Comparison game for early expansion

[Catwalk]

I'd like to do a little test to figure out the effects of playing with different races and strategies in the early game. Specifically:
1) When should you start researching?
2) How much are poor planets worth?
3) How much do you benefit from using drawn-out rather than crash research?

One turn has passed, default race is Sakkra and Advanced Space Scanners have been turned on to remove the randomness of scouting. Only one opponent that shouldn't have much effect on the early game.

If somebody wants to try it, I'd like you to make a detailed report for the first 60 turns, and make sure to explain what your general strategy is. Remember that you can use RACEED to change races, so you can try this out with any race.

 

[Catwalk]

Sorry, forgot to mention that this is for MOO1.

 

[Zed-F]

I'd think we should all start with the same race for at least one trial. Comparisons are most meaningful if as many variables are eliminated as possible, it makes analysis of the remaining variables more useful. Perhaps picking a race like Bulrathi which has a pretty standard startup curve with no special bonuses might have been better. Sakkra's cheap planetology tech tends to skew things slightly in favour of early tech. Bulrathi will give a better picture for the majority of races.

I guess I will try RACEED out and change to Bulrathi since I have some time tonight.

 

[Zed-F]

60 turns of Bears, round 1:

2301: We have a 100 Ultrapoor Ocean to the SW and a 50 Steppe to the NE. Not great choices, conventional wisdom suggests the Steppe is the better second planet. We have a ridiculous amount of room to expand. This means we will want to be building lots of colony ships and can afford to delay on building them until we can produce at max rate. Plus we have multiple directions in which we can expand so we don't even need seem to need range tech. Waste reduction could be a bonus though. Initial thoughts are build factories then colony ships at the homeworld, let the second planet take care of the research as it can. We have lots of poor worlds around, but aren't really in any danger of losing any to the Darloks anytime in the first 60 turns, I assume. I never build factories at UPoor worlds, but I probably will at the poor worlds the first time through. Mind you in 60 turns they won't have begun to really contribute to my empire yet, so in some ways that's a long term play rather than a short term one.

2304: I colonize Rigel and send 13 pop, which will put it to ~15/50, or 30% capacity. It should be able to grow fine on its own from there.

2322: Aros is at 170 factories and 86 pop. Rigel is at 39 pop and 28 factories. I could max Aros from here and send pop from Rigel ot top it off but since Rigel doesn't have all that much to spare and I will need what it has for jumpstarting other worlds, I decide to halt factory construction on Aros for now and go ahead with an initial colony ship. I can build one in 5 years, leaving a single tick in factories to slowly work on maxing them out. I notice that I can get to Nitzer, a 150 Jungle Fertile, in just 2 colony ships, so that's what I will do.

2327: Colony ship constructed and enroute to Gorra via Rigel. Since Rigel will be providing most of the initial population, and is at 48/100 factories right now, I stop building factories there, and start building scouts at a trickle (1/turn) and put the rest to research, which right now is 35 RP. First I want to see what I have.

2328: I am unimpressed with my selection, my only option is in computers where I take Jammer I over Deep Space Scanner. I doubt I will need to bother researching computers within the first 60 turns on this huge map. All research goes to planetology for terraforming +10, since I don't need range 4.

2331: Terraforming +10 at 7%.

2332: Gorra settled, Rigel transports half its pop (24) to kick-start it. Colony ship constructed, bound for Nitzer via Rigel. Terraforming +10 at 14% (would have been more but for Rigel tossing pop on transports.)

2333: Terraforming at 25%.

2334: Terraforming +10 comes in. Since there are a couple barren worlds nearby, one rich, even though Terraforming +20 would give the most benefit over the short term because I have lots of non-hostile worlds, I elect for Tundra bases next over Terraforming +20 and Dead bases. Plus it's the cheapest, and might get me closer to a waste reduction tech sooner. Even so, I opt to switch research to IIT 9 as it's much less expensive. Rigel's research budget goes to Terraforming so for the moment research is off.

2336: Rigel's done terraforming so research goes back on at 29 RP per year. I terraformed mainly so it could grow pop a tad quicker. It is still building scouts at a trickle.

2337: Aros completes a colony ship, which heads for Denubius, the next colony ship being due in 4 years.

2338: Nitzer colonized. Gorra sends 17 pop (half) to Nitzer. I would like to send more from Rigel but Denubius is due to be colonized next turn. This sometimes happens when you start colonizing in a new direction, but there's no help for it.

2339: Denubius turns out to be fertile as well as ultrapoor. I swap it from factories to research, as I don't build factories at ultrapoor worlds generally speaking. Transport 13 from Rigel to Denubius.

2340: Complete a colony ship, send it to Uxmai so I can offload some pop from Gorra and Rigel more easily, and allow Denubius to grow a bit before I send a colony ship to Centauri.

2342: IIT 9 at 2%.

2343: IIT 9 at 9%.

2344: IIT 9 at 17%.

2345: IIT 9 completes, IIT 8 is next. Research shifted back to planetology. Colony ship enroute to Centauri via Denubius.

2346: Uxmai founded. Rigel sends 13 pop, Gorra sends 8. That should be the last pop I need to send out of Rigel so it can now start maxing. Most of its research budget is set to factory construction.

2349: Centauri colonized, colony ship completed and sent to Aquilae. Next colony ship due in 4. Denubius sends 13 pop to Centauri.

2353: Next colony ship completes and is bound for Spica. Next colony ship due in 4 turns.

2355: Gorra sends 15 pop to Rigel to help it max out while Rigel continues to build factories. Aros finishes terraforming.

2356: Aquilae settled, Centauri sends 5 pop, Denubius sends 20.

2357: Colony ship completed, bound for Hyades.

2358: Spica settled. Uxmai sends 18 pop.

2361: Last turn. Aros completes colony ship, bound for Quayal. Rigel finally maxes, and can pump out 84 RP per turn. We have a total of 9 planets, plus 2 colony ships outbound. Still no contact with the Darloks.

Planets screen:

Save file:
http://www.civfanatics.net/uploads8/rbo-comp1-bulrathi-2361a.zip

 

[Zed-F]

60 turns of Bears, round 2:

2301: We have a 100 Ultrapoor Ocean to the SW and a 50 Steppe to the NE. This time I opt for the Ultrapoor Ocean. Knowing it is Fertile as well as Ultrapoor helps. This will give us a planet that can pump out more pop as our second world at the cost of our second planet ever being able to push out more than about 50-odd RP per turn (more realistically, less, given it is a pop source not a sink.) This move also facilitates expansion west and hinders expansion east. I will continue to build factories at poor worlds for this trial.

2303: Colonize Denubius, set production to research. I transport 10. Since I know what techs are available I set research to Terraforming.

2314: Terraforming +10 is at 6%.

2316: Terraforming +10 comes in, Tundra bases again selected next and research set to IIT 9.

2321: Aros is at 165 factories and starts colony ship construction slightly ahead of where it did the previous time around (as a consequence of sending fewer people offworld initially.) The first one will again be done in 5 years, with excess production going to factory builds.

2322: IIT 9 is at 3%.

2325: IIT 9 comes in, research shifted to Tundra bases. So far we have traded about 50 factories on Rigel for 2 1st tier techs. So far, so good. Research set to Planetology once again. My first colony ship is nearing completion; I opt to go Centauri -> Aquilae with my first two ships this time, then start spreading them out a bit more after that. I hope to get to Nitzer again this game.

2326: Colony ship completes. I elect to take a couple extra turns away from colony ship building to max Aros, using Denubius to backfill pop. I didn't have this option last time since Rigel was too small to support it and still properly kickstart other new worlds. I also take this opportunity to build a few spare scouts at Aros out of leftover ship construction credit. Unfortunately I don't really have anywhere to build scouts out of like I did last time, other than Aros itself.

2328: Aros maxes both pop and factories (ignoring terraforming.) It goes back on colony ship construction, with leftover production going to terraforming. Colony ships come every 4 years.

2330: Centauri colonized. Denubius sends 43 pop; might as well, since Denubius will regrow them quickly enough and I can always send extra on to Aquilae once that colony ship is done.

2332: Colony ship completed and sent to Aquilae, next colony ship due in 4.

2336: Colony ship completed and sent to Rigel. Aros pauses to build more factories and scouts. Next year it will start another colony ship, due in 4.

2339: Aquilae founded. Rigel founded. Centauri transports 23 pop to Aquilae, Denebius transports 25 pop to Rigel. Next turn it will send some more to Aquilae.

2341: Aros remaxes. Colony ship complete and sent to Gorra, next one will go to Hyades, then Nitzer.

2344: Tundra bases reach 2%.

2345: Tundra bases at 3%. We complete a colony ship and send to Hyades, next due in 3.

2347: Tundra bases at 6%. We colonize Gorra. Rigel sends 10 and Denubius sends 20.

2348: Colony ship completed, heading to Nitzer via Rigel. Next due in 4.

2350: Tundra bases at 11%.

2351: Tundra bases complete after having been at 11% the previous turn. We pick enhanced eco restoration next, but that will take forever to come in at this rate. Let's pick something a little more manageable in the available timeframe. I go for propulsion tech just on the off chance we can make partial progress toward nuclear engines or something before time runs out.

2352: Colonize Hyades. Denebius sends 20, Centauri sends 10. Aros builds a colony ship which heads for Rotan. The next one is due in 2! That's a fluke, it shouldn't be less than 3.

2354: Colonize Nitzer. Rigel sends 10, Denubius and Gorra send 15. Complete a colony ship, next due in 3. This one heads for Uxmai for lack of anyplace better that's reasonably close.

2357: Colony ship completes. There's nowhere close I can send it so it goes to Lyae. Next colony ship in 3.

2358: Hydrogen fuel cells at 5%.

2359: Well that was lucky, Fuel cells come in. I take nuclear engines next. Rotan and Uxmai settled, I send 10 pop from each Rigel and Gorra to Uxmai and 20 pop from Denubius to Rotan.

2360: Colony ship completes, sent to Quayal. I elect to stop here since nothing significant will happen in the next year. I have 10 planets and two colony ships outbound at this point.

Did I do better this time? Yes, I think so. But I think the reason I did better has more to do with the fact that the critical early need is population, and Denubius was a much better planet to provide for that need than Rigel was. Did having more/faster tech help? Yes, but not really all that much. On the other hand, in a game where more of the valuable early techs were available in the tree, it might have made a more significant difference.

One significant factor that might be otherwise overlooked in the comparison is that my total research production at the end of the game was higher in the game where I settled Rigel first than in the game where I settled Denubius first. I had more tech in hand in the second game due to having invested in tech over a longer period of time and more consistently, but I would have been starting to catch up somewhat, at least until the population advantage I had in the second game matured a bit more.

http://www.civfanatics.net/uploads8/rbo-comp1-bulrathi-2361b.zip

 

[Catwalk]

Thank you for participating, and thank you for pointing out the flawed racial selection. I've uploaded a game using Bulrathi instead. Is the game format okay? I don't think scouting would serve any purpose, as we're not comparing scouting or guessing ability here. Are 60 years enough? I think it'd be a good idea to list the result as #of planets, production, population, #of tech levels and extra colony ships, in order to get a general feel for how well you're doing.

As I listed on the RBO forum, there are three issues to examine:
1) Early research vs industry
2) Industry on poor worlds
3) Drawn-out research vs crash research
4) Racial impact on early strategies

It would probably have been better to create a game where we could examine just one at a time, this map deals with the first three because of all the poor worlds. But since you started it already, here is an update with bears for anyone else who may be interested. Thanks again for your interest!

 

[Catwalk]

I just did a quick run-through without taking notes using a research-oriented approach, I'll do another one more carefully. I got 484 population (you got 454 and 466), but only 641 production compared with your 774 and 757. I had paid 85% of the base cost on Enhanced Eco Restoration, though. It would hit in 6-7 years and give an instant boost of ~250 BC per turn. I had two colony ships en route, it looks like you had 3 in the last game.

Ultra-short report was that I researched IT+10, IIT9 AND Tundra bases before building my first colony ship, having stored up enough credits to send out three right away after maxing Aros.

Research wasn't very fruitful in this game. Neither Improved Eco Restoration nor Reduced Industrial Waste were available, and fuel cells weren't needed.

Sorry about not giving you all the necessary information, knowing the tech trees and the fertility of Denubius would have helped out quite a bit. Of course there's tons of stuff you have to make educated guesses about in a real game, but for purposes of comparison all info should be available.

I'll try an approach that is less heavy on research, focusing more on expansion and industrial development. This game was indeed a case of not really being interested in research.

Is 60 years too short a time-frame after all? It covers early expansion, but it doesn't cover early development. And then there's problem of research in progress.

Another consideration is whether to max the home world and use the colony for population or to stop maxing the homeworld at 80% and use both for population. The latter would provide more population, but would divide your production among two planets.

EDIT: I just tried it out with Klackons for fun, as the map is perfect for them. I got 1500 production on turn 2350...

 

[Zed-F]

Did you use Bulrathi as I did? That would make a big difference on how much of the way through Enhanced Eco you managed to get. I'd also like to see your detailed report. Possibly I had less pop than you because I had just sent a bunch off to newly settled worlds right before the game ended in each case.

I had thought that 60 years might be a bit too short as well, but now that we've chosen that figure I'd rather stick with it for a while.

I plan to do at least one more examination using the poor worlds for research and not production. When you use poor worlds for research rather than factories, what you are effectively doing is trading more research later for less research now. Whether that is a good tradeoff depends on what research you need to be doing immediately. In general my approach has been that I would rather have more research later, but in the very earliest game when techs are cheap (1st and 2nd tier techs anyway) then if there's a tech that will make a significant difference, it might in some cases be useful to delay poor world development purely to get that tech. It will depend on whether you have other maxed worlds to do the research. Nevertheless, looking purely at production understates the value of factories on Poor worlds, IMHO. Part of the reason it is valuable to get a good industrial base on poor worlds early precisely because it does take so long to build them up. If you don't have a good industrial base on a poor world before RC III hits, you will have a very long wait ahead of you when you do try to stand up that world. If it is already well-developed, the hit is not nearly so painful.

I certainly would not hold off on building colony ships just to wait for Tundra bases to come in. What you gain by having cheaper colony ships you lose and then some in terms of delayed starts for your other planets. Plus it is hard to find enough pop to stand up 3 planets simultaneously when you just have 2 planets, regardless of how good your second planet is, and you can't take advantage of range extensions to get the best planets as easily.

 

[Zed-F]

60 turns of Bears, round 3:

2301: We have a 100 Ultrapoor Fertile Ocean to the SW and a 50 Steppe to the NE. This time I again opt for the Ultrapoor Fertile Ocean. For this trial, I will perform research at poor worlds rather than build factories.

2303: Colonize Denubius, set production to research. I transport 10. Since I know what techs are available I set research to Terraforming.

2314: Terraforming +10 is at 6%.

2316: Terraforming +10 comes in, Tundra bases again selected next and research set to IIT 9.

2321: Aros is at 165 factories and starts colony ship construction slightly ahead of where it did the previous time around (as a consequence of sending fewer people offworld initially.) The first one will again be done in 5 years, with excess production going to factory builds.

2322: IIT 9 is at 3%.

2325: IIT 9 comes in, research reallocated to planetology.

2326: First colony ship completes, sent to Centauri. Again I decide to max Aros and backfill with pop from Denubius, as well as building a couple spare scouts from Aros with leftover ship production.

2331: Colonize Centauri, send 43 pop from Denubius as last time. Centauri set to research.

2332: Colony ship completed at Aros, sent to Aquilae, next due in 4.

2336: Colony ship completed, sent to Rigel, next due in 4.

2339: Aquilae settled, 26 pop sent from Centauri. Rigel settled, 25 pop sent from Denubius.

2340: Aros completes a colony ship, sent to Gorra. Denubius sends 14 more pop to Aquilae. Tundra bases at 3%. Aros is now completely terraformed and will spend another turn maxing factories and building a few spare scouts before starting another colony ship, due in 4. Tundra bases at 6%.

2341: I get lucky and Tundra bases complete. Cost of a colony ship drops from 546 to 469, a significant decrease, meaning it will be done a year sooner. I still think getting an earlier start on new planets is a better deal than waiting for Tundra bases to come in before building colony ships, though, and even more so on a smaller map where competition for stars is more intense. I elect to go for Enhanced Eco to see how far into it I can get.

2344: Colony ship completes, sent to Hyades, next due in 3 years.

2346: Gorra settled, Rigel sends 10, Denubius sends 15. Next turn a colony ship will complete and head for Nitzer via Rigel.

2350: A new colony ship completes, but my ship to Hyades hasn't quite arrived yet. I could wait a turn and send it to Rotan or go ahead and send it to Uxmai now. Since Rotan will open up the way to the Artifact planet I elect to send this one to Rotan and the next to Uxmai.

2351: Hyades founded. Centauri sends 18 to Aquilae which sends 18 to Hyades. Denubius also sends 10 to Hyades.

2353: Colony ship complete and sent to Uxmai, next one done in 3. Nitzer founded, Nitzer sends 10, Gorra sends 10 (still waiting for a block of pop from Denubius) and Denubius sends 25.

2356: Colony ship completes, sent to Laan for lack of anywhere closer to send it. Next due in 2.

2358: Rotan colonized, Denubius sends 20. Uxmai colonized, Rigel sends 10, Gorra sends 12. New colony ship completed and sent to Quayal.

2361: Aros completes colony ship, next one due in 3. I stop here. I just had a wad of pop arrive at Nitzer that was a turn or two out last time I stopped, so I expect my total pop will be a bit higher. Plus I got Tundra bases slightly earlier which allowed construction of one more colony ship.

How does this compare to my last game? I'm about halfway through Enhanced Eco restoration, going by the lightbulb. I figure I would have only been about a quarter of the way through if I had gone with that instead of propulsion tech last time. So, a small advantage there. More significantly, I got Tundra bases slightly earlier and was thus able to put out some colony ships more quickly. Since the name of any 4X game is early expansion, it's clear that getting a second-level planetology tech is useful from the point of view of speeding expansion, at least in those cases where many colony ships will be needed, such as on large and huge maps. Early research at poor worlds obviously helps with this.

The downside is that my poor planets are completely undeveloped. This cost is difficult to evaluate since we never got to the point where any of my core worlds got maxed other than the homeworld, let alone any of my poor worlds. Recall that investing in factories on a poor world is trading more research later for less research now; it's hard to make a fair comparison if we never see the cost of factories amortized over the course of a whole game. This cost is more than just purely in terms of production -- it also makes it harder to develop them later, as I alluded to earlier, given that RC controls will be arriving at some point.

I think it's fairly clear from this study that the value of tech vs. factories on poor worlds depends on circumstance. On maps where you expect a long slow drag out game, where you expect the cost of early factories to be amortized over a long time of researching at a higher rate, more factories makes sense. In games where you expect to reach a decisive point relatively early, building factories on poor worlds will probably never pay off in a big way. Perhaps a good general approach would be to do a mix of both factory building and research on poor worlds. One possible rule of thumb might be to run 50% research and 50% factories by default, and adjust up or down depending on how long you expect the game to run.

Save file: http://www.civfanatics.net/uploads8/rbo-comp1-bulrathi-2361c.zip

 

[Catwalk]

Impressive play, I'll sharpen up and try to do better. One note: I forgot that the planets screen doesn't take waste into account when displaying production. Considering that you had more factories than me, the lead wasn't quite as big as first assumed. Going for Tundra so soon was a mistake, I figured I wanted to colonize the rich Tundra and Barren planets. They would have been nice, but with so many other good planets nearby they were by no means crucial.

One other thing I've been pondering is to account for transportation costs when calculating the value of colony ships. In the last game I discovered that I was paying 23 BC per turn for a single colony ship AFTER the cost had gone down from 570 to 469 or so. It took the colony ship 7 years to arrive at its destination, meaning that it effectively cost me another 150 BC.

And yes, I did play as the Bulrathi.

 

[Catwalk]

2301: Open up Planetology and Construction, move to Denubius
2302: Start up research in both fields with 14 RP
2303: Send 10m to Denubius (did a little reloading here to figure out how many I should send, since research was not a high priority and colony ships were a high priority, it didn't pay off to send very many).
2304-2314: A little research at Aros to take full advantage of the interest rate, no industry at Denubius

Status 2311
pop 87
prod 79 (after accounting for waste)
RP 156

2316: IT+10 hits at 8%, Tundra next
2318: IIT9 hits at 3%, I swear I'm not cheating Then again, you have had similar luck
2319: I start sending colonists back to Aros which is at ~80m

Status 2324
pop 166
prod 144
RP 208, two techs researched
one colony ship en route

2327: Colonize Centauri, send 14m from Denubius.
2328: Colony ship done, send to Aquilae. I plan on building Aquilae up rapidly to let it begin colonizing the western systems, saving a lot on travelling costs. Aros goes back to building factories (had 104)

Status 2333
pop 215
prod 196
RP Tundra at 5%


2335: Aquilae settled, send ~50 colonists from Centauri and Denubius
2337: Aros resumes building colony ships after reaching 196 factories
2338: Tundra hits at 23%, pick Enhanced Eco Restoration. Considering that IT+30 is not available, it might be better to go for IT+20. However, I find that terraforming is not as useful when expanding rapidly, it doesn't do much to a planet that is half-full and still exporting colonists. I wouldn't bother with Enhanced Eco Restoration if Improved Eco Restoration had been available.
2339: Colony ship built at Aros, goes to Rigel
2342: Rigel colonized, send 10m colonists from Aros and 30m from Denubius

Status 2347
pop 320 (+30 en route)
prod 336
RP forgot to note down
2 colony ships en route to Gorra and Uxmai(I would have had to wait 3 turns if I wanted to send it straight to Nitzer, and getting closer to the artifact planet is also good. It works well to take turns sending colony ships in different directions so that new refuelling stations are available when you need them).

Forgot to take notes for the last part, but nothing unusual to report. After sending off a colony ship to Hyades, Aquilae goes all research to take advantage of tech interest. Aros does the same after sending a colony ship off to Quayal, having already colonized Spica.

I manage to colonize Quayal on turn 2361, and Enhanced Eco Restoration hits at 16%. All colonists also manage to arrive for the report, making my numbers look better in comparison. This is always in problem in comparative games, unless there is a simple goal to strive for, like victory date.

Status 2361
pop 545
prod 688
no extra assets


I have a grand total of 259 factories at this point, but now I'll start churning out high-tech industry at all planets but Denubius and Aros. No poor planets have factories at this point. Aquilae only has 45, as I would rather take advantage of all the excess population with a colony ship (had 120m when the colony ship for Hyades was completed). I managed to reach 11 planets with the skin of my teeth. I think heavy research was worth it here after all. With fewer poor worlds, I might have stopped at IT+10 and IIT9.

Sorry about not posting the planet screens, I'm having trouble figuring out how to dump pictures on this computer.

EDIT: I'm still not sold on building factories on poor worlds. Drawn-out research usually pays off with standard worlds barring unusual conditions, on poor worlds they almost always pay off. Even in the unusual conditions of this game, which were fairly anti-tech. If I can invest my production at an 8% interest rate through tech, then I see little point in investing in factories at a 3% investment rate.

I think the value of factories goes down as you approach the population limit. That is, building a factory on a size 100 planet with 60 pop and 100 factories gives a better return than building it on a size 100 planet with 100 pop and 180 factories, as you lose more emmigrants.

I agree fully that research priorities are heavily dependent on the map. In order to examine this properly, we'll have to create a Bulrathi game that favours tech more (environmental tech required, waste clean-up available, fuel cells required) with one or two poor planets in the initial batch. Adding more opponents might add or detract from research importance, I'm not sure which it is. I find extra range highly useful in deterring opponents from colonizing my systems, and Propulsion/Construction techs lower the cost of blockade ships greatly.

 

[Zed-F]

2318: IIT9 hits at 3%, I swear I'm not cheating Then again, you have had similar luck

Not on TL+10 or IIT9. Later on Tundra at one point in one game of 3. Most of the time I had tech %s in the teens or twenties before they hit.

If I can invest my production at an 8% interest rate through tech, then I see little point in investing in factories at a 3% investment rate.

The problem is that this 8% interest rate can be an illusion, depending on what you are researching. If you are researching something that directly helps your early expansion, then yes it's worth it. If you're researching something in the consolidation phase to help you come up with a defensive force (bases/fleet) worthy of the name, then it's only worth it if you actually need to build such a force. If you've gotten to the point where you're reasonably well defended and ready to go on the offensive, then it's definitely not worth 8% anymore; it doesn't usually matter if you go on the offensive a couple turns later or a couple turns earlier. An investment in factories is constant so long as you are confident the factories will not be destroyed and doesn't depend on what you happen to be researching.

EDIT: The corollory to this statement could be that another simple rule of thumb might be not to invest in factories at poor worlds so long as there are techs you still need for early expansion, and start investment in factories at poor worlds only once you get to the consolidation phase (e.g. when you are ready to turn on computer tech research for the first time.) This will ensure that you will get the full benefit of the RPs for the techs that most need them, and that you will have at least a few factories in place by the time robotic controls hit.

I actually think this game is a good one to examine the impact of research on the early game in the sense that having more valuable tech to research early will only tell us that it's more important to do early research with more low-hanging fruit to grab. That is, of course, a trivially simple conclusion to draw. The fact that there were only a couple techs that were worth researching this time around kept us focused on a few critical points of which techs arrived when that mattered. Fewer data points to look at makes it easier to make comparisons with less noise.

 

[Zed-F]

I just checked your save file and noticed that you never built scouts and in fact scrapped your starting two. That makes a difference as well -- that's something you couldn't afford to do in a real game. I did build at least some scouts in my games. In my next game I will forego that and forget scouts entirely, we will see how things come out.

 

[Zed-F]

60 turns of Bears, round 4:

2301: We have a 100 Ultrapoor Fertile Ocean to the SW and a 50 Steppe to the NE. This time I again opt for the Ultrapoor Fertile Ocean. For this trial, I will perform research at poor worlds this time once again. I'm also going to try to take more advantage of research interest by spending a turn with high research at Aros when I am about to start a new tech. I'm not about to fiddle with the sliders every turn, however.

2303: Colonize Denubius, set production to research. I transport 10. Once again I start with Terraforming +10, though I thought about trying IIT9 first for variety's sake. I will start research once my initial block of pop arrives at Denubius.

2305: Spend 40 RP from Aros and 6 from Denubius to kickstart Terraforming +10. After this Aros will go back on factories.

2311: Terraforming+10 at 3%. Aros+Denubius: 60+28 pop, 67+0 facs, 98+14 income.

2313: Terraforming+10 hits, was 9% last turn. Research shifted to IIT 9, I spend 74 RP from Aros to kickstart it (plus 19 from Denubius.) Next turn Aros will go back on factories.

2318: IIT 9 at 3%.

2320: IIT 9 comes in after being at 10% last turn. I start research on Tundra bases and kickstart it with 110 RP from Aros (plus 34 from Denubius.)

One thing I don't know is whether what techs you have completed researching have any bearing on how far your RP can decay in any given field if you don't put any spending into it. That would be a minorly useful thing to know; otherwise you pretty much always have to leave a tick in any field you have started.

2321: I can build a colony ship in 5 years from Aros so I decide it is time to start colonizing.

2324: Aros+Denebius: 89+76 pop, 138+0 facs, 187+43 income. Colony ship due in 2.

2326: Colony ship complete.

2330: Centauri settled. Denubius sends 40.

2331: Colony ship complete, sent to Aquilae via Denubius. Aros goes back on factory construction and also invests in terraforming.

2333: Planets - Aros+Denubius+Centauri 98+63+42 pop, 166+0+0 facs, 220+35+23 income. Tundra bases at 4%.

2334: Tundra bases comes in, Enhanced Eco is next. I add Aros's 142 RP to Denubius and Centauri's 42 and 27 respectively for this turn and will go back to factories next turn. Denubius will send a few pop next turn to help max Aros post-terraforming.

2336: Aros maxes this turn at 220 factories and 110 pop, leftover production to colony ship construction.

2337: Aquilae colonized. Denubius and Centauri each send 15.

2338: Centauri and Denubius send 15 and 10 more respectively to Aquilae. I spread it out to lessen the (apparent) research hit somewhat.

2339: Aros completes colony ship for Rigel, next due in 3.

2343: Rigel settled, Denubius sends 30. Aros completes colony ship for Gorra, next due in 2.

2345: Aros completes colony ship for Hyades, next due in 3.

2347: Total pop 291+30 enroute, facs 220+25 (Aros/Aquilae), income 441. Colony ships enroute to Gorra and Hyades.

2348: Aros completes colony ship for Nitzer via Rigel, next due in 3.

2349: Gorra settled, Rigel sends 17. Gorra goes on research. Denubius sends 25 pop to Rigel in anticipation of colony ship enroute to Nitzer.

2351: Aros completes colony ship for Uxmai, next due in 3. Centauri backfills Denubius with 10 pop.

2532: Hyades settled, Aquilae sends 40. Centauri backfills with 10.

2353: Denubius sends 25 pop to Rigel in anticipation of future needs, such as the imminent colonization of Uxmai.

2354: Colony ship completes at Aros for Rotan, next due in 3. Colonize Nitzer, Rigel sends 12.

2355: Rigel sends 13 to Nitzer, Gorra sends 6.

2356: Colonize Uxmai, Rigel sends 13.

2357: Aros completes ship for Spica, next due in 2.

2359: Aros completes ship for Laan, next due in 3.

2362: Aros completes ship for Quayal, next due in 3. Spica and Rotan colonized. Enhanced Eco is somewhere around 2/3 done. I think the industrialization at Aquilae is worth the delay on Enhanced Eco. When it is done it will be a powerful research engine, and I'm not concerned about needing it to provide pop. There are lots of other developing worlds in the area to provide pop. Alternatively I can swap colony ship building there and put Aros on research.

Save file: http://www.civfanatics.net/uploads8/rbo-comp1-bulrathi-2361d.zip

Priming the pump with a turn of tech spending did seem to help a bit. Not a ton, but every bit counts I suppose.

 

[Catwalk]

The problem is that this 8% interest rate can be an illusion, depending on what you are researching. If you are researching something that directly helps your early expansion, then yes it's worth it. If you're researching something in the consolidation phase to help you come up with a defensive force (bases/fleet) worthy of the name, then it's only worth it if you actually need to build such a force.

By 8% interest rate I do mean tech that will actually boost your production by 8% of your investment, ie. Improved Eco Restoration that gives you 64 BC per turn all in all: 24 from extra colonist production, 40 from reduced cleanup spending and a little deducted from the cost of the tech because of miniaturization discounts. So yes, I'm referring to things that directly benefit your expansion.

EDIT: The corollory to this statement could be that another simple rule of thumb might be not to invest in factories at poor worlds so long as there are techs you still need for early expansion, and start investment in factories at poor worlds only once you get to the consolidation phase (e.g. when you are ready to turn on computer tech research for the first time.) This will ensure that you will get the full benefit of the RPs for the techs that most need them, and that you will have at least a few factories in place by the time robotic controls hit.

Hey, we're starting to agree here Now we only need to determine who's agreeing with who

I actually think this game is a good one to examine the impact of research on the early game in the sense that having more valuable tech to research early will only tell us that it's more important to do early research with more low-hanging fruit to grab. That is, of course, a trivially simple conclusion to draw. The fact that there were only a couple techs that were worth researching this time around kept us focused on a few critical points of which techs arrived when that mattered. Fewer data points to look at makes it easier to make comparisons with less noise.

True, there were some useful techs and there was indeed less background noise. Still, what should be done with the problem of gauging the value of techs in progress? One way to do it would be to set the end date to 2370 at the latests, by which time you must have researched Enhanced Eco Restoration. The player who finishes sooner then takes additional turns to reach the other player.

Remember that you don't have to be playing with sliders when playing the interest rate, although a calculator helps out greatly. We're only researching one tech for the most of this game, meaning that you just need to change total spending to make sure that you're putting in at least 7.5% of your progress in tech points. The real problem is counting your progress, which I admit that I have used EDMOO for, although I do keep manual notes as well just because it's faster.

Not building scouts saved some ressources, but the upkeep from my two scrapped scouts never made a difference. And I didn't spend the 4 BC i got from scrapping them On the other hand, you played up until turn 2362

My vacation is done, and I may not get access to a computer that can play MOO in the near future. It was nice doing this analysis with you

Med venlig hilsen,
Catwalk

 

[Zed-F]

I think we would have to set a specific tech which counts as the "last useful tech for expansion" and from which point we would say we are out of the expansion phase and into the consolidation phase as far as tech goes. After that tech was researched everything would have to go into say weapons tech so we could discount it. I would still rather have the game finish by a set date, +/- a year or so, but 2400 might have been better than 2361. OTOH, keeping it short means we can run more trials. 2380 maybe?

By playing with sliders, I include planetary spending as well as which fields I'm researching by how much. I'm not really interested in counting my progress with any degree of precision or adjusting planetary sliders to ensure my research spending is above a certain level. Nor am I interested in using 3rd party programs during a game (especially a competitive one) as that feels like cheating. Besides, it takes away from the immersion. I would rather have a simple rule of thumb that I could use to play with sliders once and then leave alone for the balance of the time.

There are a couple things I would like to understand better about tech, mostly pertaining to the "staged" nature of techs but the frequent references I have seen to RP totals:
- how does decay work relative to having researched a given tech. Can your RP total decline past the level needed for that tech if you leave it unfunded long enough?
- When putting in 7.5% of your progress in tech points, how exactly do you calculate what 7.5% is? Is it 7.5% of the points you put in since you started the specific tech you're researching right now, or 7.5% of the points you have ever put into the field?

 

[Catwalk]

I understand your feelings about detailed paperwork regarding research. A useful rule of thumb is this:

1) When you start a new tech field, kick-start it with a sizable amount of research
2) Next turn, scale down to ~1/14 of that amount
3) Every turn after that, increase your research by alternately 1/4 and 1/5 (giving an average increase of 22.5%)

This still requires some work, but once you get used to it you will do it automatically. I'm a bit of a perfectionist so I usually write down exact numbers in the early game. All you have to do is look at the research total on the tech screen and up it a little every turn. This is a lot easier when researching just one tech at a time.

As to how research works:
1) Only tech progress in the current tech being researched can decay.
2) I'm fairly certain that the manual is correct when it states that decay is 10% of the current progress in the tech currently being researched
3) You take the 7.5% of your current progress in the field, RP spent on techs already researched are irrelevant.
4) It is 7.5% of the BC you have spent on the current tech, but remember to account for tech interest. Whether you take full advantage of it or not, you'll usually end up getting at least 1.5 RP per BC spent. For drawn-out projects you can typically reach 2.5 RP per BC without delaying the project too much.

Example: You open up research in Weapons with 400 RP and pick Gatling Laser. Because this is the first time you spend RP on weapons technology, 10% of the amount spent is lost, meaning that you only get 360 RP. This only applies to the first time you open up a particular field, so you should always spend 10 BC or less on opening a field.

You now have 360 RP in your Gatling Laser project. An easy way to calculate 7.5% is that it corresponds to 3 points for every 40, meaning that 27 is the max amount you can invest next turn at the full interest rate. Those 27 BC = 81 RP, every BC after that = 1 RP. Assuming you spend 27 BC, you have 441 RP next turn. You can then spend up to 33 BC at the full rate, giving 99 RP. All numbers are rounded down, meaning that it makes quite a difference whether you start out investing 13 or 14 points in Planetology early on. All numbers are also rounded down on the tech screen, meaning that 23 BC split evenly between Weapons and Propulsion gives 11 BC to each field.

I agree that the techs eligible for research could be pre-determined. The time frame should then be set so that it's realistic to get there, and not much further. You would have gotten Enhanced Eco around 2370 in all your games, so that would be a reasonable time frame for this game (2380 sounds reasonable as a standard figure). After researching it, I would go on an expansion spree no matter what, so my stats wouldn't suffer from not being able to keep researching.

I think I have you beaten on that last game. Our stats are almost identical up until 2347. I think the main difference is that you send colony ships westward from Aros, whereas I settled Aquilae earlier and produced a colony ship there instead. I think the transportation cost is quite a significant factor on large and huge maps, making nuclear engines a better investment.

I don't think I'll be able to play anymore comparison games in the near future. The computer that I use for playing MOO at home is an old laptop with a floppy disk drive that has stopped working. The computer still works, but nothing can get in or out

I would be happy to comment on any comparison games that you should wish to run yourself, and I hope you'll be able to convince others to join. I think it would be useful to develop a simple shorthand code for reporting, to make comparison as easy as possible. I'm not used to reading reports, and I had some trouble comparing our results. Your reports were easier to read than my own, leaving info out near the end turned out to be annoying after all. I propose the following as a standard:

1) A status is done every 10 years listing pop, factories, actual production (meaning that you discount the amount that is used on clean-up), colony ships and approximate progress in techs currently being researched. I think the four stages on the light bulb correspond to 20%, 40%, 60% and 80% of the base cost, I'll check this out.

2) Comments are listed below the report
3) Col, XXX - YYY - Z = Colony ship built at XXX, sent to YYY, arrives in Z turns
4) Settle XXX = XXX colonized
5) Send X from YYY - ZZZ - Q = send X million colonists from YYY to ZZZ, arrives in Q turns
6) XXX Y% = tech XXX has a Y% chance of hitting
7) XXX hits - YYY = XXX hits, pick YYY next
8) As little background noise as possible should be on the map, meaning that the choice of planets to colonize should be fairly obvious unless that is what we wish to compare. Tech choices should be examined, to make sure they fit the examination topic.

 

[Zed-F]

I think I have you beaten on that last game.

I don't agree with that conclusion, but since we didn't play to 2370-80, I guess we won't find out on that one soon. Did you play the imperium while you were on a computer that you could use to upload your results somewhere?

 

[Catwalk]

I must admit that I got discouraged mid-game, and stopped playing in frustration. I wish I'd kept the save files so I could finish and report, if I run through it again I'd have an unfair advantage. I'm too much off a perfectionist (as you can guess from my microing in our comparison game), and I figured that I'd pretty much thrown any lead away I might have had up until then. Personally, I'd prefer to eliminate or minimize major luck factors. In normal games, no tech will ever be truly crucial. Again, my apologies for not finishing.

EDIT: Sorry for spoiling I have a few hours to spare, I'll try to see if I can get MOO running on this computer.

 

[Zed-F]

Normally we have a "no spoilers" policy in that we don't talk about games prior to the reporting date, success or failure. I'll only say that the scenario is quite possible to finish with 5 colonies and no Thorium.

You also have the option to play again and report a shadow game, if you like -- i.e. one that doesn't count for scoring purposes, but could be informative for comparison purposes.