Fun Hypothetical

[Flubber]

Originally posted by binTravkin


You didn't get my point.

I was talking about XED.
It makes your units move like worms so you can accompany a worm with cheaper unit easily.

True and actually then it can be your worm slowing things up .

It just seems by the time people build that, airpower is in play which makes any ground only force vulnerable

[Petek]

After thinking about this issue overnight, I see that I was wrong about battle resolution. If the attacker's weapon and the defender's armor have equal value (and all other factors are equal), but the attacker has a fusion reactor and the defender fission, then each unit has a 50% chance of winning any mini-battle. However, the attacker has more hit points and so has a greater chance of winning the overall battle.

I'm still of the opinion that the battle odds are not computed correctly. Suppose that the attacker has level 1 weapons and a fission reactor, and that the defender has level 1 armor and a fission reactor. Morale is the same for both units and no other factors influence combat. In that case, the battle odds will be shown as 1:1 and that is correct.

Now suppose that everything is the same as above, but the attacker has a fusion reactor. This doubles the attacker's hit points, so the battle odds will be shown as 2:1 in favor of the attacker. Those odds are not correct. To see this, suppose for the sake of simplicity that the defender has 1 hit point and the attacker 2. That means that the defender will lose the main battle if it loses one mini-battle, but the attacker has to lose two mini-battles before it loses the main battle.

There is a 50% chance that the attacker will win the first mini-battle, hence winning the overall battle. If the attacker loses the first mini-battle (also a 50% chance), then it is reduced to having one hit point. The attacker and defender now have equal odds of winning the second mini-battle and whichever wins that one wins the main battle.

Thus, the attacker's odds of winning the main battle are 50% (its odds of winning the first mini-battle) plus 25% (50% of the 50% odds that the second mini-battle will take place). That adds up to a 75% chance of winning, which are odds of 3:1, not 2:1.

I again set up a position to test this. I created an attacker with level 1 weapons, and two hits points remaining. The defender had level one armor and one hit point remaining. As expected, the battle odds were shown as 2:1 in favor of the attacker. I played the attack 60 times. The result was 45 wins for the attacker and 15 for the defender, or exactly 3:1 as predicted.

So, I withdraw the claim that the displayed battle odds reflect the predicted outcome of the mini-battles, but still hold that the displayed odds are wrong in most cases.

[CEO Aaron]

Centauri Ecology. I know it isn't glamorous or really high up there, but I really do believe that factions that begin with the ability to field formers have a strong competitive advantage against those that don't.

[Flubber]

Originally posted by CEO Aaron
Centauri Ecology. I know it isn't glamorous or really high up there, but I really do believe that factions that begin with the ability to field formers have a strong competitive advantage against those that don't.

Disagree strongly since you can pick anything else and still have centauri ecology as your first tech to research

[Kataphraktoi]

eudominia,you could have no faction bonuses but to start with this(and 40 ec) and that would be bonus enough imo

So, I withdraw the claim that the displayed battle odds reflect the predicted outcome of the mini-battles, but still hold that the displayed odds are wrong in most cases.

i find that when using locusts with 50%+ SP i often get 6 to 10 or worse odds but win anyway,vs elite aaa with aerospace with actual odds of 4v1 with 90% damage

[Net Warrior]

Darsnsn, I'm playing one of those PBEM's you set up with rover formers and it's real eye opener. Since then, I always make it a high priority to build a few rover formers ASAP.

[Senethro]

Rover formers are not good for their mineral cost until fusion power.

[Chaos Theory]

Hypothesis regarding combat odds:

The chance of winning a "mini-battle" is calculated based on attack and defense values, as well as any combat modifiers such as terrain, special abilities, and morale. This ratio (call it R), is an accurate expression of the chance the attacker will win a mini-battle.

The game calculates the ratio of hit points, and multiplies R by this to yield R', a poor estimate of the chance the attacker will win a battle. This number is converted to odds and presented to the player.

The actual chance the attacker will win a battle (call it B) is based on not only the ratio of attacker hit points to defender hit points, but also on their actual values. For example, if R = .5 and the attacker has 2 hit points to the defender's 1 hit point, B = .75, as shown by Petek. If R = .5 and the attacker has 2000 hit points to the defender's 1000 hit points, B > .99. Here's how to calculate B in the general case:

A = attacker's hit points
D = defender's hit points
B = sum(i from 0 to A-1, (i+D-1)-choose-i*(1-R)^i * R^D)

where x-choose-y = x! / ( (x-y)! * y! )

I have not tested to see if this matches game results, but this would be the result of a straightforward implementation of combat.

[Darsnan]

Originally posted by Net Warrior
Darsnsn, I'm playing one of those PBEM's you set up with rover formers and it's real eye opener. Since then, I always make it a high priority to build a few rover formers ASAP.

Glad you like it!

Originally posted by Senethro
Rover formers are not good for their mineral cost until fusion power.

Probably correct for the most part. However, if your popping a pod near a base which cannot "instabuild" anything better, then selecting a Rover Former in hopes the pod will instantly build it for you is not a bad choice.


D

[binTravkin]

Rover formers are not good for their mineral cost until fusion power.

I disagree with this.

A usual former costs 20, rover former costs 40 IIRC, and it makes up each of the extra mineral if you need to build a road to somwehere quickly.

When playing Sparta I usually try to instabuild a couple early (before other, better instabuilds are available).

[#endgame]

What's the cost of a rover former as compared to an inf former + inf transport? It'd accomplish the same thing.

[Ka Plewy]

Infantry transport costs 3 rows.
You'd spend more to build an infantry transport and an infantry former than just building the rover former, although, you could use the transport for other things.
Bringing some Impact infantry to your neighbors doorstep and lettit it attack. That 25% bonus for infantry is very effective!

[Sikander]

Originally posted by Chaos Theory
Hypothesis regarding combat odds:

The chance of winning a "mini-battle" is calculated based on attack and defense values, as well as any combat modifiers such as terrain, special abilities, and morale. This ratio (call it R), is an accurate expression of the chance the attacker will win a mini-battle.

The game calculates the ratio of hit points, and multiplies R by this to yield R', a poor estimate of the chance the attacker will win a battle. This number is converted to odds and presented to the player.

The actual chance the attacker will win a battle (call it B) is based on not only the ratio of attacker hit points to defender hit points, but also on their actual values. For example, if R = .5 and the attacker has 2 hit points to the defender's 1 hit point, B = .75, as shown by Petek. If R = .5 and the attacker has 2000 hit points to the defender's 1000 hit points, B > .99. Here's how to calculate B in the general case:

A = attacker's hit points
D = defender's hit points
B = sum(i from 0 to A-1, (i+D-1)-choose-i*(1-R)^i * R^D)

where x-choose-y = x! / ( (x-y)! * y! )

I have not tested to see if this matches game results, but this would be the result of a straightforward implementation of combat.

The game system is rather elegant in this regard even if the "reactor level" seems to be a bit silly. In fire combat IRL assuming that every soldier in a battle is equally skilled on both sides and equally equipped and that both sides soldiers can fire at any soldier on the other side, but won't overlap fire until every soldier on the other side has been targeted and finally that there is a certain fixed percentage chance that any shot will miss...

then the true odds of a particular side prevailing (another assumption is soldiers are not affected by morale for better or worse) can be obtained by squaring both sides numbers and comparing the squared totals. Thus in tank battles etc. once the tanks are exposed and a general firefight occurs the side with the greater numbers will have a much greater advantage than a simple comparison of the numbers might suggest.

The game seems to simulate a series of small battles where friendly units are not close enough to one another to support each other directly, but must be fed into the meat grinder serially. This takes away much of the advantage of the numerically superior (or in SMAC better reactor side as well) side.

[Commy]

Originally posted by Petek

The question is whether or not the odds shown on the battle odds screen reflect the chance of the attacking unit winning the battle. I say that they do not (unless the odds are shown as 1:1).

You suggest that they do and that this can be proven (if I understand your above comment correctly) by starting with two units of equal battle value (so the battle odds will presumably be 1:1) and then giving the attacking unit a higher reactor (for example). Clearly the battle odds screen will then show that the attacker now has more favorable odds. However, this doesn't prove that the battle odds screen display the odds that the attacker will win. You would have to play the attack multiple times and note how often the attacker wins.

I thought I said this earlier, but no, they do not...battle odds and actual odds are two very, very different things...think of it this way...

Say you have two regiments that each hate eachother...one regiment has 1000 people, the other has 500...

The battle odds are 2:1...by probability mathematics, this indicates that out of three battles, 1000 men would win twice and 500 men only once...civ 3 works this way...

However, the actual odds are different, so in reality, the 1000 would probably win almost every time, not two-thirds of the battles...

Arguements on the Internet are so interesting, especially when people who think share opposite views really share the same and prove eachother correct...