First of all, "complement" of something is everything that something is NOT. Cassel's complement on the field are the 10 guys NOT playing QB.
If I have a probability of 25% of failure, then I have a 75% probability of success. The complement of failure is success, so the probability of success is 100% - 25% = 75%. In decimal notation: 1 - .25 = .75. Conversion to decimal form is easy: 75% is 75 per hundred (per cent - per century - century is 100, hence 75 divided by 100). It's the decimal representation that lets us do the math. The "1" corresponds to "100%".
Product principle: The probability of 2 independent events occurring simultaneously is their product. Product means multiply. The probability of 2 players in a row being a big hit, if the probability of success were 25% is .25*.25 = .125. Very small.
But the purpose of this strategy is to get a HIT on AT LEAST ONE. One will get you where you wanna be. More than one is just fine, too.
At first blush, a 25% chance of a good OL on any one try seems like a shitty proposition. Why bother? But what if you bring in 5 guys with that same shot? The way you calculate the probability of one hit is to find the (easy-to-find) probability that THEY ALL FAIL. The complement of this calculation is the probability of at least one hit. Just subtract (ALL FAIL) from 1 and you have it.
The probability of all failures is the product of all the failure probabilities. Let's go with 25% or .25, just because...:
Probability of all failures is the product (.25)(.25)(.25)(.25)(.25) = 0.000976563. A very SMALL probability. This means that the probability of ONE guy being a quality starter is, in theory, 1 - 0.000976563 = 0.999023438, or 99.9% (approximately). That's virtual certainty! (keeping in mind that 75% probability of any individual player being "the guy" is fairly generous)
So a guy with some cap room to play with, MILES (and months) away from adding a player directly through the draft, SHOULD prefer to have a large number of prospects, rather than sinking a large amount of dough into ONE guy, even if he's the best player available, with, say a 75% probability of solving all your problems. (Probably generous, even for the big name FA tackles who keep getting mentioned - even Tauscher, if he suddenly can participate fully is close to 50-50, imo)
That, I believe, is the genius behind a Tom Brady pick. It wasn't that he was the guy they knew would lead them to a SB. He was a pretty good player, and they had quite a few pretty good players come through, and the ones that did better stayed longer, and a great player emerged from the chaos. New England NEVER had their heart set on the guy, and, until he wowed as Bledsoe's replacement, they could've parted ways with him and few tears shed.
Nothing dramatic. Not genius-level talent evals (although certainly want that to be as good as possible). Just simple math.
Maybe, just maybe, this simple math applies to bringing in Goff, Ghiaciuc, and now these two guys from Miami. On the face of it, a less-than-50% shot for any ONE of these guys, taken individually, to be a hit. But the probability of AT LEAST ONE bein' the guy we need (or who will serve well enough) grows with every guy thru the turnstile. Furthermore, the modest salary hit for each one leaves them plenty of flexibility to make yet more moves in the future. Seems a lot more sensible than just putting all your eggs in the Levi Jones or Jon Runyan basket.