(v19, linux)
Abstract: Created two antiallies (i.e. allies with a negative skill). And collided them with ordinary allies. Results - see attached screens.
Problem #1: Point allocation buffer
It looks like you want to give us some space to reconsider skill / stat allocation on lvl up, so the game is supposed to remember where the points were distributed lately. However, seems it doesn't take into account more party members, and remembers only the last skill / stat operation, regardless who. Naturally, one wonders, where this can lead into...
How to produce an antially:
1. Need more than one character and lvl up when allies get 6 skills to distribute (or I think any lvl up could work too)
2. Using Ray and Elena: Ray +6 to CC and then Elena +6 to Firearms, in that order. -> Tha last is +6 to Firearms...
3. Go back to Ray and observe, can't subtract 6 from CC, but can from Firearms despite him not putting any points there.
4. So ok, let's subtract 6 from Firearms for Ray and put it all to CC. -> Now the last is +6 to CC...
5. Go to Elena, -6 CC & +6 Firearms... Go back to (4) and cycle till each has at least one skill negative.
6. Two antiallies created! In the end it's just about putting enough energy in the process...
The same goes for stats (by one point). I was actually able to test both at the same time.
Problem #2: Annihilation of portraits
For some unknown reason the game, as observed so far, contains only ordinary allies. Such asymmetricality usually means what happens, when an antially is floating around, is mostly untested. And indeed, its portrait collides with the nearest ally you switched from, and is annihilated. The problem is, the result of reaction aren't two photogenic portraits as one would expect, but we're now getting two portraits of the same ordinary ally! In attached images, for example, see what happens when you switch from Sam to antiRay - his portrait also becomes Sam.
I'd like to ask a question whether an antially existence (skill<0) is intended? As it doesn't look immediately as a bad thing. If yes, then you need to better safeguard their portraits. Played a mission with them. Haven't actually encountered any problem yet, apart the full-body portrait issue, again.
However, AFAIK there aren't any natural ways to lower skills, and I suspect some further consequences may occur. So in case you think the negative skill shouldn't be reachable (in other words must I test it?:)), then just to be safe I suggest making sure that your character accelerator (i.e. char. system) won't even allow the antially in principle (->no negative skills allowed). Not just solving problem #1.
As for base stats, their negative numbers seem more stable. Portrait annihilation wasn't happening for allies with a negative stat. Isn't it possible for Emilia (Leadership 2) through injuries, after all? Perhaps that should be alright then?
