Meklars vs. Klackons

[Zed-F]

A good question (that I'm not very sure about) is what techs are the best for early defense. NPG, deflector II and nuclear engines are among my top choices.

Depends on what you are planning to use for defense. Shields are not particularily useful for small fighters, which are the cheapest way to get weapons into space and keep it there (if the most subject to attrition.) NPG is good for fighters, but you may find that medium or large missle boats work better for attriting enemy forces via shoot-n-scoot, though certain specials can change that calculus in some cases. Do you favour a heavily mobile defensive force or mainly use missle bases, which have to be built per planet but have the advantage of being upgradable? Do you favour a more proactive or reactive style of defense?

I find that missles tend to be the most cost-effective option in the early to mid-game. Once warp 4+ engines start becoming available, beams and bombs start to take an increasingly large role.

-----

The question of interest for research is a very interesting one. Certainly it is less expensive to take advantage of research for any given tech in terms of absolute RP cost. However, this is counterbalanced by the opportunity cost of having to spend cash on research when it could be going to other needs. For instance, say you could spend X money on factories and Y money on research over T turns to finish maxing a planet and get tech A, or you could spend X money on factories at the start and then crash tech A, spending Y+y1 amount of money on the tech. However, because you finished all the factories first, you actually have earned more money over the total amount of time it takes to build all those factories and research tech A. So, which is better, to be more efficient on the research side by researching slowly, or be more efficient on the production side by building the factories first? If we're talking about factories + colony ships + research, then the equation is balanced even more heavily in favour of early production. Still, it's not always a clear-cut answer.

Moreover, there is another opportunity cost associated with research. Clearly you obtain more dividends from research interest if you research all fields slowly than if you dump everything in one field. However, one shouldn't ignore the benefit obtained from the tech being researched itself. If I spend all my research on an early game range tech and as a consequence settle an extra planet as opposed to what I would have gotten if I spread my research evenly, doesn't the benefit of having an extra planet outweigh the interest I gained from researching evenly?

I break up the early game into two phases: REX and consolidation. In REX, I am very selective about what techs I want -- I generally want specific techs to increase my ability to grab non-hostile planets, with a few exceptions. This means range 4-6, Improved Terraforming +10, and a pollution cleanup tech (preferably improved eco.) If there is a particularily juicy hostile world nearby which a low-level base will let me colonize, or several decent ones, then I will go for that as well. LR Colony ships can be a priority, depending on the map. On a large-huge map, possibly an engine tech, if the time spent researching the engine is justified by the amount of time saved getting colony ships to my worlds, without reference to a defensive fleet I haven't built yet. Guns? Shields? Computers? Forget it, I don't have time for those. The opportunity cost is too high; I need to focus solely on expansion at this stage. Similarly, the savings of spreading research across multiple fields and and/or of waiting for interest to accrue is eclipsed by the opportunity cost of delaying factory and colony ship builds, which is where just about every cent I can spare is going.

In the consolidation phase, I have sent colony ships to most of the non-hostile worlds I intend to settle, and while I may not be completely done my colonization effort, colonization is starting to tail off as much of the galaxy is already occupied and empire borders are largely set. Now is the time for research, building a defensive fleet, and the occasional colony ship or batch of such to settle hostile or particularily remote worlds. At this point, research across all fields starts to become more urgent, as the AIs are also starting to get out of expansion mode and start thinking about building up their own fleets. At this point of the game, research interest starts playing a larger role, and the opportunity costs of spreading research across all fields start to wane. There may be some techs that merit a particular focus, for instance if you think you are in a race with the AI to get the tech to settle a batch of hostile worlds, but for the most part slow and steady research can be a winning proposition at this point.

I tend to put an emphasis on Planetology techs, so I don't think you'll find much discrepancy there. I suspect that you start researching techs I would consider non-essential sooner, while you are still in what I would consider to be the REX phase. By the time I reach the consolidation phase, I will have at least a couple planets maxed out and ready to do the necessary research. Hence, whether to spend money on research or production is rarely an issue for me, even at (defensible) poor worlds. Either the research spigot is mostly off anyway (REX) or it's on but being taken care of at mature worlds.

[Catwalk]

Shield II was intended for missile bases only. I'm not very keen on using shields on my ships unless I have a major tech lead. I also use missile boats a lot, although I find them overpowered (as I mentioned in an earlier discussion). One problem with them is that they don't seem to do very well against transports.

Slow research vs extra factories and later crash research is indeed an interesting question, but I actually do believe that there often is a clear-cut answer, especially considering that the interest triples your investment rather than doubling it. The main question is whether you want to do all the paperwork required to keep track of your research, since I'm not allowed to use the editor that lets me track it (which I still find silly, but I'm trying to get used to doing without it).

I don't quite agree with your opportunity cost argument in favour of crash research. If you're able to foresee needing a certain tech at some point in the future, then you should start a trickle right away. If something unforeseen comes up (like too many asteroids on red and yellow stars, foiling your expansion plan) then you should go for rapid research. But even in this case I'll still take advantage of the tech interest. Instead of using a large part of my production for 8 turns, I'll use ALL my production for research on the first turn, excepting only that which is needed for clean-up, and production on rich worlds. Then I'll scale down to 7.5% of that and rise by 22.5% every turn. You'll get there in the same time and spend a lot less ressources. Only very rarely is it necessary to spend full production on research on consecutive turns, and it's usually a result of a poor expansion plan, poor scouting or unrealistic defense capability estimates.

I have no disagreement with your distinction between the REX and consolidation phase, and it doesn't look like we go for different techs. So we agree on two points out of four, namely expansion priorities and tech choices. That leaves drawn-out research vs crash research, and factories vs research on poor worlds. I'll use a spreadsheet to support my arguments.

Example 1, simplified and using abstract numbers only. I'm assuming that you can build an endless amount of factories, and that you will be needing a tech that costs 1000 RP in 20 turns for expansion purposes. I'm assuming standard eco restoration. I'm assuming that you need to get up to a 40% chance before succeeding. You have a total production of 100. First, a crash research version. The approach used is to spend all on factories for x turns, then switch to spending all on research until the research cost is met. The table below shows production and research time as a function of how many turns you spend building factories before switching to research. I have used fractions of turns to get a proper comparison. Format [turns spent building factories, production, research time]

0 100 12,09
1 105 12,68
2 110 13,29
3 116 13,91
4 122 14,55
5 128 15,21
6 134 15,89
7 141 16,58
8 148 17,29
9 155 18,01
10 163 18,74
11 171 19,49
12 180 20,25
13 189 21,02
14 198 21,78
15 208 22,54
16 218 23,32
17 229 24,10
18 241 24,89
19 253 25,66
20 265 26,45
21 279 27,25
22 293 28,06
23 307 28,86
24 323 29,66
25 339 30,47
26 356 31,28
27 373 32,11
28 392 32,94
29 412 33,76
30 432 34,58

No surprises here, total research time goes up the more factories you build, which is balanced by the fact that you have more factories after you're done researching.

Here is a similar table with the same parametres, but using drawn-out research. The approach used is to make an initial investment of x, then scale that down to 7.5% of x next turn and letting the annual investment grow by 22.5% every turn. This means that all research except the initial investment is tripled. Format [initial investment, factories, research time]. Note that I do not round. This means that the numbers won't be 100% precise if you try to duplicate them in the game, but the overall comparison will still be viable because the errors even themselves out.

100 168 17,05
90 174 17,55
80 181 18,14
70 190 18,80
60 200 19,55
50 213 20,44
40 229 21,54
30 252 22,99
20 287 24,98
10 340 28,38

This lets us compare the results (format [Crash production, Crash research time --- drawn-out production, drawn-out research time])

307 28,86 --- 340 28,38
252 25,66 --- 287 24,98
208 22,54 --- 252 22,99
163 18,74 --- 181 18,14
148 17,29 --- 168 17,05

The drawn-out method wins out every time. You get the tech faster AND you get more factories. Now, there is a third approach that this didn't cover, namely full factories for a number of turns, then the drawn-out approach. Also, these examples only apply for basic restoration, construction and waste techs, I'll account for those as well in the next example.

[Zed-F]

I found your example hard to read. Can you update the format so that all 3 sets of numbers follow the same format? Ideally, one indexed by time would make it most easy to follow.

OTOH, it also requires a ludicrous amount of micromanagement. I am not willing to spend the effort to (a) figure out what the optimum amount of research I should be doing at every world is and (b) adjust production empire-wide to ensure I meet that target. The improvement in game results is not worth the management headache.

I would be more interested in an example where you don't use an obscenely micromanagement intensive approach to research. As well, it would be better to create a test game and draw an example directly from that. Take a save game at a particular point, make the save game available so everyone interested can experiment, and describe those scenarios you are interested in describing. For example, save the game the turn before you are about to settle a regular world and a poor world, and follow what happens to your empire as those two worlds grow and you research stuff.

Even that will be an example with limited application since one needs to consider one's empire as a whole, but it would be much better than a column of abstract numbers for making a point.

[Catwalk]

I agree, this is not a very productive way of illustrating my argument. The problem is that the only computer I can play MOO on has no access to the outside world (the floppy disk drive failed 3 months ago), so I can't share files If you're interested, you could create a game and post the details here, we could then compare two different approaches. Save the game after initial scouting is done, so an expansion plan can be made.

The method I'm suggesting does not require as much management as you think, at least not early on. Especially not when you only do research in 1-3 fields, and you go for a long time span on the projects. Once you set the tech sliders properly related to one another, they stay that way and you only need to change the total amount of research. I don't normally use it middle and late game, unless I'm facing severe opposition. I'm storing many of the numbers I need on a spreadsheet so I don't need to recalculate everything all the time. Once you get used to it, it really isn't much work. I could create a spreadsheet to track the numbers, automating much of it.

Here is what I do:
1) Decide if I want a tech or not, and when I want it
2) Calculate if I have the industrial basis to support the research project all the way
3) Figure out how large my initial investment should be
4) Next turn, scale down to 7.5% of the initial investment
5) Every turn after that, increase my investment by 22.5%, until it reaches 10-20%, depending on how urgently I need it

[Catwalk]

I have access to a computer that can run MOO during the holidays, would you like me to set up a comparison game? It could be against one player on a huge map, in order to isolate the economic components from the strategic ones. I propose that we test Psilons and Klackons first, then Mrrshans and Meklar.

[Zed-F]

By all means, go ahead. I will be spending most of my time on Half-Life 2 over the next few days but I can probably arrange for some time on MOO as well. Post what saves you are interested in examining, and what specific scenarios you are interested in testing.

[Catwalk]

Will do. As I understand it, there are four issues to examine:
1. How much you benefit from drawn-out research projects
2. Whether to build factories on poor planets or not, and how soon
3. How much tech to go for in the early game

My claims are:
1. The benefit is substantial, and it is not impractical to take advantage of it in the early game
2. Poor planets should only build factories after there is no room for factories on normal planets, assuming that you have research going
3. I assume that we'll both be researching fuel cells when we have to, so any disagreement will probably have to do with how many planetology techs to go for

[Factoid]

This is an investment in growing the normal / rich / artifact planets faster (less the opportunity cost of transporting pop).

What do you mean by this?

By opportunity cost I mean the cost of trade-offs in general. If I am interested in the fastest possible economic growth of a planet I can either build factories or force-grow population. If I choose to build factories I sacrifice the opportunity of growing population, and vice versa.

This is classical economic theory, though I am probably being too loose in the way I use the term. Apologies for the jargon, the idea itself is obvious.

In transporting population from planet to planet it costs a few BC (can't remember, maybe 0.5 BC/pop) to transport population, and while the population is in transit it isn't contributing to the economy of any planet or helping grow population. So there is a sacrifice of short-term growth in transporting population around, even if the long-term result is overall faster growth.

When you say "grow population", are you referring to natural growth or eco pumping?

Eco pumping, specifically in preference to building factories.

Cheers,
Factoid.

[Catwalk]

By opportunity cost I mean the cost of trade-offs in general. If I am interested in the fastest possible economic growth of a planet I can either build factories or force-grow population. If I choose to build factories I sacrifice the opportunity of growing population, and vice versa.

I'm not too sharp on academical jargon. I'm struggling to keep up with you, but don't let that stop you

In transporting population from planet to planet it costs a few BC (can't remember, maybe 0.5 BC/pop) to transport population,

I believe this isn't the case. The actual productiondisplayed on a planet that is about to export colonists is incorrect, as it always deducts 1 BC per colonist, no matter what your planetology tech level. It is pretty much accurate early on, but try it out with planetology TL 50. It will correctly deduct the production of the colonists leaving, but it will display former production - 1 BC per colonist. I can't remember if it correctly deducts production lost from factories being inoperable, I think it does.

[Factoid]

I believe this isn't the case. The actual production displayed on a planet that is about to export colonists is incorrect, as it always deducts 1 BC per colonist, no matter what your planetology tech level. It is pretty much accurate early on, but try it out with planetology TL 50. It will correctly deduct the production of the colonists leaving, but it will display former production - 1 BC per colonist. I can't remember if it correctly deducts production lost from factories being inoperable, I think it does.

Hi Catwalk,

Yeah, the in-game display is a bit simple-minded. The green/yellow planetary production numbers are based on production with the current population and factories (capped by "full employment" - must have enough pop to run the factories). The *real* production is calculated after population increase and other stuff, so you get a small bonus in most cases.

The planetary production sliders *estimate* results based on the displayed yellow net production numbers. This is sometimes very inaccurate.

When half the population is about to be transported away the numbers are very wrong, ALL the sliders tell you they will produce roughly twice as much effect (ships / missile bases / factories built, waste cleanup, terraforming, forced pop growth, research) as they actually do. The Eco slider does tell you (even on a currently full planet) that population growth can happen, but as I said the real growth is roughly half the estimated growth.

I haven't worked out the order of all in-between turn events (IBT, what happens when you hit "next turn") but my planets-eye view of the IBT order is:

• Outgoing transports leave (not harassed by current in-orbit enemy space fleet).
  
  
• Waste cleanup??? Somehow happens before natural population growth, but depends on production, which depends on pop growth???
  
  
• Natural population growth. The non-ECO slider stuff.
  
  
• Current empire-wide costs (for espionage, ship and missile base maintenance, trade, empire-wide tax) are calculated.
  These tend to disproportionately affect low-production planets (truncation effect?).
  
  
• Ship and missile base costs drop, due to any technology learnt / traded / stolen this last year.
  (This can mean an almost-built ship is produced even if no planetary production is devoted to ship building this year.)
  
  
• Gross and net production is calculated and applied (ships / planetary shields / missile bases / factories built, terraforming, forced population growth, research).
  
  
• Technology advances made (notably, improvements to ground combat (gropo) take effect immediately).
  I'm fairly sure missile base tech advances (shields, missile types etc) DON'T happen this IBT.
  
  
• All ships for all races move. Ships built here but redirected elsewhere are not harassed as they depart.
  
  
• Space fleet/missile base combat resolved (which can cause population and factory losses), ships from the losing side that retreated leave orbit. Ships from the winning side that retreated (during combat) stay in orbit.
  "Third party" space combat (against other hostile forces) happens before combat against owner of the planet?
  What happens when several mutually hostile fleets (and transports too) arrive the same turn? Maybe player always fights against the maximum number of opposing races?
  Odd things can happen if the combat is a draw (times out with both sides still present, 100 combat moves?). An opponent Colony ship may get a chance to settle when it really should be forced to retreat.
  
  
• Bombardment. More pop and factory losses.
  Beware of non-hostile opponents (NAP/allies) glassing an enemy planet after you render it defenseless and before your gropos arrive... In Imperium 4 I was blamed for genocide (even by the offending ally) in exactly this case. Yes, I *intended* to commit genocide but it is still a game flaw.
  
  
• (If all the planet population was killed, all arriving attacking and defending transports are automatically lost.)
  
  
• Defending transports arrive, fight through in-orbit opposing ships (if any), added to defending population, truncated to maximum current planet population (after this IBT terraforming, but not counting this IBT waste or bio-weapon effects?).
  
  
• Attacking transports arrive, fight through missile bases and in-orbit defending ships (if any), consolidated survivors fight ground combat against defenders.
  
  
• If all the planet population was killed (also if the planet is not settled) and a surviving (non-retreating) fleet has a suitable colony ship, they can settle the planet.
  Player gets first dibs, useful when you and a NAP (non-aggression pact) opponent arrive at an unsettled world on the same turn. NEXT turn you may be fighting over your new colony against the opponent's colony ship (and other ships, and incoming transports), but at least you got the choice...

I have probably forgotten a thing or two and the order may not be 100%, I typed the list from memory (how sad is that?). Bio-weapon effects on maximum population may interact with defending transports arriving? Waste cleanup / generation stuff above is very likely wrong

There is a lot of weirdness related to fleets being in or out of range (and vanishing towards the END of the IBT, after combat but before Colony ship settling) if other colonies are destroyed or captured. Settling a nearby colony may happen after the fleet is declared out of range and vanishes.

As Zed-F recently said, each arriving attacking or defending transport "fleet" has to fight through individually then the survivors are lumped together for the ground combat. You can have more than one "All your ground transports were lost" message and still have survivors get through. I think the 300-maximum limit for incoming transports applies before losses from opposing ships or missile bases, but I'm not sure. Fewer, larger "lumps" of population (transported from fewer, larger planets) survive better than more, smaller lumps. Increased warp speed technology is a huge help (but inertial stabilizer has no effect). Ground combat transport technology still leaves transports exposed to missile base attacks, I believe it has a 50/50 chance of avoiding opposing fleet attacks for each lump (rather than halving losses). GC transport is sometimes useful defensively, to sneak population to a planet of yours past an in-orbit enemy fleet.

There are plenty of other IBT events that affect planetary production (and I don't know where they slot in). GNN events turning a planet rich or poor happen at some stage, also rebellion, plague, nova, radiation accident and others. A race might ask for (and you agree to) a higher trade agreement, or a race might declare war, wiping out current trade. Sabotage could destroy factories. The net trade benefit increases (can even decrease) at some point each turn. Does successful espionage of a gropo tech take effect in the same IBT? And so it goes.

I hope you appreciate this classic Factoid-style posting:

• Took hours to write.
  
• Kinda hard to digest (like Christmas fruitcake).
  
• Drifts off-topic to a whole lot of stuff you may not be interested in.
  
• Still doesn't address the point you raised.

The really important thing is the statement

The planetary production sliders *estimate* results based on the displayed yellow net production numbers. This is sometimes very inaccurate.

Cheers,
Factoid.