Debunking the Chieftain Research Myth

[nbarclay]

There is an old myth (or at least I believe it's a myth) going around that playing on the lowest difficulty levels results in a dramatically lower research rate than playing on higher levels. I've never put much stock in that belief because while low difficulty levels do mean that a good player has to do almost all of his own research rather than trading to AIs for techs, there are several offsetting advantages:

1) Settling or conquering enough land to build a large, highly productive empire is a lot easier.

2) More citizens are born happy, so less income has to be diverted away from research for the luxury slider.

3) Tech costs are cheaper and corruption is a little lower, so an empire of the same size can research more quickly. That's not important when a civ is on a 4-turn research pace, but it helps until then and during the times when a civ would be temporarily thrown off a 4-turn pace on the higher difficulty levels.

4) Playing on lower levels makes it a lot easier to build a lot of wonders, and perhaps to capture other wonders.

In order to test what is possible, I've decided to do something completely absurd and play a game on Chieftain. Keep in mind that I'm normally an Emperor/Demigod level player, so the only real challenge in this game will be to see how well I can do in research. (And I may stop playing and merely extrapolate how long it would take to finish researching the entire tech tree once it is clear that I could stay on a 4-turn pace for the rest of the game.)

Game parameters:

- Huge map, 16 civs, 80% water. A huge map has slower research, at least initially, and 16 civs and 80% water are designed to make it a little harder to grab enough land for fast research quickly. (Also, 80% water may help keep the game from taking quite as long to play.)

- No barbarians, so there won't be any techs from huts to speed things along.

- Random age and climate.

- Random civ, but I'll restart if I get a Scientific civ or an Agricultural civ on a river. I'll also restart if I get a starting position with a wheat or cow on grassland.

I ended up with Carthage, which is Seafaring and Industrious, and a fairly decent river start shown below.

[nbarclay]

And here's the 4000 BC save, in case anyone might actually be interested in it.

[nbarclay]

I decided to move due east (“6” on the keypad) before settling. My worker started irrigating the flood plains, and then moved on to the next one. My capital started work on a barracks as a prebuild for a granary, and I started researching Pottery.

My research into Pottery was finished in 3250 BC and I then started research into Writing. On Emperor level, I would expect to be showing 50-turn research when I first start, and I’d have to do well to get it down very much below a 50-turn pace. But on Chieftain, with only one city and three population, the display was showing a 36-turn research pace when I first started.

I built a granary and then a settler, and discovered, to my good fortune, that the place where I wanted my second city had wheat. That meant I'd have two fairly good (albeit not first-class) settler pump cities. I then went on to build another worker, a settler, a warrior, and two additional workers. I discovered Writing in 2030 BC, at which time I had three cities. Code of Laws was estimated to take 20 turns, but would presumably take somewhat less as additional cities were built.

I might mention that I used a few micromanagement tricks to help out a little. One was using both flood plains part of the time and only one part of the time, partly to reduce the risk of disease and partly for extra production. Another involved slowing down growth in order to speed up granary construction a turn and grow after the granary was completed instead of before.

Also, there were brief periods when Carthage grew to size 5 and I actually had to use the luxury slider momentarily in spite of the four citizens born happy on Chieftain. But once I get my incense city hooked up, I won’t have to worry until size 6 even without military police.

[nbarclay]

I discovered Code of Laws in 1525 BC and started work on Philosophy. That research took seven turns, completing in 1350 BC. I took Republic as my free tech and began a revolution.

This starting position has turned out to be a bit better than I'd really been aiming for. I'll end up with several settler/worker pump cities, and without the Despotic tile penalty, they'll be mostly very good ones. I also have three luxuries nearby.

The only real down side is that I'm starting fairly close to the southern edge of the map, at least at my longitude. That means I won't have a well-centered capital once my empire becomes large, at least unless and until I get a good chance to move it.

[joncnunn]

Re: Debunking the Chieftain Research Myth

This was indeed the case in the store bought Civ 3 Classic.
An early patch fixed it.

Bascially what was going on was that the AI had a massive tech cost penalty on Chieftain level while the human had a constant cost. (And the AI having a massive reduced tech cost bonsus on Diety)

An early patch reversed it to AI having a constant reserch cost, and the human having a massive reduced tech cost bonsus on Chieftain level. (And the human having a massive increased tech cost penalty on Diety)

A human that normally plays on Emperor level or above should be able to on chieftain level research just about every tech in the game on a small to standard map in 4 or 5 turns each.

Originally posted by nbarclay
There is an old myth (or at least I believe it's a myth) going around that playing on the lowest difficulty levels results in a dramatically lower research rate than playing on higher levels. I've never put much stock in that belief because while low difficulty levels do mean that a good player has to do almost all of his own research rather than trading to AIs for techs, there are several offsetting advantages:

[vmxa1]

Well actually you won't be able to research techs in 4 or 5 turns for quite some time at any level. You just do not have the production.

[vmxa1]

Map of huge is 400 vs std of 240 and the cost facor is 20 chief vs 10 for Regent.

So BW will be twice as many beakers on Regent (76).

Nathan, this can be a great instructional tool. I am not sure what it will prove as far as the myth or lack of same. You would need to take the same map and play it at say Regent to make a comparison. Otherwise all you will prove is that you can research better than others.

You may also need to avoid any trades for tech or even gold. I say this, because you may just be a better trader than others at this level.

Just for your consideration. OTOH, I may be distracted and be full of crap.

[Aeson]

DaveMcW chose Chieftain for his try at a fastest spaceship launch. It was a HOF setup that saw him launch in 570AD. Quite a different setup in that game though.

[Kuciwalker]

Interesting... what's the fastest SS win on sid?

[Aeson]

At the CFC HOF the fastest (and only so far) Sid Spaceship is 1695AD. Could be done faster no doubt.

Sid tech rate is actually somewhat slow. At least in conditions where it's possible to consistantly win. Archipelago starts offer the best chance, and so there isn't as much early trading going on. Tech rate is still pretty fast through the Middle Ages, but Deity/Pangaea is faster. Later the Sid AI's production bonuses eat into their commerce in huge unit support costs. Plus the player almost has to go Monarchy instead of Republic to be able to handle the wars on Sid.

I think a sub 1000AD Spaceship is possible on Sid. How far under I have no idea, but would be suprised if it's anywhere near DaveMcW's date. Generally I'd say Deity would be faster than Sid and pretty close to Chieftain. Deity games are definitely easier to control than Sid, and you can get a lot of help from Deity AI's that you won't get on Chieftain.

[Tattila the Hun]

I'm feeling inferior...

[nbarclay]

Once the new government emerged, it was barely possible to research Mathematics in four turns. Currency took a bit longer and was not finished until 950 BC. My biggest problem with research at that point was that my worker pumps were working a bit too efficiently, so much of my financial bonus for being in Republic was eaten away by worker upkeep costs.

Construction took until 710 BC to research, by which time I’d finally been able to trade for Ceremonial Burial. I decided to go ahead and research Mysticism myself on a four-turn pace to build up gold reserves for other research. With that completed in 630 BC, it was possible to use gold reserves to research Map Making, Literature, and Polytheism in four turns each. Carthage entered the medieval era in 390 BC and could research Feudalism at a 5-turn pace, albeit having to run a deficit to do so.

The REXing situation for an Emperor/Demigod player playing on Chieftain with a good starting position is absolutely unreal. I have twenty-seven cities, compared with three to five each for the four AIs I’ve met thus far. And my tech lead is perhaps even more absurd. If I’d play the game out, I would almost certainly be in the industrial era, and maybe even the modern era, before my neighbors could get out of the ancient era.

[vmxa1]

1695 is snappy, best I have is 1800's for a launch, but I was not trying for time.

Was the 570AD done in C3C, I suspect not and hence can't be compared to Nathans attempt. All aspects of Civ are easier on vanilla and PTW.

Anyway I would be shocked to see any SS sub 1000AD at Sid on a huge map.

Most sid games are designed to try to slow the tech race down as much as possible. Even with a massive empire it is hard to match the KAI in research, because you are likely to be in Monarchy form of government.

I know of one game that was done as a research run by Mountain Sage and IIRC he was in Republic.

[vmxa1]

The huge map aspect is going to make it very hard to want to continue on. So much land to grab and workers to manage.

I think I see 4 settlers of on the NE end in addition to the 27 cities.

[Aeson]

It was a C3C game. Definitely not comparable though, as it was a HOF setup where you play out maps until you get the perfect start. He had an early SGL for the Pyramids, most of the Ancient techs from huts, and it was on a standard map with 8 Scientific AI.

In C3C there is the free Philosophy tech, and the Radio is no longer in the tech tree, so that's at least 8 turns of research cut out of the tech tree. On Chieftain you can still get to a 4 turn tech rate just as quickly as in PtW, and late game Palace isolation in PtW won't do anything that C3C Scientists can't do.

I agree Sid is best approached to try to slow down the tech rate, but that doesn't make it a necessity. In a HOF type setting you can do things that often will lose you the game on Sid because if you do lose you can just start another attempt. You can fight wars to help AI's keep their unit counts supportable or play on a map where the AI's aren't all segregated to boost the tech rate. Normally that will make the game very difficult to control, but in this case if you lose control you can just try again.

"Safe" Sid SS will be much later of course.