It’s about how they derive that value. Ba had 8 tackles, which they value highly. Thing is, ba wasn’t even playing in the box. His tackles were mostly down field after first down yardage was gained. Like when wes lysack was leading the league in tackles and was cut, with no one picking him up. Biggie also had 8, and Thomas 4. So that’s 20 tackles with just 1 being for a loss. They are over valuing tackles because it’s easy to extrapolate.
Value for a db for instance should be some thing like pick, ff/fr, pkd, incompletion to their cover target, tfl, tackle no gain, then tackle short of yardage on 2nd/3rd, then tackle on a completion equal or less than 5, then less than qb avg per completion, then less than wr average. Finally you’d have negative value plays. Some times you get beat and some times another player makes a great play, but in order to be reasonably accurate you have to skew to a low score. A pick might be a score of 1, with a tfl a value of .5, a tackle that yields a first but with in the avg per completion a .1 and a long td surrendered being +1.
what I’m sure they are doing is int, tackle=tfl, ff,fr where the value doesn’t change much between each stat.