Benimoto said:
Yeah, but you could reverse that and legitimately ask why you would complicate a system just because of an assumption that the current rules don't meet the precision needs of the target audience.
The current rules do meet the precision needs of the target audience.
You could do things to make the 3.5 system more precise. You could break everything down to precise measurements in feet and say that diagonals are 7 feet, or even 7 feet, 1 inch long. Or, instead, you could require a player to keep track of whether he's moved an odd or even number of diagonals, so that the 1-2-1-2 count could start up correctly on the next turn instead of resetting.
A) The five foot squares have been accepted in gaming well beyond D&D or even RPGS.
B) You still have not addressed the fundamental difference between accepting that some error exists and an intentionally addition of new error onto the system
C) The drumbeat of 4E is "more simple", so new complexities are a non-starter
D) I'm not convinced that an error of 6% is at all significant being as it seems very safe to assume that all humans vary their rate of movement in combat situations anyway. 6% is probably high, but it is reasonable. 40% is absurd.
The thing is that the system is already fairly imprecise. As I mentioned earlier, isn't a little weird that most of the entire human race moves at a speed of exactly 30 feet? Is that really how far someone can walk in 6 seconds? Isn't it weird that all difficult terrain exactly halves your speed? We could easily make these systems more precise as well.
Again, there is a fundamental difference between accepting implicit error and adding error that the system doesn't require. With all the D&D assumptions in place there is a "best" answer. The intentional addition of error is a whole new issue.
They're all abstractions, and all compromises between precision and usability. Your complaints are legitimate, but why should your ideal system get preference over all the people who had been counting 1-1-1-1 all along and liked their system?
If it is an abstraction worthy of ignoring 40% error then why the hell does anyone care if they forget whether the next move is a 1 or a 2 and they might add another 5% of less error? Seriously, the entire reason for doing this is to avoid an error in counting 1/2/1/2. And the "solution" is to make everyone a lot more wrong always.
Anybody who truly believes that the 40% error of 1/1/1/1 is ok because it is an abstraction can't possibly care if their movement was off by 5% because of 1/2/1/2 mistakes. The solution contradicts the very foundation of the problem it claims to solve.
And if you are just going with a majority rule answer, then I firmly believe that 1/2/1/2 would win that handily.