions are generally going to be wrong and are using the mock as a method of exploring angles that other analysts might overlook. But I suspect most readers use these mocks as a starting point, not as a piece of the larger draft puzzle.
Still, there’s a difference between George Mallory and mock drafters. Mallory tried climbing the world’s tallest mountain just because it’s there, while mockers try to predict the world’s least predictable sports thing because of at least one more fueling motivation.
“They get beaucoup pageviews,” Kurtenbach says.