Wow, choppers are grossly OP !!!

[Yitzi]

That's good stuff to ponder, Yitzi.

Hmm...

I do like your term, "rocket tag", to describe combat between unarmored Missile Infantries.

I've first seen it in the context of high-level D&D, where offense tends to be overly powerful in comparison to defense, and what defenses exist tend to be complete no-sells...but it does apply here too.

I'm thinking about your example of 1-5-1 vs 1-5-1. I'm struggling to see how this pillow-fight can result in serious destruction, or be the bloodiest based on winner's final damage.

Because eventually enough damage will be done to kill one side (even if it takes 2-3 turns of slugging it out), but by that point there'll have been enough fighting that even the winner will have taken almost enough damage to kill him.

I think 8 rounds is plenty (i.e. 4 attacks, 4 counter-attacks) for Fission units or psi units. However, for Fusion reactors, maybe increase to 12 rounds? Some adjustment would need to be made for combat between differing reactor levels.

Sounds like a good idea; more attacks is more rounds.

Should combat become more destructive as reactor values increase while Weapon/Armor ratios remain similar? I can imagine advanced-reactor units staying in the fight, pressing for a victory this year, not next year. Both sides are willing and able to take more damage.

If so, you'd need to adjust weapon-armor ratios if you want a substantial number of "both sides survive" battles.

[TarMinyatur]

In more detail, the probabilities of the 6-1-1 vs. 6-1-1 battle would come out roughly as follows:

41.4%: Initiator wins in first round, takes no damage.
24.3%: Defender wins in second round, with median loss around 20%.
17.7%: Initiator wins in third round, with median loss around 20%.
8.58%: Defender wins in fourth round, with median loss around 50%.
4.69%: Initiator wins in fifth round, with median loss around 50%.
1.96%: Defender wins in sixth round, with median loss around 60%.
0.881%: Initiator wins in seventh round, with median loss around 60%.
0.333%: Defender wins in eighth round, with median loss around 67%.
0.203%: Both combatants survive, taking damage around 67%.

How do you roughly calculate damage? For example, in Round#2 the victorious Defender suffers 20% damage. Would it be common to lose 10% or 30%? Is this distribution in the form of a steep bell curve?

[Yitzi]

In more detail, the probabilities of the 6-1-1 vs. 6-1-1 battle would come out roughly as follows:

41.4%: Initiator wins in first round, takes no damage.
24.3%: Defender wins in second round, with median loss around 20%.
17.7%: Initiator wins in third round, with median loss around 20%.
8.58%: Defender wins in fourth round, with median loss around 50%.
4.69%: Initiator wins in fifth round, with median loss around 50%.
1.96%: Defender wins in sixth round, with median loss around 60%.
0.881%: Initiator wins in seventh round, with median loss around 60%.
0.333%: Defender wins in eighth round, with median loss around 67%.
0.203%: Both combatants survive, taking damage around 67%.

How do you roughly calculate damage? For example, in Round#2 the victorious Defender suffers 20% damage. Would it be common to lose 10% or 30%? Is this distribution in the form of a steep bell curve?

What I would do is approximate a bell curve (via flipping a coin a lot of times and counting heads minus tails then scaling appropriately), then square the result and multiply by attack/defense.

For the 20% median I mentioned, 10% and 30% would be fairly normal, though it would tend to be bottom-heavy for one round.  (Two would be less so, and starting at three rounds you'd start to see a peak.)

It is not a bell curve, but rather a different distribution; the following image (copied from Wikipedia) shows the distribution of damage taken after k taking-damage rounds in the case of 4-1-1 vs. 4-1-1.  (6-1-1 vs. 6-1-1 would therefore be 1.5 times as much damage.)



It may be more helpful, though, to consider the probability of taking at most a given amount of damage; that is shown by the following image (also from Wikipedia, and again 6-1-1 vs. 6-1-1 is 1.5 times as much as shown):



Of course, this can result in more than 100% damage, which will mean that that unit will not actually be the winner.  My 20% value was achieved by taking the yellow line in the bottom image, seeing what the chance is of survival (0.586 as it turns out; remember, it's 6 vs. 1 so the unit will be killed when the value is over 2/3), and then finding the damage taken at half that probability (it turns out to be .141X1.5, or 21.15%).  So the distribution will be the yellow line along the first 2/3 of the first column of boxes.

In defense-heavy combat, there'll be a lot more rounds, resulting in something more bell-curve-like.

[Yitzi]

By the way, the discussion (for anyone following this thread) has migrated to here.

[Question]

Im not impressed by choppers. I mean yes they can wipe out a whole invading force in one turn...but at that point, the AI never really tries to attack me anyway, and their range is just way too short to mount offensives. Im usually bombing the AI's units in his territory with needlejets, so defence isnt an issue.

[Yitzi]

Yeah, choppers are more of a concern in MP games than vs. AI...

[Question]

I think part of the problem is that air units dont need armor so they are much cheaper than typical land units. There is a risk with building unarmored rovers in that if they get caught, they will likely die without putting up a fight. The risk just isnt there with air units so you get super cheap air units. For choppers taking out an entire army by themselves, im not sure how that works in MP when human players bring AAA, but perhaps they should have a penalty for each additional attack in a turn, or limit them to 2x attacks a turn?

Its interesting to note that choppers and needlejets are actually CHEAPER than rovers....

Also without any movement bonuses choppers have very limited range. 8 MPs basically means you can move 3 squares, attack 2 times then head back to base.

[Yitzi]

I think part of the problem is that air units dont need armor so they are much cheaper than typical land units. There is a risk with building unarmored rovers in that if they get caught, they will likely die without putting up a fight. The risk just isnt there with air units so you get super cheap air units.

Why is the risk not there with air units?

Also without any movement bonuses choppers have very limited range. 8 MPs basically means you can move 3 squares, attack 2 times then head back to base.

Which is why I feel that removing the reactor-based movement bonus on choppers is one valid solution.

[Question]

Because air units spend most of their time in a safe base or airbase. Unarmored rovers trekking across the map are really vulnerable once spotted. Infact i tried this earlier and the AI was inflicting high casaulties with needlejets....i would have lost my entire force if he was using choppers. I had no way to actually block his needlejet's attacks so i was losing one unit to every needlejet.

Air units are only vulnerable to SAM and they have "first strike" capability. While you can spot enemy rovers coming to get you and potentially kill them before they get into range with your own stuff, you cant touch needlejets that are 6+ tiles away in a base. So they are guaranteed to inflict some casaulties (probably killing one unit per needlejet unless you have high armor AAA units to block them).

Also since they are cheaper than rovers, you can out-spam an enemy rover force.

But choppers dont get bonus MP from reactors as it is?

[Yitzi]

Because air units spend most of their time in a safe base or airbase.  Unarmored rovers trekking across the map are really vulnerable once spotted.

Unless you stack them with higher-armor units.

Also since they are cheaper than rovers

Barely.

But choppers dont get bonus MP from reactors as it is?

They do, but my patch allows that to be changed via alphax.txt.

[Question]

Unless you stack them with higher-armor units.

Sure but for every attack rover + defence rover you make, your opponent can make more than 2 air units focused soley on weapons without worrying about armor.

So for example if you have 2 attack rovers and 2 defence rovers with AAA, i have 4 (or more) choppers or needlejets. Since i have first strike, i also get the benefits of any offense abilities like sophoric pods, nerve gas, etc. You may or may not have terrain bonuses (you are going to have to cross open ground sooner or later) and even with a 50% defence bonus, its 4 attackers v 2 defenders so i can win via attrition. Once your defenders are gone, your weapon rovers are defenceless and can be picked off at my leisure.

[Yitzi]

Unless you stack them with higher-armor units.

Sure but for every attack rover + defence rover you make, your opponent can make more than 2 air units focused soley on weapons without worrying about armor.

So for example if you have 2 attack rovers and 2 defence rovers with AAA, i have 4 (or more) choppers or needlejets. Since i have first strike, i also get the benefits of any offense abilities like sophoric pods, nerve gas, etc. You may or may not have terrain bonuses (you are going to have to cross open ground sooner or later) and even with a 50% defence bonus, its 4 attackers v 2 defenders so i can win via attrition. Once your defenders are gone, your weapon rovers are defenceless and can be picked off at my leisure.

Except that with a 100% defense bonus from AAA, even without terrain bonuses he'll lose two of his air units just damaging your defensive rovers, and then badly damage the other two killing them.  (With terrain bonuses and tech parity/balance, it's pretty much a toss-up, except that if he wins he's got badly damaged units whereas if you win you've got 2 healthy attackers.)  So air still counters rovers, but not very strongly.

If, on the other hand, you use infantry instead, they're slower but a lot cheaper, so you can get enough units to completely no-sell his attack.

[Question]

? Wouldnt a missle air unit vs a plasma steel AAA defender be roughly 50-50?

[Yitzi]

? Wouldnt a missle air unit vs a plasma steel AAA defender be roughly 50-50?

But that's not even levels of tech; missile is level 4 requiring 8 techs total (counting itself), and plasma steel is level 2 requiring 3 techs total.  A better comparison would be missile vs. silksteel (level 4 requiring 10 techs total).

Of course, there is the issue of air power requiring Synthetic Fossil Fuels and nothing of significance until much later requiring Silksteel, but that can be fixed without my help.

[Nexii]

You also have to look at expected mineral loss, not just the combat odds.  If an air unit costs 40M and infantry 20M, and it's 50-50 combat odds, then the defender is coming out ahead.  This is why infantry should cost less than air.  Air will still have use for scouting, bombarding sensors, or multi attacking lone units.  AAA although strong isn't a given on all land units as there are many other good abilities.  For sea I'll agree that AAA is pretty much a given in one slot once you get it.