The 2023 NFL Draft is less than a week away and the number of methods used to evaluate draft prospects continues to grow. With millions on the line, teams continue to find an edge in evaluation and one of the latest methods is something called an S2 cognitive test.
We’ve seen the Wonderlic IQ test fall out of favor while the S2 test has possibly replaced it in evaluation. Not every team uses the S2 test and it’s new enough that there isn’t a ton of data for contextualization, but it is supposed to be a step up from the flawed Wonderlic.
Comparing the Wonderlic and S2 isn’t exactly fair though, as one is an “IQ” test, while the other tests different cognitive skills.
- Perception Speed
- Search Efficiency
- Tracking Capacity
- Visual Learning
- Instinctive Learning
- Decision Complexity
- Distraction Control
- Impulse Control
Soon to be drafted Florida QB Anthony Richardson explained the test like this:
“They want to see how fast you can recall things and notice certain things. On one of the [questions], you had to look at six balls that they highlighted and they move all over the screen. You had to pick out the balls and highlight them. I’m like, ‘How can I focus on six balls at once?’”
Statistics on the S2 Cognition site show that the test has a 28.7% predictability rate for a quarterback’s career passer rating, while the Wonderlic was around .01% Of course, much of that ability to process in real time is seen on the field, but if the numbers correlate with a scout’s thinking, it can only reinforce their evaluation.
The test has seen some recent attention in the NFL Draft world due to the test results leaking through Bob McGinn’s substack.
Multiple sources said Young’s total score was 98% whereas Stroud’s total score was 18%. Some other total scores in the class of quarterbacks this year were 96% for Fresno State’s Jake Haener, 93% for Kentucky’s Will Levis and Brigham Young’s Jaren Hall, 84% for Houston’s Clayton Tune, 79% for Florida’s Anthony Richardson and 46% for Tennessee’s Hendon Hooker.
The percentages are percentiles and not the way you think of grades from your time in school. CJ Stroud’s 18% isn’t him getting 18 questions right out of 100. But, in comparison to Bryce Young’s 98%, it’s of course easy to say Young did much better on the test.
UPDATE: How accurate the scores that have been leaked are, is being disputed by the people at S2 Cognition.
What do we make of @S2Cognition saying all the top QBs in this year's class (including CJ Stroud) "scored really, really well"? Is he just being nice?— Manufacturing Consent...and Nonsense (@trayNTP) April 15, 2023
Which is more likely: S2 is lying, or the media (that also claims that CJ Stroud is "uncoachable") are the ones lying? #Panthers pic.twitter.com/IAMb5IRm3r
How much this test sways teams in the range to draft Young and Stroud is completely unknown. More data needs to accumulate before anyone can feel confident that the test has a meaningful ability to predict how a QB performs in the NFL.