"One unit per one tile" strategy thoughts

[Ahriman]

2 Strength 7 units should be as good as 1 Strength 14 unit

Why? I dispute your premise. It is totally arbitrary.

This is how it worked in Civ1 (basically) and it was a very poor combat system. The current system is much superior.

It also means that in order to significantly change the combat odds from higher strength, you would have massive "strength inflation". A rifleman wouldn't be strength 20 vs a strength 2 warrior anymore, it would have to be much higher.

 

[da_Vinci]

You see a Strength 5 unit (under current circumstances)
You have a Strength 6 unit. (under current circumstances)

What do you think will happen?

I can tell you a few 'thinks' that would be Sensible (ie make sense to the average person)

1. The Strength 6 wins and that is it... it is stronger so it wins

2. The Strength 6 has slightly greater chance of winning than the Strength 5... since one has to win it should have a 6/11 chance of winning (this is how Civ 1 did it)

3. The Strength 6 wins, but only has 1 Strength left.


All of those make sense...

73% chance?!?!?!??

Is... 5 and 6 -> 73%...self explanatory?

Do the 5 and 6 accurately represent the strength of those Troops?

NO, that 5 and that 6 are "window Dressing" they Disguise the actual combat mechanics

The fact is the Strength 5 unit should actually be closer to Str 125 and the Str 6 to Str 216

Those would be slightly more accurate Strengths for those units.

I think what is missing here is the notion of rounds of combat, within each unit vs unit encounter. And the idea that "unit counter" conceptually represents a collection of armed men or machines.

If six muskets encounter four muskets and square off, (assume 50% kill chance per musket), then after one set of volleys, what is left is 4 vs. 1, and then either 4 vs or 3 vs 0 ... as the expected result after second volley. Two left out of the six is less likely. Six and four represent accurately the strength for a single volley, but not for multiple rounds of volleys. In multi-round combat, the net effective strength becomes a function of the difference between the strengths, and thus you can never assign a net effective strength to a unit in isolation.

Do we prefer combat that is decided in one round? Who wins, and at how much damage sustained, would be more predictable (is that desireable?).

If we prefer multi-round combat, then that would need to happen all in one turn, wouldn't it? Do we really want single engagements consisting of multiple rounds at one round per turn? That would distort the temporal relationship between combat cycles and rebuild and resupply.

dV

 

[KrikkitTwo]

It also means that in order to significantly change the combat odds from higher strength, you would have massive "strength inflation". A rifleman wouldn't be strength 20 vs a strength 2 warrior anymore, it would have to be much higher.

No you would be removing Strength inflation

If I said my Wealth factor was 2 and Bill gates Wealth factor was 20, then I would be inflating my wealth.

On the other hand if I said mine was 2 and Bill Gates was 20,000,000 then I REMOVED the inflation

It would require Bigger numbers whoopdedoo

I'd far rather have a
"4000 Str" Tank that 2000 times better than a "2 Str" warrior
than a
"28 Str" Tank that 2000 times better than a "2 Str" warrior

Average person will think:
4. The Strength 6 is stronger than Strength 5 so a Strength 6 is better period.
Hardcore person will read civilopedia, Civ 4 combat mechanics is explained there.
Average person will not think about 6/11 or something complicated like that, what is it for?

That was option #1, 6 is bigger so 6 should win

I don't need to do research to determine if New York or Chicago is more productive...

I add a bunch of numbers and multiply by the bonus production

If New York has 20 production and Chichago has 10, then New York will finish the project in 1/2 the time chicago would.

So, there is no problem with game mechanics and we don't need to dumb mechanics down. The problem is in you - you want to know more about combat mechanics but you don't want to do your research.

Mechanics are Dumbed down when there is nothing Smart the player can do with them.

The Civ 2-4 Strength->Combat results Are a Dumbed down system.

They are Dumbed down because they do complex things without player input

Changing 5 and 6 into 73% is Complex but there is no player input (at least in Civ Rev you could retreat)

Imagine if production worked so that New York's "20 production" actually meant that it had 20 tries per turn to get a random possibility of finishing something. And that it would probably finish a project 10x as fast as Chicago which had 10 production.

Would that be a "deeper" game mechanic... No it would be mind numbingly dumbed down...
20 production is better than 10 production duh
but how much more... well let me consult my tables, and ask the in-game calculator how much 20 is better than 10

ok 20 is 8 times better than 10.

If New York is 8 times faster at producing things than Chicago, and Chicago has a production of 10... New York has a production of.....20 WTH??
No, If NY is 8 times as good, NY has a production of 80

Basic rules are simple. It's not that simple to calculate exact results, of course.

Yes it IS simple to calculate exact results, I can tell you Exactly how long it will take to finish the Temple of Artemis with a particular city without looking at in game calculations. (assuming I do nothing to change the situation, ie connect/disconnect marble)

Adding up the Hammers might take a while, but that is counting/adding.

If food/getting/losing marble come into play that might be a complicating factor, but even that is easy to calculate. as seperate situations

I should be able to easily tell what the In game benefit is of moving my Longbow onto the hill.
You may say duh he gets a +50% total bonus to Strength.. but what does that mean, Strength does NOT represent actual combat results in any meaningful way. (it is used tocalculate them, but does not represent them)

Also, while results of combat may be straight forward, a bigger picture isn't straight forward anway so you have a moot point. So what if you completely dumbed down combat and units have only a strength value? If you want to know the exact details, you should calculate if you can spare all that production you'll waste attacking enemy unit. What if he's on the hill? Do your spoils of war compensate for your combat losses? There is no way to have an easy answer to these questions (in general).

There should be
The principle is
Basic parts should be VERY simple a 4th Grader should be able to understand and predict basic parts

Interactions between those simply calculatable/predictable parts can add complexity

so
5 Str v. 10 Str... should be VERY simple

so that way the Strategy of DECIDING
I get a +100% Str by moving to this terrain, or building this building or getting this new more expensive/less flexible unit? is it worthwhile
becomes important

Just like Chicago prod 10 will take 2x as long to build it as NY; prod 20... but
Noone else is near building it and Chicago will get a better benefit (also easy to calculate)
or
NY is closer to the war front and in danger (it also needs to concentrate on military) [NY military production, easy to calculate, NY v. Chicago distance to from both me +enemy easy to calculate, New York defender chances against enemy attackers... also easy to calculate *... assuming you know the enemy attackers]
or
Chicago is in a better location to build this
or
I'm just about to get a food tech that will explode Chicagos population(in an easily calculatable way) leading to an increase in its production (in an easily calculatable way)
allowing it to be built faster (in an easily calculatable way)


* whoops no, Well I have an Archer inthe city giving me a Str of 6... the enemy has one unit of str 4 approaching and another 3 hexes behind it... what is the risk? Let me pull up the calculator on my desktop...

NO. I should be able to look at that and say...
IF the enemy attacks with 1 each turn the chances are X
IF the enemy waits and attacks with both of them the same turn the chances are Y

Parts=Easily calculatable
Player controlled Interactions=easily calculatable

Allows for Strategy.




As for multi round combat.

Can the player DO anything between one round and the next to affect the battle... if not it is just stupid complexity.

"multi round combat" is achieved by attacking with multiple units...
Unit A attacks Unit B=Round 1
Unit C attacks Unit B=Round 2, etc.

That is how it should be.

Who wins, and at how much damage sustained, would be more predictable (is that desireable?).

VERY VERY MUCH (note this should be predictable in a particular circumstance... ie Musket v. Musket in general should not be predictable... but Combat 1 musket on Hill attacking Fortified Musket on plains Should be predictable without anything beyond 4th Grade math)

 

[Ahriman]

I'd far rather have a
"4000 Str" Tank that 2000 times better than a "2 Str" warrior
than a
"28 Str" Tank that 2000 times better than a "2 Str" warrior

That's a valid opinion, but its not the only opinion. I'd rather deal with the latter.

In almost every turn-based strategy game I can think of, a unit with twice the "strength" of another is more than twice is good.

 

[KrikkitTwo]

That's a valid opinion, but its not the only opinion. I'd rather deal with the latter.

In almost every turn-based strategy game I can think of, a unit with twice the "strength" of another is more than twice is good.

Why... really why?

I can see it if strength is logarithmic

ie Str 3 is 2x as good as Str 2, str 4 is 2x as good as str 4

indeed that might make a lot of bonuses simpler (they don't add +100% Strength, they just add +1 Strength)

But why the connection to a number that only has the most convoluted relationship to what it is supposed to represent.

 

[Tlalynet]

I wish they'd kill the randomness entirely, Civ II had the option to do that and it wasn't game breaking at all. IIRC the defender always won 50\50 battles with 1 of 10 (or 20) hp left. Otherwise combat was predictable.

With a game that is not random in most important aspects why do they hand something so important of over to a RNG?

The civ IV str weirdness is cause STR increases the chance to hit, increases the damage done, and reduces the damage take. It's an all in one attack\defence\aim stat. I call that dumbed down.

 

[cardgame]

ie Str 3 is 2x as good as Str 2, str 4 is 2x as good as str 4

what are you blathering about? Make sense!

 

[KrikkitTwo]

what are you blathering about? Make sense!

A log scale....

This would be far simpler than Civ 2-4 but would avoid big numbers

so
Strength listed: actual combat value

1:1
2:2
3:4
4:8
5:16
6:32
7:64
8:128
9:256
10:512
etc.

That way you could fit a Str 2 warrior alongside a Tank (Str 13) that was actually 2000 times better.
Fortifying wouldn't give +25% Str it would give +1 Str, etc.

Now I'd prefer the strength to be what the combat value actually was... but a logarithmic scale might not be too complicated for people afraid of 3 digit numbers.

 

[Yakk]

Decibel scale!

Every +10 is a factor of 10.
Every +1 is a factor of 10^0.1 =~ 1.26.

Strength 2 vs Strength 3 = nearly even.
Strength 2 vs Strength 5 = 2:1 strength ratio.
Strength 2 vs Strength 8 = 4:1 strength ratio.
Strength 2 vs Strength 11 = 8:1 strength ratio

The nice thing about exponential scales (of which decibel is a special case) is that each unit is a uniform distance away in "perceptual space" in a sense.

A strength 2 vs strength 3 is the same edge as a strength 99 vs strength 100.

In civ4, power was roughly with the ^1.5 of strength (very roughly), so an end-game 40 strength tank, vs a 2 strength warrior, was a factor of 90.

To remap to a decibel scale...

-1 strength scout
2 strength warrior
4 strength archer
6 strength spearman
8 strength axeman
9 strength swordsman/longbowman
11 strength maceman
12.5 strength knight
13.5 strength grenadier
15 strength rifleman
17 strength infantry
19 strength tank
21 strength modern armor

Now, the neat thing about this is it makes clear why axemen are such a huge technological leap. A 2 unit gap in a decibel scale is a gap that is large enough to really feel in game, regardless of what scale you are playing on.

And you can see the points where technology breaks the game: axemen, grenaders/riflemen, tanks and modern armor. At each of these points, the attack unit is (A) not countered by lower-tech unit defence bonuses, and (B) about 4 decibels above the strength of the previous era's defense unit. The only exception being modern armor (as the previous era unit, the anti-tank, ends ends up being a serious pain).

Of course, what really ends up happening is that artillery wins the game, but that is another question entirely.

 

[da_Vinci]

As for multi round combat.

Can the player DO anything between one round and the next to affect the battle... if not it is just stupid complexity.

"multi round combat" is achieved by attacking with multiple units...
Unit A attacks Unit B=Round 1
Unit C attacks Unit B=Round 2, etc.

That is how it should be.

VERY VERY MUCH (note this should be predictable in a particular circumstance... ie Musket v. Musket in general should not be predictable... but Combat 1 musket on Hill attacking Fortified Musket on plains Should be predictable without anything beyond 4th Grade math)

Can we assume you want the odds (or perhaps we mean probabilities) of winning to be predictable from an inspection of the strengths (or the strengths modified by damage, terrain, fortifications, unit-specific bonuses), without having to resort to a calculator of iterative rounds of combat within a single unit on unit encounter? I hope you don't want the stronger unit to be guaranteed to win each time?

You seem to be asking to take the complexity hidden in the iterative strikes that make up a combat, and bring that out in plain sight in the strenghts, and remove the iterative strikes ... have I got your intent right?

The iterative system seems to be the means by which differing amounts of damage emerge from combats where the odds are the same. Would you have a substitute system for varying the combat damage, or do you want to also be able to predict the combat damage for a given matchup?

I fear your idea may make combat way more predictable than it is in reality ... in your system, does Henry V ever win at Agincourt? Or Lee in most of his early campaigns in Virginia?

dV

 

[Tlalynet]

@ Kirk,
That pretty much kills subtlety doesn't it? Even your suggestion from fortifications quadruples the value of fortifying. You would either have to add decimal places and defeat the purpose of a log scale or have everything run off of percent bonuses. In most cases better units are between 33 and 50 percent better, not 100% each step.

 

[Ahriman]

It is worth noting that the current system (multiple combat rounds, each with a draw from a random distribution) mean that we get a well behaved probability distribution over different possible outcomes (final strengths of the two units).

Any system that has a simple strength ratio for win/loss means its going to have a really weird distribution over the different possible ways in which you can win or lose (ie the distribution over residual strength for the victor).

 

[Johnny Bravo]

I like the one unit per tile idea! I was never of fan of stacking. I think the new combat system is a more realistic take on REAL life combat. That is, weaker units in the front lines and stronger ones in the back guarding your city/leader. I’m also excited about the long-range attacks.

 

[KrikkitTwo]

Can we assume you want the odds (or perhaps we mean probabilities) of winning to be predictable from an inspection of the strengths (or the strengths modified by damage, terrain, fortifications, unit-specific bonuses), without having to resort to a calculator of iterative rounds of combat within a single unit on unit encounter?
You seem to be asking to take the complexity hidden in the iterative strikes that make up a combat, and bring that out in plain sight in the strenghts, and remove the iterative strikes ... have I got your intent right?

Yes

I hope you don't want the stronger unit to be guaranteed to win each time?

No... although I wouldn't be opposed to that

The iterative system seems to be the means by which differing amounts of damage emerge from combats where the odds are the same. Would you have a substitute system for varying the combat damage, or do you want to also be able to predict the combat damage for a given matchup?

No, the differing amounts of damage happen because there is a random factor.

For the Simplest system that is still random

Each unit has a "Displayed Strength" going into combat (taking into account terrain v. unit bonuses, health, etc.)

so Unit A v. Unit B=5 v. 6 (That's what I see when I look at their positions, etc.)

When the combat starts, a 'coin' is flipped and one sides strength is doubled
so the matchup is now
A v. B=10 v. 6 OR 5 v. 12

so the two possible results
A wins but takes 6/10 damage (from its current health) [so if it was at 75% health, it loses 45% health for a total of 30% health.] Basically its Strength goes from 5 to 2
or
B wins but takes 5/12 damage (from its current health)... basically Strength goes from 6 to 3.5


Now this system would only maintain some randomness... if the displayed strengths are 5 v. 12... then the 12 Will win.

But
the 12 could win and only take 5/24 damage (a minor hit) OR
the 12 could win and take 10/12 damage (a nearly fatal wound.. probably forcing a retreat)


Now you could provide more of a continuum, but that would add extra complications... and wouldn't really be necessary since the unit will be in multiple combats with other unts, will heal in between, etc. and that will provide the continuum.

I fear your idea may make combat way more predictable than it is in reality ... in your system, does Henry V ever win at Agincourt? Or Lee in most of his early campaigns in Virginia?
dV

Yes, due to
1. the effect of positioning and tactics (which should be major... none of this +25%, no you move onto a hill you have 2x strength defending)
and
2. the effect of luck

Finally EVERYTHING in civ is infinitely more predictable than it is in reality.

Techs for a really big one... who knows how long it will take for us to get fusion research
Rebellion... what are the odds of North Ireland being flipped by Dublin
Production...can you Really tell me how long it will actually take to build a new airport
Culture... what is the cultural value of the Getty museum compared to Vatican City
Revolution... how many more turns of Anarchy before Somalia gets its new civics, or before Afghanistan recovers from the invasion.

Note: the game would be practically unplayable if any of those were anywhere near as unpredictable as they were in real life. Combat is Only allowed to be random because it is not that important what happens to individual units.

It is worth noting that the current system (multiple combat rounds, each with a draw from a random distribution) mean that we get a well behaved probability distribution over different possible outcomes (final strengths of the two units).

Any system that has a simple strength ratio for win/loss means its going to have a really weird distribution over the different possible ways in which you can win or lose (ie the distribution over residual strength for the victor).

Not in the system above...

You either have
0% chance of winning (you are less than 1/2 strength)... and you can do 1/2 or 2x damage to the enemy

50% chance of winning (you are between 1/2 and 2x strength)... If you win you will take 1/2 of the enemy in damage if you lose you will do 1/2 damage to the enemy

100% chance of winning (you are more than 2x strength)... you can take either 2x enemy or 1/2 enemy strength in damage.


This allows a very well behaved distribution... 2 'peaks' 1/2 damage, 2x damage.. directly proportional to enemy str v. your str.

Not weird at all, 2 possible results: you did well in battle or you did poorly

Now.. this means that increasing your strength might not increase your win chances, but instead mean you 'survive' a win better and your enemy 'survives' a win worse.
This means that those 'health points' you lost should be important (because the enemy got more strength to make sure that happened)

The current system gives a nice... sort of since its not perfectly smooth... statistical curve, that would be great if this was Civ: statistics and probabilities

But this gives straight lines...
Like the line that says how big a project you can complete in 5 turns if you have a certain amount of production.

 

[tom2050]

No, it would be worse. The entire point of Rock/Paper/Scissors is that you cannot choose all three.

One-unit-per-tile will allow the proper use of combined arms. You must position your infantry to prevent the enemy from reaching your archers and you can use your cavalry to flank their archers if you find an opening.

No, 1 unit per tile means you put ranged behind whatever best front line unit you have in front.

It's exactly the same as everything else you said, you could still position your infantry in front, with a cannon there... and have 1 cavalry flank the others units. It's not a whole lot different. Would everyone always have 1 of each, if they could, probably.

But for the current situation, it will be quickly obvious what always works best, and it will always be used, just like you said, melee in front, ranged behind, flank with cavalry. Only time it changes is based on terrain, still same concept applies.

EDIT: for a unit that loses, they have already confirmed that the unit does not 'Die' in several articles, interviews, w/e. I suppose it can eventually, but it doesn't from an initial encounter I'm guessing, perhaps it needs to be attacked multiple times in order to actually defeat a unit.

Tom

 

[KrikkitTwo]

No, 1 unit per tile means you put ranged behind whatever best front line unit you have in front.

It's exactly the same as everything else you said, you could still position your infantry in front, with a cannon there... and have 1 cavalry flank the others units. It's not a whole lot different. Would everyone always have 1 of each, if they could, probably.

But for the current situation, it will be quickly obvious what always works best, and it will always be used, just like you said, melee in front, ranged behind, flank with cavalry. Only time it changes is based on terrain, still same concept applies.

EDIT: for a unit that loses, they have already confirmed that the unit does not 'Die' in several articles, interviews, w/e. I suppose it can eventually, but it doesn't from an initial encounter I'm guessing, perhaps it needs to be attacked multiple times in order to actually defeat a unit.

Tom

Hopefully what that means is that combat goes through multiple "rounds" wth player control in EACH round.

Ie
Attacker Fires on Defender (damaging it)... Defender then can either Halt combat or Fire back (damaging the attacker)... and then process repeats

If the Attacker out ranges the Defender, the Defender will stop combat after the 1st round. [classic bombardment limitation... Defender gets bombarded and they hide]
If the Defender is faced by a more powerful attacker, they can hold out longer.

This has the potential of meaning
Units are like cities... you build an occasional new one when you need it, and spend most of your resources developing/maintaining repairing the ones you have.

So new tech=/= build new units
new tech= upgrade units

Hopefully not with Gold, but with some Other resource that gets sent to an 'imperial stockpile'

 

[tom2050]

I think that, from the interview that mentioned this.. is the unit's will have military maintenance, so money can be spent to repear them. So however combat works, I don't believe that it will be down and dirty (controlling turns in the combat itself), but it sounds like it will be tough to actually kill a unit (always retreats perhaps, and must be attacked again for the death blow).

BUT, with 1 unit per tile; one of the ultimate warmonger strategy is going to be to either use military alliances (if they are in), or simply attack another civ that is already at war. Regardless of the way combat works, numbers still have advantage, so this will become a vital and necessary war strategy of the game. A civ at war with 2 or more civ's will have a massive disadvantage with all else equal.

Tom

 

[Grimlin]

Are there any information about air- and naval forces yet?
It makes sense to me having only one land unit per tile... But where do I put my airforce? On an insane amount of forts? In the air? Same goes for fleets... Not even thinking about the tedious task of moving a carrier battle group in single ships.

 

[Thormodr]

I sincerely hope we don't end up with "ping ponging" like in EU3 where you have to keep beating down a unit again and again.

 

[KrikkitTwo]

I think that, from the interview that mentioned this.. is the unit's will have military maintenance, so money can be spent to repear them. So however combat works, I don't believe that it will be down and dirty (controlling turns in the combat itself), but it sounds like it will be tough to actually kill a unit (always retreats perhaps, and must be attacked again for the death blow).

Well did they say Gold would be spent to maintain the Military? I'm sort of hoping not.

Hoping for something more like production (but an Imperial 'Military Production Pool')

Also I think Round by Round is the way to go. (assuming that each "shot" does about ~30% damage (so If shoot the enemy 3 times, they are almost dead... if they are as strong as me... if they are weaker, I might kill them in one shot, if they are stronger they might kill me in one shot)

That seems about right. 3 : 1 ratio->one round resolution

I sincerely hope we don't end up with "ping ponging" like in EU3 where you have to keep beating down a unit again and again

Well if it is expensive to repair the unit, then they aren't likely to ping-pong back.