New Combat System

[Arathorn]

I'll add my support the TNO's (warpstorm's?) suggestion to have it be a variable that can be set by the user. As long as the option is available to continue to play the old way, I will have no major complaints. (Yeah, if I have to play everything as a scenario, my High Score table will be trashed, but that's a piddling detail, not a major issue.) I think very few people will complain about MORE choices.

I would hope/ask that the default be one roll per combat, but if the code for multiple rounds is essentially complete, why not make it available to the modding community?

Arathorn

 

[Sandman2003]

This preference idea where you can selectively choose whether to use the new combat calculator or keep with the old is not a bad compromise. However, I would still prefer no change, primarily for two reasons.

1) The current combat system works. If it is changed then there is a risk of introducing bugs, and however slight this risk may be, it is a totally unnecessary risk.
2) There is no possible comparison possible between two games where all settings are otherwise equal except the combat calculator. These two games are essentially different.

Also, I note that the primary advocates of this preference idea actually seem to be suggesting that their own preference would be the default setting of 1 (ie the current system), although they might, maybe, possibly once ..., try another setting but only 2 etc. ie they are not really convinced of the benefit of the change either.

Yes this is a beta release, but, firaxis, you would make a lot of happy civ players by releasing everything else in the patch, especially the fp and gpt fixes, and leaving the combat calculator concept to a future date, and more discussion...

 

[alexman]

OK, as I said before, I like the idea behind the new combat model, but here are some changes that must be made to some units to compensate for the change in the RNG:

Remove combat bonus for Armies: Otherwise a single leader in the ancient age will affect the outcome of the entire game.

Reduce defensive bonuses for terrain and fortification: Otherwise defense has much better odds than offense for units of comparable strength.

Increase retreat odds for fast movers: So horsemen are again an alternative to swordsmen as an offensive unit in the ancient age.

Increase cost of Immortals, Mounted Warriors, Hoplites, and other UU that have a higher strength than their contemporary units. Any bonus in strength gets magnified with the new combat model.

Probably more things like that. What I really hope is that they don't give us a new combat system without changing anything else.

 

[The Last Conformist]

Thanks for the link, Bomber Escort. Now, I just wish I hadn't voted "don't understand the question"; this one isn't huge, it's frikken humongous!

My take on the subject:

i) The old system didn't need fixing.

ii) The new system throws the present unit balance out of the windows. You can just as well forget all your present combat rules of thumb if/when this is implemented.

iii) The new system looks like it's deliberately designed to be confusing and non-transparent. I don't want to believe the developers purposely make things more complex or confusing than there is any call for, but it certainly looks like they've taken the German Bundeswehr's supposed motto to heart: "Why simple, when you can make it complicated?".

iv) This sort of change in a patch?!? It's rather of the order of magnitude you'd expect between different games in a series.

And a final question; is this new system also going to apply also to bombardment? If so, that on the whole would make artillery more powerful in the field and less against units in cities. The "trench war" strategies with Arty and Cav are going to need considerably Arty to be similarly effective, for instance.

Needless, to say, could I change my vote, it would be to "no".

 

[Nad]

Originally posted by alexman
OK, as I said before, I like the idea behind the new combat model, but here are some changes that must be made to some units to compensate for the change in the RNG:

Remove combat bonus for Armies: Otherwise a single leader in the ancient age will affect the outcome of the entire game.

Reduce defensive bonuses for terrain and fortification: Otherwise defense has much better odds than offense for units of comparable strength.

Increase retreat odds for fast movers: So horsemen are again an alternative to swordsmen as an offensive unit in the ancient age.

Increase cost of Immortals, Mounted Warriors, Hoplites, and other UU that have a higher strength than their contemporary units. Any bonus in strength gets magnified with the new combat model.

Probably more things like that. What I really hope is that they don't give us a new combat system without changing anything else.

The changes you propose here might be good for *balance* under a new system but they would be horrible from a *strategy* point of view, especially removing defence bonuses. IMO, defence bonuses are a *crucial* part of combat strategy, if you remove them you simply do not have any tactics to combat whatsoever, it's just units vs units. Not only is this completely unhistorical, it is also mindbogglingly boring.

EDIT: just read your post more closely and I see you say "reduce" defence bonuses, not "remove". That is better but how much can you reduce them by without negating their effect?

 

[alexman]

I didn't say to remove the defensive bonus. I proposed to reduce them.

Specifically, you can reduce them to the point where the odds of victory for units of the same strength attacking each other under the new system are approximately the same as with the old system.

 

[Nad]

Originally posted by alexman
I didn't say to remove the defensive bonus. I proposed to reduce them.

Specifically, you can reduce them to the point where the odds of victory for units of the same strength attacking each other under the new system are approximately the same as with the old system.

Yes, sorry, see above, I'd already edited the post.

The point, however, is still this: if you're going to change the combat values to have the same chance as the old system, why change the old system in the first place

Perhaps one option (just a thought) would be to introduce *attack* bonuses in certain situations to counter the new changes? Eg, if a unit attacks *from* high ground it gets a combat bonus for strategic territopry advantage. Another option would be to give fast units a combat bonus if they move and then attack in the same turn (to simulate a "charging" bonus, perhaps?), or give attacking units a bonus if they spend a turn fortified before attacking (to simulate "preparation")?

 

[Pounder]

There has to be a chance of a spearman beating a more modern unit. Like others, I have played countless games and I don't remember a spearman beating a tank with full health. It just doesn't happen with any frequency.

But there has to be a chance, that there is a smarter spearman that used something other than his spear (maybe he dug a pit and the tank fell in ), or maybe someone used a grenade rather than his rifle. Or maybe the tank operator (not so smart or slow) left the hatch open and the operator was killed by the spearman (in that case maybe the spearman should capture the tank ).

These variables obviously would not be animated during the fight, but there has to be randomness.

Can't have a fight be an almost certainty, the fun would be gone.

 

[Park Ranger]

I read the thread about the patch and posed a question hidden in a rather lengthy post there. I'll try again here in a clearer form.

Is the new system going to average 4 rolls to get the result (outcome for one HP) or take the results of 4 rolls and average that? I think most have been assuming the former since Mike B. dismissed the idea of a tie-breaker roll, which you would need for the second case.

Example (using dice):

The A/D values indicate the attacker wins on a 5 or 6. Roll 4 times.

Say you roll a 5,5,5, and a 1.
Case 1: average first, then determine results. Average is a 4, so attacker loses 1 HP.
Case 2: average the results. Three wins (the 5's) vs. 1 loss means attacker wins, defender loses 1hp.

In case 1, as has been shown, forget about spear vs tank. Spear barely has a chance vs. Knights (or their contemporary nemesis, the Persian Immortal).

In case 2, well, I have not the maths to model the combat results. But my intuition says its less drastic a change, more like (but not quite) lumping HP in groups of 4 and always taking off 4 at a time.

The posts from the Firaxians have led us to believe it's Case 1. But they also say they've been playing this for months...surely someone would have noticed the dramatic effects predicted in Case 1.

Oh well, either way I'm not too keen on the changes. My spearmen already always lose to tanks, no need to rub it in by letting the tanks off without a scratch.

 

[Grey Fox]

Originally posted by Park Ranger
Case 2: average the results. Three wins (the 5's) vs. 1 loss means attacker wins, defender loses 1hp.

Actually, that wouldn't be so bad! Or atleast I think it wouldn't.

 

[SirPleb]

Originally posted by Park Ranger
Is the new system going to average 4 rolls to get the result (outcome for one HP) or take the results of 4 rolls and average that?

Case 2 was what was described as being in the patch near the beginning of the patch thread, and was what some initial math was based on. The math for it came to frightening results, quite imbalancing. Then a bit later on Firaxis corrected us, stating that case 1 was what was actually happening. More math was done and case 1 turned out to also be frightening and quite imbalancing, in pretty much the same way and magnitude as case 2

 

[The Last Conformist]

The discussion now appears to be moot, for the time being; according to reliable sources, the new combat system has been withdrawn from the BETA patch. Check out the stickied thread on the patch for details.

 

[Portuguese]

According to latest news, this is back:

 

[SoRHunter]

I'd like to play under the new combat system. It seems better than the old one (which is one of the reasosn I stoped playing CIV3...).

I want the new combat system!

 

[The Last Conformist]

Portuguese: It was never gone.

 

[Portuguese]

But I hoped the chances would get very very low!!!

Bow, they will just stay the same. Using the CombCalculator:

77,13% Win
15,10% Loss
7,77% Draw

With this:

(Is this fair?)

 

[Sandman2003]

It is great news that the new combat calculator has been removed from the patch! This enables time for a much more considered approach to dealing with the issue.

Personnally, while I still maintain that nothing needs to be done to the combat calculator, there is no harm in proposing alternative solutions that would have a much less game unbalancing impact than the original changed combat calculator proposal.

One way you could significantly reduce the odds of extreme events from occuring, while only affecting close matchups minimally, would be to eliminate the tails from the normal distribution curve. This would require a percentage value to be truncated from both ends of the curve, eg 2%. I am assuming that at some point the RNG generates a number between 0 and 1, or for simplicity 0% and 100%. Using a truncation factor of 2%, this would have a consequence on the RNG result as follows:

If the RNG produced:
0 - < 2% - a redetermination (a) would be required
2% - 98% - RNG uses this value
> 98 % - a redetermination (b) is required.

The simplest way to redetermine the value would be (a) just to multiply by 50 and check again, or (b) first subtract 98% then multiply by 50. The impact would be significant on extreme possibilities, but small on closer calls EG

A 1 in 20 proposition is equivalent to a 5% versus 95% scenario, becomes a 3/96, or 3.125% proposaition versus 96.875%. Or about a 1 in 32.

A 40/60 proposition becomes a 38/96 versus 58/96, or 39.6% versus 60.4% proposition - virtually the same as before.

This would be per hit point, so obviously the differences are further magnified, and the truncation factor can be adjusted to achieve the desired result 2% is just an example.

 

[Dragonlord]

Originally posted by kring


Same here. We don't have enough information to know how balanced/unbalanced the change is.

If people complain enough about stuff before testing the change(s), Firaxis/Breakaway are going to be less likely to add or change other things.

I agree. Let's try it and see before screaming bloody murder!

In principle, as I've stated on the other thread, I'm all for reducing (not eliminating) randomness in the game. I agree with some of the other posters that there are alternatives possible, though, especially simply increasing hitpoints, as many modders do. It would be great to simply increase hitpoints by era, for instance: all medieval units would have one more HP than those from the ancient age e.g.

IF FIRAXIS WOULD ONLY ALLOW MODDED GAMES IN THE HALL OF FAME, WE COULD EACH MAKE OUR OWN DECISIONS HOW WE LIKE OUR GAME. It would be so simple: just mark modded games with an asterisk or something in the HOF, then those wanting to compare 'pure' unmodded games could still do so.

 

[player1 fanatic]

Originally posted by Dragonlord
IF FIRAXIS WOULD ONLY ALLOW MODDED GAMES IN THE HALL OF FAME, WE COULD EACH MAKE OUR OWN DECISIONS HOW WE LIKE OUR GAME. It would be so simple: just mark modded games with an asterisk or something in the HOF, then those wanting to compare 'pure' unmodded games could still do so.

Even better, have MODs have their own Hall of Fame files.


Or have them with * as you said, but add some filter too (only moded/only unmoded games).

 

[BomberEscort]

Originally posted by Portuguese
But I hoped the chances would get very very low!!!

Bow, they will just stay the same. Using the CombCalculator:

77,13% Win
15,10% Loss
7,77% Draw

With this:

(Is this fair?)

True, these are the odds. But remember, this spearman is closer to the defense of a rifleman than that of a spearman. In this case, this battle resembles a Tank-Rifleman battle more than a Tank-Spearman battle... But I understand your point...