Damage mechanisms: damage and multipliers, is that all?

[Alex_aiwfor]

I've searched for the way that damage is calculated; I've found all the info (I think), but I would very much appreciate confirmation:

- Ships deal a base damage. If the shot hits, then this, multiplied by specific multiplier, is exactly the damage dealt to the target.

- The multipliers are stored in a huge array, with a specific case for each possible attacker/target pair. In particular, there is no damage/armor tpye.

- When determining if the shot hits, only the following are taken into account: _initial_ distance between the ships, range of the attacker, shileds of the target (with a few more subtilities which make the shot automatically fail or win, or give an automatic 5% chance, as explained perfectly here: http://arcengames.com/mediawiki/index.php?title=AI_War_-_Do_Shots_Miss).

- In particullar, there is no "accuracy" or even (?) projectile speed (meaning every projectile goes the same speed. Not sure and haven't played much, though).

- A _lot_ of other special abilities exist, but they are always specific (stored in a huge explicit array).



The following opinion applies only if what I just said is correct:

I am not a huge fan of the "huge explicit array" approach. I mean sure, it works, and every other sophisticated system can be compiled to this one, but it has quite a lot of drawbacks, mostly in the way that you may learn how the system works. This is true for beginners, but also for others everytime a patch modifies the balance. It also makes it very difficult to chose a new unlockable ship (the simulation percentage helps, but keeps you from understanding what is happening). Allow me to make comparaisons to two other games:

- warcarft 3: every unit has an armor and damage type. I actually hate this system, because it does not appear if you do not read every description; and it is not always obvious what kind of damage a certain attack will do (think hero's abilities). I think pro players eventually rebuild the "huge array" thing in their head. But at least it is a start (and buildings always have building armor, heroes hero attack,...).

- Gratuitous space battle: I think damage/armor type is well done here, because you can SEE the difference onscreen. Damage types are missiles/lasers/plasma. You cant miss it. You can see shots bounce from shields, and shields dramaticaly go down. But the problem (aside from not being an RTS, or scaling to 30k units, or...) is that the balancing is really not that well done. And of course the "huge  array" is more flexible for rebalancing.

I think Ai war needs _some_ damage/armor types. Starship armor type, for example. It's really like w3's heros, and it seems really obvious. I saw in an other thread that, by pulling all the raw modifiers for the dreadnought, you could see it had bonuses against (almost) each starships. You know what would be nice? A "strong against starships" line. Or even better, a "heavy cannons" weapon type, with an explanation "heavy cannons are strong (x10) against starships, but weak against (x??)". Same with the million special abilities that some important ships have: " immune to tractor beams, immune to reclamation, immune to the common cold, etc". What about "Class: Super heavy", then add to beams, reclaimers, etc that they dont work on super heavy. And if this particular super heavy ship can catch the common cold (because it's really necessary for balancing), add it as an exception: "Though super heavy, the data center is not immune to the common cold, because of the extrem cooling already in place for the servers". In this case though, it might be really useful to add a visible effect showing that exceptional behavior; sneezing, perharps?


That may sound like criticism, but I wouldn't spend that much time hoping to improve a game I don't already love! I especially love the emacs-like keybindings of death, and the easy access to the raw data. Keep up the excelent work!

[Lancefighter]

Honestly, it comes down to that x4000 has been doing it this way since *forever*... well, almost forever. He tried to balance everything without special bonuses, but since then decided it was pointless anyway.

I kind agree though - armor types and damage types are nice when you are dealing with an RPG of some sort, but honestly I dont see it working well for a large scale RTS like this.

[Alex_aiwfor]

Honestly, it comes down to that x4000 has been doing it this way since *forever*... well, almost forever. He tried to balance everything without special bonuses, but since then decided it was pointless anyway.

That's very interesting (I'm new here, so obviously I didn't know). I wonder if there was a specific reason appart from many different balance change which couldn't be "explained" by the damage type system. Out of curiosity, what was the system of damage/armor type like?

I kind agree though - armor types and damage types are nice when you are dealing with an RPG of some sort, but honestly I dont see it working well for a large scale RTS like this.

That's weird; I'm not an RTS specialist, but I can't think of an example where the "huge array method" is used. Unless of course the game is simply doing it behind your back. But in this case (if you can't feel it without extensive testing) it's generally only minor ajustements; in Ai war you have x40 multipliers... Which means you absolutely MUST care about them (or at least the sim percentage).

I still think that a damage/armor type system, with a few explicit exceptions would work well while keeping the same level of control for balancing.

At least do it for immunities. How many big ships are "Immune To Tractor Beams, Immune To Minor Electric Ammo, Immune To Attack Power Boosts, Immune To Tractor Beams, Immune To Reclamation, Immune To Nuclear Explosions, Absorbs EMPs, Immune To Being Insta-Killed" ? All the physical immunities should be a consequence of the size ("super heavy"). You could maybe distinguish a level for the technological system ("Super fast main computer") for some others. Then, if you have a "tactical class" (or whatever) ship which is not immune to reclamation, it such an unusual feature that I want it to be very visible on the description (which would read: "tactical ship, BUT: not immune to reclamation").


Slightly off-topic, about balancing: you can add some kind of dynamic (or evolutive) balancing. For example, if you have statistics on which tech is the least often unlocked (with a "send repport" button at the end of a game, for example), you could use these stats to automatically lower the cost of these ships (you can be much more precise if you state, at the creation of the ship, which parameter should go up, and by how much, in case of underuse).

Even better, you could add a research function which specifically gives you the edge against a particular enemy ship. Enemy electric shuttles giving you trouble? Purchase 30% damage reduction against them, for all your ships (and limit the total number of ships you can research against). In warcraft 3 you can resarch upgrades for your units, but that pushes you to stick to a single strategy; here it pushes your adversary to change his (in PvP, or if the AI also uses this; of course the opponent must be warned when you do this so that he has to change his strat). And if you gather statistic about the use of this feature, it gives you excellent input for future game-wide balancing (see previous point).

[x4000]

For more information, I suggest this thread:  http://arcengames.com/forums/index.php/topic,1956.msg12485.html#msg12485

It has a lot of the historical details about why this approach was chosen and why it isn't something I am inclined to change.  Hope that helps!

[Lancefighter]

i recall reading the wiki one day on the history of shields or something, and that was really the only 'history' i properly knew of.
As it is, shields is basically a range maulus for the attacker - 2k shields means the attacker has to get 2k closer than his max range to hit.

But basically, shields was an attempt to make the bomber -> fighter -> cruiser triangle work -

Bombers had high shields, short/midrange, low hp
cruisers had long range, no shields, high hp
fighters had short range but good dps, mid hp

Fighters couldnt shoot through bombers shields (as intended), but would chew through cruisers easily (no shields)
bombers could kill fighters because of their low hp and no shields, but not cruisers because of their higher hp
cruisers could kill bombers, because their long range could go through shields.

I believe, however, that there were inherent problems with this pure-shield based approach, although it might appear to work on paper.
So, x4000 decided instead to just hand out damage bonuses, which ended up being an easier approach overall.

Also, as per the immune to minor electric, instakill, etc, There is a reason they are all set up like that...
minor electric, for example, is the weapon of teleporters and anti-starship arachnids. Teleporters are immune to tractor beams because they teleport, so tractor beams were made immune to minor electric to balance this out.
However, super-heavy wont work all the time, as for instance starships are NOT immune to minor electric, but they are immune to all of the rest. You would just be creating multiple extra classifiers for the player to remember/for the dev to program, and it would be counterintuitive.
For instance, have you ever looked at something that said 'lolomgawesome' in an rpg of some sort, and wondered 'what does this mean?'
Obviously not exactly that, but you should see that the way it is set up now, at least the immunities are transparent. The damage multipliers might not be clearly listed for anything and everything there, but at least we are told those when we go to attack (mouseover with units gives damage range and chance to hit)

As for dynamic balancing, it is done that way (moreso before than now). If you feel something doesnt work, you post on the forum. if people agree (particularly x4000..) then it will be changed.

For unlockable upgrades... uhh.. no.
A unit is a unit. It can be affected by munitions boosting and shield inhibiting (by nearby ships/whatnot), but its always the /same exact ship/. Adding random 'my ship won because I have *xyz bonus*' is just.. RPGish.
The AI has their different types (cloaking type, raider type, bomber type, etc) that gets their 'goodie' ships, but in the end its always /the ship itself/ your dealing with, not knowledge based bonuses.

[Alex_aiwfor]

This post for answering x4000:

Thanks for pointing http://arcengames.com/forums/index.php/topic,1956.msg12485.html#msg12485 ; read it and a bit more.

I'm very sorry that I dug up that thing since it's so old (though I don't think there is any way to avoid it from being dug every so often). So I won't go into the meat of the argument; and by the way I agree that fun rules > realistic simulation. After all modern warfare has pretty bad balance: everyone is going missile (overpowered weapon type), all aircrafts have about the same (optimal) size, etc.

But we all agree that your objective, which is to balance a game with a high number of really different units, is *extremely* hard to attain. And you seem to agree yourself that it is not working perfectly. The best example would be the triangle reversal, but also this quote:

x4000:

Regarding the fact that you can't memorize all of the complex interrelationships, I'm not surprised.  I'm the game designer and set all of those relationships, and I can't remember 99% of them, myself.  I don't think anyone does, and for my own part I don't even try.  The tooltips on the screen are extremely handy for when you see an enemy ship coming in -- quickly check what is strongest against it in the current game, and see if you have any of that handy.  If not, look on down the list of best-against that type, and see what is the best solution.

Again, I won't go into the meat of the argument there, but obviously this isn't the way it should be: it would be like playing chess, but having to ask the chessboard everytime wether that peice can eat that other one in that situation (and having the chessboard's advice still be wrong quite often).


I'm going to post a few things that I think might really help (in my answer to lancefighter); for what it's worth, for a long time my hobby was to make rules for RPGs. You will probably not see them as satisfying solutions given your objective, but hey, I would sound even more like a prick if I critisized that much without suggesting anything.

I also take the occasion, again, to say how much I love that game. I bought "defense grid" and it was ridiculously unbalanced (cannon tower + concussion for masses ftw, according to the forums).

[x4000]

Thanks for the well-reasoned and fair response, Alex.  And, I'm glad you're enjoying the game so well, also.  I don't want to get into a big debate about this (again) with folks, but my points boil down to this at their simplest:

1. Either the game has a lot of complex interrelationships, which people cannot remember because people can't remember but so much.

2. Or the game is simplified into having fewer interrelationships, thus changing the fundamental nature of the game.

There's not much middle ground there, because any attempts to make various ship classes or armor types is generally going to make ships all simpler and more homogeneous, or it will make them out of balance.  People have debated this with me at great length in the other thread and beyond, imagining a paradise middle ground, but I think it's a misconception.

If you simplify it so that you have fighter-type armor and guns, and cruiser-type armor and guns, and then apply that to a huge number of ships, then what you wind up doing is making them all very similar aside from their special abilities.  That would fundamentally change the game, which I'm not interested in doing.  Right now, overall, the game is very well balanced and I'm quite happy with it.  I'm not going to muck with it in a major way anytime soon, so all arguments are pretty moot for the moment, anyway.

Lastly, the reason I changed the core triangle was because it drifted out of balance.  At the time of the original debate, bombers needed to be weak to cruisers for a lot of good reasons.  However, the game balance in general shifted a lot post-2.0, and bombers wound up being completely useless in general for a couple of months of prereleases.  When I looked at remedying that, it was clear that the old relationship was no longer needed and the way forward was paved to go ahead and redefine those and stop people from fricking hassling me about it all the time.  That was one where, since the balance was broken anyway and had to be completely redefined, I went with the common request just to stop hearing about it.

In this case, armor types are much less common of a request, and I'm fundamentally not in agreement with them.  It's nothing personal against anyone who wants them, but I feel like the arguments that people tend to make in favor of them are overlooking the second-order effects of such a change, and how that would either negatively affect the game, or not affect it in a terribly meaningful way.  Either it fundamentally simplifies the game (which I don't want to do), or it doesn't (in which case it is pointless, and actually adds complexity in terms of showing some new stat -- armor type -- that has little significance).  Right?  That makes debates fairly circular on this particular topic.

Anyway, this turned into a longer post than I had meant it to, but I guess I figured it was a good time to go ahead and get a new post on this topic that has all of my points in one concise place.

[Alex_aiwfor]

This post answering Lancefighter:

About the "ship class" idea for the immunities:
The idea is not to have the class be the reason for the list of immunities; but instead to group a number of them, quite arbitrarly. This is why it would not be a problem that "super heavy" implies immune to electric, and then that starships (though super heavy), are not immune to it; they would just have the text "super-heavy, but not immune to electric".

You can see it simply as a way to save space; of course, when balancing the game, it would be better to add as few exceptions as possible (balancing is a multiparameter problem, anyways; you generaly have more than one way to balance a ship); but it would always be possible to add some.
No matter how complex the interactions are, they are never going to be completly random, and so some space (in the description text) can be saved by grouping of abilities.

Grouping, then specificaly stating the exceptions focuses the player's attention on the most meaningful aspects of the ship; it actually makes them look more unique IMHO. (entropy of the information, etc...)


About the armor and weapon classes:
You can apply the same idea to a certain extent. Group multiplier by armor/speed/whatever class, then add exceptional modiffiers, listed. List them as simulation percentage, if you want, though I really hate this metric (did you know it depends on unit cap? Most mark IV ships "loose" to Mark IIIs because of that! see http://arcengames.com/communitywiki/index.php?title=Inter_Mark_Level_Relative_Ship_Strength ). A better metric would be "number of seconds to destroy one target when at range", and the number to be destroyed (not counting overkill). But at least give the multipliers.

x4000 said that multipliers were best hidden, because they didn't make any sense. If they still don't make any sense with a system of classes + exceptions, then my humble opinion is that the system as gone astray on that particular ship. It's like balancing an equation (it is that, by the way, with an equation system). If your process to find the right coefficients iterates for too long, all the coefficient become huge. Ideally, a new ship design would mean chosing two or three classes (armor/damage/computer, or whatever), then adding two, at most three exceptions (which in today's system would be a x10 or plus modifier against or from a certain ship (or class)). If you absolutely can't convey your design idea within those parameters, it is probably best not to add the ship. You still have space for way more complexity than most RTS (that I know).

An example (with the immunities):
I rembember seeing that since [unit A] could teleport and do lightning, traction beams were pwned by them. This prompted the solution to make traction beams immune to electricity, which I guess is unusual for the general "turret" class. What I think is that this kind of "prompted" balancing should be avoided. If one of the two ships has a low number of exception, then fine, add it. But otherwise, the rules of your multiverse cannot support both ships. One of them has to go, or be heavily modified. Traction beams seems like a well established thing in that game; electric teleporting unit A is the newcomer here. Either it learns to do something else when it teleports, or it takes the bus. There are other option for complexity. My personal opinion being that a powerful, immediatly-firing ship, if it can be made, indeed makes the whole concept of movement blocking obsolete (the multiverse can't take both). Maybe the teleporter can do soemthing else than shooting, like uncloacking.


About adaptive (automatic) balancing
I really thinks this is a good idea. Basically, it is a way to automate some of the painful work which balancing is.

Here is how: when you design a new ship, you generaly hesitate on some parameters. Health should be somewhere between 20k and 30k; attack between 600 and 800; speed you're sure should be 15; cost? depends on how useful it will turn out to be, right? So you have an other [x..y] range.
Now, you introduce the ship in the game with the mean values (your best bet). See if players actually use it (statistics on ship usage, assuming some way to get hose), and if they keep unlocking it in the next games or not. You can also factor the number they make, whatever. Anyhow, you had a set expectation for these numbers, and you get a positive or negative percentage from that expectation. If it did 30% less well than you thought it would, you increase all the number in ranges from 30%.

Lancefighter :

As for dynamic balancing, it is done that way (moreso before than now). If you feel something doesnt work, you post on the forum. if people agree (particularly x4000..) then it will be changed.

This method means less work for whoever grinds his teeths on the balancing, and is much more precise/reactive. It would be best suited of course for a PvP game with a lot of players and a master server (say, warcraft 3), but I still think it would really help.


Researchable nerfing
The idea: in each game you get two or three researchable techs (first one almost free) each of which nerf a particular unit. This is different from researchable upgrades, which affects one of your unit (and thus tends to limit the diversity). This gives a general way to counter a "strategy" (making a lot of a unit). In other RTS, it might be a problem (low number of units), but here the enemy can simply switch to another one (assume he gets an warning, and also that you can't wait for the end of the game to research it; it could even be a starting option, costing a small percent of all future resources for example). It could work in PvP (just saying), and the AI could use it. Of course you can always counter a type of ship with an other, at less of a cost... except if there is a balancing problem!
Basicaly that's like a vote for a limited balance change, which you pay with ingame resources.
And guess what... it would be a really good statistic to collect to feed the adaptative balancing of the previous point.

Lancefighter :

For unlockable upgrades... uhh.. no.
A unit is a unit. It can be affected by munitions boosting and shield inhibiting (by nearby ships/whatnot), but its always the /same exact ship/. Adding random 'my ship won because I have *xyz bonus*' is just.. RPGish.
The AI has their different types (cloaking type, raider type, bomber type, etc) that gets their 'goodie' ships, but in the end its always /the ship itself/ your dealing with, not knowledge based bonuses.

It's not a bonus to your ship, but a malus to other's (of one type). You'll still be fighting the ship itself, except one or two of them are you know how to deal with. If you are looking for an "intuition" justification, consider how much sense it makes to study ennemy's design so you know where the fuel tanks are to shoot it.
And also, the rules could be so that it would be only worth it if there are no "normal" counter strategy with other ships.


I'm done, thanks for reading that far. By the way, it's fun to think about these kind of problems, but I know it is a lot of work to make anything happen in that kind of project. In the end, ideas are a dime a dozen, and finished projects are a rare commodity. Thanks to all the devs for their perseverence, we appreciate it! PS I am so sorry for all the typos, I hope you all have good deductive skills to decipher it.

[RCIX]

Lastly, the reason I changed the core triangle was because it drifted out of balance.  At the time of the original debate, bombers needed to be weak to cruisers for a lot of good reasons.  However, the game balance in general shifted a lot post-2.0, and bombers wound up being completely useless in general for a couple of months of prereleases.  When I looked at remedying that, it was clear that the old relationship was no longer needed and the way forward was paved to go ahead and redefine those and stop people from fricking hassling me about it all the time.  That was one where, since the balance was broken anyway and had to be completely redefined, I went with the common request just to stop hearing about it.

Well. Someone got upset about that whole thing. I''m sure we're all glad it's fixed and don't have to deal with 5 page arguments every month!

[Kjara]

But 5 page arguments every month are fuuuun!  (this way we just get to argue about different things

[jordot42]

     Well, and this from a simple player who is very engrossed in this game:  if I want to know what unit to sent after another unit, I look at the strong/weak data on the enemy, click my ship(s) that best can defeat that enemy, and send them on their way.  I then watch the results.
     The way that X has things set up works very well in my opinion.  This game in no way claims to simulate any kind of realism, so to demand that is should do so is a stretch.  I still chuckle when I see 30-40 missiles from my frigates impact a bomber and it still keeps coming.  When I sent my fighters after that bomber, on the other hand, I do a different kind of laughter.  So the system works imo because it's tons of fun, very engrossing, and I even find myself thinking about the game when I'm not on my computer.

[x4000]

Okay, a very last few points about my motivations and general beliefs.  Alex, you make some great points and do some great analysis -- and it is indeed an interesting hypothetical discussion -- but in reading your explanation, a few things became clear to me about why I tend to get into disagreements with a certain type of highly intelligent (not un-coincidentally, often math-oriented) person.  So, my addendums:

1. I'm not a math-oriented person.  If I had to pick, I'd say my orientation is split between engineering and storytelling.  This is important.  When the math-oriented people come in and start talking about automated balancing and trying to keep the equations in line, I see where they are coming from; I come from a mix math/science/engineering-oriented family, and I can sympathize with those motivations, though I don't share them.  This is to say, I design my games from an entirely different perspective, and some ways that bothers the math people a bit, even if they otherwise enjoy the game.  The system can't be expressed in simple enough mathematical terms -- it's not "elegant" in a mathematical sense -- and this is a problem they unanimously think should be fixed.  I disagree.

2. I keep the game balance in a semblance of disorder and least slight-imbalance on purpose, and always have.  If there are major exploits, of course I fix those, but those are comparably rare.  But when I introduce a new ship, I intentionally seed it with values that will put it just a bit off kilter with other similar ships, and against the expected foes.  Usually this is via special abilities, but it's also in terms of costs, attack powers and attack bonuses, shielding, and health.  I like for all of the ships, even amongst fighter-type ships for example, to be slightly unique in subtle ways in addition to the major ways.  In a second, I'll get to why.

3. In some ways, I guess I am a math-oriented person, despite my claims otherwise.  It comes from the upbringing.  When I play boardgames such as Descent: Journeys in the Dark, I'm always the rulekeeper/game master, and I'm the one that keeps up with everything and even helps my "foes" reason through all their possible attacks and scenarios, and I'm extremely, extremely quick at coming up with the optimal solution.  This is also important.  I'm also very, very good at doing this against AIs in RTS games, and with the mechanics in general with RTS games.  When I start playing a new strategy game, like many of you I'm sure, I'm learning as much as I can about it and crunching the numbers mentally and collating and comparing in-game data against my hypotheses in order to find my personal optimal strategy.  And when I eventually do?  I've lost all interest in the game.

4. What all of those other games have in common (the good ones) is that they are designed by math-oriented folks.  They use all sorts of techniques to make sure that the balance is as pristine as possible, they try to keep things simple while also being complex, and generally they are huge proponents of the sort of advice that Alex is giving.  It's good advice, generally speaking.  Certainly a must for a competitive pvp game.  But all those games bore me to tears after 6-12 months of biweekly play.  I'm tired of being bored, and I certainly don't want to be bored by my own game, so this goes back to why I designed it the way that I did, in a lot of senses.  Escaping this eventual fate with AI War is my core motivation, and why the idea of making things simpler or more orderly strikes me as hugely undesirable.

5. Therefore, I keep things muddy with the ships.  There are a lot of them, they are complex, they have attributes that interact in varying funky ways, and in general there is too much for even me to remember despite the fact that I spent so much time with the game.  Rather than seeing that "obviously this isn't the way it should be," this is exactly the way I very intentionally made it.  I can't understand it all in one go.  I haven't been able to find a best strategy in the 13 months I've been working with the game now.  I win!  I have kept myself entertained for 13 months, and I suspect I will keep myself interested for several more years if not far longer.  If I'd made this simpler, or more orderly, I'd already be done with it and bored.

6. I've pretty much already made my point, but going along with this, a few other related observations.  You'll notice that everything about this game centers around variety.  It forces you to choose either ships or starting positions you may not prefer in the lobby; the AI types have some pretty powerful and game-altering abilities in many cases; no two games have exactly the same ship mix; the expansion is geared around adding as much variety to an already-huge game as possible; the maps are randomly-generated only, and with the added styles more recently they have all been focused on creating new game modes, rather than just looking cosmetically different; and finally, wherever possible I go out of my way to make the complexity multiplicative, in terms of letting players have as many minor factions as they want in games, and more capturables, and so forth and so on. 

All of those overlapping, crisscrossing, muddy systems create an environment that feels increasingly real to me from a strategic standpoint, and increasingly less meta-gamey.  I can't remember all the attributes and stats and strengths and weaknesses of all the ships, and so every time I start a new campaign I am presented with practically a fresh game.  Maybe autocannons will be great; maybe they will be total crap.  It all depends on the relative ship mixes, and I'll have to decide in that game.  Some ships might not have any use in one game, or might be marginalized, and in others they might hold one key role or a different key role.  If you always play your favorite ships, you start falling into a rut, which is why the game so ardently encourages experimentation at every turn. 

It's to save the math-oriented people from themselves.  If they are anything like me, they want very badly to understand everything around them, and in general they do a great job of it -- but especially when it comes to games, that is the game.  They come, the experiment, they understand, they leave.  AI War, for purely selfish reasons, is built around an entirely different premise that I can probably claim no other RTS has ever tried before.  Evidence is that it's working.

I've never really been able to articulate this before, though it's been a core design principle from the start.  So Alex, I particularly thank you for your post there, because it helped me express something that I think needed to be expressed, and that I've never done a good job of expressing before.  More than anything else, if AI War had a mission statement, this post of mine is it.

[Lancefighter]

Alex, Id love to keep arguing with you further, but at this point, our cases have been made; Neither is going to properly change the other's minds. There is no point to argue really (keep in mind that even if we do reach a conclusion, unless x4000 wants it, too bad :p )

Oh, and speaking of that.. Its nice have a mission statement /after/ 3 major releases xD

[Alex_aiwfor]

x4000:
I find myself completely convinced by what you just said!

When I start playing a new strategy game, like many of you I'm sure, I'm learning as much as I can about it and crunching the numbers mentally and collating and comparing in-game data against my hypotheses in order to find my personal optimal strategy.  And when I eventually do?  I've lost all interest in the game.

I can absolutely relate to that. of course, this is only true for games against an AI; basically when you find the optimal strategy, you "beat" the AI. In PvP games (as in chess), the enemies will have counter strategies to find, then counter-counter, etc. But this also leads to more importance for micromanaging (warcraft 3), or thinking way ahead (chess): in AI war, what you want (I guess) is to extend that period where you just discover the basic game's rules (like when picking a new flash-based tower def, for example: it's over when you understood which two towers are best). After this, you still have to discover that kind of thing, but it becomes increasingly less important compared to meta-game/micromanaging.

I aslo perfectly understand the notion of leaving some unbalance on purpose: after all, it is fun to find the two or three overpowered tower in a simple towerdef. What you're basically saying is that Ai war gives you a new "simple" (not perfectly balanced) tower def for every ship mix.

One of my reasons for suggesting automatic balancing (apart from my natural tendency to add some PvP to it) was my fear that it would be a colossal task for you to balance the game. But from what you say, you are satisfied with you ability to do so (and you don't aim for "perfect" balancing). To me, seeing that you yourself had to rely on the simulation percentage, and the triangle reversal, was proof that it was not satisfactory (for you); I see I'm just wrong here.

If I might just keep one request, it would be to add the multipliers in the descritption (of course, that should be an option in the menu, desactivated by default). Just a few lines of code, and it means I can double guess the simulation (if I put my mind to it, I can actually predict what will happen if the battle starts at point-blank through a wormhole). Two arguments for this:
 - It gives the player the full information on the ship's mechanics. I think we all agree that's a great thing (in some games you have to experiment just to produce the spreadsheet; by hand).
 - Since the game is calibrated for you as a player (which also means everyone with hundreds of game hours), there is obviously no need to "hide the magic", or the complexity, from the players.
 - People would still have the simulation, when they dont want/have time to think that much (and its the only thing there by default). But most people dont know how excatly how the sim works (for example, the factoring of the ship cap was a real surprise to me).


In the end, what is certain is that your game is fundamently different from all the others (that I've seen), which is probably one of the best compliment you can make to any game, indy or not. Probably the only problem left is that people will start playing it without realising what kind of game it is; I think displaying the modifiers will really help here. I know I will keep playing it and buy any expansion coming, when I was almost ready to give up on it (when I thought it was trying to be a simulation, and failing at it). And of course, thank you tremendously for answering that nicely to what was some rather frontal criticism; you're awesome, and I really mean that!

[x4000]

Thanks, Alex -- glad that makes sense.  I need to add a Game Design Philosophy section to the wiki, I think, where I can maintain certain things such as this in case people want to know why I decided to do this or that.

The ability to see the multipliers is a request I've gotten often enough that I think it makes good sense to do.  I've added that to the interface tweaks request section of the development discussion area to make that part of the considerations for future enhancements.  Thanks!