User blog comment:FixSRwiki/100 or more problems with SR/@comment-31720212-20200102212301/@comment-26544499-20200806165750

Yeah, they were probably me, although I'll have to see if anyone else chimed in. I'll contribute what I can when I've got the time.

Don't worry too much about getting testers; Build it and they will come. We've even had people do streams of the DHR mod.

技能データ初期化 is the place to start for any ability stuff. Once you've found the relevant variable - 行動戦果 which, as you've said, takes some really weird values - it's normally a just a case of reverse searching. You've rightly pointed out that you can't find the relevant stuff because the game never actually triggers it, but proving that is a bit harder. When you look through what uses 行動戦果, you'll find that the only relevant one is the 合戦／行動後処理 function, and that looks like it works just fine. The next step is to check what calls that. Sadly, the answer to that question is "five functions that nobody is likely to understand". But on the bright side, all of those five are called by 合戦_行動 and that's actually our broken function! If you look through the ability types of the stuff that doesn't work (see 技能データ初期化) and compare them with the switch statement in 合戦_行動 then you'll see that it's skipping over checking if melee or passive skills have the relevant sorts of special effects, and that's our glitch!

Just be careful if you actually patch this stuff. For example, Kenshin's passive wants to add +30 to the Battle Rating, which on top of the +5 that it's already coded correctly to do, is many times what Initial Battle Rating already does. +30 is like annihilating six units, so it's clear that these values were never tested by anyone.