Will Levis’ manufactured hype exploded in everyone’s faces.
Will Levis is still on the board after the first round of the 2023 NFL Draft. This should not be a surprise to anyone who watched Will Levis play football. The only way the Kentucky QB was being taken in the top-five of the 2023 NFL Draft is if a GM chose to ignore every single red flag, and blindly throw a dart at a board hoping Levis would be the second coming of Josh Allen.
This was definitely enough for a lot of mock drafts to put Levis early in the draft. Hell, we did it too, but mock drafting is always educated guesswork. However, ESPN put up one of the most confusing, nonsensical stats ever after the first round on Thursday night.
Will Levis had almost zero chance to *not* be picked in the first round, according to ESPN Analytics.
Where could he land tomorrow? pic.twitter.com/LVEsZZjmh7
— SportsCenter (@SportsCenter) April 28, 2023
To anyone who appreciates real analytics this graphic is maddening enough to cause physical pain. By nature, analytics involves studying data sets to determine trends, or in the case of predictive modeling you’re using data to project an outcome which is unknown. This isn’t new stuff, either in sports or beyond — but when you’re touting something with a 99.9% probability, like Levis being drafted in the first round, it’s a sign something is extremely wrong with your formula.
We really don’t need to dive into the workings here and try to assume what ESPN was doing. The only way a 99.9 percent Levis selection makes sense is if their “analytics” were evaluating betting trends or mock drafts on the internet. We know this because looking at the actual known quantities about the draft, there’s no way this figure could ever have been true.
Here’s why:
Let’s work off what we knew entering Thursday. It was widely accepted that we had a pivot point in the draft between the top two QBs in the draft, Bryce Young and C.J. Stroud — and then Anthony Richardson and Will Levis. The former represented the safer, NFL-ready duo, and the latter based on promise and upside.
If you liked an upside quarterback, then Richardson was easily worth rolling the dice on over Levis. Richardson out-performed Levis at the combine, represented a true dual-threat QB, which the draft didn’t have otherwise, and better fit a team like the Colts (who took him at No. 4) where new head coach Shane Steichen just came off making a Super Bowl as an offensive coordinator with Jalen Hurts.
If we look at the teams that absolutely had to take a quarterback in the 1st round this year because their room was so bad it’s not even sustainable for a season, then you truly only had four teams: Panthers, Texans, Colts and Buccaneers. With Tampa Bay at No. 19 already commenting that they weren’t targeting a QB in the first round.
Levis being QB4 was now trying to fit into a three-team picture. So we next move down to the teams that need a QB in the next year. That’s the Titans and Vikings — both or whom had some rumors about being interested in Levis.
From there, the last remaining subset of teams are those who don’t need a QB in the next year, or even two years, but the value would be too good to pass up. This year that’s Las Vegas, Seattle, Detroit, Washington and New England.
In totality you now have 11 teams total who would entertain taking a QB. With Young, Stroud and Richardson likely going before Levis that number is reduced to eight. Five of these eight teams don’t even really need a QB, all of whom have bigger fish to fry.
We can’t use predictive modeling to tell us the likelihood Levis would be drafted by one of these teams. There is no functional analytic method which can give us a 99.9% probability, because to reach that number you would either need absolute statistical certainty that at least one team was taking Levis, or significant enough figures to indicate that 28 teams might take Levis.
There were only eight. If we really buckle down at look at the factors in this draft the only teams who reasonably could have taken Levis were the Texans, Colts, Seahawks, Lions, Titans, Buccaneers, Patriots, and Vikings. Nobody else made any real sense to take a risk on him.
In the most rudimentary way, with the Seahawks and Lions having two picks each, that’s a 32 percent chance Levis is selected on Thursday night, a 68 percent chance he’s not. Keep in mind this extreme disparity between the most basic possibility that 32 percent of teams would think about drafting Levis, with “analytics” showing that number is 99.9 percent.
I’d love to know how ESPN Analytics arrived at their figure, because it makes no damn sense.
