Caster of Magic Release thread : latest version 6.06!

[Seravy]

Fixed it, please re-download.

...we could have avoided these bugs if someone actually bothered to test the release candidate. Too bad.

[Seravy]

By the way, of the reasons you just gave for why the starting settlers are necessary, the only one I really agree with is 'otherwise the AI already have all the good spots'. But another solution to that is.. Reduce AI cheating.

I think we've got to that point.

It's not the cheating. The AI always does settlers first, which means the player has to do the same otherwise they miss out on getting any good place (unless map is huge or distance from others is very large). If you delay your first settler by 20-25 turns to have a few swordsmen, a granary or something, you can forget about having good cities.
Do note the AI doesn't buy settlers for gold unlike sawmills, so all they have for them is the production bonus. And they will actually have a "gap" between them for the fighter's guild (well, used to, now the AI is a bit more detailed in picking when to do that) so if the player keeps to settlers they can actually make up for the production speed difference on the third/fourth etc settlers.
But this is about more than just the AI, anyway. Delaying your first settlements 2-3 years means you'll be that much behind in developing those cities, which is pretty bad. And it takes a lot of time for those to turn productive even without such delay.

[Kaiku]

If you delay your first settler by 20-25 turns to have a few swordsmen, a granary or something, you can forget about having good cities.

This makes me so sad. Are there no other ways for the AI to stay challenging? I feel this 'AI must spam cities' will keep haunting this game and possibly get in the way of many further changes.

Maybe something like giving some personalities big resource bonuses, so they don't have to expand as much?

[Seravy]

Are there no other ways for the AI to stay challenging?

No, there isn't. The same applies to the AI, if they don't build the cities early enough, they lose. Unless we discard diplomacy entirely and allow them to always declare war and conquer from others which wold be horrible for other reasons. You need to get cities one way or another...cities are like 90% of the possible resources you can get from a map, nodes and treasure might be strong early boosts but on the long run they won't compete with a city that's producing 50 gold, 20 power, 20 research and whatever else on its own.

It's like a non-violent war, settlers are ultimately fighting for territory that will turn into large cities with massive profits.
Most other 4x games have some sort of a mechanic that prevents people from settling the entire map early. You need new tech to reach another continent, or just a certain distance in parsecs, or you need the tech to be able to build somewhere, like on a desert or toxic planet. MoM has none of that, the entire map is up for grabbing on turn 1, and that means everyone has to push settlers out as much as they can without risking to be unable to protect all of them. But since the AI doesn't even know the concept of protecting settlers/settlements, they don't even have that to worry about - and they have no reason to, there are 4 of them so if one loses cities to another, or even the human, the others will still be able to provide an entertaining game.

[Nelphine]

But the cheating is what prevents the human from getting any of the good spots. Yes, I understand the principle behind what you're saying, but if the AI is settling too fast, and the Hunan can't jeep up, that's partially be a use the AI cheats.

In.just saying, that is a contributing factor, and you didn't list it.

[Seravy]

If you mean not needing to scout then yes. If you mean the production bonus, then no, as I explained above. (except maybe on Master and Lunatic but probably not even there due to the "fighter's guild" gap.)
But "AI can't scout" is not something we can "reduce".

[Nelphine]

Except I'm in the camp that thinks all AI cheating is too high currently. This includes starting good, production bonus, and gold bonus, all of which contribute to AI settler building. I don't think the amounts need to change a lot, but I do think they exist, and doing a change like adding 2 free settlers is a far more radical change than reducing several bonuses by small amounts that all add up.

Given that a fee months ago we were extremely happy with where the game was, despite outstanding issues (yes I argued for fixing some if those changes myself, so I'm part of the problem) I'd be much happier with smaller incremental changes than drastic ones.

[Seravy]

Third and last time I'm saying this :
The starting gold and gold bonus DOES NOT CONTRIBUTE TO SETTLER BUILDING.

If you want to keep saying that, pinpoint the part in AI production purchase that tells them to spend gold on settlers. I believe the sources were posted when the new procedure for that was implemented.

Small incremental changes won't cut it where there are fundamental problems. Not having choices in the early game and being forced to do the same thing regardless of difficulty is a fundamental problem. Also we'll never find out if free settlers are good or bad if we don't try.

Speaking of AI cheating, does that include Advanced and Expert? I believe we agreed they are fine and only Master and Lunatic might be excessive?
My point is, I have to build settlers for the first 20 turns on every difficulty level, even Normal, assuming I picked a wizard that's suitable for the difficulty when playing anything but an early offense strategy.
I never realized until now but having Horsebowmen isn't the only thing that makes Nomads competent in the early game unlike most other slow races. It's also the 100-150 early production you save by having cheaper settler costs. 2-3 free Horsebowmen compared to that other races could afford.

Here is what Hadriex said today : "Free settlers help slower strategies over faster ones."

That's definitely the case, and faster strategies were too powerful, we had at least one thread for that issue but couldn't find a solution, so now having one should be a good thing.

[Nelphine]

The AI buids more than 2 settlers. It certainly takes far more than 6 settlers to get every good city on any map beyond tiny. 

Therefore, gold (which contributes to other buildings, like sawmills in secondary cities) absolutely contributes to AI building settlers, at a time that is relevant to human cities.

If they didn't, there wouldn't be an issue with trying to get the 13th city on the map, and it would he fine to build settlers late.

And yes this is for every difficulty, assuming you use appropriate wuzwrds for the difficulty.

And you can disagree that small changes don't fix the problem. I'm fine with that. We already know you and I have different opinions of exactly what the problem is.

But that is my point, I consider the problem you're fixing to be subtly different than you.

[Seravy]

and it would he fine to build settlers late.

Building settlers late is never fine. All your troops and resources comes from cities and they took a while to develop.
Even if there was no opponents to compete with, you'd still want settlers as early as possible if your plan is to maximize your empire in a certain amount of time.
There are other mechanics that make this core 4X mechanic less obvious here, but whoever has the most cities wins, and you get those cities by building a settler early enough to have time to grow it into a relevant size. (Or by conquest. But then you can't play your chosen race. Which is sad. Plenty of races rely on having more than just 1-2 cities of that race to work well. I still remember how playing on high difficulty meant "never build a settler. Just take whatever the AI builds for you". That definitely works but why do you get to pick a race if you're forced to play whatever your enemies picked instead? Where is the fun in that?)