Accuracy Experiment

[Roustabout]

I haven't seen much information in these forums about modifying the the accuracy value in the land unit tables. Moreover, since nobody responded to a recent post I made asking about the effects of modifying the default ranged accuracy of 5, I'll conclude that little is known about if and how it impacts the lethality of ranged units. As such, I conducted a little test to see if modifying a ranged units accuracy actually does anything. I've listed the procedures I used and the results of my test below.

Map and Units:

1. Custom battle on Nemetocenna w/ me playing as Parthia (attacker) vs. computer playing as Egypt (defender)
2. My unit = Noble horse archer general
   
   • Modded to have only 5 ammunition so it would not kill the enemy unit before its ammo ran out
   
   
3. Computer's unit = Egyptian Slinger General
   
   • I chose the slingers b/c I conducted the test myself and needed a way to keep the enemy unit fixed in place (i.e., ranged units don't charge into melee as long as they have ammo)
   • Modded to be usable as a general in custom battles
   • Moral set to 99 so it would not run away before my unit expended all its ammo
   • Sling damage set to 0 so it would not kill any of my horse archers

Procedure:

1. I tested noble horse archers with 0, 5 (default), and 10 accuracy. Five trials with each accuracy setting were conducted.
2. At the beginning of each trial, I spread my horse archers into 3 ranks and double clicked an attack order on the enemy slinger general. The trial ended when my general's ammo was spent.

Results:
Mean number of kills: 0 accuracy = 48; 5 accuracy = 52; 10 accuracy = 62
Minimum # of kills: 0 accuracy = 42; 5 accuracy = 48; 10 accuracy = 58
Maximum number of kills: 0 accuracy = 54; 5 accuracy = 60; 10 accuracy = 66
Mean kills per volley: 0 accuracy = 9.6; 5 accuracy = 10.8; 10 accuracy = 12.5

Conclusion:
Accuracy appears to have a small effect on the lethality of ranged units. For instance, doubling accuracy from the default value of 5 to 10, resulted in a 9% increase in the number of kills over the course of 5 volleys. If you are looking to reduce the effectiveness of ranged units in TWR2 (and you should be! ), then reducing accuracy from the default value of 5 to 0, would modestly reduce the lethality of ranged units. I'm not sure if accuracy can take on negative values, however. If so, that would allow for quite a bit of customization of ranged units without having to mess with a bunch of other attributed like ammo, damage, and ap. I'm also not sure if the effect of accuracy is modified by distance to target. In all of the trials I conducted, the two units were approximately the same distance apart.

Cheers!

[Magnar]

each projectile gives an accuracy bonus of around 30 for arrows i think so a difference of 5 is really only a 1/6 increase in accuracy. Then it is also relative tothe base calibration area.

[Roustabout]

So bonus accuracy (from the projectiles table) adjusts the overall accuracy of different types of projectiles, while accuracy (from the land unit table) adjusts the accuracy of individual units? If so, lowering bonus accuracy across the board would reduce the overall lethality of ranged units, and then you could adjust accuracy to make certain units slightly more accurate than others (e.g., elite units could have higher accuracy than their non-elite counterparts). Is that correct?

@Magnar - what do you mean by "base calibration area"?

Thanks!

[Magnar]

look in the projectiles table

thats where all the stats for missiles is. If you want to edit missile units in anyway you should always start here. We are currently working on a missile overhaul ourselves