Chances are this is really dumb

The attempt to quantify the likelihood of Will Levis's draft-day tumble into the second round made us all dumber. I'll explain.

According to ESPN, there was a less than one percent chance of Will Levis going drafted in the first round.

A zero-point-one percent chance according to the big brains of the ESPN Analytics department, and if I was a normal person, I would be able to shrug this off as just one more piece of drivel that gets spooned out in the name of content, giving it no more credence than whatever inanity Skip Bayless is going to offer this week about Lebron James.

But I am not a normal person. I am someone who gets entirely too wound up about entirely too many things, and one of the things that causes me to clench my teeth in frustration is when someone uses demonstrably flawed, monkey-brained methodology to construct an equation that produces a number that is presented as if it was an actually insightful measurement. I hate it. I hate it almost as much as I hate someone driving the speed limit while they’re traveling in the left lane of the highway.1 I hate it so much that I want to come up with a new name for this particular type of fraud in which stupidity is dressed up with math and disguised as a percentage. A stupistic? Statupidity? Stupalytics?

I am very pro-analytics. I want to make my bias clear right off the bat. I like math. I trust math. I was better at math and science in high school than I was at English, and I believe that computer science and statistics can be utilized to reveal game-changing insights into the sports that I have followed since I was a wee little lad.

I like “Moneyball” and Big Data and the incorporation of quants into the architecture of not just professional sports franchises, but the people who cover those professional sports franchises. But we’re reaching the point that people are trying to put percentages on things where it’s really not appropriate or applicable. The result is providing a less clear understanding of the situation, which strikes me as the exact opposite of what information is supposed to do.

If you’re not familiar with Will Levis, he was a quarterback from Kentucky who had previously been at Penn State. He was expected to be one of the first five quarterbacks chosen in the draft. There were some people who thought the Houston Texans would choose him No. 2. There was a moment when an anonymous report on Reddit goosed the betting market of him going No. 1 overall.

However, Levis was not one of the three quarterbacks chosen in the first four picks. In fact, he was not chosen until the second pick of the second round, No. 33 overall. Before the second round began, ESPN quantified exactly how improbable it was that Levis would go unchosen in the first round.

OK, that’s absurd.

A truly accurate measurement of the chances that Levis would go unchosen would require a number of repetitions of this specific event. A meaningful measurement of the chances that Levis would go unchosen would require access to the draft board of every team that picked in the first round. If all 32 teams had Levis evaluated as a top-five talent in this draft, and he went unchosen, then you could say this was wildly unlikely. However, ESPN Analytics did not have access to the draft board of every team. It did not have access to the actual draft board of even ONE NFL team.

Instead, it had access to the countless mock drafts that are generated around the country, and something that is often referred to as the “Consensus Mock Draft” which seeks to average all of these coming up with a general consensus on the ranking of the players available in this draft.

I’m not going to bag on the value of mock drafts here because there are a lot of smart people who spend a lot of time not just evaluating the players who are available, but talking to the NFL executives who are also evaluating those players. I think that a consensus mock draft does – by and large – reflect the general evaluations held by NFL teams. That reflection is far from perfect, though, and every year there are players who are thought to be top-10, maybe even top-five picks who start plummeting and everyone starts asking why.

In my experience, this is generally because there’s something that NFL teams know or understand about this player that has not filtered out to the people constructing mock drafts. Other people choose to see this as a sign of the stupidity with each team that passes this prospect that everyone has decided is a top-five talent.

Sometimes, the reason for a slide is obvious as it was on Thursday when Jalen Carter went ninth to Philadelphia. When the college season ended, many thought Carter would be the best player available in the draft. He had an arrest warrant issued on a charge of reckless driving over his actions on the night of a fatal crash and had gained weight and appeared out of shape at his pro day. Something similar happened to Warren Sapp and to Laremy Tunsil when a video of him using a gas-mask bong was posted to his social-media account the night he was drafted in 2016.

Most often, an inexplicable plummet relates to a player’s health. A knee or perhaps a shoulder that teams fear will impact that player’s longevity. Examples of this would be Myles Jack, the linebacker from Bellevue’s Academic Institute who went on to UCLA. Many people expected him to be a top-10 pick, but he went in the second round to Jacksonville. Chris Polk, the running back from Washington, went undrafted entirely.

Mock drafts are specifically susceptible to catching this sort of development, though, for the very simple reason is that there’s no incentive for any team to correct the overestimation of a certain player’s draft stock. In fact, there’s a very big incentive for NFL teams to hope that player does in fact get drafted that high.

But if you’re relying on the mock drafts, you’re going to think it’s wildly unlikely that this player had slipped. You’ll say, “This is crazy!” or, “I can’t believe this is happening!” when all the guys inside the NFL draft rooms around the country will quietly nod, and think, “I guess other teams came to the same conclusion we did.”

I’m going to wrap this up with a story from a different sport. My first job as a pro beat writer was covering the NBA, and the first draft I covered was 2003 when the Sonics had two first-round picks. That year, there was a Polish player named Maciej Lampe, and ESPN’s Chad Ford was absolutely smitten with him.

Now, the NBA Draft has never had draftniks quite like the NFL. There was no Mel Kiper equivalent. Ford was carving out that kind of niche, though, and he had an obsession with foreign players, specifically Europeans, which was understandable because the NBA had an obsession with former players, specifically Europeans. Some 7-footer named Nikolas Tskitishvili had been drafted No. 5 overall the year before.

Well, Ford was insistent that Lampe was a lottery pick, and as beat reporters started rolling out their mock drafts, there was Lampe up in the top 10. Several places had him going fifth overall to the Miami Heat. If not fifth, the expectation was that Lampe would be drafted No. 9 by the Knicks.

Well, the Heat ended up taking Dwyane Wade, which Ford criticized them for. The Knicks took a guy named Michael Sweetney from Georgetown with the ninth pick, and then – in the second round – chose Lampe with the 30th pick.

“I've seen Lampe play enough to know that his slide wasn't warranted. Lots of teams dropped the ball on him. If he can work out his contract issues with Real Madrid, he'll make Knicks fans happy.”

— Chad Ford, 2003

Now, the Sonics had the 12th and 14th overall pick that year, which they used to select Nick Collison and then Luke Ridnour. After the draft, I asked Seattle’s GM about Lampe and what happened. He said that he never understood all the buzz about that player, but he hoped like hell that some other team was going to pick him in the lottery because he was very worried whether Collison was going to be available at No. 12 so he wasn’t about to throw any cold water on that buzz for Lampe.

Sure enough, one of the local guys who is now a play-by-play announcer for an NBA team, wrote a column in the paper I worked for and stated bluntly that, “It is possible for for Lampe to slip all the way to No. 12. The Sonics would grab him in a second.”

Well, Lampe did slip all the way to 12. The Seahawks didn’t grab him. They didn’t grab him with their second first-round pick at 14, either.

If ESPN had an analytics department back then, it might have very well said there was a 0.1% that Lampe slipped all the way to the second round. A number of fairly plugged in reporters seemed certain he was going No. 5 overall. Everyone seemed to say he was a lottery selection.

But those mock drafts and projections were not based upon the view of the people actually making the picks, and to accurately quantify the odds of a player going unchosen, you would actually need to know that information not how a bunch of people covering the draft thought it would go.

Lampe never saw the floor for the Knicks, was traded to the Suns midway through his rookie season and averaged 3.4 points in the 61 games he appeared in during his four seasons in the NBA. Ford caught heat a fear years back when it became clear someone was retroactively altering the grades he applied to draft picks.

Join the conversation

or to participate.