First attempt of Immortal

[noontide]

Guys,

Thanks all for the advice. I played till 3am this morning but I beat that Korean Emperor game. Now I started an Immortal, rolled English. Attached is T0 and a screen shot of a few turns after.

Here are my first few attempt:

1. I decided to move one tile to the right to grab that Gold, or should I move one left up to settle on the hill? I think it's important to have some gold to push along the early tech.
2. Teched Agriculture first, then afterwards decided to go with AH instead of BW. Shall I opt to chop and crack out a settler first? I just don't think I have a very appealing spot for a second city while I have so many good tiles in the capital.
3. in terms of second city, I want to settle direct up to grab the wheat. Is there a good spot for a third city? The horse and gold to the far left and even the two golds and flood plain look really appealing but not sure i can afford settling that far from Capital.

I want to Construction rush Justin to the right, is that a good plan? I think I'll have three cities at most otherwise can't reach Construction fast enough.

Please advise, Thanks again!

 

[noontide]

I restarted and this time moved the settler to the upper left to leave room for City 2, now in the attached save on turn 100 I just finished Construction. Cased out Justin and he seems very strong, had a 11 population capital and about 10 or so swordsmen and axemen. I would probably take me about 10 turns to crack out enough catapults to start a war. Am I still too slow? All my population are working improved tiles and this seems to be the fastest I can get.

 

[Fish Man]

I restarted and this time moved the settler to the upper left to leave room for City 2, now in the attached save on turn 100 I just finished Construction. Cased out Justin and he seems very strong, had a 11 population capital and about 10 or so swordsmen and axemen. I would probably take me about 10 turns to crack out enough catapults to start a war. Am I still too slow? All my population are working improved tiles and this seems to be the fastest I can get.

Much better but still quite slow, with that many gold mines I would've expected an attack date of t100. At this rate if Justinian gets feudalism, which he can in 10 turns probably if he wants, and you bot into him without elephants, it's gg. I honestly think SiP was better, and a city could've been later founded to claim the northern wheat. Also cottaging non-riverside tiles is honestly kinda a waste. Stop building useless buildings too; aqueducts are unnecessary and barracks are questionable in hammer (food)-poor cities.

Your problem here, I think, is mainly overinvesting in military early on. You captured Bactrian but that's a pretty meh city tbh; cow is only good workable tile. And in 1120-ish BC you razed another barb city. That's probably 100-200 or so hammers down the drain, all for maybe 50 gold and...nothing else. Let the AI throw units away on these things and focus on teching construction early on. You could've made it plenty early with even 2 gold mines and some better cottages, as well as trading for alpha and building research.

 

[noontide]

Played some more. So I was able to take three cities from Justin, including its capital with some wonders, but didn't seem to do me much good. I don't have enough army to wipe him out, and he researched Machinery and now has crossbow men. As soon as I DoW on Justine both Germany and Rome DoW on me, I was able to fend them off and eventually can suit for peace. But with the cities I took pinched in I'm vastly behind in tech, and it's just matter of time another wave of attack will come from both Germany and Rome. Lost cause?

 

[Fish Man]

Start over and 1. Make a commitment to not go after a single barb city, and 2. Settle your capital where York is, York one north of where London is, and the other 2 cities in the same place. Oh, and start running research as either 0% or 100% to save on the rounding.

 

[noontide]

Start over and 1. Make a commitment to not go after a single barb city, and 2. Settle your capital where York is, York one north of where London is, and the other 2 cities in the same place. Oh, and start running research as either 0% or 100% to save on the rounding.

If I settle my capital at where York was, will the other two cities be too far away from the capital?

I built so many axemen so early because I ran out of thing to build. Should I trade for Alphabet and run research instead? Thanks.

 

[Olafeson]

Well depending on your gameplan i do not really consider non river cottages too poorly. If you plan already for a Liberlism game with a late attack, then it can be worth it, if you have worker turns to spare. Also if i am not wrong Bactrian is a captured Barb city?

I think you should have prevented the spawn of a barb city in this location with fogbusting units from the start. If you went for AH early, then this spot would have been my 2nd or 3th city. A nice river, a cow, and some forests and silk tilers which you can farm and later change when you got calendar.

Probably the city to the far east with double gold and floodplains should be a very high priority city. As 2nd or 3th city before someone else can get it. Farm some floodplains and work the gold mines.

With lots of early gold mines you also do not need to attack someone with axes and catapults. Because if the captured land is not as good, you slow yourself down and let other A.I. players run away in tech. My gameplan with such good land would probably have been Liberalism, and attack later with Cuirassiers or Canons.

Anything else i agree with Undefeatable.

 

[Olafeson]

Definitly trade for Alphabet and run research, also start running 2 scientists in some cities with extra food. Building Research early over forges and other buildings is often times also very useful to get to the early key techs faster. Code of Laws, Philosophy, Currency and Civil Service. And if you are running out of things to build, just get another worker and settler.

 

[sampsa]

While on emperor expanding with a catapult rush is rather easy, you need much more precision on immortal. Non-riverside cottages are certainly ok, if there is nothing better available. You should definitely cottage riverside calendar resources btw, free . You should aim to chop all forests pre-attack, otherwise you are just leaving lying around. And yes, as pointed out, always 0%/100% slider.

 

[noontide]

While on emperor expanding with a catapult rush is rather easy, you need much more precision on immortal. Non-riverside cottages are certainly ok, if there is nothing better available. You should definitely cottage riverside calendar resources btw, free . You should aim to chop all forests pre-attack, otherwise you are just leaving lying around. And yes, as pointed out, always 0%/100% slider.

Does this mean stop research until gold accumulates, and then slide to 100% at the loss of gold? How much of a difference will this make? Thanks.

 

[krikav]

Thats what it means yes.
The differance is most of the time 1 beaker per turn.

 

[Fish Man]

Thats what it means yes.
The differance is most of the time 1 beaker per turn.

Replying more to @noontide but: 1 bpt might not seem like a lot, but if you save 1 bpt per turn, that's 100 bpt saved, roughly, by turn 100. Not an insubstantial amount. Also having more gold at the beginning of every turn where you run 0% research increases the chances an AI will give you a deal because you have more money to make up for the mildly ripoff trades that the AI usually proposes. And hoarding gold until you get, say, a library up will increase the efficiency that you use your commerce with. Small things but they do add up.

 

[AcaMetis]

Do be weary of the possibility of AIs demanding your entire treasury, if you hoard too much and research too little. Not that you should be in such a situation very often, but it's something to keep in mind.

 

[elmurcis]

Thats what it means yes.
The differance is most of the time 1 beaker per turn.

There is one situation where 0% research slider might need to be moved to other % position (with smallest part of beaker lost if can't find one with precise .00) - when running 2 scientists (like early for 1st GS) and have clear 7.5 beakers. 4 scientists are great with 15 beakers/turn etc.

 

[unas876]

Why would you research something in 10 turns when you could save gold for 5 turns and then research at a 100% and get the tech you need in the next 5 turns?


What hasn't been explained yet is that if you run at 50% every turn, you pretty much need to decide what to tech on the turn you finished the previous tech. What if you put three measly 50% turns into Music and then bang, Mansa gets it next turn? You pretty much wasted your money there. Think about what to tech, accumulate enough gold until you can run your slider at 100% until you can finish the tech, and then tech it. Leaves maximum room to decide what path to take. A lot can change within a few turns of waiting.

This as well as the waiting for science multipliers (eg waiting for library, oxford, bureaucracy switch, etc... to finish) takes some nuance and some experience to properly do, but its important when trying to improve.

Besides AI demands for saved up gold, forced civic switches (demands, AP, UN,...) it is known to happen to people with no experience in this matter that they forget to put their slider up again for lots of turns and inevitably fall behind hopelessly.

Especially after watching Lains Deity LPs I have tried to make it a routine to explicitly check again whats going on with the research slider before I end every single turn.

 

[krikav]

What hasn't been explained yet is that if you run at 50% every turn, you pretty much need to decide what to tech on the turn you finished the previous tech. What if you put three measly 50% turns into Music and then bang, Mansa gets it next turn? You pretty much wasted your money there. Think about what to tech, accumulate enough gold until you can run your slider at 100% until you can finish the tech, and then tech it. Leaves maximum room to decide what path to take. A lot can change within a few turns of waiting.

Yes, this is the real issue.
A few beakers here and there is no biggie, but the increased flexibility and control over the flow of the game is whats crucial.

 

[rah]

And it gives you the chance that two known AI will complete a tech and make it cheaper for you to complete.
And yeah, if it was just 1 beaker a turn I probably wouldn't bother unless I was record shooting, but there are so many other advantages.

 

[sampsa]

There is one situation where 0% research slider might need to be moved to other % position (with smallest part of beaker lost if can't find one with precise .00) - when running 2 scientists (like early for 1st GS) and have clear 7.5 beakers. 4 scientists are great with 15 beakers/turn etc.

Indeed, running 2sci via library at 0% loses half a per turn and doesn't lose any (0% never does). While it's possible to find a slider position that loses less +, we are talking about fractions already, not 1 per turn. Meanwhile, other benefits of binary research remain.

For example, an empire creating 19 per turn at 80% slider leads to 19*0,8=15,2 and 19*0,2=3,8, meaning 0,2+0,8 is lost due to rounding, thus leading to a full "" lost per turn. (unrelated to library, just trying to clear it up for everyone)

Edit: got interested enough on the subject to make a spreadsheet.

Spoiler :

I came up with following rather random assumptions: Capital has library and is running 2*sci, creating 14, other cities have no -modifiers and are creating 14 in total.

0% slider: 28 7,5 (loses 0,5 per turn)
50% slider: 14 23,25 (loses 0,25 per turn)
100% slider: 0 39 (perfect, as capital output happens to be divisible by 4)

Thus here running 50% slider loses 0,25pt and running 0/100% slider loses 0,5 half of the time, so they are equal. Other slider positions lose clearly more, some even more than 1,5+ pt (for example, 30% creates 19,6+16,95)

Will edit this post after more testing.

Edit2:

Spoiler :

After more testing, the result is very inconclusive. A change of one changes everything (and it's possible this happens every turn).

Random findings:

• running 2*sci, it seems impossible to find outputs that allow you not to lose any fractions due to rounding
• when the total of the empire is an even number, 50% is attractive for rather obvious reasons
• when the total of the empire is divisible by 10, all non-binary slider positions become more attractive (because all 10%,20%,30%... of for example 40 are whole numbers) since then the amount of you produce will always be whole number. For this is not true because of the library +25% bonus
• when the library city (that is also running 2*sci) is producing an amount of which when added to the 6 that the scientists are producing is divisible by 4, binary slider positions become more attractive. Say 14+6=20 because then the library +25% bonus is a whole number. Say 17 would be especially bad since (17+6)*1,25=28,75 as the 0,75 is lost.
• even if you manage to find slider positions each turn that cause you to lose the minimal amount of fractions every turn, you need to take into account the findings of my earlier thread "hidden research bonuses?". You gain +1 for every 24 you produce in a turn for every AI that knows the tech. This certainly favors 100% slider. Also, you can't ignore prerequisite bonuses as for example if you are researching a tech that has one prerequisite you gain one if you generate at least 4 raw per turn, 2 if you generate at least 9 raw per turn, 3 if you generate at least 14 raw per turn and so on. So, not hitting these integers causes you to lose fractions again...

At this point we have to just admit that there is no way it makes sense to try to not lose a fraction of here and there. For these reasons, I'd recommend binary research even when running 2 scientists. I did find situations where 10% slider was better than 0% slider, so it really is rather random. However, the best chance of improving on binary when one city is running 2 scientists seems to often be 50% slider.

Apologies to noontide for de-railing your thread. I'll play this start soon.

 

[lymond]

@noontide Not sure if this was mentioned by anyone before, but it does not appear you have BUG/BULL installed for UI improvements. You can find it here:

https://forums.civfanatics.com/threads/download-bug-and-bat-here.274636/

The SP (Custom Assets) installation installs the game in your Custom Assets folder..no need to load a mod. It will work for you current game as well. Great option for forum games as anyone can load your games.

You can also find the BAT mod there as well, which is another great option. Basically BUG/BULL plus some graphic enhancements. It is standalone so would not work for current game, but great for personal games.

 

[elmurcis]

Indeed, running 2sci via library at 0% loses half a per turn and doesn't lose any (0% never does). While it's possible to find a slider position that loses less +, we are talking about fractions already, not 1 per turn. Meanwhile, other benefits of binary research remain.

Hmm... big thanks for this - mind was so on research part that it forgot about gold (rounding) too....
(sorry for off-topic )