KarinsDad said:
How so? You’ll have to illustrate this with an example. The directional constraints are a lot less than squares and the distance constraints are small.
Maximum deviation from direct line is 30 degrees on a hex grid and 45 degrees on a manhattan grid. Thus, if we call the distance moved 1 (conveniently making it the hypotenuse), then the straight-line distance is cos 30 (0.87), 13% less, for hex and cos 45 (0.71), 29% less, for manhattan.
Yep, hexes are more accurate. I've also noticed 5 or more sided regular polyhedrons tend to be approximated by the brain as "sorta-circular" while 3 or 4 sided figures are not.
However, both hexes and manhattan feel - at least to me - restricted, that I'm moving in allies rather than freely. The 3.5 metric felt - to me - like "real" distance. Whether that was because the deviation was only 22.5 degress or that the uneven metric obscured the grid I don't know. Deviation is trickier, since (for example) moving 2 up and 1 across costs 2.5 squares of movement (1 diagonal, 1 straight) rather than 2.41 - the straight line is thus 10.5% shorter that what you actually pay, which is pretty close to that for hexes.
KarinsDad said:
And, DND is not played with orthogonal movement only.
Quite true. Nor is it played in only 3 axes. Any game on a gride ultimately fixates around the axes of the grid, which is one reason that more axes (even with non-simple metrics) provides fewer restrictions to play.
Beyond that, I find it easier to think in the 8 cardinal directions, which is why deviation on a diagonal square grid feels less "annoying" to me than deviation on a hex grid. The inherent limitation of the grid is off-axis and thus not as visible.
KarinsDad said:
Draw a line or hold a string up. Cover exists if a hex has 50% or more of it on the side granting cover. That’s just as easy of a rule as "if a line drawn through any" of 3.5 for squares.
Huh? How does that work if I run a long line across two adjacent hexrows, where it spends a large amount of time clipping alternate hexes? Yes, you could rule on the 50% crossing point, assuming you can eyeball this accurately, but that isn't nearly as simple as "if it clips it, it's in".
KarinsDad said:
I fail to see this as hindering game play in any way.
Obviously you're better at on-the-fly trig than I am; I struggled with it for about 6 years of playing Battletech, and is why I loathe mixing hexes and LoS.
TwinBahamut said:
Really, I know this has very little to do with the change from 1,2,1,2 counting and diagonal=1 counting, but it rubs me the wrong way. However, in "manhattan geometry" or a hex grid this situation is not present (or at least a lot better hidden), which I think is a decent enough argument for using one of those systems.
It's a function of degrees of freedom. With 2 axes, you only need 2 people to block half the movement options. With 3, you need 3. With 4, you need four. Try your example again with hexes by rotating onto the hex-row - you either need to block along a hex-row (and thus not truly normal to movement) or have a double row in the same way as diagonals need a double row.
The number of "extra" people compared to a "true" example is dependent on the error of grid-row to actual path. The 1.5 case is a little odd, due to the difference in movement metrics in each axis.