Bonus Ship Ommission File

[Toranth]

Second, saying "Someone else did it, therefore it must be right" is not a useful claim.

This is a straw man argument, or a caricature of what I have said.  To claim, "It has worked well before, therefore it can work well again", is much different than saying, "It has worked before, and therefore must be right".  You've built a straw man implying that my entire argument hinges on the fact that it has worked before, and you're wrong because I also provide plenty of other reasons why as well.

Actually, I was pointing out that your claim was a variant of the 'Appeal to Emotion' fallacy called 'Appeal to Popularity'.  Warcraft 3 was a commercially successful product.  This does NOT mean that Warcraft 3's armor system was good.  The two items are completely unrelated.
That said, let's please drop the fallacing-pointing-out-arms-race here.  That's the just "Fallacy fallacy" in action, after all.

If you were seriously suggesting that we replace the wide range of numerical values with just two or three simple ones, I might agree with that.  But you're not suggesting that, you're suggesting that we take a simple and intuitive system, and make it non-simple, obtuse, and inconvenient - aka, not fun.

What is obtuse and inconvenient about a system that can actually show you the ACTUAL damage reduction of each attack right on its bar?  The system we use where 7 steps of calculation are required to reach a result, instead of 2 steps, is much more complicated than what I'm proposing.  This is more fear mongering.

I've presented no doomsday scenario, nor made any claims of one.  I've pointed out that your suggestion is non-intuitive.  That's it. 
Additionally, you said that the damage reduction is on the bar... but, it is not.  It can be computed from what is on the info screen, if you know the method to convert the 'Armor' value to a 'Damage Reduction' value.  But the reduction value itself is not listed.

Now, your 'effective health' system could sufficiently obvious if instead of having an unestricted armor value arbitrarily map to a percentage-based damage reduction, you just flat out used a percentage damage reduction.  That's simple, and obvious.  Apply the same 'minimum percentage of damage' we have now, and you've got your armor system. 


Next, I have indeed been misunderstanding you.  I did not realize that when the discussion shifted from 'hull-type vs hitpoints' to 'armor systems', you were including getting rid of the hull types in your armor system.  That had never been made clear in my mind.
Now that I understand that you want to replace all five offensive and defensive factors (hull bonus, hull multiplier, damage per shot, target hull type, target armor) with two (Damage per shot, target Armor), I object even more.

The hull types and damage multipliers are complicated, and at the moment rather arbitrary.  However, they are useful in differentiating the ship types.  Because of multiplier bonuses, each ship can be customized as a counter for specific ship types without becoming better against non-counter units.  This is not possible if the only differentiator is Armor. 

EDIT: Keep in mind, no matter what formula we go with, the game can tell you how much damage is going to be done. Both in the actual matchup (the damage display when you hover over a ship when you have a ship selected) and the reference tab for "hypothetic" match-ups you don't have in front of you at the moment but you would still like to know. So people who don't want to do "math gymnastics" to figure out damage don't have to, the game can tell them. They would only need to work through the formula if they want to understand how the damage number is gotten, or for what stats to look to min/max when finding the best unit to counter another unit.

I had forgotten all about this.  Yeah, the game will do the math for you for each given matchup.  Usually, I just do it myself as part of strategic min/maxing.

[Fluffiest]

Sorry for this mammoth post...

Ship A shoots ship B. What factors are involved in the calculation of effectiveness?

Firstly, ship A has an ammo type (such as Flame Wave, Fusion Cutter, Missile, and all the rest). Ship B may or may not have an immunity to this.

Secondly, ship B has a hull type (such as Command Grade, Polycrystal, or Ultra-Heavy). Ship A may have a set of attack multipliers which make it more or less effective against targets with certain hull types. These are, as far as I'm aware, completely independent of the ammo type.

Thirdly, ship A has a damage value and an armour piercing value, while ship B has an armour value, which together determine how effective ship A is at damaging ship B.

Finally, there are a few other considerations: Armour rotting, Zenith Polarizers, Impulse Reaction Emitters and so on.


Which of these FOUR systems give us unique advantages that we can't replicate with the other three? Which cause complexity and confusion that we would desperately like to be rid of?

• Immunities are listed on the target ship, and care about ammo type. Attack multipliers are listed on the attack ship and care about hull type. You can have a x10 multiplier against something, but find out later that it's immune to your ammo type.
• Attack multipliers vary from 0.001 to 30 (admittedly, these are extreme cases on unusual ship types - a more typical range is 0.1 to 8 ). Modifiers are not consistently good, or consistently bad. Ships can, in theory, have weak damage but be very valuable because they have high multipliers against common or dangerous hull types, or vise versa.
• Attack multipliers are not consistent with ammo type: for example, Spire Siege Engines and Fortresses both use flame wave ammo, but only the latter has a massive penalty against Polycrystal. In fact, most of the time, ammo type does not matter - it's only used in the calculation of immunities, and most ammo types have nothing that's immune to them and are therefore completely irrelevant.
• Hull types include Ultra-Light, Light, Medium, Heavy, and Ultra-Heavy. These do not usually correlate to strength of the attack multipliers against them: many things have higher damage against ultra-heavy than against ultra-light. They do not correlate with health: there's a starship with an ultra-light hull. They do not correlate with amount of armour. A heavier-sounding hull type is not necessarily an advantage and a light-sounding one is not necessarily a disadvantage.
• Hull types do correlate approximately with role, for example there has been an attempt to ensure most stuff that turns up in swarms has the Swarmer hull type, most cloaked units have Refractive hulls, and most big scary stuff is ultra-heavy or command grade.

We want to keep hull types in some form, because armour alone is probably too simplistic. We could not, for example, have bombers that are very effective against a heavily-armoured starbase but weak against an heavily-armoured fleet ship.

Armour is nice to have, since it gives you a clear indicator of unit strength: high armour is strictly better than low armour. It is NEVER a disadvantage to have more armour, unless the AI has an absolute tonne of Zenith Polarizers. The same is not true of heavier hull types!

Currently, and ammo types are only used to determine immunities. This seems bizarre.

My recommendation would be to pick one of the following:

1) Remove Armour entirely. Revise the distribution of Hull Types and bonuses to ensure the same things are still good against the same targets.

2) Use armour as the main measure of which ships are effective against each other. Remove the Ultra-Light, Light, Medium, Heavy, and Ultra-Heavy hull types altogether, and replace them with an appropriate amount of armour and a new "standard hull", which nothing has bonuses against. Use the other hull types sparingly in order to give ships specialist roles or specific weaknesses. Advantage: Most of the time, you don't need to care about hull type. Armour and armour piercing values tell you most of what you need to know. Nonstandard hull types and multipliers can be reserved for specialized ships since armour is doing most of the heavy lifting, and standardization should be easier.

2.5) As above, but we could remove hull types altogether - instead of giving ships a list of hull types that they get multipliers against, give them a list of ammo types that get multipliers against them (some of which will be a multiplier of 0; i.e. immunities).

3) Switch from per-unit damage bonuses to per-ammo-type damage bonuses. For example, Flame Wave always gets a substantial penalty versus Polycrystal, energy bomb ammo type always gets a bonus against Structural. This would require changing the hull and ammo types of a lot of units to keep the balance right. Advantage: It's much easier to memorize what's good against what. The other big advantage here is that you could improve the unit display. Instead of "Ammo type: Energy Bomb; Hull type: Polycrystal; Multipliers: 6 Heavy, 6 Artillery, 6 Command-Grade, 6 Ultra-Heavy, 6 Structural" it would just display"Ammo type: Energy Bomb; Hull type: Polycrystal". Mouse-over "Energy bomb" and it will give you the multipliers list. Mouse over "Polycrystal" and it will give you the inverse of the multipliers list; all the ammo types that are good against this unit. We might end up keeping armour and armour piercing in some incarnation, but hull types become less of a memorization fest. I don't know if we'd keep immunities as separate from the bonuses system or not; we'd probably have to.

4) Completely overhaul how we think of "hull type". Let units have multiple hull types, stacked on top of each other, each giving them a specific advantage in the form of a damage reduction against certain attacks. More hull types - or possibly call them "Defenses", because one possible Defense might be "I'm very small, fast, and hard to hit" - is always better than few, but some units get to ignore or reduce the bonus from some Defenses. Remove armour entirely since it's not really needed at this stage. "Heavily armoured" can just be another Defense. This is the idea Hearteater came up with, but I wouldn't make all the defenses reduce damage by 50%.

As for the argument of armour as it stands (subtractive, capped at -80%) versus EHP armour, I'm going to stay out of it for now.

[keith.lamothe]

it would just display"Ammo type: Energy Bomb; Hull type: Polycrystal". Mouse-over "Energy bomb" and it will give you the multipliers list.

I may have missed it in there somewhere, but:

How do you mouseover something that only shows as part of a mouseover tooltip?

who watches the watchers?

[TechSY730]

Oh yea, I forgot about the proposal to make hull multipliers per shot type rather than per ship type. How did I forget that? Didn't Keith also originally propose this one?

[keith.lamothe]

Oh yea, I forgot about the proposal to make hull multipliers per shot type rather than per ship type. How did I forget that? Didn't Keith also originally propose this one?

Yea, I would love it if you could look at two a unit and say "Ok, this fires 'siege plasma', that means it's good against anything that's 'Shield' or 'Titan Plate'" or "Ok, this fires 'pulse laser', that means it's good against anything that's 'Polycrystal' or 'Unarmored'" or whatever.

But I'm not sure that would actually be the right direction for this game.

[LaughingThesaurus]

I've avoided this thread because it involves a whole massive mess of an issue that I don't really have the time to indulge in... but I do like the idea of Ammo Type > Certain Hull Types rather than it being on a ship by ship basis. I think going that direction, however, will lead to a big complicated mess of rebalancing everything to work the way that it's meant to. Sounds really impractical to implement to me, given how you already fit ships into particular roles.

Regarding the big armor/HP debate, I don't like the idea of the armor value being something other than a straight reduction. I want to look at a number and know the difference easily, rather than having to ponder what percentage of damage reduction there happens to actually be and calculate percentages of numbers in my head. I'd rather just subtract and be done with it. Now, what would be the problem with taking HP back down to its significantly less inflated state? The idea of doing hundreds of millions of damage is cool and all, but it gets a bit old after enough time. Aside from it just being impractical to do, where would be the problem? There could be a whole lot more difference between every ship without reaching some kind of critical mass of HP that causes the galaxy to collapse in on itself.

[chemical_art]

Torn on this whole mess.

On the one hand, the easy to read ability of the current armor system is inice.

On the other hand, that is kinda moot when armor is only very rarely matters for a select few units sometimes, and multipliers stomp armors so hard that their already weak state is even more niche.

[Wingflier]

I've presented no doomsday scenario, nor made any claims of one.  I've pointed out that your suggestion is non-intuitive.  That's it. 

And I've kindly pointed out why you're wrong again and again.  It's a very intuitive system.

Additionally, you said that the damage reduction is on the bar... but, it is not.  It can be computed from what is on the info screen, if you know the method to convert the 'Armor' value to a 'Damage Reduction' value.  But the reduction value itself is not listed.

The game can compute it for you.  The damage reduction is simply based on armor and max health, therefore "damage reduction" will become a new stat that is easily labeled on the ship's bar.  All the player has to do is mouse over to see the damage reduction against raw damage.

In addition, the game already has a method of telling you "damage to do" when you have a ship selected, and mouse over an enemy ship.  With the new system, we could simply change "damage to do" to "damage per shot", and by mousing over an enemy ship with your own, you could quickly tell how much you would do without even having to do the quick math in your head.

I want to look at a number and know the difference easily, rather than having to ponder what percentage of damage reduction there happens to actually be and calculate percentages of numbers in my head. I'd rather just subtract and be done with it.

In the EHP Armor System, the damage reduction IS a percentage based on the max health of the target.  The actual damage reduction in percentage could be listed in a unit's information bar.  This meets your criteria.

Now that I understand that you want to replace all five offensive and defensive factors (hull bonus, hull multiplier, damage per shot, target hull type, target armor) with two (Damage per shot, target Armor), I object even more.

Actually you're forgetting armor penetration in both scenarios.  So I want to change the damage system from 6 factors to 3, making it much more elegant and robust than it was before.

Please understand that "less variables" does not automatically equal "less deep".

I'll give an example using standards of measurement:

The Metric System vs. The U.S. Customary System. 

NOBODY says that because the U.S. Customary Measurement System has more variables, that it is a deeper or better way of measurement.  The Metric System uses many less variables, and uses many of the same variables over and over (such as the prefixes, centi, milli, deci, etc.), yet the system is much better, deeper, and more elegant all at the same time while using less variables.  More variables does not automatically equal better.

The hull types and damage multipliers are complicated, and at the moment rather arbitrary.  However, they are useful in differentiating the ship types.  Because of multiplier bonuses, each ship can be customized as a counter for specific ship types without becoming better against non-counter units.  This is not possible if the only differentiator is Armor. 

Well, you've taken my suggestion and changed your argument from "your system is too complicated" to "your system is too simple", which is a better tactic than before.  Unfortunately, even by shifting your direction 180*, you're still wrong.

However, they are useful in differentiating the ship types.  Because of multiplier bonuses, each ship can be customized as a counter for specific ship types without becoming better against non-counter units.  This is not possible if the only differentiator is Armor. 

Once again you're forgetting armor penetration which is an equally important part of the process, and aids greatly in creating said "counters" that you seem to think are impossible to create using the EHP system.

I'll make a quick mathematical demonstration of how we can keep the Triangle (Fighter, Bomber, Frigate) as they pertain to each other almost exactly the same using the EHP system in the next post - debunking your argument that you can't create effective counters using the system.

[TechSY730]

I've avoided this thread because it involves a whole massive mess of an issue that I don't really have the time to indulge in... but I do like the idea of Ammo Type > Certain Hull Types rather than it being on a ship by ship basis. I think going that direction, however, will lead to a big complicated mess of rebalancing everything to work the way that it's meant to. Sounds really impractical to implement to me, given how you already fit ships into particular roles.

Regarding the big armor/HP debate, I don't like the idea of the armor value being something other than a straight reduction. I want to look at a number and know the difference easily, rather than having to ponder what percentage of damage reduction there happens to actually be and calculate percentages of numbers in my head. I'd rather just subtract and be done with it. Now, what would be the problem with taking HP back down to its significantly less inflated state? The idea of doing hundreds of millions of damage is cool and all, but it gets a bit old after enough time. Aside from it just being impractical to do, where would be the problem? There could be a whole lot more difference between every ship without reaching some kind of critical mass of HP that causes the galaxy to collapse in on itself.

You mean something like divide all HP and weapon damage by 10 or something, but leave armor as is, as a "first step"? (and tweak any anomalies that might be produced afterwards)

[chemical_art]

I think changing to a % based damage reduction over time is easier to balance over all.

The current system simply cannot be balanced due to caps.

A MK I bomber on normal caps cannot pierce 100k armor of hardened shields or the armored golem. That same bomber on ultra low caps has over half its damage piercing the armor.

If you made armor scale with caps even for units that don't other wise scale with caps (golems, structures, starships, etc.) then you have to increase armor piercing to units as well...and in this process, units that have little to no armor piercing are more gimped in ultra low cap games due to a greater part of their damage being reduced (for their damage doesn't rise with the greater armor)...and if you increase damage then you pretty much have to boost all parts of the game so every unit scales in cap.

However, if the units had the same % piercing / reduction regardless of caps, then caps no longer dictate the varying interactions of armor with unit caps.

[chemical_art]

You mean something like divide all HP and weapon damage by 10 or something, but leave armor as is, as a "first step"? (and tweak any anomalies that might be produced afterwards)

That really doesn't address that in fleet ship interactions their armor is so low on average compared to even armor piercing unfortunately. That solution is a band aid, and I don't feel it addresses any of the underlying problems.

[TechSY730]

You mean something like divide all HP and weapon damage by 10 or something, but leave armor as is, as a "first step"? (and tweak any anomalies that might be produced afterwards)

That really doesn't address that in fleet ship interactions their armor is so low on average compared to even armor piercing unfortunately. That solution is a band aid, and I don't feel it addresses any of the underlying problems.

Divide all HP and weapon damage by 50 then? 100? And divide all armor piercing values by 10?

You are right, it does not address the root imbalance, but it could be a feasible "stop gap" measure, and/or a good first step. It would also have the nice side effect of combating the ridiculous HP inflation that has crept into the game over the last few major versions, which alone makes it seem like something to at least consider.

[keith.lamothe]

The current system simply cannot be balanced due to caps.

A MK I bomber on normal caps cannot pierce 100k armor of hardened shields or the armored golem. That same bomber on ultra low caps has over half its damage piercing the armor.

Have you actually observed this experimentally?

There are some quiet rules in there that handle the "scaling shot vs non-scaling armor" (and vice versa) interface.  I'm not 100% certain they work because people don't really notice them, but they're there.

The need for stuff like that is one reason I'd like the system to just go away.

[Toranth]

I've presented no doomsday scenario, nor made any claims of one.  I've pointed out that your suggestion is non-intuitive.  That's it. 

And I've kindly pointed out why you're wrong again and again.  It's a very intuitive system.

How?  As an anecdote, I just asked three people how much more effective 100 armor should be over 20.  All three said 5 times as effective.  That's intuitive.
You've declared your suggestion is intuitive, but I fail to see how anything that requires a mapping equation is intuitive! 

Additionally, you said that the damage reduction is on the bar... but, it is not.  It can be computed from what is on the info screen, if you know the method to convert the 'Armor' value to a 'Damage Reduction' value.  But the reduction value itself is not listed.

The game can compute it for you.  The damage reduction is simply based on armor and max health, therefore "damage reduction" will become a new stat that is easily labeled on the ship's bar.  All the player has to do is mouse over to see the damage reduction against raw damage.

If the game will compute the damage reduction percentage, and display it for you (instead of the Armor value) then why use an armor value?  Especially since hidden equations lead to unexpected behaviors. 
Also, the GUI issues mentioned:  how does your Armor Piercing work with this display?  Would mousing over a ship show different numbers depending on what you have highlighted?  What if you have multiple ship types selected, or none?

Now that I understand that you want to replace all five offensive and defensive factors (hull bonus, hull multiplier, damage per shot, target hull type, target armor) with two (Damage per shot, target Armor), I object even more.

Actually you're forgetting armor penetration in both scenarios.  So I want to change the damage system from 6 factors to 3, making it much more elegant and robust than it was before.

<snip irrelevence about metric>

The hull types and damage multipliers are complicated, and at the moment rather arbitrary.  However, they are useful in differentiating the ship types.  Because of multiplier bonuses, each ship can be customized as a counter for specific ship types without becoming better against non-counter units.  This is not possible if the only differentiator is Armor. 

Well, you've taken my suggestion and changed your argument from "your system is too complicated" to "your system is too simple", which is a better tactic than before.  Unfortunately, even by shifting your direction 180*, you're still wrong.

The hull types and armor are two completely different, unrelated systems.  When I talk about Armor, I'm talking about Armor.  I'm not talking about hull types, or ammo types, or to-hit percentages.  My argument is still the same:  Armor should be a direct, no-hidden variables, no secret equations, damage reduction.

When you conflate some of the major differentiators of ship types (Hull type and target multipliers) into your 3 factor armor system, you lose most of your flexibility.  Additional variables allow for additional categories. 
How do you differentiate 100 different bonus ship types, if you only have 3 categories to fit them in?  Hull types, multipliers (and, as Valtiel pointed out, ammo types) are a very flexible way to do so.  So, when we discuss fixing the Armor/damage reduction issue, any suggestion that incidentally eliminates that ability to differentiate is 'too simple'.  Unless you are suggesting that most ship types be eliminated...?

However, they are useful in differentiating the ship types.  Because of multiplier bonuses, each ship can be customized as a counter for specific ship types without becoming better against non-counter units.  This is not possible if the only differentiator is Armor. 

Once again you're forgetting armor penetration which is an equally important part of the process, and aids greatly in creating said "counters" that you seem to think are impossible to create using the EHP system.

Armor Piercing would apply to all armor values, though.  That doesn't create a counter to certain ships, it creates a category - the third leg of your Rock-Paper-Scissors triangle (DPS, Armor, Armor Piercing, IIRC).

I'll make a quick mathematical demonstration of how we can keep the Triangle (Fighter, Bomber, Frigate) as they pertain to each other almost exactly the same using the EHP system in the next post - debunking your argument that you can't create effective counters using the system.

I'll look forward to reading what you show - but to skip a step in the argument that I know I'll make, after balancing your three triangle ships, try throwing in a bonus ship or two, and show how they fit into your system, as different ships, from the triangles.

[chemical_art]

The current system simply cannot be balanced due to caps.

A MK I bomber on normal caps cannot pierce 100k armor of hardened shields or the armored golem. That same bomber on ultra low caps has over half its damage piercing the armor.

Have you actually observed this experimentally?

There are some quiet rules in there that handle the "scaling shot vs non-scaling armor" (and vice versa) interface.  I'm not 100% certain they work because people don't really notice them, but they're there.

The need for stuff like that is one reason I'd like the system to just go away.

If there was a way I could get a bomber of mine to attack a shield of mine (because that is about the only way I could reliably get 1 v 1 combat) then I would check it.

The fact you already have "behind the scenes" math though to compensate for caps dilutes the pro of the current system of armor being easy to understand, for now I'm a stage of "well, I don't know" in regards to how the armor for non scaling things works.