"Race Design for Force-Gen Games" by Paul Vernon Lydiate & Thorsten Haupts 1997 v2.6/7

From Stars!wiki
Revision as of 03:52, 11 April 2011 by Gible (talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Author: Paul Vernon Lydiate

When I'd completed my first 2 pbem games I signed up for a third which would be force generated for 100 years before play began. Both the other games had finished after 100 years, and I was just getting to like commanding BBs with Doomsday missiles . It soon became apparent, however, that race design for force-gen games needed a completely different approach to normal race design.

To illustrate this, I did four separate testbeds using the races listed on Jeff Hudelson's web page (which can be found at http://www.geocities.com/Athens/Delphi/5722/stars.html). All these races were designed for normal (as opposed to force-gen) games, and where there were multiple designs for a PRT I used those which would need the least optimisation for force-gen games thereafter (number in brackets is the number in which they appear on the web page). The races used in the tests were CA (2), IS (2), IT (1), JOAT (2), SS (2), WM (2), HE, PP (1), SD and AR.

For the first test, (and those which followed) the production queues for each world were set at Factories Autobuild 300, Mines Autobuild 300, Max Terraform 10% (no point in building defences if nothing is going to be able to attack you for 100 years…), only excess resources used for research. All fleets were given scrap orders. Research was set at 0%, next field lowest, with the first field of research being the lowest field nearest the top of the list (energy at the top and Bio at the bottom) for the first test. Where GR was chosen a LRT, a normal field was chosen throughout ('same field' as next field to research). A Large/Dense Universe was used with maximum minerals (so that minerals couldn't be a factor), Accelerated BBS, No Random Events (so that comets smashing into homeworlds wouldn't be a factor) and public scores. One hundred years were then force-generated (100yrs seeming to be the most popular figure for this) and the results appear in the first two columns of the table below.

Optimising for force-gen play

The whole point of force-gen games is to allow players to start with the higher levels of technology needed to built the more 'interesting' ships etc. Naturally, the higher the levels of technology you achieve during the force-gen the better placed you will be when the game reverts to one move per year. The problem is that you have only one or two planets to fuel your research, so the first thing to do is ensure that your planet(s) generate(s) as many resources for research as possible.

You will see that the JOAT was the most successful of the 'one-planet' races in the original test (2nd in resources - a whisker ahead of the PPs - but 4th in tech levels thanks to the spying abilities of the SS race). The higher population ceiling which JOATs enjoy is a factor here, of course - especially since OBRM was a LRT of this particular JOAT too. This isn't the whole story, however. The JOAT's factory organisation was 15/7/25 with less G checked which, after 100 years, allowed their single planet to produce 5556 resources per year!

The other side of the coin is that the fewer the resources it takes to research in any particular area, the more tech levels you are going to end up with. Although +75% extra research fields can work well in normal games, it does require the early acquisition of colony planets to provide the extra resources needed. Having only 1 or 2 planets for the first century means that this isn't going to happen. Those races not having TT can get away with BIO +75%, but for force-gen games everything else should be normal or -50% less if possible. Since any more than two -50% fields gets very expensive on points, I gave each race two fields at -50%, the rest normal and Bio +75% (unless TT was taken) before undertaking the later tests.

Finally, LRTs need looking at, not only as a means of getting the points to allow the above modifications, but also because some otherwise advantageous LRTs are useless for force-gen play whilst other, less popular ones, can be vital. I'll go through the LRTs in order, leaving out those which have no bearing on force-gen play one way or the other.

IFE: Incredibly useful for games which start in 2400, but totally useless for those which begin in 2500. After 100 years of research ramscoops will have been available for years, and those races taking NRSE should have Interspace-10s. Even the AR race (which did by far the worst in all of the tests) ended up with Trans Galactic drives. Consequently, none of the optimised races in the later test had IFE.

ARM/OBRM: Generally, ARM isn't a good LRT to pick for force-gen games. Usually the potato bugs initially received will have to be scrapped immediately anyway, and it prevents you from taking OBRM. By the end of 100 years your initial world(s) will be swimming in minerals in any case, so it's probably better to make your mines more productive/numerous than take ARM. OBRM is an excellent LRT to take, however, as it gives you more population (and thus more factory capability) to crank out resources for that crucial first century.

GR: Although usually not a particularly popular LRT, this is a must for force-gen games unless (possibly) you are taking SS as the PRT. With GR, 15% of your research resources go to the each of the 5 fields you are not actually researching (totalling 75%) whilst 50% go to the field you are (for a total of 125% - despite what the manual says). Consequently, GR gives you 25% more resources going on research over a 100-year period. There is another decided advantage too. GR allows you to concentrate solely on your most important 'normal' field of research for the whole force-gen period and still end up with reasonable levels in all fields - rather than having to set 'lowest field' as the next field of research. This means that you don't spend any years concentrating on useless (to most races, anyway) research into the higher levels of Biotech. All races in the second test were given GR.

CE: Not a good choice at the best of times, and certainly not for force-gen games. When the force-gen is over, the scramble for free planets is exacerbated by the fact that scanners are so much better than they would otherwise be and ships can move a lot faster. Having engines which are prone to not starting in such an environment is not a good move…

NAS: Although a useful LRT in normal games, again it can be a severe handicap in force-gen games. By the end of 100 years, all races not having this LRT will have 320X scanners, and some will have 400Xs. These races can colonise those planets close to their homeworlds immediately, without having to send scouts out first. Even a SS race, which can build ships with penetrating scanners anyway, will have to spend a turn (which could still be more than one year at this point) building them. The only PRT which can possibly manage with this is the JOAT - so long as one of the starting ships can be left in orbit around the homeworld for a hundred years instead of being scrapped. Even with a JOAT I'd be tempted not to take NAS, personally. In the second tests, however, the JOAT was the only race to have this LRT.

LSP: Again, not usually a popular LRT in normal games as it puts you at least a couple of years behind your rivals. In force-gen games, however, by the end of a hundred years your beginning world(s) will be almost maxed-out regarding pop anyway, whether you take this LRT or not, and the points come in useful elsewhere. You will remember that the JOAT in the initial test was the best performing 1-world race despite having LSP? All races in the second test had this LRT.

To fund the above amendments in order to optimise the original race designs for force-gen play, I tried wherever possible to get the needed points from the factories and mines page - increasing the number of colonists need to produce 1 resource, bringing factory costs up to 8 if lower, and increasing the cost of mines. As mentioned above, the HW will be swimming in minerals at the end of 100 years and should easily be able to provide the minerals necessary for building factories and starbases on new colonies.

If this wasn't enough, I took growth rates down as far as the 15% mark. As the HW will have almost maximum population after a hundred year force-gen, it is possible to get 2 hi-hab worlds to 33% immediately and use these as breeder worlds - both to populate further colonies and get the HW back to maxed out levels fairly quickly. In this sort of environment, a very high growth rate is not so important.

I tried to leave the hab ranges of each race alone, if possible, as this is such an integral part of a race's 'character'. The exception was the AR, whose hab range was increased for the purposes of the third test (see below) so as to enable the ARs to find a second planet to colonise within easy reach.

As you can see from the results of the second test (columns 3 and 4 in the table), virtually all races did much better in both areas after being optimised (the JOAT suffered a slight loss in production in return for the 12 extra tech levels it received). The HE and AR races still did badly in comparison to the others, however, due to population limitations (integral for the HE and because it wasn't possible to upgrade the starbase in the case of the AR). The IT and PP still did far better than the rest by virtue of having 2 starting planets. The ITs excelled because their second planet had a higher hab percentage than that of the PPs - the luck of the draw.

It follows that, in games where only 1 turn is allowed before a hundred year force-gen takes place, the smart money is always going to be on ITs and PPs (or maybe SSs or JOATS with OBRM if the player feels like a challenge…). Since ITs are far more popular that PPs it is likely that the game universe will be largely populated by the former. Conversely, no-one thinking things through (or having read this article) would ever consider a HE or AR race for a force-gen game of this nature.

Levelling the playing field

The remaining sections of this article are primarily aimed at people intending to host force-gen games in the future (and who don't want to end up with universes populated almost solely by ITs and PPs).

Race Test 1 Test 2 Test 3 Test 4
Tech Res Tech Res Tech Res Tech Res
CA 72 3682 84 4392 93 8601
IS 72 3616 82 4412 91 8701
IT 80 6529 93 8086 93 8086
JOAT 75 5556 87 5315 93 10147
SS 78 3448 92 4402 98 8552
WM 70 3136 84 4420 88 8172
HE 61 1703 73 2171 86 8150
PP 76 5537 88 7457 88 7457
SD 72 3448 81 4381 86 7893
AR 59 983 65 1261 72 2011 73 3262

One answer to this - already used by some hosts - is to exclude IT and PP races from force-gen games, but although this gives a wider spread it still only leaves 6 serious candidates out of the 10 possible PRTs. A better approach, IMHO, is to allow non-IT/PP races to submit moves at both 5 and 10 years in addition to their initial move - but not to allow them to build any ships over and above those they start with and insist that their ships are scrapped or stop moving from year 10 onwards. This allows each race to have 2 planets (or 4 in the case of HEs, but this evens out as their total planetary pop is half those of other races) for the vast majority of the force-gen period, and evens things out considerably - as columns 5 and 6 of the table (where this approach was used) illustrate. In fact, where this approach is used, non-IT/PP races can often do better than the IT/PPs, since they have the opportunity to choose their second planet whereas the IT/PPs are stuck with the ones they are given. This is the reason for the relatively poor showing of the PP race during this test. The PRT always likely to do best under this system is a JOAT with OBRM. For this reason, it might be an idea to stipulate that JOATS, if used, must not have OBRM as a LRT if it is intended to force-gen a game on a 2 (or 4 for HE) start-planet system…

The only race still at a severe disadvantage here (despite the race design being amended to widen the hab range and thus make the finding of a second planet nearby possible) is the AR - because it was not possible to build the bigger starbases this race needs to increase population during the 90 years of force-gen. The only way around this is for the host to follow the force-gen year by year and add empty space-dock, starbase and Ultra stations to the AR planet production queues when they become available and/or the planets' pop reaches 25%. Naturally, this increases the time it takes the host to do the force-gen, but at least it should make it feasible for AR races to be included in force-gen games. The results I got with AR even doing this weren't particularly good (see columns 7 and 8), but could probably be improved if Energy and Construction were both -50% (I took weapons at -50% instead of Con with the amended AR race) and if Energy was researched exclusively (leaving GR to provide advances in the other fields).

Remarks on PreGen games - by Thorsten Haupts

Pre Gen games are a totally different thing to normal games - especially those pregenerated over a period of 50y or more. In basic hints to race design I agree with Paul despite some differences. But there are some special needs for these games (Setup in the first turn, 100y pregen, no movements allowed for the initial ships if not forced to scrap them).

  1. Research
    GR is highly recommended in these games, but there must be a focus on two fields: electronics and weapons! Why these two? The first one is a no brainer, the better your initial planet scanners are, the easier to find habitable planets from the beginning. But weapons may be crucial to your success. If you are on Jihads and the other guy you unfortunately are a neighbour to is on Juggernauts you´ll get killed if he gets to know it! Your homeworld is a precious thing to get - tons of minerals and (using neutron bombs) all fabrics built :-). Given nearly the same ressources and both on Super computers a player with Juggers just builds the better warships no matter what you do. Remember the high costs to jump from 12 to 16 in weap. And you will not have the time to research in the first "normal" turns after the PreGen - you´ll need the ressources to get out colonizers, privateers and freighters as fast as you can. Almost every other player after the PreGen must be expected to be on warp 9 if not 10 engines than so there´s no time to lose.
  2. LRTs
    While a high growth rate is not that important (in fact you may lower it to 13% using nearby "good" planets colonized with 100k of colonists or more as breeders) I recommend two LRTs to take first: NRSE and LSP. With GR you may be on warp 10 engines even if you set the engines research to "costs 75% more" AND your main research field is weap or elec cause the Interspace 10 comes at tech 11! Even more useful if you take the IT as PRT cause the stargates allow you to build freighters equipped with Alpha 8 engines to move around fast. LSP is the second nobrainer - the 100y PreGen results in a disadvantage nearly unvisible. The points you gain here are better spent on cheap and effective fabrics.
    OBRM seems to be a nobrainer but is not! Remember battleships are heavily expensive in minerals, especially ironium. If you want to (or are forced to) build many early on, you get short of ironium earlier than you might think. In my testbeds before PreGen games (settling on mines effectivity between 11 and 14 and No. between 10 and 15 I got "only" from 18 to 30000 tons of Ironium when game got going. And this may cost you when you have to build warships early on (to take advantage of your better weaps or if attacked). Cause when time brings better tech you may be unable to build the better warships you will no doubt need.
  3. Expected results of race design
    At first the combination of "25 fabrics cost 7 or 8 each effective 15" coupled with expensive mines and ineffective colonists seems the obvious combination. But if you take PP or IT this may cost you too! In testbeds nearly half the time I got the second planet germ poor. Result was sometimes I could not build 2/3 of the fabrics possible on this planet in the PreGen thus heavily harming research! The effect was a difference of up to 20 tech levels meaning the main field to gain only a level of 15 instead the expected 20. Even worse the effect on colonization: Given expensive mining you´ll have to ship Germ with your colonizers to give the colonies the start boost needed. If the home world was not that good on mineral concentrations too you may have a hard time trying to develop your colonies. Not this high but not neglectable was the effect of the second planets initial value. If this is bad (15% at beginning) may be you´ll find it only at 55% after PreGen, a difference of up to 2.000 ressources compared with a 95% second planet for another guy who got more lucky. My testbeds showed a big stress on luck in minerals and hab in PreGen games overall.
  4. Judging the PreGen conditions closely
    I found combinations from 25 to 2600(!)y of PreGen in game announcements over some months. Whilst the first were to speed up the start the last was to get tech maxed out regardless of the players settings. But the difference between 50 and 100y of PreGen is higher than the number. Your research mainly is done after maxing out the one or two starting worlds meaning the main gain in research comes after the first 50y. The effect of a second world which is bad (15% for example) is much higher than in a PreGen of 100y. If not taken TT as a LRT you will find yourself unable to bring those worlds to much better values cause you don´t have the techs for high terraforming. In this case the CA or the JOAT may do as well as an IT or PP or even better. Therefore TESTBED a PreGen game using different race approaches before joining. Remember the "bad guy" coming out of the PreGame much better than you - how much time will he give you to colonize and develop? The result of the PreGen determines wether you´re in the ballpark or not - failures in choice or design of a race may cost you any chance in the game afterwards.