civ5 Combat and Odds (Video analysis)

[PieceOfMind]

Yes, in general when I say "attacker" I mean the unit that initiates the combat, unless I say otherwise. I agree it's pretty awkward for describing civ5 combat, but it was pretty straight forward for civ4 so I'm sort of stuck with it / used to it for the moment. I prefer to think of the roles of attacker and defender staying fixed during combat, rather than changing each round.

If you were to allow the attacker and defender to change during combat (i.e. units keep switching between attack and defense) then it would not be correct to say:

Flipping this around to get the amount of damage the defender expects to deal to the attacker, and everything is still symmetric between attacker and defender.

anymore, because it wouldn't really make sense.

 

[Polobo]

I hate this kind of thing. I hope there is some kind of advanced option (or at least some willing modders) that will provide this info.
I want to know what the probability distribution is of the various possible outcomes, much like Battle for Wesnoth.
I can't plan an optimal strategy unless I know what's actually going on in a statistical sense. There's just no good reason to hide this data from the player.

Giving the player too much information can be a bad thing, especially for beginners. Plus, some people like the idea of immersing themself into a make-believe world but one that still resembles the world we live in. You never truly know odds and probablilities but do have a general idea as to what outcome to expect. Life is never optimal but more "as good as possible given what I know".

I do agree that making the added info available as an option would be nice; and since they likely have the info available already for debugging purposes putting some effort into a GUI widget and giving the player an option to view that screen would be great.

 

[weakciv]

I like knowing the odds. I also like knowing there is a random aspect. As long as that "catastrophic" outcome is VERY VERY rare!

 

[Brawndo]

I hate this kind of thing. I hope there is some kind of advanced option (or at least some willing modders) that will provide this info.
I want to know what the probability distribution is of the various possible outcomes, much like Battle for Wesnoth.
I can't plan an optimal strategy unless I know what's actually going on in a statistical sense. There's just no good reason to hide this data from the player.

Ha ha oh my. Listen, we're playing a video game, not working on Excel. Combat should have some randomness thrown in for spice, just like in real war. Many times one side looks better on paper, but when the **** hits the fan, things work out differently.

I liked how in Panzer General (another 1upt series), sometimes your tank would be predicted a 6:1 kill/loss ratio against an enemy infantry unit, only to have a "Rugged Defense!" modifier be thrown in at random. The valorous infantry take a stand and suffer only 2 losses while your tank limps away with 5/10 strength.

 

[Ahriman]

Giving the player too much information can be a bad thing, especially for beginners. Plus, some people like the idea of immersing themself into a make-believe world but one that still resembles the world we live in.

So hide the data unless I hold down the alt key, in which case you show it.
Easy.
But there's no reason to hide this data from someone who wants to see it.

Ha ha oh my. Listen, we're playing a video game, not working on Excel. Combat should have some randomness thrown in for spice, just like in real war.

Ha ha oh my.

Listen, did you even read what I wrote? I'm fine with there being randomness, I just want to know the probability distribution of that randomness.

I'm fine with unlikely outcomes occurring - I just want to know how unlikely they are.
Its not good enough for the game to just tell me "you will probably win", I'd prefer to know "with 50% chance the enemy unit will be destroyed, with 25% chance it will be on 2 hit points, with 25% chance it will be on 4 hit points".

There is no contradiction between having randomness and displaying to the player how that randomness works.

 

[PieceOfMind]

Ha ha oh my. Listen, we're playing a video game, not working on Excel. Combat should have some randomness thrown in for spice, just like in real war. Many times one side looks better on paper, but when the **** hits the fan, things work out differently.

I liked how in Panzer General (another 1upt series), sometimes your tank would be predicted a 6:1 kill/loss ratio against an enemy infantry unit, only to have a "Rugged Defense!" modifier be thrown in at random. The valorous infantry take a stand and suffer only 2 losses while your tank limps away with 5/10 strength.

The irony is that if the game hides info from the player, the guy using the Excel spreadsheet is actually the one who is rewarded.

The reality here is that this is a game. You can talk all you like about what it's like in real battles but it's irrelevant. There is a distribution so you have to expect that the distribution will be discovered and used by players. Of course, players who don't care about it will continue to not care about it - nothing wrong with that.

 

[Xetal]

The frigate-on-frigate naval combat confirms my fears about the failings of attacker-only ranged combat.

 

[Jaybe]

Ideally, ranged combat between units which are both in range (and sight) of each other should be the same as a normal combat, though with tactical naval combat sometimes someone gets to "cross the T" (be able to fire a broadside while the opponent gets to fire only bow OR stern guns (often very little with sailing vessels)).

The other objection I have at this point is that barrage weapons should not inflict as much damage on weakened units as full strength ones because there is less chance of something being hit. Inflict losses proportional to the REMAINING strength (if damage would be 20%, but defender is only at 60% strength, actual damage ought to be 12%). I REALLY wish in civ4 that unit density would have had much more impact on siege effects!

 

[Xetal]

It doesn't concern anyone else that "he who shoots first, shoots last" with ranged bombardment being so strong that the first strike is the only strike that makes a difference?

 

[Polobo]

It doesn't concern anyone else that "he who shoots first, shoots last" with ranged bombardment being so strong that the first strike is the only strike that makes a difference?

It is a concern but as of right now I figure that play-testing and balancing will address it such that it would generally take multiple (3-5) hits to bring down a "normal" unit. Even with range it will still be somewhat difficult to field ranged units in such a way as to be able to quickly decimate an entire front - just because of the added base movement and the need for infantry units between the ranged units and the opponent. Plus consider that even though you can attack does not mean you are going to significant damage each turn - ranged attacks + defensive promotions means that actually hitting the enemy could be quite tough.

 

[Xetal]

It is a concern but as of right now I figure that play-testing and balancing will address it such that it would generally take multiple (3-5) hits to bring down a "normal" unit. Even with range it will still be somewhat difficult to field ranged units in such a way as to be able to quickly decimate an entire front - just because of the added base movement and the need for infantry units between the ranged units and the opponent. Plus consider that even though you can attack does not mean you are going to significant damage each turn - ranged attacks + defensive promotions means that actually hitting the enemy could be quite tough.

The frigate on frigate ranged combat we saw linked in this thread shows predicted loss of 50% total health (with no chance to fire back) with actual damage of 70% of total health.

That was a battle of two identical units, which is the best metric you're probably going to find.

Other bombardment attacks show 30%-50% of max health when the bombardment strength is roughly equal to the defenders base strength. Again, with no counter attack possible.

If bombardment attacks do as much as regular attacks with no real drawbacks to them, I can see the strategy being single faceted and boring. Bombard attacks need to be secondary, a way to soften the enemy up and make them act or be whittled down... not as a way to completely decimate an enemy in two attacks.

 

[Thyrwyn]

I am not too concerned.

With regards to naval combat: It makes sense to me that given two evenly matched fleets, the fleet that chooses the time and place of battle should win. There is no terrain to consider. Given that (from what we have seen) naval units are unable to move after attacking - it means that the fleet that fires first will be stuck there until their next turn, so the "defender" will get their shots in with surviving ships. Individual promotions could swing the outcome.

With land battles, terrain will be the big factor. Units in the open should be vulnerable to ranged attacks; units in rough terrain or with fortification bonuses will be less vulnerable. That is as it should be, too. In sizable battles, the attacker will have to concede the first ranged shots to the defender and will probably have the added disadvantage of being in less favorable terrain. All things being equal, the defender will have the advantage, which is at should be, too.

Given what we know, the system could be broken, but there is no reason that it has to be.

 

[Schuesseled]

It doesn't concern anyone else that "he who shoots first, shoots last" with ranged bombardment being so strong that the first strike is the only strike that makes a difference?

we've seen ranged attacks completley obliterate an enemy unit, and we've seen them bounce harmlessly off the enemies shields, so for right now its a wait and see approach to range combat for me. Although i suspect its gonna rock my socks off.

 

[Stalker0]

Maybe this is why in alpha centauri you got the big ships with Doctrine: Initiative

It pays to fire first.