Discussion of HOF power plays

[SirPleb]

Originally posted by God
How do you guys catch up to the AI in techs?

An Expansionist Civ can use the goody huts to get a boost in Ancient Times. In my current game I'm playing Militaristic/Religious. As a non-Expansionist with 15 rival Civs I found the goody huts of little or no value at deity level. They certainly don't help much with the tech problem for this kind of Civ.

What I've been doing is taking advantage of the flip-side of the 1.17 tech trading. Let the other Civs get way ahead in tech. Set my research to zero until after building Theory of Evolution and generated cash instead. Most of the time I've happily lagged behind in tech, sometimes way behind. I've picked up tech when I could get it cheaply using a number of techniques:

1) Trading maps. I watched every turn to see when the AIs learned Map Making. On the first turn it was available I traded my maps (which didn't have much ) around the world for communications with everyone, all the tech up to that date, and almost everyone's territory maps.

2) Trading more maps. I continued exploring areas none of the AIs had visited yet. Occasionally it became possible to trade my updated maps for another tech.

3) Beating it out of them. Every early peace agreement I made included getting some tech.

4) Intimidation. After my first successful war there were a couple of AIs which were weaker than I. I renogotiated peace with them once in a while and tried to get a tech thrown in. (Even though they were weaker, with 1.17 trading they all magically got every new tech as soon as anyone knew it )

5) Resource and luxury trading. I traded my extra resources and luxuries. I don't do a lot of this since I don't like to strengthen the AIs. But sometimes there's one it would strategically pay to strengthen. Might as well get a tech for it at the same time.

6) Other deals. I made one MPP which was part of my overall strategy anyway and got the other guy to throw in a tech as well as 2000 gold. In the right situation (they're at war with someone already and have another MPP in place, both of which can work for you at the very right instant) they'll pay nicely for an MPP.

7) When there was no other way, and I REALLY wanted a tech, I bought it at the going rate. I think this only happened in my game twice. Chivalry and Scientific Method. Those are the only times I remember when I wanted an up to the minute tech.

After all that catchup it was another story completely to get ahead and stay ahead, more conventional approach to that. Try to tie the most powerful AIs in knots with wars, drain their coffers, and research like mad.

One other suggestion I want to note: On a continents map there is another crucial trading moment I watch out for. The very first contact between the two continents can be milked for a small fortune, by being the first to trade maps and communications. On a deity continents map I make a big effort to have a scattering of galleys positioned around the continent, hoping I'll be the first to see a ship arriving from the other land. Can't guarantee success of course. One of the local Civs might go sailing first. Or a remote Civ may still contact a local rival first. There isn't much chance at all at deity of me actually making the first contact directly. But it is worth some investment to improve the odds of being the first one contacted.

 

[Aeson]

Something I haven't mentioned, but is very important for the Expansionist game. No barbarians! They will eat up your Scouts even if you're careful. The HOF allows any barbarian settings, so just huts in this case.

Deity huts are less kind, there is some link to difficulty level. I've had awful Deity games where my first 10 or so huts were empty, that's when Ctrl-Shift-Q comes in handy. At least it's something you know from pretty early on. The exploration stage usually takes me an hour or so to complete. When compared to the overall number of hours needed, replaying this hour several times to get the best start makes sense.

I probably started 50-100 maps, played through about 10-20 turns until I could tell the general lay of the land, and how the huts were treating me. About 10 of those maps I played even longer because of promising starts, until I ran into one that just worked out well. My Iroquois game certainly didn't have the best geography of them all, but the timing (settler and contacts) was right.

On anything other than Deity, it's almost a given that you'll make it out of the Middle Ages just from huts. You just have to produce enough Scouts.

As far as getting this working on Deity, it takes a lot of luck. Like I said earlier, the window of opportunity to demand Greek cities was only a couple of turns. They built their 4th city right after I demanded their 3rd. If they had gotten their 4th a turn earlier, I wouldn't have had a powergraph advantage.

SirPleb's notes on tech match mine just about exactly. Only with an Expansionist civ I can beeline a few techs into the Middle Ages and get there a little before the AI. After that I'm almost always trailing. My usual target is Gunpowder, so I can hopefully cut off the AI's supplies before they know they are there. In 1.17f Deity it's almost impossible to get to even Gunpowder first though.

 

[Aeson]

I'm in the milking stage of my current monster game. I ended up deciding to milk this one hard. I really wanted to get past 40000 if possible and that looked debatable if I just ran it off quickly. I've been production-shifting and managing entertainers. It now looks like I may end up around 45,000.

With 15AI no less... What was your score at 10AD? I must admit that my 50k guess was just that, a guess. If you can come from behind without the Expansionist jumpstart and crowded by all 15 AI... 60k for a "perfect" game? The late Industrial and early Modern era's (Infantry without many Tanks) tend to trip me up a bit.

Aeson, for a change of pace between sessions of milking I've tried a few quick starts using your Settler Flood / Expansionist strategy. I hope you won't mind if I end up going a full game with a variation of it, it really is cool!

Keep the saves seperate! Or am I the only one who loses saves like that?

And I don't mind at all, a lot of the things I've been doing in my games are just different combinations of other peoples strategies. I don't know if the Russians will do very well on Deity though, at least not without a good luxury setup. It would take at least 4 or 5 early luxuries to work best, and early Horses. They miss out on the quick workers and quick revolutions/temples, both which help speed up the expansion rate.

I've been thinking about trying the English. With so many cities early on, maybe the Commercial trait and it's reduced corruption would make enough difference. And Orange is cool! I rather think that Commercial won't add as much as even Scientific though.

 

[Aeson]

I was just playing a start to a 1.17f Deity game. Americans on a Huge Warm Wet Pangaea map. Nice starting position, Settler from a hut on turn 6. Met the Iroquois early on, but hadn't built my 3rd city yet. They only had 2 at the time, they had a sizeable powergraph lead.

Met up with the Chinese a few turns later, Babylonians shortly after. Both had an advantage in power and number of cities. Then I hit a string of lucky huts. Up to that time I had gotten about 4 or 5 advances. The next 5 huts were all advances, and even after that it was a steady flow with a couple Warriors mixed in. In 3 turns I had hit the Middle Ages.

I traded my map for the last tech actually, Horseback Riding from the Iroquois. Met the Indians the same turn. Traded maps to everyone for all their gold and maps. Looked at the powergraph the next turn, and I was now ahead of or tied with all but the Chinese! I had built one city that turn, but I had been at about half the power rating of any of the others. The only real differences were hitting the Middle Ages, and getting everyone's maps and gold.

It does say that the power graph is based partly upon gold and advances, but I never thought it was that much! Always before it seemed number of cities and units was the main factor. Also it doesn't seem like my Scouts count. In 1.16f I am certain that they did.

Anyways, hows this look for a start? I've never even come close to reaching the Middle Ages so soon. It kind of hurt getting all that tech so fast though, I actually had Horses nearby to build Chariots with... 8 cities (2 I just demanded).

EDIT: Attachment deleted as it's no longer needed.

 

[SirPleb]

My score in 10AD was 905. I'd finally climbed out of last place, was 13th in score! It sure is a long climb just to get to the middle of the pack in this kind of game. From the power graph in 10AD I'd say that's just about where I was. 5 Civs were significantly weaker, 5 stronger, and 5 about the same. I was at 9th place in land, 8th in population, and 2nd in mfg.goods. (Which means I still had a long way to go to out-produce most of them with their 60% production requirement.) A funny thing I noticed while looking at my 10AD save - I had just 3 workers on the payroll. I'd bought some from rivals earlier on and had just started capturing workers by this date. I did have 31 nice Samurai. Not a killer force, but enough to do some damage as long as I picked my fights carefully. Most of the world had just learned Military Tradition. I didn't have saltpeter but neither did my neighbor England, so that's where my Samurai were heading at the time

I think you're right Aeson, 60k is probably the right area for a "perfect" game. I'm at 1906 and my per-turn score has nearly stabilized at about 93,000. I would guess that a perfectly milked huge 60% water map could reach a per-turn value well over 100,000. A perfect game might reach the maximum between the 1/2 and 2/3 mark? (I.e. between 1400AD and 1790AD.) The per-turn scores up to the max for a game aren't linear of course, they are a pretty harsh curve. That makes it really hard to guess what the max score could be. My gut feel is that somewhere around 60k is probably right.

Your settler floods might flatten the score growth curve noticeably and could be a key to even higher scores

I am starting to think that it won't ever work to factor an early win bonus into a maximum score. With a settler flood approach the ramp-up of score could be somewhat flatter. But even if it were possible to reach Domination at say 10AD (???!!!) I don't think the scoring potential would be there. Bonus at that date would be 12,000. As a Domination win it would of course be at its max per-turn territory score component. But I doubt there's any hope of getting a substantial population by that kind of date. And the population component would account for (very roughly) 1/2 of a perfect score, well over the 12,000 bonus. It seems likely that there's no cross-over in the curves at any later date, that the ongoing averaging of a subsequently max'd out population would continue to outweigh the rapidly diminishing early win bonus.

Your latest American start sure looks awesome! Quite amazing! Going to see if you can break that 60k barrier now that we've established it?

 

[Grey Fox]

Originally posted by SirPleb

I think you're right Aeson, 60k is probably the right area for a "perfect" game. I'm at 1906 and my per-turn score has nearly stabilized at about 93,000. I would guess that a perfectly milked huge 60% water map could reach a per-turn value well over 100,000.

How do you calculate the Value? And that's the tur-score right?

By counting all territory and citizens? I know that's not the case as you have a Calculation in your Ecxel file (please spare me from checking that... ugh...).

Originally posted by SirPleb

Your latest American start sure looks awesome! Quite amazing! Going to see if you can break that 60k barrier now that we've established it?

Yea, it sure do. I always thought Americans would make the best civ for Huge maps. Both their Traits are Better when the map is larger. And Best at Huge ofcourse.

You can get an very late Golden Age too. With Rocketry(And a successful win with an F-15, can be hard) or the following Combinations of Wonders:

Great Lighthouse + Pyramids
Great Lighthouse + Hanging Gardens
Great Lighthouse + Great Wall
Great Lighthouse + Hoover Dam
Collosus + Pyramids
Collosus + Hanging Gardens
Collosus + Great Wall
Collosus + Hoover Dam
Seti + Pyramids
Seti + Hanging Gardens
Seti + Great Wall
Seti + Hoover Dam
Copernicus Obs. + Pyramids
Copernicus Obs. + Hanging Gardens
Copernicus Obs. + Great Wall
Copernicus Obs. + Hoover Dam
Magellans + Pyramids
Magellans + Hanging Gardens
Magellans + Great Wall
Magellans + Hoover Dam

I have a G.Age in the GOTM @ 1866, I think it started 1862. With the completion of the Hoover Dam. (I had captured the Colossus before...).

Anyway, something strange in my game. Well it's not strange really, but a totally corrupt city 28 Tiles(counting horizontal as 1.5) from the Capital, has a production of 2 (!) when discarding the Waste. It's a 6 size city with 6 working, and it gets 22 Shields when counting even the Wasted ones.

 

[Aeson]

I keep flip flopping over whether to use the Iroquois or Americans. I'm back to thinking cheap temples and quick revolutions outweigh the fast workers. Also the Mounted Warriors are much more valuable than the F-15! I keep trying new things out in the games I've played, too much reloading to really submit any of them. Especially reloading to figure out exactly when and what to demand from the AI. Still gotta work some wrinkles out.

In 1.17f it will be key to keep track of the AI's population in potential demanded cities and when they are built. Otherwise I keep getting these hopelessly unhappy cities which really don't add anything to the cause.

You're right that the early win bonus would never really be a factor in maxing score. Even at 3950BC the 36060 score isn't big enough.

As far as BC scoring goes, the difference between my score of 3300 and yours of 900 only makes about 560 points of difference in overall score. The main scoring improvement would be in reaching the max score per turn earliest. It's hard to predict just when the curve takes off, and how many turns of max score it would add.

 

[Aeson]

Anyway, something strange in my game. Well it's not strange really, but a totally corrupt city 28 Tiles(counting horizontal as 1.5) from the Capital, has a production of 2 (!) when discarding the Waste. It's a 6 size city with 6 working, and it gets 22 Shields when counting even the Wasted ones.

Is this with an Industrial Civ? I know they get a bonus production, probably at size 12 though. I think the bonus is applied to the city tile, and thus isn't affected by corruption.

 

[SirPleb]

Originally posted by Grey Fox
How do you calculate the Value? And that's the tur-score right?

Yes, that's the per-turn score. Here's how I worked out how to calculate it. (Skip to the formula at (5) if you don't care why it works.)

Definitions:
S(x) is the game score at turn x
P(x) is the per-turn internal score for turn x.
T(x) is the sum of all P(y) with y from 1 to x, i.e. the total of all per-turn scores up to and including turn x.

Assumption, the game score is the average of the per-turn scores, i.e.:
(1) S(x) = T(x) / x

Say we want to calculate the per-turn score at turn "n". By noting the actual game score at the turn before and the current turn we know the values of S(n-1) and S(n).

Since T(x) = T(x-1) + P(x) by definition, we can derive from (1):
(2) S(n) = ( T(n-1) + P(n) ) / n

Also from (1) we can derive:
(3) T(n-1) = S(n-1) * (n-1)

Substituting (3) into (2) we get:
(4) S(n) = ( S(n-1) * (n-1) + P(n) ) / n

And finally solving (4) for P(N) we get:
(5) P(n) = S(n) * n - S(n-1) * (n-1)

(5) in English : On two consecutive turns, multiply the turn number by the score. The difference between those two calculations is the second turn's per-turn value.

Of course there's some inaccuracy because we aren't given S(n) and S(n-1) with much precision, we just get integers. But it gives a pretty good idea.

Having done that, it wouldn't surprise me at all if there's an easier way to have worked it out

The spreadsheet I posted which does it has been replaced by an easier to use HTML/Javascript page which is available at http://forums.civfanatics.com/showthread.php?s=&threadid=17550

 

[God]

Well my big regent game ain't going to well. I'm leading at the moment, but the Chinese have had a surge in cities. I guess mainly because they got alot of good land in the south and pretty far from other civs while i was squashed between Greeks, Romans and Babs.
Right now i'm playing as the Egyptians who currently are my favorite civ.
The problem is that i just got into middle ages, we have entered AD times. Tech wise i'm good because of the great library which let me get a bit lazy. I was thinking of attacking and destroying the romans and getting some nice land but i realized that somebody had given them iron so i backed off. Warring with China will take too long and i can't beat hoplites at the moment.
I'll probably go after the Babs. They are weakest and with knights i can take them out.
I need atleat 2500 to submit to HOF which i can reach only if i manage to take half of my continent which means taking out the Greeks, Romans and Babs.
Babs will fall easily, and romans wont' be too difficult if i can mangage to take their saltpeter if they got any.
Then i'll wait for cavalry and then replaceable parts and get infantry and artillery and destroy all ruber sources for the greeks and take them out.
On the long run the Greeks are in a bad starting spot. They started in the tundra region and are trying to expand south but i'll get in their way.
There's a massive jungle to the south of my frontier cites which i plan to inhabit. I'm really glad i have industrious to chop down that jungle.
Oh well i hope i can get Sun Tzus and Leonardo's and set some city for Bachs. Then also try to get Copernicus if possible and Newtons. Sistine is not a big deal since i never bother to build to many cathredrals(waste of money). Smith's Trading is nice but Wall street will do. Theory of Evolution is essential to get replaceble parts, atomic theory and sanitation which can give me a big boost.

I'm looking at a score of 3000+ hopefully. Try spaceship but most likely end up with cultural.

I'm still not sure which civ is better, Egypt or Iroqoius. IMO they both have the best traits. Religious is amazing and i like industrious but expansionist is so attractive alongside those mounted warriors to quickly demolish a few civs.

 

[Aeson]

Those formulas give me a headache SirPleb! I have a hard time thinking in terms of one letter variables.

I think I've been doing it almost the same way, just with the subtraction first and then multiplying. Your's is probably more accurate.

CurrentScore: histograph score for current turn
PreviousScore: histograph score for previous turn

This is the equation I've been using:

TurnScore = [(CurrentScore - PreviousScore) * TurnsPlayed] + CurrentScore

SirPleb's equation in my language:

TurnScore = (CurrentScore * TurnsPlayed) - [PreviousScore * (TurnsPlayed - 1)]

 

[SirPleb]

I do tend to degenerate to single letters when I do algebra

I think the equation "TurnScore = [(CurrentScore - PreviousScore) * TurnsPlayed] + CurrentScore" is indeed slightly off. Consider the case where there have been just two turns, first turn internal value 10, second turn 20. So CurrentScore = 15, PreviousScore = 10. That calculation would work out the current internal value as 25. As TurnsPlayed gets large the error becomes quite small I think. I think what it is meant to be is "TurnScore = [(CurrentScore - PreviousScore) * (TurnsPlayed - 1)] + CurrentScore".

Funny thing, I had a similar slight error but I think it worked in the opposite direction in my Excel approach. I think I got it right in the HTML version at last

 

[Grey Fox]

You should both come to the chat more often!

I would love to discuss more freely with both of you, and I would like to listen in on your discussions.

So come to the CFC chat, very easy to do... you don't even need an IRC client if you don't have any. Just click on the Chat button on the main page.


Hope to see you there!

 

[Aeson]

I think what it is meant to be is "TurnScore = [(CurrentScore - PreviousScore) * (TurnsPlayed - 1)] + CurrentScore".

You're right SirPleb. My original formula would always be off by the increase from last turn, plus any rounding error. I was multiplying the increase by all the turns, but then adding in the CurrentScore which already contained the increase for that turn, counting it one too many times. It could also read:

TurnScore = [(CurrentScore - PreviousScore) * (TurnsPlayed)] + PreviousScore

Rounding still throws things off. Mine would be subject to rounding error for the increase, which would be multiplied by turns played, then have the rounding error for the current histograph score added. Yours would be subject to rounding error for the current histograph score, multiplied by turns played, plus rounding error for the last histograph score, multiplied by turns played - 1. When dealing with whole numbers, both should be right I think.

 

[Aeson]

What time is the chat room most inhabited?

 

[Grey Fox]

Originally posted by Aeson
What time is the chat room most inhabited?

American and European Evenings/Nights.



I've been learning to use Infantry units a little better, by not building Tanks in this GOTM, I used only the combination of Marines + Infantry + Artillary + Field Spy's (As I like to call the Explorers, would you all be interested in a little New graphic and chance in the Bic file to use that name and grapichs, if I made em?)

 

[Aeson]

I think I found my map.

- Wines just a tile north of my capitol.
- Settler from a hut in 3800BC.
- Horses appeared 2 tiles away from my starting point when the wheel was aquired.
- 23 cities in 1275BC, 7 of which were demanded from the AI.
- Was able to hold off on getting Horseback Riding until 20 Chariots had been built.

 

[God]

@Aeson: I see how closely congested you build your cities. i realize this means less corruption and fast production early in the game, but what will you do into the Industrial times when your pop is going pretty high?


Anyway i've noticed that the Iroqoius always have horses near where they start.

Oh well my monster regent game didn't go to well. I had about 2000 score about to enter industrial era in the 1300's but China declared war on me, at a really bad time and before i could get musketmen their cavalry hit alot my southern cities and got frustrated so quit.
Oh well the game was going really well until then. I could probably have got 4000 points at the rate i was planning to invade the continent, and then milk it to 2050.

Started another game with same conditions, same civ, Egypt, and this time i started on an island with Germans and Babs. I managed to get both pyramids and great library in this game, and at the moment i only know 2 civs. but from demographics i figured out that i'm 1st in pop, area, and a number of other big things. I'm around 300 Ad and entering the Middle Ages. I'll invade Germany soon with swordsmen and take their land which is pretty good and close to my capital. Then the Babs with knights and longbowmen.
I plan to own the whole continent which seems to be a mid-sized continent on this world. Then probably go for spaceship. I probably take a few islands but leave the mainland.

BTW- @Aeson, SirPleb and others, i'm finding this topic really helpful even though i'm playing at regent. You guys have alot of interesting techniques, and i'll probably use them more on Monarch.

 

[SirPleb]

At last, I've finished my current game. Final score 44313. No surprises in playing it out. The remaining Persian town stayed VERY quiet. I didn't work in all six victory conditions this time. Didn't get Diplomatic. I set up the UN at an appropriate date in case I got an easy shot at it but the opportunity didn't come up and I didn't go out of my way (in the quest for score) to get it. So it didn't happen. It is possible to vote for Xerxes in 2050 and lose the game though. The other five victory conditions are all there.

Your latest start looks awesome Aeson, great potential in that one!!

 

[Aeson]

Congratulations on finishing your game SirPleb!

I haven't played too far into my game, but it's looking good so far. Just narrowly avoided a painful situation this morning, and decided to make a history of it.

The Japanese Iron War

I'm working on a write-up of the American "invasion" that happened shortly thereafter. Worked out very well for me...