Let me start off by pointing out that the bolded part here is simply inaccurate, WV. The dropoff in talent is most drastic after the first 8-10 elite prospects are taken. The talent pool from from the 11th player or so on my board, all the way to the 50th best player on my board is typically very similar. The same applies to the next 40-50 players, and so on. There's a lot of hair splitting that goes on when distinguishing these prospects from one another.
By the same token, the talent level from round 4 to round 6 is also very similar. The first steep dropoff in talent in any draft is after the best 8-10 players come off the board, not necessarily the first 8-10 picks. The next steepest dropoff in talent is found after the first 150 or so players are taken. The talent pool for the final 100+ players selected extends all the way to the UDFA pool after the top 150, is just as closely together as the talent from 11-50, just less of it (talent). The talent begins to become extremely watered down in this range. You're hoping to find players that can simply make your roster and provide depth in this range.... which is a longshot.
This is why getting a player of 3rd round ability in the 7th round range is MORE valuable than getting a 1st round player in the 3rd round if the expected linear value is the same. A team got a potential starter here in the range where selecting a guy who can even make your roster is a longshot. That's value.
If you go back to Awsi Dooger's comment regarding how according to my board, Sharrif Floyd should be the best value pick simply because he was the #1 player on my board, the explanation I just gave you should answer his question as to why Floyd wasn't the best value pick of the entire draft. It's because the talent pool where Floyd came off was still populated with elite talent. However, Floyd was the best value pick of the 1st round in my opinion. Which is why there's only one player selected in the 1st round that made my top 64 value picks.... it was Sharrif Floyd.
The biggest reach using my rankings was not Travis Frederick. I had a 3rd round grade on Frederick (99th overall) on my board. Fact is, there were probably bigger reaches in the draft that didn't even make my top 26. The reason they didn't make my list is because I had not scouted those players, therefore I can't comment on how big of a reach that selection was. There were players taken that I had never even heard of.... and if you're taking players that I haven't scouted or heard of before, that's saying something. Trust me.
Selection value simply based how they were ranked according to my board is best explained using the absolute difference in a linear fashion that I provided. Asking me who I think are going to be the best pros regardless of where they were selected is an entirely different subject. Corey Lemonier is a fine example.
This can't be right. The talent pool almost certainly follows a normal distribution (bell curve) as WV suggests and not the linear distribution you described. That means that the difference between the 22nd ranked player and the 42nd ranked player is always going to be greater than the difference between the 42nd ranked player and the 62nd ranked player.