Spoonwood's Hall of Fame Attempts

[Spoonwood]

In 900 AD Dyrrachium gets autorazed. Now just to finish off Korea's ships.

910 AD - We replace the city with the Heroic Epic with a new one. Core irrigation gets underway.

920 AD - We join workers until a city reaches size 27. We then abandon our capital, since it's also in a not so good spot, because of it's culture. And it works:

Score up to 5681. 25 turns left on Scientific Method. Then we'll use mass specialists to research.

The 940 AD score is 5935, and the previous was 5808.

I checked with SirPleb's score calculator, and it estimates 22127 as the final score. That consists of an underestimate though. Rysingsun's Deity game had a score of 4844 in 940 AD and 4942 in 950 AD. That estimates to 17,436, when his final score was 25,358. So maybe we have about 8,000 points underestimated, and we can break 30k?

Honestly, kind of hoped that this map could break 40k. But, I haven't looked into what's possible score wise.

 

[Spoonwood]

950 AD - It looks like we can finish off Korea's ships with our current fleet. So, the luxury slider now gets bumped up to 70%. I did go back to having workers plant forests, chop, and then re-irrigate. But, not so carefully now as to use combinations that end up as optimal as possible, since turns still end up time consuming (though not as bad as the turns when ferrying around troops and workers to new islands!). I've also started taking notes to sell off temples after cities reach 10 culture.

970 AD - Almost all of Korea's ships have gotten sunk. We had gotten rid of the infantry a few turns ago. Really, like usual, I think I ordered more built than we needed, but I like to have 3 or 4 defensive units in cities at Sid. Also, a few turns ago Korea landed 4 units, which we managed to take care of with one army. But, now the units we have even more seem like overkill. So, we disband them in cities to pick up the building of infrastructure. Granaries where needed, temples where useful for cultural expansions, aqueducts, harbors, and marketplaces for the artillery or cavalry, though mostly granaries. Armies get disbanded only in cities building hospitals.

Also, stall out at least one city for this turn, so it has more food after growing:

Actually, the superior arrangement, which I change to, lies in using coast squares for a turn, since happy citizens give more score than specialists. Now I realize I made a slight error by getting rid of some of the galleons a few turns ago. But, oh well. Score 6328.

980 AD - Score of 6462. Sirpleb's calculator now estimates 23,635 as the final score. Oh wait... the bottom of the page says:

"(Estimated current internal 'per-turn' score is 34736)"

That's probably still an underestimate, since I still have a good amount of city growth to go, but it seems more reasonable.

We sink the last of the know Korean ships and make pizza with them.

I think we had a little war weariness which now has gone. But still, we have cities like this:

Though we have 70% luxuries, and it has 14 tiles worked currently, if one of the civil engineers got put onto a tile, then it become content, and the 2nd one would become unhappy. People have recently discussed courthouses and police stations as reducing corruption by 10%. Well, 10% of 35 is 3.5. So, 3 more commerce from the courthouse. At 70% luxuries, that should give us two more luxury commerce for more happiness. Then also the police station. And Cure for Cancer later.

Hence, optimal city placement doesn't work so well for happiness, because all tiles worked runs into happiness issues probably, or needs cultural buildings. We don't want to trip over the cultural limit. O. K., a few temples and cathedrals might not hurt us there. But also, we don't our borders to expand and trip over the domination limit from culture.

Also, the above shows why Republic (or Democracy) ends up better for maximizing score than Fedualism or Fascism.

The last screen shot also shows some of the original core. Need Cows Keeper sits where the original capital did. 002 is the 2nd city, and after putting a granary up during the despotic phase, it also put out settlers and workers using one of the grassland cows to make cities and improvements. Later, the capital become a 1 turn worker Pump from size to 6. Also some settlers, when I accidently had another nearby city use the mountain. And then later the original capital produced a settler every 2 turns during the golden age, what I call the "half-factory". New Epic was the city that built The Heroic Epic.

I also try to use civil engineers during this phase when cities have surplus food for growth or content or unhappy citizens. As one example here's bora using a civil engineer to get it's aqueduct built. Though, we could use it the citizen for happiness, and maybe that's better for score I don't know, I try to do setups so that the city can keep on growing or finish an aqueduct or hospital and then grow:

Here's Troy building a hospital:

Actually, I think last turn I gave it 7 food. But, 7x2 = 14. It can finish the hospital now without running into food overrun at size 12. And maybe I change it slightly before the hospital finishes with something a littler more optimal looking. Though also, I might not push optimizing this.

I haven't found a great way to handle the infrastructure phase, other than to cycle through cities and look at least for happiness, or something which looks like it could work better.

 

[Spoonwood]

Score of 6739 at 1000 AD. We passed Mad2rix's Large Chieftain histographic game in score a few turns ago.

 

[Spoonwood]

I took a look at Drazek's 1000 AD save last night among some others (the 93k game). I noticed him doing 100% taxes. During one of the turns I've played so far today, I set all citizens with that setup. I could make something like 900 gpt instead of 100 some gpt with that setup. Then, more of the infrastructure could get cash rushed. But, I figured out that it basically isn't great for size 13+ cities. It seems fine for sub size 12 cities. Some of them might require an extra civil engineer earlier before their marketplace gets put in. That sounds kind of o. k. to me. However, large cities with hospitals start to need more use of the luxury slider. If they don't have it, then specialists need to become entertainers more often. That doesn't leave any plan for research also.

In 1080 AD I think we have 9 turns left on The Scientific Method with a score of 7897 and had a score of 7166 in 1030 AD. I've had a lone scientist study it the whole time. I thought we wouldn't be big enough to have scientist only research and get it in less than 50 turns, but I might have gotten that wrong. Anyways, it also sounds more entertaining that a lone scientist discovers some new method he worked on for a very, very long time with little success. He got dismissed as a crackpot repeatedly even. But, eventually, everyone took up his method, and research became much quicker.

No city will get set to wealth until we have mass transit systems up. Cities build marketplaces, temples, harbors, aqueducts, granaries, hospitals, courthouses, police stations, banks, stock exchanges, explorers, cavalry, or tanks until we have mass transit centers. Then cities will get set to wealth. Cities have used civil engineers to get up buildings for the most part so far who were excess citizens (no tile to work) or would have worked unhappily or contentedly before. Courthouses and police stations definitely have proved useful, as I expected. But, once we learn the scientific method, we won't use civil engineers to build banks or stock exchanges. Perhaps those buildings end up as prebuilds on mass transit centers. Though, I might take a detour to the technology for Cure for Cancer first.

We had the ability to build Wall Street at some point, just from captured Persian cities I think. But, lost it at some point. That may well change soon, and Wall Street will get us a little more cash sometime.

 

[Spoonwood]

1110 AD - Score of 8337. That puts ahead of PrinceMyShkin's standard Warlord histographic game. SirPleb's calculator estimates 27,348 if nothing changed. The internal 'per-turn' score is 41,187.

Did I mention we had no horses, iron, or saltpeter in our original core before? I think I mentioned that we had the opportunity to trade iron with Korea early, but we waiting since we didn't have enough gpt from the AIs yet to cover it. Then they traded it away to someone else. Persia though got an extra iron source hooked up before we got to Steam Power, since we didn't need to get into using a military alliance to disrupt Korea's iron trade route.

There's coal and ivory in our original core. I suppose there could exist modern era resources also.

 

[Spoonwood]

I stopped taking notes on cultural expansions. I forgot that MapStat can get set to automatically update on a new save. So, I just try to sell off building when cycling through cities. If any still have culture, then I find them and sell it off.

1140 AD - For the last few turns, I've been at the phase where I abandon some coastal cities here and there to avoid the domination limit trigger. I haven't pushed getting there to the max. I learned in this tiny Sid game that one needs an exact awareness of how borders shift exactly with the cultural expansion, and it's sometimes more complex than expected. I've even ended some turns at 64% of the domination limit instead of 65%. Once all temples have gotten sold, there will still exist plenty of turns to grab any last good looking tiles.

Looking over the old games I had played, I also changed some smaller cities to purchasing workers every other turn for faster growth.

Occasionally, I've also found cities like this and micromanaged it exactly for what I think as optimal.

Do you understand how it got 7 food on the first turn, and then we could use another civil engineer on the 2nd? I'm not sure that such cuts down on build time, but this game seems to run into fewer turns where I see the "we have food we can't use to grow" than other games I recall.

I found what I believe an unexpected common multiple (hint... @BlackBetsy maximizing one's gameplay, in my opinion, is often about finding common multiples to hit breakpoints on builds and then prioritizing other cities to hit more breakpoints... a high food settler factory needs exactly some 4-turn combination of 7-7-8-8 shields, or 7-7-7-9 shields, or 10 shields per turn... then your worker goes and develops territory for another city... you need more workers than I've seen in your intermediate saves that I've seen to hit more breakpoints sooner for common multiples sooner or more growth so that you can hit common multiples with forestry once cities have reached size 12... and getting all cities to size 12 early has priority over common multiples). This is a minor case, unlike say getting common multiples for military builds that stay around on horseman to upgrade to cavalry or just build cavalry, but perhaps this example still ends up instructive:

O. K. now working out the reasoning. We could build an explorer in 4 turns to disband for 5 shields elsewhere. We get 1.25 shields per turn that way. Or a cavalry in 14 turns to disband for 20 shields. That's 1.42 shields per turn. Now perhaps an explorer would end up more efficient, since we can disband more quickly.

But, a 6 shield build on an infantry produces exactly 90 shields in 15 turns. The infantry disbands for 22 shields. That will give us for 1.46 shields per turn. Whether this case ends up improving my score at all seems kind of doubtful (or maybe it's 2 points overall?), but I think it helps to illustrate the concept of common multiples as useful in civ III.

Another rare case where entertainers actually seem optimal... for a turn:

 

[BlackBetsy]

found what I believe an unexpected common multiple (hint... @BlackBetsy maximizing one's gameplay, in my opinion, is often about finding common multiples to hit breakpoints on builds and then prioritizing other cities to hit more breakpoints... a high food settler factory needs exactly some 4-turn combination of 7-7-8-8 shields, or 7-7-7-9 shields, or 10 shields per turn... then your worker goes and develops territory for another city... you need more workers than I've seen in your intermediate saves that I've seen to hit more breakpoints sooner for common multiples sooner or more growth so that you can hit common multiples with forestry once cities have reached size 12... and getting all cities to size 12 early has priority over common multiples). This is a minor case, unlike say getting common multiples for military builds that stay around on horseman to upgrade to cavalry or just build cavalry, but perhaps this example still ends up instructive:

This is a really useful post. If there is a defect in my MM skills is that I don't MM shields / food as much as beakers. I think I'm really good at beakers but I don't always try to MM to get efficient shields or food. In general, when MM I am trying to get cities to 3 citizens on tiles that will produce 12 food w/ the city square and then taking the other 3 and making them scientists. For a city on a river / fresh water, I am just trying to keep enough extra food to grow while siphoning off food for science.

Often when building military units, I am focused on one type of a unit versus another (usually bombardment weapons vs. offensive troops) and the decision is usually based on where we have a barracks. In the case where barracks are not an issue, I typically go for the more efficient build.

 

[Spoonwood]

If there is a defect in my MM skills is that I don't MM shields / food as much as beakers.

I looked over your 890 AD save from your Iroquois histographic game the other night and saw some things that for sure jumped out for me with your core cities. That's what lead to my comment here. I could say more, but feel kind of hesitant to do so for some reason. If you want more tips on what I might do or advise, maybe post a screenshot from some inter-mediate save and tell me what your plan overall for the game is or was.

Back to my game, I had two cities...

So, the worker near Bonn was on a hill. The extra hills and lakes aren't good score wise (no idea how I planted that one earlier). So, I order it abandoned. Then I look over where the right city was. It could have two sea square that motivated it's plant, though Cor can use one of them. But for some reason I thought about it needing a temple. But then I realized that the cultural expansion in Bonn and a cultural city expansion nearer to Cor would grab those lakes. And that's no good.

There's a free sea square from Bremen and Cor. Also, there's two unworkable coast tiles. So, now I'm thinking that the right abandoned city was a mistake. The more I've played histographic games the more I've started to think that winning by 2050 is easy. But, optimizing use of tiles is not easy. I do think @EMan has a better handle on this than me. However, I wonder if his calculations account for all these subtle cultural expansions from two cities producing more workable. Maybe they do.

A moment later I see Gordium:

It's cultural expansion would put a forested plains in use. I thought about selling off it's temple. But maybe the better play is to abandon Polka? There's 3 extra bonus grassland for 4 food. Then 8 extra coast squares, and 1 sea square. The bonus grassland and 8 extra coast squares average out to (((3x4) + (8x3) +(1x3)) / 12) = ((12 + 24 + 3) / 12). That's 2.25 food per tile counting towards the domination limit. The free sea square though produces 2 or 1 more point. So, Polka has a little more value, but I don't think it's enough here, correct?

Also, I think average calculations perhaps a bit misleading for upper level games. Here, Persepolis could move one tile closer to Apathesis, but then out would go The Pyramids! I think I mentioned before Athens sits on top of a grassland cow (and it's not coastal), but it has J. S. Bach's Cathedral. Now I see that Polka could move one tile south, and still use that sea square. That does seem like a better city placement.

 

[BlackBetsy]

I looked over your 890 AD save from your Iroquois histographic game the other night and saw some things that for sure jumped out for me with your core cities. That's what lead to my comment here. I could say more, but feel kind of hesitant to do so for some reason. If you want more tips on what I might do or advise, maybe post a screenshot from some inter-mediate save and tell me what your plan overall for the game is or was.

That was a while ago....and I haven't played Histographic much but would appreciate hearing anything you saw.

 

[Spoonwood]

Also, here's something I like to see:

The cultural expansion from Trebizond will force a free sea square for Smyrna. So it can use that sea square without needing a temple to grab the bonus grassland hill also. I guess these sort of idea suggest how @Drazek got the idea to use AI cities to control borders. That sounds good if no flips, and the AI won't build military and become a threat.

 

[Spoonwood]

That was a while ago....and I haven't played Histographic much but would appreciate hearing anything you saw.

From memory you had a city at size 10 producing 9 shields per turn. My first thought was to forest a tile to make it produce 10 shields per turn. Also, it had no harbor. It couldn't produce good commerce, and commerce can be useful for upgrading horsemen to cavalry. It also had all sorts of unusable sea squares.

The trick with upgrading lies in either setting builds to horseman before reacquiring iron and saltpeter from AIs after pillaging, OR which I did first, pillaging out one's own source of saltpeter and iron. Also, the game has to get set to ask for "Ask for Build Orders After Unit Construction" in the Ctrl+P/Settings screen. After all builds get set to horsemand (and you won't learn a technology that turn), then one trades for iron and saltpeter or reconnects them. Then, when a horseman build completes, you zoom to a city, and upgrade to your cavalry or knight type unit. The upgraded unit then comes as ready to move on the next turn. When you can zoom to cities is also a natural feeling time to look at micromanaging tiles (the production phase has completed by that time).

As an alternative or in combination with upgrading is short-rushing. If you have a city with say 10 shields per turn, it could in principle produce 10 shields on turn one. Then on turn two you short rush with a granary or temple or muskemen. Finally, reset the build to a cavalry. Then, your cavalry will come out in 3 turns for (50 x 4 gold per shield to shortrush) = 200 gold instead of in 3 turns at (50 x 3 gold per shield to upgrade) = 150 gold. If a city has 9 shields per turn though, I feel it's better for it just to produce a cannon or trebuchet, at least for when artillery possibly come into play later or picking off stragglers. I hadn't realized before why, but calculating the "waste" for a cannon comes out with 45 shields having gotten psuedo-produced. The waste is only 5 shields. A short-rush on a cavalry with a temple or granary, though ends up with 7 wasted shields.

Oh... that wasn't the original reason, I finally remember. Originally, my plan was for sub 10 shield cities to produce artillery type units only, because they would need barracks also. They end up taking longer to get barracks up, then something with 10 or more shields. Also, the 10 shield city can shorush barracks with a settler or explorer I suppose. But, shortrushing with 9 shields leads to that waste issue. If you have a city with 21 shields instead of 20, the waste doesn't have as big of an effect, I think. But, the lower the numbers of shields per turn, the more impactful waste becomes.

Anyways, I also noticed that your capital had 10 excess food per turn while trying for Newton's! You had rails by then, but I could tell that you never seemed to have any sort of plan to optimize shield production before rails became available. You didn't have sufficient numbers of workers to do that either. Salamanca was also making 17 shields. If it were just making cavalry, 27 shields is 3 turn per cavalry instead of 5 turns at 17 shields per turn. 27 shields is also the breakpoint for any size 12 capital to reach 40 shields during a golden age (I didn't realize this until after playing many, many 20k games and trying to put in a library during a golden age).

Even more glaringly, I saw some city one tile off the coast. There existed so much food there given up and commerce also.

I also noticed that you didn't come around to jamming cities into coasts later I think. If cultural expansions are not too much of an issue, those can be worth it.

 

[Spoonwood]

Oh, now I'm just learning that 100 culture can actually end up more useful than 10 culture.

If Sip has 100 culture, I think, then that will force a cultural expansion for the two sea tiles near Saltp. Then Salp won't need the temple. 90 culture will take 45 turns with a temple, or 2 more for building the library, and then 30 more turns, for 32 turns total (actually, I think it's 31 turns, since culture gets calculated after production?). So, the library ends up more efficient for culture here. Or would saltp not use those sea squares in between turn 32 and 45 turns? The library ends up different, and thus better fits with the idea of leaving it until we have 100 culture in Sip. I have a bunch of histographic games in the HoF, some other attempts, and I'm still learning about these such here and there.

In contrast consider Bonn and uml:

Bonn can grab that sea squares near uml. But, cultural expansion to grab two more grassland squares makes sense.

I use to prefer a hospital before a marketplace. Where there exists enough grassland tiles though, I now prefer the hospital before the marketplace. It feels better for the city to keep accumulating food than to get stuck and not accumulate food, because the later hospital would lead to overrun or there's a food loss. The marketplace before the hospital might be more efficient score wise though.

I've done some of the forestry for some cities with police stations or courthouses. But, like I'm not sure that's the best play. Here's an example:

I think I ordered all relevant tiles chopped (and I don't want to check with civ assist II), so I mark it's name "Done". I think @EMan uses a tag "xX" for a similar purpose in some of his games. But here, maybe the police station isn't going to change anything with happiness. Forestry could theoretically wait until mass transit centers become available. A quicker finish of a mass transit center could mean a little less pollution and more score overall. Probably though any difference ends up rather minor. Maybe 2 points difference for a city in some cases but not in others? I guess that could be 500 points difference on a huge map maybe? That sounds high to me.

Anyways, again, I think it much easier to finish than to optimize such things and get the timing of everything just right. A finished histographic game I think better than attempt an optimization which doesn't finish or feels like too much of a drag to the human player.

1160 AD - Global warming appears in a forested tundra tile. Score of 9061. I notice that Korea's Ice Resort says 0 in the victory conditions screen?! I guess that's because we went to war on the same turn we gifted it away?

1170 AD - We learn the Scientific Method. We research Steel. After changing to scientists it projects to 9 turns. Espionage would project to 6 turns. Researching Espionage first could mean another future tech later, but I believe mass transit systems earlier slightly better for score. Atomic Theory projects to 12 turns. We dropped from +200 gpt to +54 gpt. MapStat tells us we have 337 scientists. The game has 10 industrial techs left and 17 future technologies. At 7 turns per tech, that's 189 turns, while the game has 308 technologies left. It's 100 turns from between 1950 and 2050, so likely we will still research future technology before 1950. Contrary to rysingsun's claim who did play for the top scoring deity histographic game (which was played on a pangea map), keep in mind that even though I did use trade route pillaging, research is never zeroed out for a histographic game until the last few turns of a histographic game, since future technology counts something for score.

Also, I thought using just the capital for trade route pillaging was my original idea. But, even if I hadn't read him when I learned that, I see that rysingsun seems the first person to mention anywhere using the tiles around one's capital for trade route pillaging. So even though I don't like his attitude of "this is broken" and I think he's wrong as EMan suggested in response also, kudos to him!

Score of 9203. SirPleb's calculator estimates 27912 for final score, and we still have a fair amount of growth to go. The internal per-turn score estimate is 42,005.

One last note... from the information in this post it seems that it took 30 turns before any AI researched Elecricity. Maybe it was slower, because I had Sumeria and other AIs at war with Germany. Also, probably if I had maxed out research, we could probably have gotten it 10 or more turns sooner. Especially if we had cash-rushed a few libraries. And we might still have had enough gold to reasonably maintain a good conquering pace. Trade route pillaging would still make sense to get gpt and some lump sums of gold from AIs immediately. But a combination of trade route pillaging and doing some researching yourself I do think can work out better in some situations.

 

[Spoonwood]

I use the 'G' button to try and determine if my observations are correct. It suppress all cultural territory and shows the geography of the map. I do find patches of white which do correspond to deserts. However, as I expected I had inferred correctly, this map has no flood plains. Perhaps this means I'm using the "no flood plains exploit" (that was a joke).

I notice that MapStat tells me that Korea has 29,754 culture with +68 culture per turn in 1200 AD. The game says that they have 29,672 culture. Score of 9628. I noticed that City of Armies has Magellan's Voyage, and realize it has to be an old German city. It has 5 happy German laborers. The militaristic AI helped to build the armies.

1210 AD - Rename New Capital City as Hall of Javelinists. Score of 9771. Korea's culture in the game says 29,673. MapStat says 29669. End this turn without all workers fortified or having worked some job. I might still have some unchopped shields, but oh well, I don't want to fortify workers on the Korean island and wake them up again later.

1220 AD - Score of 9913. This game has now passed Theos's number 1 histographic game on a standard map. Instead of going with foresting one tile a turn, as we did over a dozen turns ago for a bit, we use foresting one tile with a worker or a slave. This reduces the number of finger clicks. 9 workers on 9 tiles each need one finger click for 9 forested tiles. But, 3 workers on one tile per turn for 3 tiles is 9 clicks on each turn, for a total of 27 finger clicks over three turns for 9 forested tiles. I assume that @Takeo has thought of something similar before. Perhaps because I had started thinking about the Realms Beyond people, like Sirian and his variants, I've started thinking about a game where one or a team of players deliberately tries to play with as few finger clicks for workers as possible without automating workers. This would delay advantages of railroads for human players in some situations.

 

[Spoonwood]

I wrote in another thread which I started.

 

[Spoonwood]

Imagine we had a city on the oil tile here:

Now, it may look bad to settle there for score, if you assume that score depends on growth. However, it has 7 sea squares which don't count towards the domination limit. Food wise, the city center would give us 3 extra food in the box. The whale would give us another extra food. With 4 tundra tiles, that implies all 21 tiles in the fat cross with no extra food in the box for a size 21 city. At 2 points per turn per worked tile, that's 42 points per turn. But, 7 sea squares don't count towards the domination limit. Thus, since (42 / (21 - 7)) = (42 / 14) = 3, we have an average of 3 points per turn a city location there.

In contrast, consider 019:

Now, we can already see it at size 19, with 2 food surplus. So, it can go to size 20 with no starvation. It has (19 - 5) = 14 happy citizens. At 2 points each turn per happy citizen, that gives up 28 points per turn. With 6 specialists, it will then have (28 + 6) = 34 points per turn. It has (3 + 5 + 4 + 3) = (8 + 7) = 15 tiles in use. But, 2 are squares. (34 / (15 - 2 )) = about 2.615 points per turn that count towards the domination limit.

Thus, for higher score, a useful idea lies in getting population out of this city by buying workers or settler if we want to hurry, and transferring it to the tundra location or elsewhere.

Credit to the saves of @EMan from his huge map entries (and maybe some of the other ones) which have helped me to see how settling the tundra can be useful to score.

Whether intentional or not by Firaxis (I assume that sea squares don't count towards the domination limit in Civ and PtW?), I think it's interesting that sea squares don't count towards the domination limit, since then there's motivation to settle spots in the tundra and by the sea, instead of just settling high food areas and thinking them always better.

 

[BlackBetsy]

One thing that I've discovered for myself (not new overall) that it's always better to settle low food tiles than adjacent high food tiles. The city tile gives you free, unimproved +2 food on a hill, for example, where you couldn't otherwise get it past +1 food (but you could improve the shields). I believe as an agricultural civ on a river, the city tile gives you free +2 food.

 

[Spoonwood]

One thing that I've discovered for myself (not new overall) that it's always better to settle low food tiles than adjacent high food tiles.

Often that works out better, but not necessarily. Early game it likely works out better. Except maybe you don't want to have a city on iron or saltpeter in the mid-game so you can upgrade horsemen to knights or cavalry. As a late game counter-example from my higher scoring Sid histographic game here's South Ellipi Keeper:

But it's on top a grassland cow! If we moved it NE/to the '9' tile from the city, then we'd have use of the grassland cow. But, we get three extra sea squares from settling there. And likely not every city settled in that picture there made for the best placement overall for the map.

The following might be another example, though I haven't done the calculations on this one, consider New Knossos:

If we moved it south to the banana tile we have one less food. But, then we could have 3 sea squares in use and one less coast tile in use. I think that would have made moving that city south onto the banana better for potential score. We'd lose borders for grassland tiles also, so my guess is that splitting New Knossos into a northern spot and that southern spot would have lead to higher score, if I had transferred population quickly enough, or they had grown quickly enough.

So, what you say does hold if we add the assumption that we have an inland spot. But, if we're talking about corner tiles with many sea squares (is 1 more sea square enough for a grassland cow square to get settled on usually?), then coastal jam or settling on corners can make more sense, if the number of additional sea squares in use ends up large enough.

I believe as an agricultural civ on a river, the city tile gives you free +2 food.

Agricultural city tiles always give 3 total food. There exists +1 food for the city tile in any government which is not anarchy or despotism for agricultural civs. In despotism, any adjacent freshwater source, such as a river, also gives +1 for the city tile. But, other tiles do not until the civ completes a revolution to another government than despotism or anarchy.

 

[BlackBetsy]

From memory you had a city at size 10 producing 9 shields per turn. My first thought was to forest a tile to make it produce 10 shields per turn. Also, it had no harbor. It couldn't produce good commerce, and commerce can be useful for upgrading horsemen to cavalry. It also had all sorts of unusable sea squares.

The trick with upgrading lies in either setting builds to horseman before reacquiring iron and saltpeter from AIs after pillaging, OR which I did first, pillaging out one's own source of saltpeter and iron. Also, the game has to get set to ask for "Ask for Build Orders After Unit Construction" in the Ctrl+P/Settings screen. After all builds get set to horsemand (and you won't learn a technology that turn), then one trades for iron and saltpeter or reconnects them. Then, when a horseman build completes, you zoom to a city, and upgrade to your cavalry or knight type unit. The upgraded unit then comes as ready to move on the next turn. When you can zoom to cities is also a natural feeling time to look at micromanaging tiles (the production phase has completed by that time).

I did post on that game and it was Deity level and my goal was first to conquer the AI and then do the histographic work. I was pretty exhausted from the conquest piece and I didn't build my core with histographic / sea square usage in mind. I got a little bit better on sea square usage when it came to areas outside my core. But I was so focused on just getting the win piece of it (knowing that the AI were likely to hit the cultural limit if I didn't conquer them) that I probably made a lot of histographic bad decisions.

In general, I've only posted I think 3 histographic games at this point and I really would need to study up before doing another one.

I did wind up building horsemen and then re-hooking up iron to make Cavalry in that game. But I often did so in batches for 6-7 turns at a time. I'd unhook, have a couple of cities produce 2-3 horsemen, and then mass convert to Cavalry, and then repeat. Yes, more efficient to do so intra turn to make them available but overall less time when you batch-process them in a build/spend cycle.

 

[Spoonwood]

@BlackBetsy

With the dearth of histographic entries, it may be better to just get something on a table first, and then worry about improving on it later. We can think of different perspectives.

1. It works out better to always put forth our best effort. Try to do everything as perfectly as you can the first time.

2. Just do something for your first effort. It works out better to get a game onto a table that charts than not do anything at all. Then if you ever do want to do a second game, well are there two entries on the table? If not, it could make sense to just get something on a table, since that can easily be quicker than doing things exactly right for a histographic game. Then afterwards, if you want to try for a table again, you can improve on it, and have something to approve on.

I think 2. more likely leads to a sense of fun. 1. might seem like it never leads to regret, except it can lead to the regret of not having tried a variety of things/games I think. So, 2. I think better in the abstract. But, 1. might lead better to rank (or it feels too stressful and then counter-productively does nothing for us also).

2. I think also has more potential to encourage people. Like if we had a gauntlet, I don't think we would want one of @Calis 's classic entries to lead it. If he changed his playstyle and tried to finish a game as quickly as he thought reasonable though, that might work for encouraging people in some competition.

 

[Spoonwood]

1260 AD - Score of 10,603. This game now has passed BlackBetsy's Sid histographic entry, but it's not a histographic finish. We learned Steel last turn. I went back to using the luxury slider. Change back luxury slider to 0%. Civ II had a famous neverending war post. In civ III, war until the last turn can be useful.

A new city:

Before Korea gets anything:

Goodbye to a size 8 city which was on top of furs (we had bought some settlers out of it at least). Gift Korea city with new fur source. It takes 319 gpt to tell Korea that we will give them furs if they supply us gpt. And they agree that if they do not supply the furs, they will not honor the peace treaty. Pillage out the capital with 2 armies (I had overestimated how much we would need to pillage). And the Koreans declare war on us.

I thought war happiness was just one unhappy citizen, but I found some old article said that it equals 25% of the citizens becoming happy. Our warrior razes the Korean gift city, and abandon the harbor city.

The luxury slider gets dropped to 50%. But, we still have more citizens working tiles now it seems. And Combustion has dropped to 7 turns instead of (I think) the 8 expected. So, as the upshot here, pillaging out one's trade route after trading for a luxury or resource with gpt can be useful for score at probably any level, given that the AI is so weak that you'll never get war weariness from them (if you also have workers to rebuild around your capital at least)! It can also, be useful for research if you don't need civil engineers as much it seems to rush in what looks like will be the infrastructure that you need.