Keith, for clarification, do you mean to say that each wave is 25% larger than the last, even in the same scenario?
I'm not sure what you mean by "even in the same scenario", do you mean during the same stage of the progression? Anyway, yes. To give details (which will make their way to the wiki at some point) :
* AccumulatedEventWaveCounter starts at 0.
* NextRawEventWaveSize starts at 10000.
* SpecialDifficultyMultiplier =
** if difficulty < 7 then difficulty * 0.15
** if 7 <= difficulty < 9 then 1 + (difficulty - 7) * 0.5
** if 9 <= difficulty then 2 + (difficulty - 9)
** SpecialDifficultyAndHomeworldCountMultiplier = NumberOfHumanHomeworlds * SpecialDifficultyMultiplier
* Every minute, each completed city hub adds 1300 to AccumulatedEventWaveCounter (this used to be roughly 5500 or 6000, it's been nerfed quite a lot).
* If you have a galactic capitol it adds 6500 (5*1300).
* If you've initiated the final sequence (I don't think anyone's gotten that far yet), AccumulatedEventWaveCounter is doubled every minute. Yea, it goes nuts.
* When AccumulatedEventWaveCounter > NextRawEventWaveSize:
** 1) An EventAttackArmada with total budget = AccumulatedEventWaveCounter * SpecialDifficultyAndHomeworldCountMultiplier is deployed (one battle group against a human home command station, one battle group against the refugee outpost, and one battle group against each city hub).
** 2) AccumulatedEventWaveCounter = 0
** 3) NextRawEventWaveSize = NextRawEventWaveSize * 1.25 (we actually use fixed-int math, not floating point, but that's not significant here)
In this case, you would want to complete the entire quest line as quickly as possible because you are going up against this multiplicative increase over time.
There is certainly a motivation to not take longer than you need to; the journal text is intended to communicate that the AI is taking these cities
very seriously and that this is unbelievably dangerous. Looked at another way, this is a separate AIP counter for the emerging Spire forces, and the Spire just aren't good at avoiding attention
But bear in mind that during a given stage (meaning a given number of cities), the "velocity" of AccumulatedEventWaveCounter does not increase, even though NextRawEventWaveSize is increasing. This means that the time to hit the threshold is increasing by 25% each time too.
Briefly, the reasoning behind the 25% increase rather than just leaving the threshold as-is or having the threshold only increase with new story stages:
1) Thematically, if the AI's attack fails, is there a point in it sending exactly the same size attack again and again? Would it not realize that it should send something bigger next time, even if it takes longer to get it ready?
2) Fun-wise, do you really want to get hit by the same-size attack that you just defeated, again and again? Is that fun?
3) Challenge-wise, if it hits some kind of "ceiling" at each stage, that means you can just build up defenses sufficient to handle that and take your sweet time building the weapon to destroy the AI, in its plain sight.
Can you state the optimum scenario for time limit?
On Diff 7 through 8 it should be quite doable if your campaign lasts the normal length of time, which for most folks seems to be 8 to 16 hours. I'd imagine that the first city is going up at about the 4 hour mark, and another city every 1-2 hours thereafter (so fifth one between 8 and 12 hours in). Once you've got that and have built the special structure available then and researched the advanced stuff then available, you should have a sufficient "bonus fleet" from the Spire stuff that you can take out the AI homeworlds yourself via direct attack. Depending on how long you've been playing, difficulty, etc, you might not be able to pull it off. You can then either use the advanced technology to build more cities, build more "bonus" ships, etc, until you can manage it. Of course, the AI attacks will keep getting worse as you do that, but depending on your defensive setup (chokepoints, etc) you may get to a point where it doesn't really matter what it throws at you.
But if, for whatever reason, that avenue of victory is not possible, you can continue into the last part of the progression, which culminates in the "alternate" victory. It's quite challenging in itself, and during the last phase the AI will be going absolutely coo-coo-for-cocoa-puffs with the event attacks, but if you succeed you've won the game.
Of course, it's quite possible that the way the numbers are balanced right now (whether it's the point-accumulation-rate, the threshold-increase-rate, the point-cost of golems and such, etc, or some combination thereof) that this is just unreasonably hard. I'm still getting somewhat conflicting data on that. I really don't want to just nerf the challenge into the ground and have people walk over it and then complain about how going fallen-spire is so OP, etc, etc... so we have to approach this in a somewhat stepwise fashion.
My next crack was going to be to reduce the threshold-percent-increase from 25% to 15%, so it would take _much_ longer to build up to the kinds of critical mass that has been roflstomping people. Looking at the numbers, with one new city going up every 1.5 hours after the first until you have 5, and completing the galactic capitol an hour after the fifth city, the current setup would yield:
First 100,000+ point attack is the twelfth attack, at 4 hours after the first city hub is finished (three cities).
First 200,000+ point attack is the fifteenth attack, at 6 hours (about to finish fifth city)
First 400,000+ point attack is the eighteenth attack, at 8.3 hours (still just 5, but with capitol)
First 800,000+ point attack is the twenty-first attack, at 11 hours
With 15% increases:
First 100,000+ point attack is the eighteenth attack, at 4.9 hours (fourth city went up recently)
First 200,000+ point attack is the twenty-third attack, at 7.3 hours (capitol went up recently)
First 400,000+ point attack is the twenty-eighth attack, 9.4 hours
First 800,000+ point attack is the thirty-third attack, 13.8 hours
Actually, looking at that, it seems that it would be harder with 15% attacks because you would get hit far more often (which point I think orzelek was trying to get across).
So scratch that change, I think, unless someone else sees something else I don't.
Well, I'm a bit wary about it, but I could just drop the per-minute-per-hub increment from 1300 to 1000 (and the capitol's increment from 6500 to 5000).
My guess that this is a bit overkill, but I'm not the one getting hammered by multi-golem attacks
Anyway, with the reduced increments:
100,000+ (12th) : 4.7 hours
200,000+ (15th) : 6.8 hours
400,000+ (18th) : 9.1 hours
800,000+ (21st) : 12.6 hours
I guess we'll see.
Anyway, I'll be starting a separate thread on this sometime in the next few days. The goal balance-wise is to provide a total difficulty similar to the "normal" game, understanding that it shuffles around the timing and type of difficulty that you experience.
