A small tweak to combat mehanics that would change a lot
Posted: Sat Sep 27, 2014 1:08 am
With the present mechanics, when combat occurs, attack/defence numbers of units are compared, a dice is rolled, whoever wins, inflicts damage on the opponent, repeat until one dies.
Proposal: change the "repeat until one dies" part.
In other words, give a reasonable chance for a unit to survive combat even if it is defeated and suffered heavy losses. Additional option: if the defeat is really, really heavy, it makes a "forced retreat" or simply runs away one tile in the direction opposite of the attacker, if possible.
Explanation:
Basically, the current mechanics is a relict from really old times when computers were less powerful and it was difficult to add more properties to units, such as "current strength" and hitpoints. This can be seen from the fact that very few games today that include combat - if any, at least among the more complex ones - have the property that in any combat event one unit ALWAYS must die. It creates a situation where there is very little continuity and in some cases wars turn into "flood contests": whoever gets more units out of his cities wins. (Also there is the issue of binary luck: a unit either wins or dies and an unfavourable dice roll ALWAYS hurts you very much.) As it is now, every unit fights for itself because, after combat, either the opponent is dead or the unit itself is dead.
Very little war strategy in all that. Some, but very little.
Implementing this would create more thought-through campaigns, would involve more strategic thinking and make good use of terrain critical (always must have access to the rear to withdraw your damaged units). It would also make wars more persistent, not just sporadic bursts of violence placed between long periods of production and preparation.
Not to mention that this would be much, much more realistic.
Less unit destruction would also create more units in the field, again, more fun for the strategically inclined and also more possibilities for outmanoeuvring your opponent. As it is now, frontal steamrolling is the only available option; with this change, concentration of forces would suddenly mean much more, turning wars into really complex events.
Also, sieges would become more viable; as it is now, you simply have to sacrifice more units to conquer a city; with this change, it would take more time, but inflict less losses on the attacker's side, again, making the whole thing more realistic.
So, the main issue here would be: how long should the combat take?
For starters, there should always be a chance that a unit dies. So, the base number of combat rounds should definitely be at least 10. Everything else is negotiable. The following are proposals and can be analysed and discussed. So, the length of combat should depend on:
- defending terrain; obviously, a unit in a swamp or in the mountain should have more chance of evading combat and surviving - even if defeated - than the one in grassland or in the plane
- damage to the attacker; in reality, very rarely, if ever, is the attacking unit completely annihilated (unless they have a very stupid commander); I'd say that the probability of the attack continuing should depend on the hit points of the attacking unit (plus a small modifier to make it possible, albeit unlikely, that the attacker *is* destroyed from time to time)
- difference between attacker and defender speeds; a defending cavalry should have a better chance of running away from attacking infantry than the other way round
- experience; veteran units should be more able to continue the assault if it suits them and disengage if it doesn't
A lot to be discussed here. Maybe even modify default attack and defence values, together with the amount of hit points and firepower.
What could be discussed is also creating three types of attack: probe, standard and "at all costs" (or even five, or simply introduce a percentage of "dedication" to the attack). However, while I presume that the original proposal would require a lot of analysis, but only a small tweak in the code, adding different types of attack would probably require a major rewrite, including additions to the interface.
Post update with a variant that came to mind later on, simplifying the whole thing (copied from one of the later posts): when a unit attacks, create a pop-up (like when a Diplomat is entering a city). Options are: 1. Attack, 2.Disengage. If "Attack" is chosen, unit attacks for one round. If "Disengage" is chosen, attack is stopped, pop-up is gone and unit loses one movement point. End of story. (Maybe create additional options in the pop-up, like "Attack for 5 rounds" or "Attack for 10 rounds", to avoid excessive clicking. Or enter the number of rounds...)
(Additional option 1: maybe create an opportunity for faster units to automatically disengage (run away, retreat) when they are attacked. For example, if Archers attack Horsemen, when Horsemen are down to 1, 2 or 3 HPs, they retreat in the direction opposite to where the attack came from, if possible. Entering enemy ZOC bans retreat, fortified units don't retreat. Additional option 2: make it possible for slower units to retreat, too, under slightly more strict conditions.)
Proposal: change the "repeat until one dies" part.
In other words, give a reasonable chance for a unit to survive combat even if it is defeated and suffered heavy losses. Additional option: if the defeat is really, really heavy, it makes a "forced retreat" or simply runs away one tile in the direction opposite of the attacker, if possible.
Explanation:
Basically, the current mechanics is a relict from really old times when computers were less powerful and it was difficult to add more properties to units, such as "current strength" and hitpoints. This can be seen from the fact that very few games today that include combat - if any, at least among the more complex ones - have the property that in any combat event one unit ALWAYS must die. It creates a situation where there is very little continuity and in some cases wars turn into "flood contests": whoever gets more units out of his cities wins. (Also there is the issue of binary luck: a unit either wins or dies and an unfavourable dice roll ALWAYS hurts you very much.) As it is now, every unit fights for itself because, after combat, either the opponent is dead or the unit itself is dead.
Very little war strategy in all that. Some, but very little.
Implementing this would create more thought-through campaigns, would involve more strategic thinking and make good use of terrain critical (always must have access to the rear to withdraw your damaged units). It would also make wars more persistent, not just sporadic bursts of violence placed between long periods of production and preparation.
Not to mention that this would be much, much more realistic.
Less unit destruction would also create more units in the field, again, more fun for the strategically inclined and also more possibilities for outmanoeuvring your opponent. As it is now, frontal steamrolling is the only available option; with this change, concentration of forces would suddenly mean much more, turning wars into really complex events.
Also, sieges would become more viable; as it is now, you simply have to sacrifice more units to conquer a city; with this change, it would take more time, but inflict less losses on the attacker's side, again, making the whole thing more realistic.
So, the main issue here would be: how long should the combat take?
For starters, there should always be a chance that a unit dies. So, the base number of combat rounds should definitely be at least 10. Everything else is negotiable. The following are proposals and can be analysed and discussed. So, the length of combat should depend on:
- defending terrain; obviously, a unit in a swamp or in the mountain should have more chance of evading combat and surviving - even if defeated - than the one in grassland or in the plane
- damage to the attacker; in reality, very rarely, if ever, is the attacking unit completely annihilated (unless they have a very stupid commander); I'd say that the probability of the attack continuing should depend on the hit points of the attacking unit (plus a small modifier to make it possible, albeit unlikely, that the attacker *is* destroyed from time to time)
- difference between attacker and defender speeds; a defending cavalry should have a better chance of running away from attacking infantry than the other way round
- experience; veteran units should be more able to continue the assault if it suits them and disengage if it doesn't
A lot to be discussed here. Maybe even modify default attack and defence values, together with the amount of hit points and firepower.
What could be discussed is also creating three types of attack: probe, standard and "at all costs" (or even five, or simply introduce a percentage of "dedication" to the attack). However, while I presume that the original proposal would require a lot of analysis, but only a small tweak in the code, adding different types of attack would probably require a major rewrite, including additions to the interface.
Post update with a variant that came to mind later on, simplifying the whole thing (copied from one of the later posts): when a unit attacks, create a pop-up (like when a Diplomat is entering a city). Options are: 1. Attack, 2.Disengage. If "Attack" is chosen, unit attacks for one round. If "Disengage" is chosen, attack is stopped, pop-up is gone and unit loses one movement point. End of story. (Maybe create additional options in the pop-up, like "Attack for 5 rounds" or "Attack for 10 rounds", to avoid excessive clicking. Or enter the number of rounds...)
(Additional option 1: maybe create an opportunity for faster units to automatically disengage (run away, retreat) when they are attacked. For example, if Archers attack Horsemen, when Horsemen are down to 1, 2 or 3 HPs, they retreat in the direction opposite to where the attack came from, if possible. Entering enemy ZOC bans retreat, fortified units don't retreat. Additional option 2: make it possible for slower units to retreat, too, under slightly more strict conditions.)