Here is the scenario I have pertaining to a combat system I am jotting down on paper.

The attacker has 100 soldiers each with 1 attack point and 3 health points. The defender has the same.

All the player has to do is press a button and combat is all computed then the player is just shown the results. From my current example, I would have the attacker's soldiers do 100 points of damage to the defender resulting in 33 defenders being killed. The same happens to the attacker's soldiers. This continues until both sides defeat each other at the same time and it ends in a draw.

I feel if I introduce a random factor, the battle could get lop-sided and the smaller side could not recover So I thought I would ask the community for their opinions of the very simple combat scenario.

The game concept I am designing deals with combat from outside the actual conflict. Sort of like a coach and a sports team. You give orders and watch as your units perform them. The game does start small, with 100 to 200 soldiers (all the same) and could grow to larger numbers as well. I was just wanting to get a system in place for small conflict that could scale into larger ones.

Each side has a unit type (soldier) a unit quantity (100). The player give the order, and the computer does the rest. After both players give their orders, the results are computed instantly. Keeping all things equal, Attack power, defensive power, etc. (I need a baseline) What would be the best way to determine damage? Static numbers, or RNG numbers?

To use or not use RNG in combat?

This sounds similar to auto-resolve in total war or such games.

As such, I think you need some mild randomness to it, to make it seem more realistic/simulation-like.

Just try and add a factor like x0.8-x1.2 for each "round of damage" one side deals to the other and see how that feels. Given all other factors are equal (terrain, morale, equipment etc) 200 soldiers should always win against 100. But not necessarily against 160 (but more often than not!).

So 100vs100 would often result in 20-25 guys alive on one side and the other side dead. For realism you could allow routing since the loosing side almost never fights to the last man in battles (or are we inside the movie "300"? )

You can also try and add bigger randomness factor, but make these uncommon and very clear to the player. Like a popup telling you the "event":

Fog - only half casualties on both sides (if your system supports not always killing one side completely)
Commander killed - your side only deals 70 % dmg for the remaining rounds
Rain - archers (if applicable) deal only 70% for the remaining rounds
You get the idea!

This can add character to an otherwise plain "press the button and get the results". Each "combat rounds" can take some moments to resolve so the player can watch the numbers tick down (the entire battle can still be over in 3-10 seconds depending on the feel you are after). I tried this for a game and it gives a bit more organic feel to the simulation.

There are other factors that will be involved in the actual resolution of combat, like weather and terrain.  I guess I should have explained that a little in the original post.

I guess I am looking for baseline stats.  What happens on a sunny day, flat land, and everyone in good moral.  I would think all things being equal and no combat modifiers, I was thinking that a soldier would do 1 point of damage.  The group would do 100 points of damage, that then could be modified by outside factors.  Just need to get a base line established.

I like the random range idea of making the group of 100 soldiers deal damage of 80 to 120 points of damage, but feel that may be too big of a swing before other modifiers.

If there's a geometric aspect (flanks and such), I think randomness is worth having just to introduce variation in outcomes along the line when two walls of units approach each other. Given that the player is sort of a coach, my design instinct is to introduce unpredictable wrinkles in combat to strategize around.

In terms of range of values, there's two levels a combat can be viewed in: at the low level, it's pairs of individuals fighting. If 100 equal soldiers per side pair off and fight to the death, 95% of the time each side will end up with 40-60 soldiers (binomial distribution). As the armies grow larger, that collapses into a tighter and tighter proportional range (1000 vs 1000 is about 470-530).

At another level, though, it's a team activity: If there's a lot of victories at one flank, the forces can wrap around and start seeing big advantages in that region. If an individual trips, he might leave his allies vulnerable. If one guy panics, it might sweep into a general retreat. At this level you can have much wider variations in outcome (although there's no standard equation anymore).

I think you answered your question yourself. As you need baseline stats you already explained that on the first post - 100 apoints vs 300 hpoints. This is your baseline stats and there comes an events/buffs/modifiers/other factors which somehow affects some sort of rounds endings. 

From a player perspective - RNG is fun tho. As you are sure you can win with 1000 army versus 100 army, you are not sure what about 100v150, how many losses you have to encouter and so on !

Countable fight ending in a long-term could come to a boring gameplay.

On 3/13/2018 at 6:53 PM, Mercury Gate said:

deals with combat from outside the actual conflict. Sort of like a coach and a sports team. You give orders and watch as your units perform them.

If this is your objective, "100 soldiers attack 100 soldiers" is a completely inadequate level of detail: giving skillful, nontrivial orders in order to win (as a coach would do with a sports team) would be impossible. The game would reduce to the strategy of moving troops to the appropriate front line locations. without interacting with actual combat.

Randomness is quite orthogonal to the choice between nonexistent, simplified and detailed tactical combat; its main purpose is limiting the usefulness of detailed planning (so players decide quickly, keep pace and don't succumb to "analysis paralysis") and providing unpredictable situations to react to, so if tactics are trivial, strategy becomes the level that offers interesting decisions and might need some randomness to work well.

Thanks for your feedback.  Randomness in the damage results can lead to players being happy when the RNG is in their favor, and unhappy when it is not.  If I know that 100 solders does 100 damage, I can plan my attack according to that and will know the result, but if I am depending on RNG to swing in my direction, my soldiers could do 95 damage and I loose the skirmish.

I guess I am looking at it more like chess.  Each piece has a set movement and attack direction and leave it up to the player to decide the best course of action based on that.

I could be wrong in my thinking and still would like to hear what others have to say on the topic.

On 3/26/2018 at 5:49 AM, Eimantas Gabrielius said:

Countable fight ending in a long-term could come to a boring gameplay.

you may be right.  But if feel it can depend on the context of how weighted the combat is to the over all idea of the game.  If the game has a combat focus whereas the combat is the "be all" and it end up boring, the game will be looked at as a boring game, but I think if the combat is just a small portion of the game, the lack of details from the combat system could be a blessing allowing the player to focus on the game play aspect they like the most.

2 hours ago, Mercury Gate said:

I guess I am looking at it more like chess.

Chess is not a terribly "fun" game, in the sense most video games aspire to be. The better player will win 100% of the time, which makes it distinctly less fun for the perpetually losing side.

I do find dice rolls to be fun when I'm actually throwing the dice. But if I'm playing a strategic game, I tend to avoid attacking unless I already have a pretty good idea about the outcome. The game isn't about counting on a lucky dice roll it's about getting the strongest pieces into the right position the fastest. I may be wrong about what my opponent is planning or encounter something else unexpected but it seems like the general idea is to see if your plan succeeds or fails on its own merits.

8 hours ago, kseh said:

I do find dice rolls to be fun when I'm actually throwing the dice. But if I'm playing a strategic game, I tend to avoid attacking unless I already have a pretty good idea about the outcome. The game isn't about counting on a lucky dice roll it's about getting the strongest pieces into the right position the fastest. I may be wrong about what my opponent is planning or encounter something else unexpected but it seems like the general idea is to see if your plan succeeds or fails on its own merits.

 

Games generally don't only depends on one factor like Attack Points. For example RPGs like World Of Warcraft, no matter how much Attack/Agility/Spell Power you have, you will never hit for the same amount but you still know that you cant beat level 3 monster if you are level 10, but there is RNG anyway. It doesn't mean you have total RNG of fight ending, but some factor where you can't count 100% of fight flow i think is a must.

I've made an experiment on Stronghold. I sent 5 infantryman vs 1 horseman  x10times. 
1st ending - 1 infantry x 0 horseman left
2nd ending - 0 infantry x 1 horseman left
3rd ending - 0 infantry x 1 horseman left
4th ending - 2 infantry x 1 horseman left
5th ending - 0 infantry x 1 horseman left
6th ending - 0 infantry x 1 horseman left
7th ending - 1 infantry x 0 horseman left
8th ending - 0 infantry x 1 horseman left
9th ending - 1 infantry x 0 horseman left
10th ending - 1 infantry x 0 horseman left

Although you know your horseman is strong enough, you can't be sure you won't be beaten by 5man. And if there wouldn't be RNG, I only have to do the math and ensure my win. Is this interesting ? Then I would prefer go do math instead of playing game