Author Topic: Ok so ... How is superhuman to be played?  (Read 12551 times)

Offline Sneak Dog

  • Sergeant
  • **
  • Posts: 10
    • View Profile
Re: Ok so ... How is superhuman to be played?
« Reply #15 on: November 17, 2023, 05:18:23 pm »
How you ramp up the research is a delicate balance. As a rule of thumb you should be able to cover the monthly salaries with your current council funding. Building facilities, purchasing equipment and enrolling workers may be something you will need to use other finances as well (except in the very early game, where - except for a few lucky streaks - council funding is net positive).

1999 is already midgame. By the end of 1998 you should have probably about 5-7 bases, with at least a science lab in each (25 researchers), possibly other labs (5-10 more researchers). So let's say this is totalling about 120-180. Sometime in the early 1999, you will hopefully have enough tritanium to start building improved labs, to get deeper research opportunities and further researchers.

I'm on veteran, november 98, kept my score above 2k every month looking at my graphs. It hasn't been perfect, I think there's been about 4 incidents of countries not happy total due to stuff. Funding is 9 million. I suspect superhuman wouldn't have increased funding?

5 bases with 125 researchers costs 6 million in upkeep. That leaves you with 3 million in upkeep for agents, some engineers, vehicles and whatnot. Agents may well take half of that. Not to mention how each science laboratory costs roughly 2 million. (7 bases of this would cost 8.5 million upkeep by the way.)
Seems really tight, especially getting the 5 million per science base seems tough. That sounds like some manipulation of which missions can spawn is going on. Because I haven't gotten aliens pop up yet beyond the military shooting them down, so no selling those sweet alien weapons yet which is when the income really kicks off.

Offline psavola

  • Commander
  • *****
  • Posts: 848
    • View Profile
Re: Ok so ... How is superhuman to be played?
« Reply #16 on: November 17, 2023, 05:50:58 pm »
I'm on veteran, november 98, kept my score above 2k every month looking at my graphs. It hasn't been perfect, I think there's been about 4 incidents of countries not happy total due to stuff. Funding is 9 million. I suspect superhuman wouldn't have increased funding?

Not as such increased funding, but because there are significantly more enemies on each mission (and each with loot), you can practically get a lot more points on each mission. Let's assume up to 40-50 % more maximum points per mission than with veteran. So I'd suspect the funding could go up in practice much faster in SH than veteran.

The funding can go up really fast once you do those missions which give you huge number of points (say 1000+ each). In my current SH game in January 2000 the funding is at some 57M and if the graph is correct, a year before it was something like 16M. So you can easily double and even quadruple the funding during 1999.

I also loaded my previous SH game that I stopped playing in June 1999. The income is at 27M. Based on the graph the income in January 1999 appears to have been around 15-16M and in July 1998 maybe 7.5-8M. So the funding appears to have been doubled in 6 months in 07/98-01/99. So if you get missions that give you good points the funding can really ramp up fast once you get on with it.

It's difficult to go into the specifics. Up until 1999, you don't really need any other crafts than 1-2 private cars, some vans, helicopter, an osprey, a land rover, and kitsune (and once you have the latter two, you no longer need the others). The rent for these is minimal. Until you get kitsune, you can do all missions either by a car or by fly-transfering from a main base, you don't need to stock up every base. You need only 50 soldiers and lots of dogs. You don't need more than 50-90 engineers ever. Though later in the game (in 1999+), once you start intercepting - and the enemies start retaliating -, you may want to deploy some more soldiers on other bases as well, depending on your interception strategy.
« Last Edit: November 17, 2023, 05:57:30 pm by psavola »

Offline Stone Lake

  • Colonel
  • ****
  • Posts: 193
    • View Profile
Re: Ok so ... How is superhuman to be played?
« Reply #17 on: November 17, 2023, 05:56:39 pm »
Quote from: psavola
I'd say earning 15-20M a month in the beginning of 1999 should not be all that difficult.

Nope, it seems rather exceptional, so good job. And most likely depends on when you start to snowball on score. I mean, getting millions of funding is matter of "when", rather than "if". So the earlier you start to get 2k+ per month, the more funding you get by 1999.

Also, I wonder how much of points is from just research? Usually most $$ mission come up later (tasty aliens), so you get more money to get extra scientists. Extra scientists get you extra points. Voila, points-funding feedback loop.

Quote from: psavola
Agents are not all that expensive, and you should not need hundreds of them. If they get killed in every mission, you are doing something wrong.
Agreed! Every agent death feels like a honest-to-god mistake. If only I could stop making hundreds of them  ;D

That being said, getting this much funding is not at all necessary to beat the campaign, as invasion proceeds very lazily nowadays. For me, pre-99 is clearly the hardest part of the campaign, the rest is just mop-up. You can loot/capture some very cool stuff and snowball real quick.




« Last Edit: November 17, 2023, 06:02:27 pm by Stone Lake »

Offline psavola

  • Commander
  • *****
  • Posts: 848
    • View Profile
Re: Ok so ... How is superhuman to be played?
« Reply #18 on: November 17, 2023, 06:07:48 pm »
Also, I wonder how much of points is from just research? Usually most $$ mission come up later (tasty aliens), so you get more money to get extra scientists. Extra scientists get you extra points. Voila, points-funding feedback loop.

Could very well be so. Looking at my previous SH game (that I stopped playing in the beginning of June 1999), it seems I had over 2K-3K from research alone every month already a year ago (so, in the middle of 1998):

researchScores:
  - 2351
  - 3532
  - 3521
  - 3486
  - 2928
  - 3691
  - 4542
  - 5070
  - 4453
  - 3527
  - 5650
  - 1790

Offline Sneak Dog

  • Sergeant
  • **
  • Posts: 10
    • View Profile
Re: Ok so ... How is superhuman to be played?
« Reply #19 on: November 17, 2023, 06:26:34 pm »
Maybe XCOM files works differently, but as I understood 2k points is all you need to make countries happy for the highest increase tier of 5%~20%. This'd mean that funding goes up a lot later on because it's compounding. Not because your score gets so much better. If you get 12.5% each month for 12 months, that quadruples your income. (1.125^12 = 4.1)

https://www.ufopaedia.org/index.php/Country_Funding_(EU)#Monthly_Funding_Change

Offline psavola

  • Commander
  • *****
  • Posts: 848
    • View Profile
Re: Ok so ... How is superhuman to be played?
« Reply #20 on: November 17, 2023, 07:12:39 pm »
Maybe XCOM files works differently, but as I understood 2k points is all you need to make countries happy for the highest increase tier of 5%~20%. This'd mean that funding goes up a lot later on because it's compounding. Not because your score gets so much better. If you get 12.5% each month for 12 months, that quadruples your income. (1.125^12 = 4.1)

https://www.ufopaedia.org/index.php/Country_Funding_(EU)#Monthly_Funding_Change

I haven't really looked at this so I working on assumptions. You're probably right. Looking at the OXCE rulesets, there doesn't even appear to be a way to change these percentages directly.  There is an OXCE ruleset giving bonus funds directly proportional to the score (performanceBonusFactor) but it isn't used in XCF.

Offline Stone Lake

  • Colonel
  • ****
  • Posts: 193
    • View Profile
Re: Ok so ... How is superhuman to be played?
« Reply #21 on: November 17, 2023, 08:57:41 pm »
As I understand it, the funding increase is effectively proportional to the score. The increase is RNG and each score point raises chance that country increases funding. There are like 100 countries without any guarantee that funding will be increased. Thus every increase in the chance counts, and there's no score limit to get "max" funding.

Offline Juku121

  • Commander
  • *****
  • Posts: 1799
  • We're all mad here.
    • View Profile
Re: Ok so ... How is superhuman to be played?
« Reply #22 on: November 17, 2023, 09:31:32 pm »
I'm not certain, either, since I don't have nor plan to have a Github account and thus can't search code very well...

But as I understand it, every month every country evaluates your performance and falls into one of these four categories: happy, satisfied, dissatisfied, pact. Your overall score generally correlates with that category, but not always (alien infiltration, lots of alien missions there, no X-Com presence, the works). Once they're happy, they'll just randomly roll on that 5-20% wheel.

Although I saw general Council happyness threshold being at 500, not 2000 'rating', which seems to be more or less score. @Sneak Dog, where does your 2000 value come from?



There are also custom score values ('monthlyRatings'), but as far as I can tell, these are just feel-good (or bad :P ) words for the Council to pat you on the back with.

Offline Sneak Dog

  • Sergeant
  • **
  • Posts: 10
    • View Profile
Re: Ok so ... How is superhuman to be played?
« Reply #23 on: November 17, 2023, 09:44:48 pm »
I'm not certain, either, since I don't have nor plan to have a Github account and thus can't search code very well...

But as I understand it, every month every country evaluates your performance and falls into one of these four categories: happy, satisfied, dissatisfied, pact. Your overall score generally correlates with that category, but not always (alien infiltration, lots of alien missions there, no X-Com presence, the works). Once they're happy, they'll just randomly roll on that 5-20% wheel.

Although I saw general Council happyness threshold being at 500, not 2000 'rating', which seems to be more or less score. @Sneak Dog, where does your 2000 value come from?



There are also custom score values ('monthlyRatings'), but as far as I can tell, these are just feel-good (or bad :P ) words for the Council to pat you on the back with.

Yeah, this seems about right? It'd make sense given how simple a lot of these small mechanics in XCOM are. Apparently the 2k rating came from a post on a forum of just some random person on the internet. My bad.
Regardless, five bases with scientists or even seven by the end of 1998 seems ambitious. Though having such a couple of months later is suddenly reasonable, hah.


Offline psavola

  • Commander
  • *****
  • Posts: 848
    • View Profile
Re: Ok so ... How is superhuman to be played?
« Reply #24 on: November 17, 2023, 10:04:17 pm »
The algorithm seems to be similar to what's described in the OXC wiki. See in particular how 'newFunding' is determined by RNG within certain limits:

https://github.com/MeridianOXC/OpenXcom/blob/oxce-plus/src/Savegame/Country.cpp#L181

Offline Juku121

  • Commander
  • *****
  • Posts: 1799
  • We're all mad here.
    • View Profile
Re: Ok so ... How is superhuman to be played?
« Reply #25 on: November 17, 2023, 11:18:41 pm »
Spoiler" Oooffftopic":
OXC wiki
What's an OpenXcom Wiki? Ufopedia has very little to do with OXC(E) as such, and the ruleset reference doesn't go into anu of this.

Offline Stone Lake

  • Colonel
  • ****
  • Posts: 193
    • View Profile
Re: Ok so ... How is superhuman to be played?
« Reply #26 on: November 18, 2023, 05:16:01 am »
Notice first that it's a per-country increase, not overall.

So, the score is made of
```good = (xcomTotal / 10) + _activityXcom.back();```
Total score /10 + local score for this country (e.g. mission was there), or, conversely, an alien base / UFO flyby creates alien activity.

There also seems to be distinction between `negative score` per se and `alien score`. IIRC, any battlescape result, negative or positive, is x-com activity and penalty for ignoring a mission is alien activity. Kinda wondering which is worse, to ignore a mission or to get wrecked there.

Notice how
```if (RNG::generate(0, good) > bad)```
there's always a chance to get 0 and no funding increase.
And if bad ~ 100, good ~ 200 (~2k score), there's basically 50% chance to increase the funding.

So it's important have a lot of score relatively to `bad` to increase the chances of funding, and to keep mistakes (ignoring missions?) and thus `bad` low.

Thus funding increase `per country` is fixed 5-20% of old funding. But the funding increase overall, like I said, is effectively proportional to the score, sans corner cases. And the funding itself is exponential (granted you get that 5-20% of old funding), up to a cap. Which is how you reach those gajillions of money.

There also could be an argument about prioritizing bigger-funding countries...
« Last Edit: November 18, 2023, 07:42:21 am by Stone Lake »

Offline Juku121

  • Commander
  • *****
  • Posts: 1799
  • We're all mad here.
    • View Profile
Re: Ok so ... How is superhuman to be played?
« Reply #27 on: November 18, 2023, 11:08:01 am »
Good analysis!

Although...
But the funding increase overall, like I said, is effectively proportional to the score, sans corner cases.
This is not really true in either the strict sense (proportional = constant ratio) nor the weak one ('strongly related').

First, since there are a lot of countries and only so many aliens/cultists/shady orgs to go around, alien activity is going to be zero or near zero for most of them. If you don't believe me, go and check a 1999 save for 'activityAlien'. You'll see a lot of zeroes.

And once you clear the 'good > bad + 30' threshold (300 score as a minimum, probably 500 to be safe), you're likely to get an increase from most of them, barring local score spikes.


Second, the score increase is far from proportional. If you have 2000 score and the aliens scored 100 in the country, that's 200 vs 105, a rougly 50% chance, as you said. If you increase your score to 4000, it's ~75% chance, a ~50% increase in funding for a 100% increase in score. Going to 8000 is 200% increase for ~75% better funding, 16k is 400% to 88%, and so on. These are pretty strongly diminishing returns, and that's one of the countries where score actually matters. For most of them, 1000 score would already make the funding increase a near -certainty (100 'good' vs a few points of 'bad').


TLDR: What matters more is not big score numbers (unless you use 'performanceBonusFactor'), but rather offsetting alien activity with your own.


Kinda wondering which is worse, to ignore a mission or to get wrecked there.
Ignoring the mission, since alien score is vastly worse for the country than generic negative score.

Incidentally, that also makes slander missions a real menace to your funding.
« Last Edit: November 18, 2023, 11:11:10 am by Juku121 »

Offline Stone Lake

  • Colonel
  • ****
  • Posts: 193
    • View Profile
Re: Ok so ... How is superhuman to be played?
« Reply #28 on: November 18, 2023, 12:02:53 pm »
Quote
And once you clear the 'good > bad + 30' threshold (300 score as a minimum, probably 500 to be safe), you're likely to get an increase from most of them, barring local score spikes.
No, because total alien activity is a thing.

Offline Juku121

  • Commander
  • *****
  • Posts: 1799
  • We're all mad here.
    • View Profile
Re: Ok so ... How is superhuman to be played?
« Reply #29 on: November 18, 2023, 12:30:01 pm »
Sure, it's not quite as good as I said, you're right.

But that still presupposes X-Com score is relatively close to alien score, in which case your overall score will suck (relative to all the action going on). As long as you have positive score, the threshold is better than your score indicates, since score is [X-Com activity] - [alien activity], but 'good' vs 'bad' is [X-Com activity]/10 - [alien activity]/20 + [local difference].

If your overall score is low, of course you're not going to get a big funding increase.


E.g. if you have 500 score, of which 1700 is yours and 1200 is the aliens' in a country with negligible activity, say, 10 alien and none X-Com, you get 'good' 170 vs 'bad' 70, a ~58% chance of an increase. And that's a kinda bad score overall, you're barely keeping X-Com's head above water.

If it's more like the average on your graph, ~700 or so, a more likely +500 score and local activity 0 would get you 120 vs 35, a ~70% chance of an increase. And I think a successful month is more like several thousands of positive score overall, so more of an 80-90% chance.

But your point is taken, I did forget about the roll vs 'bad'.
« Last Edit: November 18, 2023, 12:45:46 pm by Juku121 »