Outpost growth

[Seravy]

For similar variance problems, I suggest outpost max pop +20 instead of +10. Maybe +15. The difference between a high pop spot and a low pop spot (say 10 vs 20) is enormous, which I don't think is as intuitive as I expected. Yes, a low pop spot should take longer, but not years longer.

Or max pop +25 and drop the *2.

Pop 25 with an average race takes 10 turns. Pop 10 takes 18 turns. I see nothing wrong with that, the best spot being nearly twice as fast as the bad one. If anything, it's less effective than the intuitive "2.5 times the pop so 2.5 times faster", not more.
Below pop 10 where chance of shrinking gets mixed in the numbers can get ugly, but that's the price for building on a place where you might not even complete the city.
At pop 5 it would take 24 turns but that's only if no shrinking is rolled. With a chance of 7%, that's unlikely, so 30+ turns is more realistic. However a pop 5 spot without ores is horrible and no sane person will want to build on it, not to mention it's really close to the point where the outpost isn't growing at all on average. Fun fact : in this case Raise Volcano might improve the growth rate tremendously, if an otherwise worthless tile gains iron ore or coal - even more effect than a new grasslands tile from change terrain.

[Nelphine]

Right, but to me max pop is just that - max pop. It shouldn't determine whether you even build a city at all. A 10 pop square is a perfectly viable city in my eyes, due to granary/farmers market. But with this formula, it's far more efficient (especially with below average races!) to go halfway across the map to get the max pop squares.

To put it another way: The formula as is (with both racial and pop modifiers where they are) will make lizardman the by far best species, bar none. Because you'll be able to have easily 6 cities in the time it takes an average race to get 2, let alone a slow race. And once you get those cities, their population will increase faster as well. Admittedly, lizardmen were already a top 5 race, but for instance, with this formula there's no reason to ever play a myrran race, which is silly when you consider it costs 2 picks to be Myrran. Sure, in absolute terms, other races are better once the cities are fully developed. But lizardmen can get far more fully developed cities far faster.

The only way to play other races (at the top difficulty) is to never take inquisitor, and either conquer lizardmen cities, hope their are no lizardmen, or hope to have a lot of neutrals to conquer. (Which to be fair is roughly what lizardmen do now, and perhaps your formula is even better than the current one, but, i'm analyzing your proposal, not the current one.)

(And before you bring up 'but lizardmen units aren't that great', please recall my playstyle is to make my entire play based on the worst unit possible that can win, because then i only need to develop to that unit - i don't need to develop further. Which lets me pump out far more of them, far faster than another person making better units. And while lizardmen halberdiers are no longer as strong as they once were, I've beaten impossible with beastmen halberdiers, and lizardmen should still be sufficient to defeat any nodes in the game, and most rare summons an ai will use. Very rares should beat them (unlike beastmen), but i should be able to have far more lizardmen halberdiers than the AI have very rares.)

[Seravy]

Planned racial modifiers :

Barbarians : +30. Barbarians grow quickly and their cities are worthless after the early game.
Beastmen : 0.
Dark Elf : -33 Slow growth and Dark Elf cities are highly valuable due to power/pop so use of bad spots should not be easy
Draconian : -10
Dwarf : 0 Although the race is slow growing, their construction and engineering is excellent, so they should be more effective and building outposts. This makes poor mountain terrain with lots of ores more accessible for them.
Gnoll : 0
Halflings : +40 Halflings have the best agriculture of all races, so they should be outstanding in founding new cities.
High Elf : -25 Slow race
High Men : +10 Slightly above average
Klackon : 0
Lizardmen : +50 Fastest growing race
Nomad : 0
Orc : +20 Fast growing race
Troll : -20 Troll cities are valuable sources of regenerating units, spreading them should not be too easy

[Seravy]

Right, but to me max pop is just that - max pop. It shouldn't determine whether you even build a city at all. A 10 pop square is a perfectly viable city in my eyes, due to granary/farmers market.  But with this formula, it's far more efficient (especially with below average races!) to go halfway across the map to get the max pop squares.

To put it another way: The formula as is (with both racial and pop modifiers where they are) will make lizardman the by far best species, bar none.  Because you'll be able to have easily 6 cities in the time it takes an average race to get 2, let alone a slow race.  And once you get those cities, their population will increase faster as well.  Admittedly, lizardmen were already a top 5 race, but for instance, with this formula there's no reason to ever play a myrran race, which is silly when you consider it costs 2 picks to be Myrran.  Sure, in absolute terms, other races are better once the cities are fully developed. But lizardmen can get far more fully developed cities far faster.

The only way to play other races (at the top difficulty) is to never take inquisitor, and either conquer lizardmen cities, hope their are no lizardmen, or hope to have a lot of neutrals to conquer.  (Which to be fair is roughly what lizardmen do now, and perhaps your formula is even better than the current one, but, i'm analyzing your proposal, not the current one.)

If anything, the OLD system was more beneficial for Lizardmen because racial modifiers were independent of terrain. It used a formula of (max pop+racial modifier)% chance to grow.
Lizardmen had racial +30, average races had +15. So a lizardmen could use a pop 10 spot as quickly as another race could use a pop 25 spot.
The new formula uses race as a multiplier : lizardmen will complete the city 33% faster on any spot than another race on the same spot.

The game always gave you a significantly faster growth on a better terrain and this isn't going to change much.
On pop 10 the old formula gave you 25 growth with a normal race, on pop 25 it was 40. It also gave 5% shrinking, for simplicity's sake we can say the total growth was 20 vs 35. That's a 1.75 times difference.
The new formula has 40 growth for pop 10 and 70 for pop 25, which is also a 1.75 times difference. So what you are talking about, the max pop effect, is literally unchanged in this range. (Yes I'm ignoring the 2% chance to shrink here at pop 10, I'm lazy to include it. So it's actually true for 12-25. Shinking only has a major effect around 7 or below though)

Either way I'll need to run some tests to see how this works in practice.

[Nelphine]

True enough. If you post it, I'll do an extreme luzardman game and see if I can actually make it work how I expect.

[GermanJoey]

Sounds good to me. You should really help Gnolls out a bit though, that race is so sad. Giving their settlers 2 moves wouldn't be bad either.

[Nelphine]

I could see gnolls with speed 2 settlers
They use wolves obviously

[Seravy]

So far, done, two things remaining :
-Adjust the growth bonus/ore, the 3-7 is probably too low. Maybe 5-12?
-Decide how the AI handles low pop tiles.

There are basically three possible ways to deal with this
1. Leave it as is. The AI keeps sending settlers to low pop tiles if nothing else is left, where the outposts eventually disappear and it acts as an infinite black hole swallowing settlers. (probably worst solution)
2. Add an AI only "minimal growth" threashold that makes sure the AI can succeed at making cities on any tile - of course chance to shrink might still waste a settler or two but eventually it works. Problem with this one is, the human player will be eternally unable to use those spots, so they kinda end up as "AI only". (This is not all that different from the current system, I don't think any player ever attempts to build on a pop 2 tile, even if the chance of success was not that bad in the old system as the new one)
3. Set a limit where the AI is allowed to send settlers to, for example "at least 4 max pop". Problem with this one? Slow races will still end up going for tiles impossible for them like in "1." while fast races will overlook at waste tiles they could be using.

Probably 2 is the best? Or 3 if I could make the AI also consider the race modifier?

(PS : gnoll testing and buffs is for a later time, first I want to get this working)

[GermanJoey]

word.

Your #2 is probably best; those spots will probably be only ones that the AI settles very late in normal situations; if they aren't settled late, that means the AI probably has a really awful land situation and thus need whatever help they can get.

[Nelphine]

Ive colonized pop 2 squares. I'd much rather leave as 1, since that's exactly how it ends up happening for me as well.