Effectiveness of accuracy stat

[Shazor]

Sooo, i always wondered how exactly the accuracy stat effects the "lethality" of a bow unit and if its worth it to increase it but i nearly found any information about that so i did some research myself but keep in mind that i use darthmod so the results might differ from vanilla shogun 2.
I always set up the same scenario with three different accuracy values. 40 for the default accuracy value of the darthmod bow samurais, 100 for checking if accuracy actually caps at 100 and 400 for an extreme comparison value and i repeated the setup 5 times per accuracy value to average the results a bit.
My setup was one bow samurai unit in always the same position, formation and the same map vs one yari ashigaru unit controlled by the ai who charged at my archers.

The results:
I always wrote down how many ashigaru units were left of the 375 ashigaru after they reached my bow units and started melee combat so i could see how many got killed by arrows only.
For accuracy 40: 240,287,281,261,252 which averages at about 264
For accuracy 100: 236,204,235,212,240 which averages at about 225
For accuracy 400: 190,166,184,173,169 which averages at about 176

Sooo, this should be at least be proof that accuracy isnt capped at 100 and that accuracy doesnt mean hit percentage and actually scaled pretty high but there seems to be quite a "soft cap" but it also could be that my results are quite garbage because i only did 5 runs with each setup.

What do you guys think about that and/or what is your theory/research about this topic?^^

[YXT]

Sooo, i always wondered how exactly the accuracy stat effects the "lethality" of a bow unit and if its worth it to increase it but i nearly found any information about that so i did some research myself but keep in mind that i use darthmod so the results might differ from vanilla shogun 2.
I always set up the same scenario with three different accuracy values. 40 for the default accuracy value of the darthmod bow samurais, 100 for checking if accuracy actually caps at 100 and 400 for an extreme comparison value and i repeated the setup 5 times per accuracy value to average the results a bit.
My setup was one bow samurai unit in always the same position, formation and the same map vs one yari ashigaru unit controlled by the ai who charged at my archers.

The results:
I always wrote down how many ashigaru units were left of the 375 ashigaru after they reached my bow units and started melee combat so i could see how many got killed by arrows only.
For accuracy 40: 240,287,281,261,252 which averages at about 264
For accuracy 100: 236,204,235,212,240 which averages at about 225
For accuracy 400: 190,166,184,173,169 which averages at about 176

Sooo, this should be at least be proof that accuracy isnt capped at 100 and that accuracy doesnt mean hit percentage and actually scaled pretty high but there seems to be quite a "soft cap" but it also could be that my results are quite garbage because i only did 5 runs with each setup.

What do you guys think about that and/or what is your theory/research about this topic?^^

Thank you so much for sharing your research! I've been looking on the internet for ages about the accuracy stat! It was for me deciding if I should upgrade my Samurai Hero to 120 accuracy or keep at 100, because I was afraid I might waste the 20, since I thought accuracy was in percentage. Your research proved that wrong so thanks a lot!

[peterdavid12345]

This is a fantastic research! Thanks a lot man!

[baselhun]

Yes it isn't capped, there are also other factors such as, calibration area, spread, and marksman bonus in projectiles tables. You'll see matchlocks won't be accurate as bows because of that.

[WhiskeySykes]

Interesting topic. I'm working on unit stats now for my own mod, so this was timely.

+1 rep