Civ3 on one roll of the dice - steal techs

[Jon Speelman]

I presume it wasn't Firaxis's original intention but in my present game a single roll of the dice - whether I can successfully steal a tech - is worth about 60,000 gold and hence I presume the total outcome of the game.

Inspired by Aeson's splendid Iceworld thread "So very cold" (sorry I can't find the exact reference at the mo) I tried Deity on a Huge map with lots of water. My starting island at least had fifty percent greenery but it seems hard enough to me even so, especially after Firaxis kindly upped the science rate still more as compared to 1.21f which I believe he was using.

If I've understood the science calculation at CivFanatics (goi to their FAQ and then there's a reference from the science) correctly then originally the base rate for science was charged at Deity level at x24 for them and x40 for us - now as of 1.29f it's x40 as compared to x66.66 (400/6).

The discount has also been reduced. You used to multiply by approximately (1 - 2N/3L) where N is the number of civs you've met which have an advance and L is the total number left including yourself - now it's (1 - 4N/7L)

Anyway, after considerable work I got a position (iI'm afraid that even affter zipping it's bigger than 500K so I can't append it here) in which Greece was the tech leader with Space Flight extra and France, Russia and China had Computers and Fission.

I saved up for 7 or 8 turns, paid my 5634 gold to steal "safely" and lost. The result was that I had 630 gold was making 692 per turn and was at war with Greece. Greece is v rich with about 150K so if I can manage to capture one of their island cities I may get a few K back but I'll certainly have to wait some moves before trying to steal again and it'll probably be too late then (somebody else will have got Space Flight): essentially it seems to me to have been a one shot bet and I can probably resign - not that I'll get wiped out but I'm v unlikely to be able to win.

So I rewound, waited one move and tried again. This time I scored (does anybody know the exact probability of success when stealing "safely" from the tech leader).

The result after trading with China, Russia and France was that I had 15,781 gold
897 per turn plus 3(!) extra techs - Space Flight, Computers and Fission; and Aluminium from the Russians which I guess is worth a good 2K.

Taking the techs as Space Flight 19 K (ie about 8/9th of 66 x 320 and the other two as say 14K each (3/4th of 20K ish), Incidentally in 1690s just to check I tried the Greeks at the full 15,781 for Satellites and they would be "insulted". The total difference is well over 60K on the one roll which seems ridiculous to me.

I haven't thought yet about how the espionage system could be improved but surely this wasn't the original intention?

[ChrisiusMaximus]

Dont really understand the maths but I agre that espionage is virtually useless in civ3 just getting the spy planted in the first place is risky and if you fail then reload play a turn try again it can work.I have read here somewhere that they didnt want spys as powerful as they were in civ2 well they certainly have not this time their grossly underpowered to the point where its really not worth bothering with.Although I still use them.

[Dis]

never use them. except maybe to plant a spy.

[vmxa1]

Once I build the IA and planet spies, that is the end of it. I tried them a few times, butwith little success and much expense, so I ignore it now.

[Jon Speelman]

Actually tech is so expensive now that it seems forced to steal it, risky though it is - as I was playing
(Deity Huge board) the Modern ones seemed to be worth about 20K, God help us, especially if they were necessary for the Space Ship - while stealing still sayed
at less than 6K.

But I still don't like the system at all - it was a lot easier - and also more fun - in Civ2 when the level bonuses are in production and happiness but techs remain quite afforable.

[jshelr]

Jon

It might be my imagination, but I've actually felt that the human civ becomes more competitive in research late in the game because the AI civ fails to build the economy as well. Do you use the AU mod which tried to fix that flaw? If not, you could be approaching research parity with the AI by the late modern era, even on Deity. You are certainly correct that the gold price of stealing is the same, but the price of having to change government in the modern era after going to war on getting caught stealing is overwhelming. Can't afford that. Wouldn't be prudent at that juncture.

[Jon Speelman]

jsheir

It's probably because I didn't play the opening well enough. Actually I misrepresented my island - it was all grassland and forest (the half and half was the previous disaster) but I had no natural resources or luxuries except for one lot of horses until the discovery of rubber - though once I had map making there was iron fairly nearby across the water.

At some point in the early Middle Ages I wickedly double crossed the Egyptians and from that moment onwards until the end of time they all refused to sell me any tech on credit - which seems like sloppy and/or malicious programming to me.

I was playing catch up thoughout The Industrial Ages and although I was Persia (I should have stated that earlier I guess) I entered the Modern Ages two behind the leaders even after my bonus of Rocketry.

After rewinding and stealing from the Greeks - which seemed necessary if I didn't want to resign if a bit unprincipled - I tried later with another tech and was able to take my medicine then. By paying all of the rest of the world to gang up on Greece I managed to remain in a Republic for the full 20 turn war before agreeing peace with them - but even so some of the other Civs, notably China, were so big that they were researching quicker than me under Communism.

In the end - after I successfully stole Synthetic Fibres - there was a space race which was approximately a "score draw" - I was rushing it because I had to go out into the real world (and wanted to finiszh before real life intervened) and they sabotaged my last component two moves before completion. I rewound and won by a tempo by exposing the Chinese Spy who had done the dastardly sabotaging but I guess that if I'd taken it slower in the preparation for the race then I would have been a few tempi ahead. I finished up getting the rocket as the Chinese, now at war with me after I failed to expose another spy , smashed through my three cities on their huge continent.

Anyway, it's v tough work and I shan't try again for a while but I'm sure the lure of the challenge will get to me in the end.

[Demerzel]

it seems to me when playing that stealing "cheaply" and stealing "safely" has the same chance of success for your buck. An attempt to steal cheaply works 95% of the time whenever a "safe" attempt would, and for the price saving its not worth bothering with the options other than steal cheaply.

This really does limit the espionage game to a single dimension and really makes you wonder whether they could not have found some way to blunt the overpowering spies of Civ2 and implement that system.

As it is, I've reduced the base cost of stealing techs from 10 to 7 in the attempt to make stealing tech somewhat more useful regardless of the seeming 99% chance of failure when trying.

[Jaybe]

Originally posted by Demerzel
...
As it is, I've reduced the base cost of stealing techs from 10 to 7 in the attempt to make stealing tech somewhat more useful regardless of the seeming 99% chance of failure when trying.

Ahhh, here's to the differences in experience.
I have had a 100% success rate stealing techs, and cheaply at that.
OTOH, that was the only time I have attempted it.

[Demerzel]

One time success does not a trend make

[player1]

As far as I know, when you still safely steal tech there is no chance of detection (as repuattion killer).
Only chance to fail mission (and by my XP, it's probably less then 15-25%)


Needeless to say, from 1.29f patch, prices of tech buying are higher, but spy missions stayed the same.

And needeless to say, with my own MOD reduced steal tech cost to 70%, I had great succes.

Not to say, that it became BORKEN.

For example I could steal late industrial tech for 1500gp, or buy it for 3500gp. In modern it was 2500 vs 7500gp.

I stole at least 5-6 techs from enemy civ.
And saved a lot of money.

The trick was that when only ONE Civ knows the tech, it will be pretty expensive.

If more Civs knew it, it just doesn't pay to steal.


P.S.
I think I'll lower steal tech mission cost in my MOD to original values.

[player1]

Of course the REAL probelm with spying is SPY PLANTING.

It's cheap but fails wery often.
(that descourages a lot of people)

If it could just have risky & safe (for more gp) options, it would be much better.


Anyway, presonnaly, I CHEAT when planting spyies.
If I fail, I reload and try in next turn.

[Theseus]

PF, I did the same, reload on spy plant failure, until about a month ago. It just pissed me off. I have since decided to live with the results... which sucks. I won't try until I'm ready for war, and if I fail, I conduct the war without unit knowledge... which sucks.

Democracy and Communism, stay away. All others seem to work easily.

[Jon Speelman]

Of course you can steal tech with a diplomat rather than I spy (I don't know the relative probabilities of success and costs and guess nobody else will exactly?). But if you fail and they declare war then until you can get a spy in you can't try to steal again even if you have the cash available after the failure.

[Jaybe]

At least you can still attempt to plant a spy into the civ during your war.

[Jon Speelman]

Jaybe Yes I've always thought that the best time to plant a spy is precisiely when you are at war - again I wonder if anybody knows if the probability of success is affected by being at war?