so where is this modified exe so I can play the CU? and yeah I'm very late to the party.

It's good that we get additional time, in the meanwhile we might be able to sort out some stuff were we didn't take a look before.

***

I've taken the liberty and integrated OShee's changes into GC2Types.xml:

http://www.datafilehost.com/d/87f2e5f8

Later I aim to see through the changes in game, if someone finds something off please let me know.

I noticed that there were two weapons (Harppon + Photonic Torpedo) who had their costs reduced inside their T - this *could* lead to ships that are cheaper than their predecessor, and this shouldn't happen because downgrading stuff costs additional funds. But if that happens under a normal minia+hull setting will be seen later....

Ok, I see now why Harpoon + Photonic Torpedo have their costs reduced: The size of these weapons become only smaller by -1, whereas other weapons get -2 from tech to tech. So, in order to justify Harpoon2/3 & Photonic Torp.2/3 these weapons gain less cost as an extra.

However, these reduced costs should also translate into less costs ingame, but that might not always be the case.

Under a normal progressive setting for Harpoon; (+10% minia + medium hulls)

4*Harpoon1 = Att+12 for 344 bcs w/ 4 spaces left.

4*Harpoon2 = Att+12 for 328 bcs w/ 8 spaces left.

However, usually the AI will try to use any existing space, with the priority at defenses followed by hitpoints mods and range. In the first case the only things that can fit is either hitpoints or range. Hitpoints have to be researched but range is always available so I take these:

We get 346 bcs for the Harpoon1 ship and 332 bcs for Harpoon2.

However, the Harpoon2 ship actually has enough space left so that a defense mod could take place. Using Duranthium III = 343 bcs, I still am not able to get a more expensive ship, although it is superior. Stockraces also don't do reliable research into defenses, so this is not even a favourable scenario.

The difference in production is only -3 MP, but the upgrade costs of Harpoon1-->Harpoon2 = 324 bcs (!!!)

If I take a Sensor-mod instead of an armour it is still 288 bcs, although the difference in MP is greater. As you can see the game becomes kinda erratic when downgrading stuff (with planetary improvements it's even worse....)

However, there is a solution.

The costs of the weapons inside a Tier should normally rise a little bit, for example, Laser1 -6 MP, Laser2 -3 MP, Laser3 stays. We could then leave Harpoon & Photonic Torp. at equal costs an thus, totally eliminate all downgrade issues.

This way is also more interesting to those player that spend alot of time in the shipdesigner because now, previous weapon techs aren't totally obsolete as they currently are, eg:

@10 Minia I can put 4*Laser3 on a small hull for Att+4, costs 160 MP, 2 spaces left.

 If Laser2 would cost 3 MP less I could use the 2 leftover spaces to make my ship a bit cheaper:

@10 Minia I put 3*Laser3 + 1*Laser2 on small hull for Att+4, costs 157MP, 0 space left.

Let me see if I can integrate these changes without causing other stuff that is off....

Tier by tier weapon comparison,  using the values provided by Maiden's GC2Types.xml,  using Mass Drivers as nominal cost/power and nominal power/size.   Battles were simulated using $100k bc worth of ships in fleets,  throwing fresh fleets against fresh fleets,  then injured,  until one side was dead.  The simulation is run several times for each weapon combination, hull size, and tier.

Note that I consider the simulation "close" if the winner has <10% of their original ships remaining,  a solid win from 10-25%,  a decisive win from 25-40%,  and anything over that is crushing.

Tier 1: 

Beams are 50% more expensive and have 25% more power/size.

Missiles are 32% more expensive and have 53% more power/size.

Small hull, no mini:  Beams beat missiles.  Drivers win decisively against both.

Medium hull, no mini:   Missiles and drivers about equal.  Each win solidly against beams.

Tier 2:

Beams are 50% more expensive and have 25% more power/size.

Missiles are 37% more expensive and have 36% more power/size.

Small & Medium hulls, no mini:  Missiles beat beams.  Drivers win decisively against both.

Large hulls:  Missiles beat beams.  Drivers beat beams decisively and missiles solidly.

Tier 3:

Beams are 24% more expensive and have 28% more power/size.

Missiles are 21% more expsneive and have 9% more power/size (small hulls) up to 22% more power/size (big hulls)

Drivers have solid wins on all hull sizes.  Beams and missiles close on small and medium hulls,  Missiles beat beams solidly on large and huge.

Tier 4:

Beams are 39% more expensive and have 42% more power/size

Missiles are 27% more expensive and have 11% more power/size (small hullls) up to 23% more power/size (big hulls)

Drivers have solid wins on all hull sizes.  Beams and missiles close on small hulls,  Missiles beat beams solidly on medium or bigger.

Tier 5:

Beams are 33% more expensive and have 28% more power/size

Missiles are 22% more expensive and have 0% more power/size (small hulls) up to 13% more power/size (big hulls)

Drivers have decisive and occasionally crushing wins on all hull sizes.  Missiles beat beams on all hull sizes.

God tier:

Beams are 21% more expensive and have 38% more power/size

Missiles are 3% more expensive and have 25% more power/size (small hulls) up to 42% more power/size (big hulls)

Drivers win decisively on all hull sizes.   Missiles beat beams on all hull sizes,  but its close.

Conclusions:  

* Mass Drivers are too good (or cheap, take your pick).  They allow you to get a lot of metal in the air,  and even though pound for pound they don't do as much damage as beams or missiles,  there's enough HP left to carry the day,  quite decisively most of the time.

* Beams and missiles are close.  Missiles have the edge,  but they're more expensive to research and have the lowest utility due to their size.

* due to how sizeMod works (constant + to weapon size regardless of initial size),  beams get heavily penalized on larger ships.   This really hurts them.

* I haven't taken defenses into account.   That's an order of magnitude harder to code. 

* I owe OShee an apology for accusing him of underestimating the value of per-fleet attack.   I overestimated it at least as badly.

Solution:

Decreased size of all Lasers by -1, also Kinetic Streams I & II

Kin Str III cost to 37 (-5) because the change in size from 4-->3 would be too much.

Decreased size of all Particle Beams by -1. I've tried to decrease their costs but because Laser/KinStr are so expensive that would have automatically resulted in "negative cost shipupgrades"

Decreased size of Plasma & Ion Beams by -1 but size mod increased by +1. Decreased cost of Plasma by -3 and Ion Beams by -10%. (-6, -6, -8)

Decreased size of Phasors by -1, but increased their size mods by +1. Also slightly adjusted their costs inversely to their size because going from size 6-->5 is stronger than going from 8-->7. Costs of Phasors I -6, Phasors II -3.

Increased size mod of all Drivers by +1.

Decreased size of all Disruptors by -1, increased cost of Disr II by +2 & Disr III by +4.

Increased size mod of BHG by +1.

Decreased cost of DoomRay by -15 but increased its size mod by +1.

GC2Types.xml

T1 MD also get now the 10% increase in cost, except Scatter Blasters (got a size mod +1 instead)

I like this approach.   What does everyone think about the following:

Drivers:  cheap research,  but defense is also cheap and good.  best early tiers but peters out
Beams:  moderate research, expensive to build,  most difficult/expensive to defend against,  best midgame
Missile:  expensive to research,  avg build cost,  avg cost/ability to defend,  poor at start but best later

So,  each weapon type would get a single tier where they are the best choice,  bc wise:

Tier 1:  drivers dominate

Tier 2:  drivers solid,  beams close second

Tier 3:  beams dominate

Tier 4:  beams solid,  missiles close second

Tier 5:  missiles dominate

Tier 6:  missiles solid,  beams close second

An UN-balanced approach like this will also contribute towards breaking stalemates.

@Maiden:  I appreciate the tweaking,  but if there's call for it I'm just going to work on the whole thing from scratch until the simulator numbers look good.   I've automated quite a bit of it,  in that I just have to feed it the relevant numbers for a tier and I get a nice summary across all hull sizes and weapon matchups.

It'll probably take me a while to get the numbers working right;  not only do they have to produce the desired sim. outcomes they also need to improve tier to tier in a meaningful way.  Eep.

Just for yucks,  I think I'll also test the CU 5.7.1 values. 

My 2c on the weapons issue.

The existence of a significant difference in the "bang for the buck" between weapon systems is not necessarily a bad thing, unless the difference are so great that there is no point at all in building them.  What would that point be?  When building the more efficient systems create more damage even against the best defense than the less efficient system does against no defensive structure.   Then one would never build the weaker weapon system.

Somewhere north of this is an area where the ideal mix of weapon systems given the optimum defense will be a mix --- which could be calculated using game theory, but probably no one is going to bother with that  .    

Ideal numbers will not be necessary, just a rough notion of how much of each make sense.  

Of course a huge factor in all this is what the AI will research and build for weapon and defensive systems.    If the AI prioritizes defenses against the most efficient weapon, then building less efficient systems becomes indicated.    

My point is that "bang for the buck" in principle should be determined over a complex mix of criteria, not just cost per part per size, for example.   Allowing significant differences in some areas can add some "color" to the game (others may differ!).

I like the idea of weapon efficiency peaking at different times, because then everything is in play.    One can't just ignore a given line of research without facing a serious catch up problem later.   But I wonder how to implement this considering the huge difference between galaxy sizes and research rates that can be chosen.   Others may be able to address this.

I'm afraid this is impossible to do. Although the games manual explicitly explains that the techspeed setting is of serious importance (it is the only setting where such a remark is made...) the setting is still at total liberty in the hands of the player. I remember someone playing very fast speed under a gigantic all-abundant setting. At that, one could research tier1 + tier2 in a single turn (just wait until the colonial rush is over) and this setting would practically reduce the early-weapon advantage to 0 and the midgame to something very low.

We first need to finish weapons before we can take up defenses.

I have no idea if research into defenses (or shipdesign for that matter) is reflected on the weapon-type of neighbours/war-enemies or simply random or selected by sheer number strength of what's available, but this are really good questions that should be solved beforeahead.

Defenses are pretty hard because alot of stuff can go wrong. In the vanila setting armour was twice at strong in T3 than PD which made point defense obsolete at that stage, because even in an offset scenario Tri-Strontium gave as much defense against missiles than an ontype pointdefense (4-2-2) vs (2-1-1). Until past Tri-Strontium there was no reason to research anything else than Tri-Strontium, therefore there is a need to adjust the different defenses to almost equal numbers (in terms of raw protection)

However, we have different strength in raw attack of each weapon branch and the defenses have to reflect this as well, otherwise PD and Shields will be relatively weak against their ontype weapons and armour strong.

Hey guys,

Yarlen and I had a brief exchange and due to circumstances there's a few weeks left to tinker with the CU. I proposed implementing weapon/defense changes you guys have been working on provided they are fleshed out and properly tested.

We have a go.

If it's unanimously agreed upon the update is finished I'll integrate it into CU V5.7.1 and update the change logs. If not, we'll ship CU 5.7.1 and the weapon/defense enhancement can be released as a modification at a later date.

I can't play a part in the design of the new weapon/defense system other than providing administrative duties and hosting the file archive. My job is hectic and is about to get frenzied. So I'll be working on a new manual when I can (which isn't often). I do believe I can finish it in time for release though.

Good luck. Have fun. I look forward to testing the new weapon/defense system!

Regards,

M.

In my case this is because I don't play huge/gigantic/immense sized games, so it's natural to research only half of the techtree....

Same here. I usually play on Large / Huge. Once i get up to the Evil weapons, the game is usually in the bag or i've lost long before that.

Despite that, The first tier or two feels like it drags on and on...

Alright.. so I reconsidered my earlier "impossible".   It's only "really hard".  I can actually do 5 eras with meaningful,  albeit wonky,  progression throughout the tiers.   You get two tiers of ho-hum improvement (but worthwhile nevertheless),  then BAM -- the next tier races ahead of the other two weapon types.  Call it a breakthrough -- and if we do this we should REALLY reflect it in the tech descriptions to make it obvious that the new weapon type is,  in fact,  revolutionary.

The problem is one of scale.   In order to have meaningful differences between weapon tiers as well as weapon types within a tier,  a certain ratio of difference is needed.  I don't want to push the low end OR the high end of what we're used to,  in terms of weapon costs,  power,  and sizes,  unless I really have to.

I'll keep working on it.  Maybe with a tiny bit of expansion at both low and high ends I can fit 6 in.   I actually do kind of like it better.

To clarify,  I'm looking at (trying to) make the following happen:

Drivers most efficient in T1 and T4

Beams most efficient in T2 and T5

Missiles most efficient in T3 and T6

Any other comments about single cycle (each weapon shines in early, mid, or late game) vs double cycle (as I've described in this post) would be quite welcome at this point.   I'm leaning towards double cycle,  if I can manage it.