Slitherine

Edit: DERP , I just realized this has got to be the effect of high morale, I guess you can safely disregard this report.




While fiddling with the editor trying to figure out the exact formula for accuracy, I came across what I think is an inconsistency in the way it appears to be calculated, or at least displayed in game in the hit chance / expected damage box. The Ranged accuracy of all units seems (again this is going purely from the displayed hit chance) to be higher by 10% than what is displayed in their unit card.

Ordering a SL Destroyer Upgraded ( 80% ranged accuracy, 100% Weapon accuracy with 0% loss per hex) to target a squad of marines sitting on an Ash tile 4 hexes away with no line of sight obstacles in the way, one would expect a straightforward 80% hit chance (0.8 * 1). Such is not the case however: Now at this point I must say I had already figured by playing the game that the loss of accuracy over distance was generally off by ~10%, and thought maybe that simply meant the first hexagon of distance was free (I hadn't taken a closer look at weapons losing more than 10% per hex). But of course in this case you wouldn't expect a weapon with no accuracy loss over distance to be affected by it.

After experimenting with a couple units of Ogryns Veterans (70 Ranged accuracy, 100%/-20% ripper guns), I found that their (predicted) chances of hitting those marines in the open at 2 and 1 hex of distance were respectively 48 and 64%: Playing with the numbers a bit, it couldn't be that the first hex was free as that would instead give them 56% and 70%.
The only way those numbers make sense, is if the ranged accuracy of the shooter is artificially increased by ten percent.

The formula would therefore be ( shooter accuracy + 10% ) * (weapon accuracy * (100% - accuracy loss * distance) )
For two hexes of distance: ( 0.7 + 0.1) * ( 1 * (1 - 0.2 * 2)) = .48
For one hex of distance: (0.7 + 0.1) * (1 * (1 - 0.2 * 1)) = .64

I can't find any explanation for that extra 10%.

This increase seems to appear consistently even after you add the complications of cover. Replacing the Destroyer in the initial test with a Baneblade (90% Ranged accuracy, Baneblade cannon with 100%/-10%), adding two sparse forests in the way (two 10% LOS obstacles) and entrenching the marines in a Fortification tile (50% cover), we get 24% accuracy at 4 hexes: Updating the formula to take cover into account : ( shooter accuracy + 10% ) * (weapon accuracy * (100% - accuracy loss * distance) ) * cover * ( 100% - (sum of Block LOS values of tiles in the way) )
( 0.9 + 0.1 ) * (1 * ( 1 - 0.1 * 4)) * 0.5 * (1 -0.1 * 2) = 0.24