Aspiring Quartermaster
- Thread starter SpiffyKeen7744
- Start date
zerksees
in remission
39,144 is a good score, and should last a long time. Nice work!
Bartleby
Remembers laughter
Congratulations.
SpiffyKeen7744
Prince
Thanks everyone. This HOF community is a lot of fun thanks to all y'all.
Congratulations! I always thought this is a good community, too.
SpiffyKeen7744
Prince
It has been a long time, but I have been playing PBEM games, so I was not totally Civ3 free.
My Quartermaster score and rank have been dropping, and I feel like playing a few games. Top on my list of bad scores are Warlord (7), Demigod(5), Regent(3), Large(5), Huge(3), Small(2). So I looked at the tables, and I found some that seem interesting. Some look like I could easily get 2 or 3, but will also give me a chance at #1.
Huge, Warlord, Domination - 280AD, 900AD, 1762AD
Large, Regent, 100k - 1375AD, 1405AD, 1846AD
Small, Demigod, Spaceship - 1385AD, 1575AD, 1590AD, 1788AD, 1812AD, 1973AD
I don't know what to try for first... perhaps I will start with the Warlord game, since it is domination. Keys are fast growth (Argriculture) and a fast early tech pace (Expansionist): Inca anyone? I can hope for a SGL for ToA. I want to capture cities, not destroy them, so I will try to choose civs that build culture, and I will gift the civs government techs. Also, I should choose civs that I can start their GA early for faster growth on their part. Anything I missed?
My Quartermaster score and rank have been dropping, and I feel like playing a few games. Top on my list of bad scores are Warlord (7), Demigod(5), Regent(3), Large(5), Huge(3), Small(2). So I looked at the tables, and I found some that seem interesting. Some look like I could easily get 2 or 3, but will also give me a chance at #1.
Huge, Warlord, Domination - 280AD, 900AD, 1762AD
Large, Regent, 100k - 1375AD, 1405AD, 1846AD
Small, Demigod, Spaceship - 1385AD, 1575AD, 1590AD, 1788AD, 1812AD, 1973AD
I don't know what to try for first... perhaps I will start with the Warlord game, since it is domination. Keys are fast growth (Argriculture) and a fast early tech pace (Expansionist): Inca anyone? I can hope for a SGL for ToA. I want to capture cities, not destroy them, so I will try to choose civs that build culture, and I will gift the civs government techs. Also, I should choose civs that I can start their GA early for faster growth on their part. Anything I missed?
AutomatedTeller
Frequent poster
You might want to have agri civs for your targets over culture ones.
SpiffyKeen7744
Prince
So I looked at the Annex, and I found that Alpha was LARGE Warlord Domination. This is almost the same. Killercane won around 500BC, so it should not be unreasonable to beat 280AD. In any case, it looks like key points are
1. Pop early settler
2. Early SGL for Pyramids
3. Build ToA either SGL or hand as soon as possible
4. Get settlers out as soon as possible, and workers for roads.
5. Moderate units and conquest, just enough to capture some AI cities.
6. Gift cities to far away civs, reconquer, grab far away small cities from peace???
Is #6 allowed?
Also, I should think about starting wars so that AI civ's GAs can start. Also, when should I start my GA? I should try for Monarchy... so I should be able to pop some huts, and enter my GA after government change without losing GA turns at the end? Or is it better to get the VERY EARLY GA for faster settler factories? I'll have to see which works best. What I should look into is 80% Pangea domination limits.
1. Pop early settler
2. Early SGL for Pyramids
3. Build ToA either SGL or hand as soon as possible
4. Get settlers out as soon as possible, and workers for roads.
5. Moderate units and conquest, just enough to capture some AI cities.
6. Gift cities to far away civs, reconquer, grab far away small cities from peace???
Is #6 allowed?
Also, I should think about starting wars so that AI civ's GAs can start. Also, when should I start my GA? I should try for Monarchy... so I should be able to pop some huts, and enter my GA after government change without losing GA turns at the end? Or is it better to get the VERY EARLY GA for faster settler factories? I'll have to see which works best. What I should look into is 80% Pangea domination limits.
killercane
Deity
Yeah its allowed as far as I can remember. Fun game as I recall, I think I had 55-60 cities by 1000 BC!
Just to confirm killercane's reply that this is allowed.
Good luck with this. It's good to see you back here!
SpiffyKeen7744
Prince
Thanks everyone! I hope to get back into the HOF rhythm.
In any case, my first few tests were all failures. In some cases I could not get a free settler, I got a free town in the worst possible place, or I only had the one starting cow to use.
I did have one that looked promising, but I ended up being attacked by the Byz with all my troops out hut popping. I had a free settler in a good place, lots of cows spread out corresponding to good city placements, and I was getting settler factories up and running with forest chops for granaries. However, even without the Byz attack, I did not get a SGL. In fact, I played about 10 games without one.
I will probably try again tonight. From trying some games already, I have a few new considerations for this type of game. First, the Inca cannot build scouts, but only their UU, 20 shield, 1/1 scout. I don't think is a huge problem, but I cannot pop huts as fast. Second, a SGL would be awesome with CB. Even with chops it takes a while to build granaries. Third, I need to build some warriors for happiness and defense. There is a risk, even at Warlord, that an AI civ may stop by if I am weak versus them.
One other thought had to do with techs from huts. I really don't think I am doing this correctly. So I looked up SirPleb, as is often a good idea, and he says that he leaves some huts un-popped near home and left some scouts near huts at crucial times. He also mentioned his research order: CB, Myst, IW (at 0, so that he could pop writing with a higher probability), and then Philo. I need to try that too. Is there any information out there about WHICH techs are likely to be popped? I know that the one you are researching cannot be gained, and that only AA techs can be popped. Anything else?
In any case, my first few tests were all failures. In some cases I could not get a free settler, I got a free town in the worst possible place, or I only had the one starting cow to use.
I did have one that looked promising, but I ended up being attacked by the Byz with all my troops out hut popping. I had a free settler in a good place, lots of cows spread out corresponding to good city placements, and I was getting settler factories up and running with forest chops for granaries. However, even without the Byz attack, I did not get a SGL. In fact, I played about 10 games without one.
I will probably try again tonight. From trying some games already, I have a few new considerations for this type of game. First, the Inca cannot build scouts, but only their UU, 20 shield, 1/1 scout. I don't think is a huge problem, but I cannot pop huts as fast. Second, a SGL would be awesome with CB. Even with chops it takes a while to build granaries. Third, I need to build some warriors for happiness and defense. There is a risk, even at Warlord, that an AI civ may stop by if I am weak versus them.
One other thought had to do with techs from huts. I really don't think I am doing this correctly. So I looked up SirPleb, as is often a good idea, and he says that he leaves some huts un-popped near home and left some scouts near huts at crucial times. He also mentioned his research order: CB, Myst, IW (at 0, so that he could pop writing with a higher probability), and then Philo. I need to try that too. Is there any information out there about WHICH techs are likely to be popped? I know that the one you are researching cannot be gained, and that only AA techs can be popped. Anything else?
AutomatedTeller
Frequent poster
There is a random bit to it, but in general you'll pop the least expensive that you don't know.
RFHolloway
Analyst in the UK
or aren't researching - so if you have a hut you can block the cheapest tech by switching your research to it.
And, this might be too obvious, but you want to pick philosophy as your project as soon as you get writing or you can easily pop it and have whatever you left as your current project complete with no choice of free tech. It may be flawed, but my strategy is to research code of laws as soon as writing is popped, not pop anymore while that's finishing, and then pick philosophy and go back to hut popping, but this is on the low levels. I'm sure that won't work higher than regent or monarch.
AutomatedTeller
Frequent poster
that works on higher levels, in some circumstances. I've pulled off the slingshot on demigod before.
On very large maps, I will research philo and hope to pop CoL. On smaller maps, you can run out of huts before getting CoL.
On very large maps, I will research philo and hope to pop CoL. On smaller maps, you can run out of huts before getting CoL.
SpiffyKeen7744
Prince
Well, I decided to stick with a game that did not give a SGL. I did get a free city, but the big bonus for the map I finished was the ability to set up settler or worker factories in 5 cities, and if I had been more patient, I could have had at least one more. After the first turn I knew that this would be a pretty good one:
A few turns later, I received my free city from a hut. It was farther from my capitol than I wanted, but it was in a good place:
Early on I worked on two things. First, I set up settler/worker factories. Second, I popped huts to get the Monarchy slingshot. I ended up having to stall with 0% research on Philosophy until I found the last remaining huts on the map. I got Polytheism on the second to last one I found. Once I was in Monarchy, I started building the ToA in Machu Picchu. That was going to take over 40 turns at first, but I mined some hills and then entered my GA going against the Chinese. The final build time was probably around 20 turns.
While building the ToA, I was in the process of building a long road north into empty desert and tundra space near France. In the west, I went against China, and to the south I went against the Maya. Only later did I start building roads south:
As it turned out, I did not put enough effort into war. I tried the tactic of gifting cities, recapturing, and then peace, but it only got me 1 city each from the Germans and the Ottomans. If I ever play a Huge Warlord Domination again, I will build more workers earlier and road past opponent lands, and I will possibly even go for Knights. With popped technologies, it might be faster then horsemen and swordsmen.
I ended up getting Domination in 70 BC:
A few turns later, I received my free city from a hut. It was farther from my capitol than I wanted, but it was in a good place:
Early on I worked on two things. First, I set up settler/worker factories. Second, I popped huts to get the Monarchy slingshot. I ended up having to stall with 0% research on Philosophy until I found the last remaining huts on the map. I got Polytheism on the second to last one I found. Once I was in Monarchy, I started building the ToA in Machu Picchu. That was going to take over 40 turns at first, but I mined some hills and then entered my GA going against the Chinese. The final build time was probably around 20 turns.
While building the ToA, I was in the process of building a long road north into empty desert and tundra space near France. In the west, I went against China, and to the south I went against the Maya. Only later did I start building roads south:
As it turned out, I did not put enough effort into war. I tried the tactic of gifting cities, recapturing, and then peace, but it only got me 1 city each from the Germans and the Ottomans. If I ever play a Huge Warlord Domination again, I will build more workers earlier and road past opponent lands, and I will possibly even go for Knights. With popped technologies, it might be faster then horsemen and swordsmen.
I ended up getting Domination in 70 BC:
SpiffyKeen7744
Prince
Looking back at the game, I was producing settlers in my core that could not get to where they needed to be at the ends of the world. What I really should have been doing the whole game was fighting along lines that would give me nice lanes across the continent so that I could route settlers that way (see the pink lines in the picture). I managed to do that perfectly to the north, and I did it to a certain extent to the east. However, the land mass was larger in the south. Because there was jungle, I waited to build roads to the south. I should have made a similar line of roads through the south desert once past the jungle. There was a large region (see the red oval) that nobody ever settled, and I attempted to conquer the southern civs (Ottomans and Germans) too late. I would guess this is much like the "conquest will take as long as the civ farthest away" idea. Save the nearby tasks for last.
SpiffyKeen7744
Prince
Well, next up for me will be a Small Demigod Spaceship. I have started a game that looks promising, but I could not avoid war just before 10AD, since the Germans decided to fight. If I survive that, I should be able to do it.
However, another interesting tidbit came to me from a post in Tone's thread having to do with histograph score prediction. I will put my thoughs in further posts, and perhaps link to it over there.
However, another interesting tidbit came to me from a post in Tone's thread having to do with histograph score prediction. I will put my thoughs in further posts, and perhaps link to it over there.
SpiffyKeen7744
Prince
I had been thinking about this whole "predict my huge histograph score" dilema, and I came up with the following thoughts. I know that SirPleb (I believe) made a predictor tool that works almost perfectly. However, it works only after you reach max territory AND max population.
I think the best way to predict before reaching max population has to do two things:
1. You need to predict the internal score once max population is reached for your map. This can be done using domination limit and information from past games.
2. You need to predict the date at which you will reach max population. This is a little bit harder, but you only have to be close, not exact.
Moonsinger's thread about the map finding gods helped a lot with my ideas. The graphs that she gave were very insightful. The first one showed exactly why SirPleb's calculator works so well. Looking at the following graph (I modified it), there are two stages for "score increase per turn" during the game. During the first stage (A) the change in not very predictable. However, once max population is reached (B), the score increase per turn goes down at a predictable rate. This predictable rate allows for SirPleb's tool to work so well. In Moonsinger's 88k game (yellow) she reached max population around turn 240. In the two 80k games (blue and pink), max population was reached later, around turn 312.
The information that is more important, I think, is internal score. If you plot the internal score for a game, the final game score is the area under the curve divided by the number of turns (540). This area under the curve can be estimated rather easily, by simplifying the part of this curve that increases (A) and then using the duration at max internal score (B).
My concern for this technique is that it is not a precise prediction at all. At Sid or Deity, with a x8 or x7 modified, it could be off by thousands of points simply by a misprediction about triangle for the A region. In the case of the 80k games, around 80% of the score came from the milking part after max population was reached. If max territory and max population is reached later, that percentage changes. Also, using domination limit, you have to know the average number of citizens that will be supported per tile of usable land, which is only 66% of the domination limit.
I made a script that predicts this sort of thing, but I have only tested it against a few HOF Huge Histographic games. I would need to test it more before saying that it is a good technique to predict score.
I think the best way to predict before reaching max population has to do two things:
1. You need to predict the internal score once max population is reached for your map. This can be done using domination limit and information from past games.
2. You need to predict the date at which you will reach max population. This is a little bit harder, but you only have to be close, not exact.
Moonsinger's thread about the map finding gods helped a lot with my ideas. The graphs that she gave were very insightful. The first one showed exactly why SirPleb's calculator works so well. Looking at the following graph (I modified it), there are two stages for "score increase per turn" during the game. During the first stage (A) the change in not very predictable. However, once max population is reached (B), the score increase per turn goes down at a predictable rate. This predictable rate allows for SirPleb's tool to work so well. In Moonsinger's 88k game (yellow) she reached max population around turn 240. In the two 80k games (blue and pink), max population was reached later, around turn 312.
The information that is more important, I think, is internal score. If you plot the internal score for a game, the final game score is the area under the curve divided by the number of turns (540). This area under the curve can be estimated rather easily, by simplifying the part of this curve that increases (A) and then using the duration at max internal score (B).
My concern for this technique is that it is not a precise prediction at all. At Sid or Deity, with a x8 or x7 modified, it could be off by thousands of points simply by a misprediction about triangle for the A region. In the case of the 80k games, around 80% of the score came from the milking part after max population was reached. If max territory and max population is reached later, that percentage changes. Also, using domination limit, you have to know the average number of citizens that will be supported per tile of usable land, which is only 66% of the domination limit.
I made a script that predicts this sort of thing, but I have only tested it against a few HOF Huge Histographic games. I would need to test it more before saying that it is a good technique to predict score.
SpiffyKeen7744
Prince
Using my script, I assumed that there were about 0.9 happy citizens per worked tile at max population and about 1 specialist citizen per worked tile. Also, I assumed about 80% of the final score came after maximum population was reached. I know that this is simplified, but for now I think it is a good approximation. I might work on this later to figure out a good way to estimate how the date of max population affects the percentages.
For now, I tested it on my 39k Emperor game, and Moonsigner's 88k Sid game:
88k (4562 domination limit, max population after 270 turns, x8 multiplier) = 83,210
39k (4200 domination limit, max population after 320 turns, x5 multiplier) = 39,013
Again, there are probably ways to improve the calculations. Also, I need to re-check the domination limit for my 39k game, and when I got to max population.
For now, I tested it on my 39k Emperor game, and Moonsigner's 88k Sid game:
88k (4562 domination limit, max population after 270 turns, x8 multiplier) = 83,210
39k (4200 domination limit, max population after 320 turns, x5 multiplier) = 39,013
Again, there are probably ways to improve the calculations. Also, I need to re-check the domination limit for my 39k game, and when I got to max population.
Similar threads
