I think you guys are complicating things way more than necessary.

Here's what I need from a given weapon:

Cheapness - in terms of the most bang for buck, regardless of space used.

Basically, if a weapon is cheap, then I can mass-produce a ship design with just one or two. Though these ships may not be logistically efficient, they provide a cheap initial strike.

In other words, losing ships can be part of a mixed units strategy.

If cheapness doesn't do it for a design, then I consider Size, regardless of cost. Weapons which are size-efficient can be useful in a ship design which considers absorbing damage to be important - the less space weapons take up, the more space is left over for other systems, including defences.

If that still doesn't cut it, then I look at raw Firepower. That is, how much damage-dealing I can cram into the ship regardless of cost. This is useful when I have insane miniaturisation and/or manufacturing capabilities, and the economy to support it.

So if you want to mix it up a bit and have different tiers have a different weapon type follow one of those philosophies, then great. I'm with you there. Just don't get too hung up on which weapon is better than another, aside of course from obvious deficencies that happen when you cram the weapon scale into such a small range. If you make it, I can find a way to use it!

Maybe.  But I'm having fun doing it and learning a lot.   Something.. well.. "interesting" might come out of it.   Cost/power (ie cheapness) is THE overriding statistic if you're talking about large fleet engagements,  assuming equal logistics (which is also quite important).  After all OShee's lamentations about beams in the CU I finally built a simulator,  and lo and behold they sucked,  across all tiers and hull sizes.

So we have an opportunity to fix that before the CU goes live,  and I'm picking up where OShee left off.  Lev's double-cycle is so far the most appealing solution I've heard,  though I'm not sure I can make it work yet.

Yeah,  not even thinking of going down that road,  for all the reasons you mentioned -- I have no intention of modifying the range of power,  power/size, and cost/size,  just the path that the various weapons take to get from T0 -> T6.

Finally my moaning about weapons payed off   But seriously now, any system that you guys come up with will be an improvement over current CU weapon progression, where unfortunately, beams are just... well... broken. If there's more people working on it, there is bigger chance of getting it right. 

My concept was to stay as true with vanilla as possible, meaning that:

- missiles shining on large/huge hulls, otherwise slightly lagging behind beams/drivers, with medium hulls being just about even.

- beams being stronger then MD with extra price to compensate.

- mass drivers being weaker but cheapest, as "quantity has a quality all of its own".

Not as interesting as what was proposed later by leiavoia, but gets the job done. Numbers that I have provided must have been way of in favor of mass drivers anyway, from what I can tell looking at the simulations done by Silas (great job btw, I would love to see how you do them ). Small adjustments in price alone would suffice I guess, but there could be some inconsistencies here and there. Ideally after beam vs mass driver or missile (medium/large hulls) vs mass driver war there should be 15-20% of original mass driver ships left - like Maiden posted, stronger but less numerous ships get other benefits, from small things like experience to larger benefit gains from military resources/ability picks/racial bonuses, ships hp regenerating after battles etc. With numerous ships it is harder to concentrate them in one spot to be able to take out beam/missile ships, so that's why around 15-20% seems like a good choice. Note that % means remaining MD ships of original number participating in direct exchange, if we throw all the ships against each other at the same time. 

Sadly I hadn't had much time to work on my PC this week. Considering defenses, well, I made some small tweaks and changes here and there, but seeing that weapon values were not great in the first place, I don't think it is worth to post them, it will be same as changing from vanilla or CU, and it is far from completion anyway. Weapons have to be done first, to start any changes to defenses imo.

That said, just like Mabus, I don't have much free time lately and it will not change until the end of next week. Since new .exe release I played only 3 games myself Not enough time to finish weapon/defense system or even test AI research after Mabus lowered AI values, but I trust research is good/great/godly as not one post in few pages was about any oddities in this regard.

If you still want to continue with my concept, you have my full blessing. Change whatever you like, rip the values apart and glue them back together/better Few thoughts I had considering defenses (if it helps):

Armor has to be hardest to obtain, otherwise MD with its low attack values will suck so badly, it won't be worth picking. Big sizes or huge research cost will both work. Bc cost is irrelevant, but should stay in low-mid values.

Point Defense can be done in two ways, however you think missiles should end up. You can either set their size mod low to scale better on large/huge to counterbalance high damage of missiles on those hulls, or go opposite way so they are easier to counter on smaller hulls, but harder on bigger hulls, which will play better with missiles advantage. Problem with second option is that AI don't really like hull techs, so any AI that picks missiles without researching hulls will be at disadvantage for long time.

Shields should cost more then PD and armor per point of defense, should also be best at protection with highest possible defense achieved on ships. If beams are still too weak in their current form (after increasing cost of MDs by 10%, which I think was a good choice by Maiden), amount of protection from shields could be even or lower then PD/armor. Whatever works best. 

Most controversial changes I made to weapons where for sure Evil weapons being different weapons then CU +1 tier clone with increased price, and Photonic/Harpoon price drop within their tier.

- With Evil weapons, I done those changes because I don't know how AI designs its ships. Will AI pick Photon Torpedoes I, or will it still produce inferior ships with Psionic Missile? Both weapon modules have same damage, size, size mod, and only the cost is different. If AI drops Psionics after tier 4 level 1 is unlocked, then just like in CU, Evil weapons can be just cloned over with increased price. For sure it will be easier to balance them out. I think Maiden is the person who knows the answer to this question You sir seem to have a lot of knowledge when it comes to game mechanics and the AI.

- Photonic Torpedo I and Harpoon I had to have their sizes lowered (from 14 to 12), to evade the issue of ships with those modules being weaker on small hulls then their preceding tech. Level 3 had to stay same (size 10), to not make them too strong for their tier. Because of the lower gain per new tech, my first though was to decrease the cost of modules in the tiers, so next levels don't feel like a filler. But if this causes too many problems when upgrading those ships, then maybe level 2 should be cut out altogether. Harpoon and Photonic with only 2 levels, but at least the jump in saved space/power is more noticeable and cost per module can stay static within the tier. 

Now the rest I have to say the idea of eras for each weapon is also great. It will encourage weapon trading, help with any stalemates (which I haven't noticed in my games tbh, but again I rarely look at AI vs AI war outcomes), and for sure will create some great opportunities for the players who take their time instead of picking weapons randomly. Majority on the board agrees that this is an interesting concept, and I'm on the same ship with you :> I would love to see how this plays out.

This is better system then each weapon having only one peak. With only one peak, there are better and worse selections depending on galaxy size/tech speed. Just like Maiden plays on smaller galaxies and only goes as high as 3rd tier, I play almost exclusively on large and go to war for the first time when I reach tier 3/4 (I turtle a lot lol). With two peaks, the depth of the game will increase, and it will be harder to go and say starting from turn 1 "I'm going drivers". Double cycle is the way to go.

It will be very hard to implement and get right, but I do believe it is the weapon system that should go into official CU. It is more interesting then mine and rewards strategic play more, which should be key to a strategy game. When I go back to my usual work schedule, I will finish my system and post it as a standalone mod to the CU. Just for those who like it more oldschool. 

I might pop in now and again and write a short post, but for the most part, I won't be around until next weekend ^^ So I wish you good luck, and thank you all for taking your time to rework the weapons/defenses. And don't forget, balance brought by several imbalances is still balance! :>

So if i understand you correctly, not only were you able to get the weapon eras to work, you also changed the "acceleration" for each weapon type? That is interesting.

If you can fold that into a mod i can download, i might be able to squeeze in a game this weekend.

Thanks for the compliment , but actually this is new territory for me as well. So I've taken some time to do at least enough test to be able to give a helpful answer.

1. First off, the game sorts its available weapons in its first priority according to their raw firepower.

2. If this is equal, it will compare sizes. How size-mod comes into play here I didn't check but I assume that the game will calculate the real size (size+size mod) beforeahead. (practically shipdesigns are always bound by their respective hulls)

3. If both stats are equal, then something strange occurs: it will sort the list in alphabetic order, using the <Weapon Name="XYZ"> tag from GC2Types.xml. The shipdesigner will pick the last name, that is, the weapon at the bottom of this list . It will NOT consider alternative costs at all .

The good news is: Now that we know this, we only have to change the internal names of these weapons accordingly to our wish, and it will be fine, ie. we can trick the AI to always use the cheaper weapon  

***

However, there is never a 100% guarantee that a specific weapon is always used. During my tests I've encountered alot of oddities, most are based on the ship-template.

For example, the shipdesigner can choose to use a weapon of another branch as well, even if this weapon is greatly inferior. This happened alot in the Escort-class which is usually designed to be more cheap albeit by loss of firepower.

Defender-ships seems to primarily rely on T1 weapons, even if much better weapons are available. I've encountered Defenders with 7 Kinetic Beams although Phasors were available. The costs of these ships were extraordinarily high. Sometimes Defenders showed multiple branched T1 weapons. In a way, this even makes sense because Defenders will stay in planetary orbit and thus, gain an attack bonus, a minimum of +1 for each type of weapon. A 1-1-1 Attack Defender will become 2-2-2 and therefore sport more firepower than a 3-0-0 Attack Defender.

Some non-military designs as Constructors, Freighters etc sometimes received Tier1 weapons. It could be that this was due to the fact that KinStrIII is very small in size. BTW we need to add this to the list of hardcode changes, I've seen yet another weaponed constructor that was fleeted up and sent against enemy planets. These ships actually don't provide any strength (with 1 HP) but take away logistics, besides, armed Freighters/Miners cost alot of maintenance whereas unarmed are maint-free.

Troop Transports also can sport a weapon but these are the best available. I have no real opinion here if this is good or bad. They might wipe out defense-less ships from orbit or low-armed non-military craft but usually will commit suicide against a Defender-ship. In standard battle the AI doesn't shy away from hopeless battles, although he will usually retreat unarmed Troop Transports. IDK if this is likewise the case with armed Troop Transports.

I've also encountered alot of situations when an AI picked a weaker weapon to research, practically I've given the AI all Beams up to Phasors III and then started autoplay in order for him to create designs. With Disruptors available, he should never go for any Weapon Theory but the Yor even went so far to research right into Harpoon (!). I'm gonna try to address this problem with some tests.

As a sidenote, if the time to research a tech is too great (80 turns for sure, probably even lower) the AI will drop it after 1 turn. On lower techspeed + low planet count practically all huge techs will come solely from creativity sparks   

Minor update. V5.7.2.

Update StarbaseModules changelog. It was out of date.
Moved Iconian production modules forward by one tech.

None of these changes effect weapons. 

Sounds like that might be a good idea -- and I'll happily take suggestions.   Perhaps small values (6?) for tiers that aren't much of an improvement,  and larger values (12?) for the first weapon of a strong tier?

But then,  given that Mass Drivers are strong at T1,  if we changed the AI values wouldn't all AI dive into Mass Drivers?

e.g.

Mass Drivers are 10 (T0),  then 12-8-8 (T1), then 6-6-6 (T2)

Beams are 10 (T0),  then 6-6-6 (T1),  then 12-8-8 (T2)

So if an AI sticks with one weapon type,  they'll be fine,  but if they research both T0 Beams and Drivers?  What then?

Ok.  I think maybe it is best to give AI a small hint,  even if AIP10 and AIP11 do branch out,  perhaps they'll grab a good tier over a middling one.   Using my above example,  if they do grab T0 Driver and T0 Beam,  hopefully the "hint" of 12v6 means that they'll lean towards Drivers if they're otherwise undecided.

I think,  what I will do is this to start:

all T0 weapons keep their AI value of 10.

There must be some reason we reduced weapons AI values by 2 per tier.  So in Tier 1-3,  a "strong" tier will get 12-8-8 (most important to advance to the next tier,  somewhat important to finish it to get the size reductions).   For Tier 4-6,  a "strong" tier will get 8-5-5.

A "moderate" tier (where there is some improvement,  but it is not very much) will get 6-6-6 from Tier 1-3,  and 4-4-4 from Tier 4-6.

Thinking more,  all Tier1 weapons will be 12-8-8.   T0 is crap and only worth using if you are desperate.  Early research into first T1 weapon before tech inflation kicks in is almost always worth it.

Any objections/suggestions?  

p.s.  I am still working on defenses and tech costs.  To be honest I'm not sure how I even want to proceed yet.  Starbase attack/defense/boost modules are also on the list since I don't know that missiles need the extra attack/defense under the double-cycle system.

I implementing thse AIValues to signal to the AI that the further it progressed into it's weapon tree the more it had to research outside its weapon tree. It doesn't really work though. AIValues aren't used in decision making for weapons or defenses, just categories and branch.

AIP10 behaviour is something we cannot fix, it's hard coded. Setting AIVAlue to 10 for the Theory tech and T0 and T1 won't change it's behaviour.

Nor will adding 10% weapon bonus work. A player will will prioritize weapon research so all you're really doing is granting the player a huge weapon bonus for no other reason that to boost the AI. That was done before with starbases modules, tech cost, etc.

It doesn't really help.

If I may offer a suggestion. Perhaps it is best to fix what is broken and sticking within the current methodology of the weapon/defense system. Devising a new system takes time. Implementing it takes time. Testing it takes even more time.

I've seen multiple posts saying that beams are a problem. Perhaps it's wise to focus on that. Fix them up a bit and score, say, a 70% improvement over the current situation, release that on time before moving on to on Weapon/Defense V2.0.

There's nothing wrong in releasing something in stages but it is hard because you want to do it all.

I've written a modders resource while developing APT 1.0. There's a lot of info here and it was hard to obtain.

https://sites.google.com/site/mabusraeen/extra/modders-resource

In vanilla getting all these "Advanced XYZ Weapons"-techs took the same amount of research as to research right into Plasma II or Photonic Torpedo. I assume this is not different from now. The gain was additional +10% weapons but you have to use Tier1 weapons instead of Tier3.

The only advantage here is that you can choose from Scatter over Seeker over Kinetics, ie. cheap/lowattack to expensive highattack - but in any possible scenario the Plasma II ships will blow you out of the water everytime, and that will still stay even if we increase the weapons-bonus to a total of 30% for all three.

Besides, going this path will trigger tech-inflation at least once more than going after Tier3 weapons because you have to research 5 additional techs, so you'll always face a harder time getting better weapons....

In other words, doing so as player will weaken your game considerably! Someone who knows this will never go this route. But those who don't know it (also: AI) will be compensated by an increased weapon bonus.

Hm.  I probably should have made Seekers a bit bigger and cost a bit more.   But even so,  Scatter Blasters kill them -- are you looking at Fleet vs Fleet,  or BC vs BC comparison?

Lets look at a small hull without any mini:  a Seeker III gets 5 Missiles,  costs 24+5*30 = 174,  and has 5 attack.   The Blaster III ship has 2 Drivers,  costs 24+2*20 = 64,  and has 2 attack.    So you build (almost) three of them for the price of a single Seeker III.

So you have 5 attack and 8 hp (Seeker) vs 6 attack and 24 hp.   This is completely lopsided,  Blaster will win every time,  although they might lose a ship.   Tier 1 Mass Drivers are good.   The Iconian/Yor Tier 1 racials are very good,  for Missiles and Drivers.   Their beams (Kinetic) are bad.   This is,  IIRC,  by design.

I'll bump Seeker cost up to 35 (still cheaper than Stinger),  and space up 1 (so they're 10/8/6).   They weren't supposed to be more space-efficient than Stingers,  just cheaper.   Harpoon I was actually a bit of a downgrade space-wise from Seekers,  I was trying to avoid that.

To answer your other questions:

* Weapon bonuses are straight %,  but IIRC they are applied as a "bonus" attack all at once,  so higher space-efficient weapons have a slight edge.   I do not simulate this.

* Leftover space is ignored.  My simulations use the middle version of all weaponry.

* Hull HP is given an extra 10% per tier after tier 1

* Miniaturization is given at 10% per tier after tier 1

* Luck is ignored.   It is again % based but not biased against any particular space efficiency.   There are some crazy breakpoints with high luck,  but there always have been.   I do not simulate this.

* Repair benefits surviving fleets.  Fleets with low space efficiency will die outright more often.   Once again,  higher space-efficient weapons have an edge.

* Experience gain is rarely a factor in large engagements -- most fleets die.   But once again,  fleets with higher space efficiency will sometimes live long enough to get big experience bonuses / HP.

* I do not use average rolls.   Every single attack is rolled separately.

Drivers are close to beams,  space-efficiency wise,  at T1, T2, and T4,  and lag behind in T3, T5, and T6.   Missiles lead,  even on tiers where they aren't "strong",  but they aren't really that much better than beams (~10%,  most of the time).

After some deliberation,  I have decided how research and defenses will work:

1.  Research will be cheapest for Drivers,  because they do have some disadvantages:  low space efficiency,  and being outright crippling if the game goes on long enough.    Research will be nominal for Beams because they have middling space efficiency and are not at too bad of a disadvantage in long games.   Research for Missiles will be most expensive because they have the best space efficiency and are king of the long game.   You don't have to wait "forever" for them to be any good,  either -- they are the best T3 weapons cost-wise AND space-wise.

2.  Defenses will be identical across the board in terms of both component cost and space.   The weapons being highly variable/cyclical is "interesting" enough,  I think.   Research-wise,  I think I will reverse the costs of weaponry : chaff will be easy to research,  deflectors harder,  and armor the hardest.   I think this will be necessary to complete the balancing -- if,  for example,  Armor was easy/quick to research, it would negate a lot of the initial advantage of Drivers.

This means going in and redoing all the starbase boosts/attack/defense,  too.   They can once again be equal.

And that,  I think,  will be Cyclical Weapons v1.0 Alpha.  

Weapon Cycling v1.0 Alpha is complete.  Get it here:

ftp://guest:guest@lipford.no-ip.org:6666

Mabus,  I apologize in advance for using 5.7.1 as a base.   Hopefully it won't be very difficult to re-merge the changes from 5.7.2 if weapon cycling turns out to be more interesting and less unbalanced (fingers crossed!)

Overview of changes:

* Drivers are most cost efficient in T1 and T4.   They have low space efficiency and are cheap to research.  Armor is expensive to research.

* Beams are most cost efficient in T2 and T5.   They have average space efficiency and are average to research.  Deflectors are also average to research.

* Missiles are most cost efficient in T3 and endgame.  They have good space efficiency and are expensive to research.  Chaff is cheap to research and also slightly more space efficient than Armor or Deflectors (but not nearly enough that you'd want to use it instead!)

* all defense types are equally cost effective.

* all starbase modules now give equal amounts of attack / boost / defense

* all defenses have the same number of tiers and techs within each tier

* some attempt has been made to correct scenarios and campaigns but I'm not 100% sure I got everything.

This version shouldn't break anything but has not been playtested.   Please let me know ASAP if you find anything broken.  

Thanks guys!