Deity: how to catch up

[DRJ]

Hey guys, feel free to comment on my strategy and share your own strategies to cope with deity difficulty. I like to learn new stuff!

----catching up on tech----

This particular strategy works well for deity difficulty. I suppose trying it on lower difficulties would suck out the fun of building lots of stuff (you don't really need) in your cities.

The strategy is called "100% tech". I don't know if it is has counter arguments I didn't think about but for me, at least, it works the best.

We all know it: deity sucks in the beginning, you start with only one city compared to the two of the AI.
The game is rejecting you to conquer a secound one fast - either because you don't have a good military to do so (because you can't afford it, perhaps?), or because you started isolated on an island and don't have a possibility to.
All the while the AI just rushes away in tech and score, plus, one or two of them doing it extraordinarily quickly.

A lot of catching up to do but then, once you finally were able to settle/catch some cities, the maintenance kicks in before you are able to pay for the cities and at the same time keeping tech slider at 100%, what can you do?

Yes, you got it, you build lesser wealth (and at currency wealth) to keep it high - at 100% (or 95%, if this goes better with some balancing situations. In the beginning it might even go down to 50% or lower with lesser wealth being generated everywhere)

Once you have a bit of economical growth because new techs allow new buildings or the population works more tiles or becomes specialists, you have to unswitch certain cities from producing lesser wealth. After they completed the important building, like a school of scribes or a river port, switch them back to produce lesser wealth and the next one can construct a building.

The strategy doesn't work well with meager wealth but anticipating (I wouldn't recommend beelining it) the tech "barter" for lesser wealth will be ok, usually you have to wait for more cities until then, anyway.

The strategy will enable you to very soon get close to the #1 AIs GNP value but your production is only playing secound league. You will need a good army backbone consisting of only few units.

Applying the "100% tech" strategy will give you the chance to catch an early religion like Shinto, will help you to reach much needed civics that lower maintenance more quickly and also help you to tech to writing (and currency) soon [important if you play start as minor civs].


At the same time you have to expand your empire very strategically.
You will not have a possibility to build new cities very soon. (I stopped at 10 cities and plan to stay there, well maybe an 11th for copper ore + marble)

Having a "fish trap" slice of good land for enemy settlers and units to rush to helps you to also pillage some parts of tech as well as additional money to untune cities for some turns.

EDIT: Once you caught up to AI in terms of tech I would change the strategy and gradually unlock the cities from producing (lesser) wealth, well maybe have 2-3 wealth producing cities, still.


Here are the screenshots:
1.) The "Fish Trap" (showing score, me 4th with 266 points, 1st has 417 points)
2.) Demographics (note the nearly even GNP although I only have maybe half the cities!)
3.) City Manager (8 of 10 cities producing lesser wealth)

 

[Thunderbrd]

Always neat to read in depth strategy posts like these. Very interesting stuff. If I get sucked into commenting too much my MP games could suffer though

 

[DRJ]

continuation:

Strategy after researching Currency (and Writing + Mathematics).

Now you have the ability to trade reseources, goods and technology for money. Good.
The AI will at this point still be ahead in tech (maybe 12-15 techs instead of the like 22-25 at Sedentary) but you maybe have some techs they will still buy. AI usually researches a lot of useless not essential techs which enables you to get to the good ones first and then trade them in once the AI attempts to research them as well.
Always have an eye on their research, you must keep the monopoly on techs (if you play with no tech brokering ON). This way you can sell the monopoly tech to everyone if they have money.

You should be able to trade in workers (at Eternity gamespeed they are worth like 1000 and elephant workers like 1700-2000 )

and goods (800 for something like crabs, much much more for strategic or rare res)

When buying/selling tech re-negotiate a few times, AI makes like 4-5 different offers, ranging sometimes up to 4000 gold difference. When selling res to AI, it will often not give new prices you gotta take their 800 or leave it.

When AI tries to sell you something for like "150 and 1 gold per turn", you will often see their total income is like +10/turn and they have a max 150 in cash.

Now re-negotiate the price to for the good you want to sell to 2 gold per turn and then ask AI "what would you do to accept the deal?". In Eternity a deal lasts 160 turns, so +1 gold would be equal to 160 gold.

And voilà, the AI offers like "2 gold per turn +10 gold" for the res.

So in this case renegotiating the deal this way (more income per turn instead of bigger lump sum) brought you like 20 gold more,
[1*160+150=310 2*160+10=330 ] or about 7% more. Obviously, coping with Deity you need this very much.

But there is a downside to this: if the AI does declare war on you before the gold/per turn is over, you maybe lose more than if you'd picked the offer with the bigger lump sum.

Re-negotiating deals the way I described is very helpful shortly after a peace treaty with a civ, so after a war or if a minor nation advanced into a normal civ with writing.

With all the trading going on you will realize that while often not yet being able to catch up to the leading AI in score, you'll have gained (near-to) tech parity and also, if playing nice with the other civs (right of passage but not open borders!) you have always a place to sell some stuff to compensate your "100% tech" related income defecite, which could be about -500/turn with around 13 cities (8 overseas).
The cities should now all be out of producing wealth and catching up for infrastructure, building courthouses, libraries and such.

The next phase is the specialists/wonders strategy.

After currency and math, you're able to unlock a lot of +GP-Points civics, throwing away the old "caste" and "devine cult" stuff which had a lot of -GP.

If you're good you have your secound golden age running by now by sacrificing 2 different GP (first was around caste, i don't count the "captured fire" GA here) and can adept to democracy civics like "republic", "proletariat" and "legislature", perhaps even taking all the way to "pacifism" to keep the high GP-output even after the GA ended.

You now have to tune your cities into GP-factories and produce 3 different ones, this can be very effective if you get lucky and have a lot of advanced GP-producing sites or extremely frustrating at times when you always get unlucky.

However, even only trying to create a "never ending golden age" can be fun and if it works for 2 or 3 cycles, a lot is won in terms of catching up economically.

The next step is consolidation.

Build a ring of forts to connect your cities and to fortify important landmarks at your borders, trying to cut off enemy units to settle where they shouldn't - close the fishtraps and make it biotopes for animals before settling it later. Maybe put in an assasin to kick off scouts and such who can enter with right of passage. Assasins have a jungle attack bonus so they are suited very well for this.

If all went well until now, you should be able to enter Medieval times with sufficient infrastructure, punctual tech superiority, some wonders (rushed with captured civillians) and pretty valid trade relations to AI,
granting your economy to end a lot of turns with a high tech rate but a lot of income loss.

If you need money, "print" it, sell stuff. Only tune the slider down if noone has money left or the deals you could make are much too bad. (I sometimes sell a not so important tech for like 4000 instead of 8000 to a mediocre civ, no big deal) Just take care of what you give the leading AI.

The next focus will lie on how to prepare for the ultimate Stack of Death of the 1 or 2 leading AI, because your or their expansion will make contact sooner or later (in my case the zulu with close to three times as many cities [35:13] will knock on my doors much sooner I'd expect...)

So the next chapter will be about that.

 

[Sgtslick]

So much diplomacy abuse lol.. play with no tech trading + no advanced diplomacy (no unit trading etc) + no trading things like contacts and gold per turn. This is much more fair to ai, deity isn't that hard bro, u'll have more fun when you are playing from behind throughout the entire game, not just in prehistoric/ancient. The main advantage you have vs AI is war (apart from diplomacy abuse), so your never out of it until your dead.. just keep battling away tactically. I virtually never go for a religion on deity, I always just conquer one. I usually always run 65% science 5% espionage 30% tax. Lesser wealth isn't bad, I prefer building espionage tho since its 100% conversion, then i just steal the techs and force them to switch civics and pollute their water and stuff - meanwhile leveling up my spies. I don't mind tech trading on - coz it makes the game go faster, but I definitely don't play with unit trading, its really unfair imo.

Make sure you set reminders for when diplomacy deals expire. I tend to always sell all my resources, but I turn off gold per turn trading so its a little more fair for AI. I do do dirty sh1t like selling all my resources to neighbour on the same turn that I declare war though lol.. Then sell him peace, and when that expires.. same thing again

 

[DRJ]

Thx for your reply.

You do realize you contradict yourself, do you? On the one hand you neglect a big part of the game (tech trade & advanced diplomacy) because you say it can be abused and on the other hand you abuse constantly elsewhere by faking deals and by stealing techs and enforcing civic changes, things AI does very seldomly (never had a civic change caused by AI).

I think the main goal for experienced players like us should be to expose the flaws and we can't do that by ignoring parts of the game but by actually playing them and reporting. Thats what I do here.

One main thing is that AI buys workers & res all the time, regardless if they need them or not, if they have the money to spend. The human spends the traded-in money on income deficite and the AI has some more or less useless res or units more to spend maintenance for but not more money to have a negative deficit, probably even in the mood to try to save a little bulk, meaning anticipating a higher slider.

So in case of the AI, the old capitalistic principle of "trying to sell totally unrequired stuff to people who -originally- have no money to pay for but will accept big liabilities to do so anyway" is proven once more.

But exclusive of geo-economical principles, we must realize that aside of us playing deity, most users of the mod would not really want the AI at this stage in game to show economical scientific growth behaviour or an academic level of the best possible value in every deal. No, I think it's pretty realistic what the AI does:

Imagine people in ancient and classical era: they just came out of the late stone age and they like shiny things. Yes, they know about the concepts of power and of the benefits of trade. But often the hints to efficient and rationalized behaviour to these concepts were affected by the ritualization of society and also by the behaviour people of power tended to show (now and then): the greed for rare things because they represent prestige and therefore political power.

So why you, the deity player, would be able to argue "AI sucks in this and that", the represention of such mechanism itself is not bad per se.

What we miss here is a learning curve for the AI.

In old civs you could have people making stupid economical decisions all the time but most of the time this behaviour appeared to be ending as people understand its not sustainable to grill a cow everyday (regarding the "AI always buys workers if it has money" mechanism) and to invest all surplusses into that like a junkie. Yeah AI, that's what you are, admit it!!!

So I would argue that there could be possible other ways to play deity and still have the diplomacy bit and also the overpowered spying (although I tend not to use it that much in the beginning as I hate losing my espionage units entering enemy AI cities, later you can build a dozen and send in for certain acts) included.

Two ideas that could especially be implemented for deity:

make the trade abuse harder,
so AI will not buy unlimited stuff and only spends a certain max % of their treasury on stuff / per turn and per item (techs/units/res) if dealing with human player,

&

make the spy abuse harder, perhaps needing a great spy for important spy actions (change civics/religion/nuke).
great spy could be undetectable and would have a 85% chance to succeed and a 15% to survive. So most of the time they would be effective but gone, not always though^^


By the way 85% is also the number I propose we should cap the possible max win % that could have like another 10% by certain promos so the max win is 95 with all stuff.

Of course this only applies to certain units vs units. Situations that have a nearly 100% win like units from different eras etc sure should still have 99,9999%.

What I means is an unpromoted vs an overpromoted axeman should have a min 15%:85% chance.

Ok, if the later is fortified on a forested hill in a fort we sure could add that and he would get 99,9999%
but aside of that, during a fight axe vs axe on a tile without defence and unfortified the max you could get by promotions alone could be 85%...

 

[Sgtslick]

Yeah I like the capped odds idea that sounds great.

I do lose lots and lots of spies, over half on there first go around, but once they start getting the evade detection promotions they do better. I just like espionage, its definitely not cheap (in terms of espionage points), could be deemed cheap in terms of cheap tactics i guess.. lol

I do contradict myself, i'll agree with that, but I wouldn't say I 'constantly' abuse cancelling deals and espionage. In my current game as Tapi on GEM i've only cancelled deal once when going to war with inca. Generally you want to avoid war with AI as much as possible unless your prepared / on your terms. In terms of espionage, its not abuse. Its not easy, you need to dedicate lots of resources into espionage points and spies and it takes time - traveling and waiting in the cities etc, definately not abuse. Much less abusive than selling workers and monopolizing techs thats for sure.