[http://sourceforge.net/p/ufoai/feature-requests/416 Item 416] imported from sourceforge.net tracker on 2013-01-28 20:20:07

I know this has been discussed and is subject to change, but I would like to depict my hybrid model for reaction fire anyway:

There is no fixed value how much TUs need to be kept in reserve for a soldier to allow him to return fire. It is 'the more - the better'. When an alien walks into view, the soldier fires his weapon. He uses the same number of TUs he would need for that shot in his own turn. This model's special: The soldier can only spent half of his reserve for the reaction shot. If this value is too small to pay for the shot cost, the difference is drawn from the soldier's next turn. There is no limit how much can be 'borrowed' this way (even more than the full TUs value for the next turn), but there's a catch: The chance of getting the reaction shot in the first place depends on the percentage of 'borrowed' TUs on the shot cost.

Example: A soldier has a weapon with a snap shot cost of 10. He has 40 TUs, but uses only 15 TUs in his own turn. Now he waits for aliens to creep up with a reserve of 25 TUs.
Now an alien moves into sight. Our soldier wants to react: He can use half of his reserve (rounded up), so he has 13 TUs directly available, and needs only 10 TUs to fire. So this reaction shot is 100% payed by the reserve, resulting in maximum chance of fire. He shoots, paying 10 TUs, remaining on 15 reserve.
When the next alien shows up, the soldier can only spent 8 TUs directly (50% of 15 reserve), and needs to borrow 2 TUs. So his chance of fire is only 80% of maximum value. He is lucky after a few attemts (alien movement steps) and fires, reducing reserve to 7. If the enemy turn would be over now, he would start his own turn with 38 TUs (40-2).
The next shot would have just a 40% chance modifier (50% of 7 = 4 -> 4/10) and leave 32 TUs for the next turn (38-6).

Conclusion:
- There is now fixed reserve value, which is good. The more reserve the better, allowing more reaction shots at higher probabilities. No "Make one more step and you can't do anything".
- Small weapons (with low TU cost) are more suited for reaction fire, because they drain less reserve points and don't need to 'borrow' TUs from the next turn as fast as larger weapons.
- Each shot is payed for its exact TU price. Massive reaction fire reduces next round's TU's, but at a fair price, and should leave the soldier with some TUs left (needs testing).
- A larger reserve gives higher safety, but reduces the effective usage of TUs per turn (because half of the reserve remains unused with each shot).

Things to consider:
- The "use 50% rounded up" rule can be varied to any percentage for fine tuning.
- The "reduce reaction chance by borrow percentage" rule does not neccesarily need to be linear.
Consider: square (<BorrowPercentage> / 10) results in 10% borrowed -> -1% modifier, 20% -> -4%, 30% -> -9% and so on.

===== Comments Ported from Sourceforge =====

====== geever (2009-07-26 18:38:28) ======

I think you overcomplicated it with the "only half of the reserve" and "borrow from next turn" stuffs.
The first I don&#039;t really understand what would be good for the second I don&#039;t wanna see in the game, that&#039;s simply bad.

-geever
====== btaxis (2009-07-26 20:48:13) ======

Agreed. The current model for reaction fire is both fair and sensible. No need to change it.