On the Wealth of Nations, by Jason Cawley

From Stars!wiki
Jump to: navigation, search
This was the featured article on 21 June 2009.

On the Wealth of Nations


Why race design is the key factor in Stars!


By Jason Cawley (jasoncawley@email.msn.com)


A Note from the Compiler: This article first appeared as a response to a question on the fairness of the Hyper-Expansion race design as a game design error or an intentional sort of thing. The response is an excellent review of why race design is so important in Stars!, and how today's HP/HG scenario developed. -DB


Hotblack wrote...


Don't jump directly on the topic...let me first tell the story :)


LOL. Well, I needed to read it to even understand what the topic was. So no problem there.


>In round 20 I was 1st with 1800 resources, 10 planets - 2nd had only 900 res with 10 planets and got his points throug better tech and more armed ships


OK. First thing to understand here is that those scores are low. Yours is ok for the kind of race you were playing, since you are going to grow pretty fast later with your bi-immunity. But typical standard races (non HE, non AR) with full size planets and high pop growth (like 18-20) can easily have 2500 resources at year 20. The best races (in terms of speed to that year) can have 4000, but those peak out early because they have sacrificed on habitat. But generally, 2000 that year is the normal target for a human race, and anything above it is considered "fast" and anything below it "slow" (to start I mean).


To see that these are reasonable, notice that is you just take the humanoids and sit still on the homeworld, building factories and mines for 20 years, you will be in the 1500-2000 range. Without doing anything really. Now, it is true that you can't just do that, because you'd throw your later growth out the window by uovercrowding the homeworld so soon. With lost time from pop traveling through space to new worlds, the inputs to buy things are lower too, when you spread. But that is still just the humanoids - and human races are going to be a lot stronger than them if they are designed well. Everytime a human makes a change to his race design *away* from the humanoid defaults, he better be sure he is getting "value for money" ;-) New players often don't see this though. Some seem to expect that *any* race the wizard lets them design, that uses all the advantage points (about) must somehow be the same overall strength. Not so, of course. Race design is a competitive process, that people can do well or badly. But a new player who hasn't learned that yet can sometimes expect that it is just a matter of picking menu items as he wills, and won't have any overall consequence for his race's strength in-game.


The truth is a little harsher. Not only can you not make just any change without weaking the default humanoids, but in addition you have to make changes that end up competitive with the other human players in the game. That means the "bar" is set by how smart your rival players are in their designs. The balance is that everyone has the same options at race-creation time, *not* that their exercise of them in this way or that will have the same results.


In round 30 I had 6K res and he 2K res


OK, he is 10 years behind the curve there; you are about average, to good. 4-4.5k can sometimes be enough at year 30, if the race is a slow developer but has lots of potential later on. The very fastest can hit 10k that year (but need good game conditions, and a lot of micro-management, to get that).


Noticing anything yet about the "track" I am describing? The competitive "top" and "bottom" at the various years are about 1 factor of 2 apart - very roughly, but that is the scale. Now, at the rates people are growing on those competitive tracks, 1 factor of 2 is a matter of 5 or 10 years time-difference (7-14% growth). Think about what an attacker has to do to hit someone. He has to research parts, make designs, build them up, remember bombers and fuel ships and the like; have ships go ahead to sweep mines; gather his production from his several planets (by gate for light ships for most, even heavy for ITs, no gates for HE), cruise in to enemy territory; beat any initial defenders and their base (so needs a few years build-up for the fleet), bomb defended planets. OK? Think it is real easy to do that to more than 1-2 worlds in 5 years? ;-) Or to more than say 5, if you are lucky, in 10 years? So, the defender can have roughly 1/2 the econ the attacker does - as long as he is also growing strongly - and have time to react and match the attacker's power. Also, 1 factor of 2 difference means getting one ally to help can balance one guy's edge ;-)


That is what determines the "spread" of "playable" designs, really. There is a lot of variety among playable race designs - but naturally only a small fraction of the *legal*/possible designs will fit the bill. Things designs differ over include resource timing (lots early, or slower on that score but more later?), econ vs. tech focus (many resources but exp tech? Works if you manage to get enough planets paying the fixed tech costs; can hurt if you don't), mineral-gathering ability (total minerals usually determine *long-run* fleet strength; but resources determine *how fast* the fleet grows in size up to its mineral limit), fast-spreading ability or lack thereof (things like IFE, ISB, IT PRT, cheaper prop tech, some other PRT things like IS fuel transports and freighter pop-growth). And all of those involve trade-offs - you can get more on one of them by giving up some of another one.


Interesting was the fact, that ALL the other players had only 1/3 of my resources and they all stayed relatively close together (Their races: JAOT, SS and AR)


All this tells me is they are newbies - beginner players. Your econ performance was about standard for intermediate skill level players (which is where you should be playing BTW - leave the beginners alone or teach 'em some before playing them again ;-).


Round 35 I had 9K


That's strong. The rule/standard benchmark "track" is to break 10k on or before year 40. Anything that breaks it in the mid 30s is fine on the speed score.


the 2nd wants to give up (agreeing in my win)...I think they don't have any chance to stop me in 15 rounds - I leed in all categorys (tech, ships, planets)


Oh sure, game over dude :-) But that is because of mismatched skill levels in race design, not anything about HEs per se.


The 2nd placed (a good friend) claims that my win is a result of a false game desing (or faulty? or gamedesing-error...


Nope. He has incorrect expectations about the race wizard, to be sure - and doesn't understand its competitive nature - as explained above. There isn't pure play-balance in the race wizard, though - the best CA races are more powerful than the best of other PRTs (at least in developing quickly for an early attack). But an additional balance comes into play in real games - diplomacy ;-) When all players have decent, competitive race designs, it is *much* harder for the best race to run away with the game. Because the others gang up on him. The fastest CA might get and keep 1st place during development - but others will be within a single factor of 2 of him if the player skills are about equal (certainly the next few in order of strength will be). And they can and often do combined their efforts to clobber the initial leader. Being in 1st, and known to be such, at year 40 or 50, can be an unhealthy thing sometimes. A big bullseye painted on your chest, I mean.


His arguments:
1) It was not the intention of the designer to let the seddlers delight be used in fleets other than colonizers, giving (small) freighters the posibillity of Warp 8-9 from the very start of the game (I made freighters and tanker with it)


Yes it was. The corresponding design option for other races is the lesser racial trait "improved fuel efficiency". That gives a race that takes it the fuel mizer engine at prop tech 2. That engine can function as a warp 8 or 9 engine when enough extra fuel is used with the fleets. Example - privateer, fuel mizer, 3 fuel pods. Gives 1400 mg of fuel, enough to go 3 years or so loaded at warp 9 (and come back empty), or more like 5 years at warp 8. That means out to 250-300 LY at quite high speeds. Also, you can boost even those ranges by using "boosters" - ships of light weight and high fuel added to the fleets, then split off after their fuel is used up. E.g. scout hull, mizer, fuel pod (scanner optional). Or destroyer hull, mizer, 2 beam weapons, 2 fuel pods. Various PRTs can make other ad hoc fuel ships - like SDs can use mini-minelayer hull, no mine pods, 1 fuel pod - gives 650 mg of fuel and costs about what a scout does. ISB can also boost ranges, since the con tech 4 spacedock can be put up rapidly for about 80 resources on a young world, and refuels every ship at the same location. ITs can put up 100/250 gates on an orbital fort for a bit more, and then move pop and minerals and ship up to 250 LY a turn - or more, overgating and taking a little damage.


HEs don't need to take IFE to get good settling speed. Other races need to have *some* moving strategy, of which IFE is the easiest but costs points. But there are other options for other races too. The thing that hurts is if a design has *no* moving strategy for early settling, and faces opponents who do have one. Then one can wind up putting alone at warp 6 to get planets 150 LY away only, with lots of lost travel time; while the "movers" go 250 or 300 LY in the same period, getting 2-4 times as many planets.


But it isn't particular to HEs.


2) He thinks this is even worse: decreasing growth from 15 to 7 gives 1100 more race-points to spend (since the HE-growth is doubled, the points you get should be less then for other races)


Yep, lowering the growth rate of an HE race a lot is a great way to get points. But not unbalancingly so. JOATs get points by taking No advanced scanners for instance; CAs by narrowing their hab ranges. Yes, the points the HE gets look a lot more than that. But HEs need to keep a lot of hab width, because of their 1/2ed planet sizes, if they are going to be competitive longer-term. Once a world is crowded, HEs don't get any more pop growth out of their worlds than other races. The level the crowding kicks in at is only 1/2 has high, and then the growth is doubled - result is the same number of colonists grown ;-) The HEs get their growth edge on the lower % of cap part of a planet's life cycle - or from better hab like immunities. But you had 14% with your doubling right? Well, standard races can often afford 18 or 19%, and that helps them both before crowding and after.


3) The 50% max-pop can be ignored, if the game can end after 50 rounds


No not really. Yes a year 50 end condition can be a bit silly, but someone ought to run with the leader to that year anyway, preventing a x % or next VC that early most of the time. If the skill levels are right, that is ;-) And the 50% size really can't be ignored. Typically, races will get to crowding being important in the 30s or so, and more significant in the 40s - unless they terraform a lot those years to stave it off a bit. By year 50, good designs can be in "linear growth mode" - meaning their best "breeder" worlds held at about 50% of capacity and offloading population every year, with other greens crowded and being filled to the brim. Yellows will still be working up for non-CAs, for a couple more decades, though. An HE race doesn't get any more pop growth in that "mode" than any other race with the same settings. Not that anyone would *play* a race with the same settings - they'd never get there with 7% growth. But with 18 or 19, they are higher not lower. The HE edge if they have one will have to come from hab (and with ground to make up).


Might also show your friend the 4% tri-immune HE concept :-) That race idea was based on the same notion he has. By lowering the growth all the way to 4%, an HE race can take just about every other strength there is. And with tri-immune hab, it will still grow pop 8%, while getting say 4.75 resources per 1000 pop (15 factory eff, 25 operated, 1/1000 pop eff). But notice what that 8% pop growth rate means, down to year 50. At most the race can grow 47-fold the starting pop (less with travel time losses, and any crowding at all). From a standard start, that means 1,172,500 pop, maximum resources about 5600. In an Acc BBS start, he starts with 65,000 pop (not the full 100k, because the adjusted pop growth rate is under 15%), so his max resources are about 14,500 (4.75/1000 pop * 65 start pop * 47-fold growth). I am sure that seems like a lot to a beginner, but the usually benchmark for that year is 25,000 resources in an unopposed test. A little lower vs. human enemies sometimes, to be sure - but still in the 20k range. 18-20% growth rate standard races can get that regularly; the best can get 2-4 times that. The 4% HE will still have other things helping - good tech expense e.g., maybe high mining efficiency - but he is going to be behind in resources.


It does balance in other words - darn close.


In the next game he likes to forbid HE, since he thinks this race is to strong!


Might tell him a little bit of stars lore history. There was a period, lasted about 3-4 month methinks, several years ago when low-growth HEs had become very popular and common. Many race designers weren't all that good, and there were lots of new stars players out there. The low-growth HEs were winning games.


At around the same time, though, in two different groups of players (some overlap though), several experts were inventing new ways to design strong races. One such was a fellow named Barry Kearns. He was making Hyper-growth CA races, with somewhat narrow hab made up from by CA terraforming, factory settings that cost a lot and were geared to go up quickly - like 12/8/16 G box - and high pop growth and pop efficiency (19%, 1/1000 typically). And he was blowning the HEs away in his games. So he posted some design ideas to this newsgroup. Others suggested JOAT and IS variants (I contributed an IS one, which I had already won my first PBEM with against a low growth HE :-). People started calling those "monsters", and stopped moaning about HEs and started moaning about "monsters".


It was thought the econ-focused PRTs (CA, JOAT, IS from free terra, +20% planet size, freighter pop-growth) were unbeatable by the other PRTs. But other experts stepped forward and offered monster designs that made benchmarks like 25,000 resources by 2450 in an Acc BBS start, from every PRT except Alternate Reality. :-) Barry took the lack of an AR "monster" as a personal challenge, and singlehandedly invented a new type of AR race - high pop growth, 1 immunity, other ranges very narrow hab - that made the benchmarks too. The monster-moaners didn't entirely quiet down, but now instead of bemoaning 1 PRT (HE) as unbalanced, or 3 PRTs (CA, JOAT, IS) as unbalanced, has gone to moaning about all possible race types :-) And by then everyone understood that the real story was *skill differential*, not race-wizard balance. So we have beginner, intermediate, and advanced games (and experts, as a class above advanced but who often play in advanced games for lack of enough pure-expert ones).


Another fellow, though, named Brian Price, had meanwhile (same time as Barry - from a little before and overlapping) had invented an entirely different idea. Instead of speed, he went for capacity. He trashed his pop efficiency down to 1/2500 to pay for the best possible factories - like 15/7/25 3G. His races (then) started slower, but in the end they got about 1/3 more resources out of the same planets. These races won games - acting as "sleepers" early on, they avoided diplomatic "bullseyes", then broke out of the leader-killing "pack" around year 60-70 and exploded to huge capacity by year 100+. He called them "hyper-producers" - HPs. Others have since made many variants of those, some emphasising long-run minerals rather than resources, as the real limit on late-game fleet strength (e.g. an expert active now, "spike" being his handle, advocates these designs today).


People also showed that high growth HE races were possible, with the low-growth ones acting rather like the HPs and the high growth ones more like Barry's "HG" set (aka the original meaning of "monster"). Also, the best AR designs have been refined by one Bill Bulter, to incorporate middling pop growth and somewhat better hab that Barry's original - and in Bill's hands they have made records for mid-game performance enviable by the standards of any race.


So, now there are at least 19 different "templates" of "unstoppable" race designs :-) And when irresitable force meets immovable object...


But I don't know, how to argue...since I only played one game and only with that race How can other races compete with 3 times more resources


They can't - not with totals that low. Their designs have to be improved. With good designs by all, races within about 1 factor of 2 of the leaders, in total resources, can compete as explained above - defender's extra time, and alliances. But growing slowly and with only 3k in the 30s? No chance, even uniting. Races that poor have next to no ability to project power outside their immediate homeworld area. But an alliance of 3 races each with 10-20k at midgame, against one 30k leader? No problem. That is what makes the "timing" trade-offs in race design so interesting, and competitive. Because the design trades you can make among good races are things like "hmm, 10 years slower, but 1/3rd more eventually" ;-)


For my next race I'd like to try CA, since I think this can get much resources, too


CAs can indeed. In pure growth terms, CA is the strongest PRT and by quite a ways. The resource records for fast-growth CA races in unopposed tests, Acc BBS start, are well above 100k by year 50. Seriously :-)


One thing your friends may not understand here. Race design is competitive, but everyone can test any race before a real game ;-) And many people do, and share the results - here. By playtesting a race unopposed, which once you are used to it takes about 2 hours, you can throw out all the design ideas that don't measure up in their growth performance, keeping only the designs close enough on that score to compete. Anyone who tests, therefore, can easily avoid playing a race that doesn't have enough growth potential. Not every race will realize its potential under real game conditions of course (early war, or just boxed in, or bad luck on nearby planets e.g.) - but that factor alone means at a given skill level, almost all the races in a game will be up to snuff (unless people are in the wrong games, which does happen some of course). In your game that didn't happen because the beginners didn't test, while you used a design that was a product of such "natural selection" (whether applied by you, or by your reading about what others had done, etc). Course, maybe you dreamed it up yourself whole cloth - fine :-) I did in my first human PBEM, finding a design very close to Barry's test-selected HG races from reasoning and 1-2 AI games.


But other races, like IT or PP don't have that amount of resources...what's the trick?


PPs are harder sure. ITs aren't though. ITs work by spreading, using their gates. They also start with prop 5 and con 5, giving them a solid "move" strategy in warp 7 privateers. Then they fly in enough pop to a new place to get a orbital fort with gate up soon; then their special ability to gate colonists and minerals lets them supply the colonies with more pop and germanium easily.


As for PPs, one econ ability they have is their packets terraform worlds. That can be used to "prep" yellows for colonization, turning them green quickly before the pop lands. Often used with narrow, 1 immune habitat (for instance, grav immune, temp and rad each 30 wide - 1/10 overall, but more like 1/3 with full terraforming eventually).


JOAT races gather points from NAS, OBRM, take wide centered hab - now their planets are 32% bigger. Which means 32% more pop grown of the same high-value worlds. Couple that with a 18-20% growth rate, and off they go :-)


IS races can take 18% pop growth, and grow 9% per year on the way to their worlds, arriving with about 1/3 more than they sent. Later, they can fill large freighters over their good worlds into a flying planet that doubles in pop every 8 years, and send it around filling all their planets to the brim - and using them for invasions too.


Other races don't have special econ edges, but do have special warfighting edges. An SD race can use its defensive ability along with a "blossom late" economic plan - like the HPs above, or by taking total terraforming. Then they defend in the midgame, gumming up attackers in mines and making alliances against the fastest races, while their undisturbed worlds grow to a big capacity and win later.


WMs can take expensive tech and/or somewhat narrower hab, but cheap weapons tech. The points gained can boost growth rate, factory settings, and speed things like IFE and ISB. Then they can launch powerful attacks early while others are still developing and buying factories. Why are they so strong early? Higher starting weapons tech, plus cheap weapons tech and a heavy focus on it, can get them a weapons lead. Weapons cost them less, so they can buy more ships for the same resources. Their speed bonus can let them fight with just fuel-mizer engines and stil decent battlespeed - saving early prop tech costs, and engine costs on their "horde". Also, at con 10 they get battlecruisers, far better than cruisers as a fighting platform (especially because of much better electronics - up to 7 elec parts vs. 4 for a cruiser - can be used to ensure 1st shot with more computers for example). Later they get dreadnoughts of course, the best fighting hull til nubians.


SS races can take more expensive tech and have their spying ability make up the resource loss. Or they can take cheap tech, couple with the spying, and enemies buy them most of their tech then ;-)


Etc. Every PRT has its strengths, but they have to be "played to" in a design. A race has to have a plan - a timing idea, a way of getting competitive resource performance, a moving strategy, etc. Can't just fly throw the race wizard in 3 minutues and take things that look neat - not if you want to beat competitive, skilled humans working at it.


I hope all this or some of it (perhaps the "brief history of stars race design" portion say :-) answers your friends' questions. Stars is a pretty well balanced and very competitive game. A lot of intelligence has been expended already on making strong race designs - that is the competitive part :-) Fortunately, much of the results of all that are available freely - here, and in the strategy articles on the various stars websites - so you don't have to re-invent the wheel to have a strong race design.


Also, any new player putting up a race design here in a post and asking for comment can expect a bevy of experts to pick over it, tear it to shreds, and re-assemble it into something competitive they like, or that reminds them of the original :-) To get something to try, that is a fine procedure.


In addition to the race design, though, new players will need to pick up some of the "how to" empire managing tricks people have found over time. It is especially important to realize the importance of long-term pop growth, and avoiding crowding to maintain that by moving pop to the right places at the right times. To understand those things, I *highly* recommend new players read an article (of mine - shameless plug) on the website called "Magic of Stars" - the name of the article being "Basic Game Play". It is listed under "other topics". If you need the M.o'S. URL, I know it is listed in at the Stars Directory page-o-links. The article is all about how to get competitive economic performance down to midgame - first using just the humanoids, then with a suggested "next step" improved race design (also a Jack of all trades race). It is very detailed; you can follow it right after finishing the tutorial even. I think it will save a lot of time and effort of any new player, who wants to get good, to go through that article and try out the things it recommends - it will probably turn most beginners into intermediate players in a week or less, if you play the short test-games it recommends. Of course people will want to learn how to play their favorite race idea afterwards, but having the building blocks in place from some "plain vanilla" jack of all trades races can make all the rest of that a *lot* easier.


Hope this helps, and have fun with stars!


Sincerely,


Jason Cawley