Page 1 of 2

Feedback needed: Production & Research Priorities

Posted: Tue May 17, 2016 7:06 pm
by MatGB
Edit at top: Research now done as well, and tweaks made as outlined later in thread:
Production Priorities
==
With Add priorities and attempt to balance Industry modifiers I've used the relatively new ability to decide when an effect is processed to try to balance out various effect timings to make the Good/Bad Industry species modifiers more effective. Before doing this virtually all effects were added after the species modifiers making them effectively irrelevent after the first 20 or so turns.
Early is for effects that are then modified by species traits, default should be just the species trait, late and very late for items we don't want modified by species (for either balance or flavour), and then the concentration camp set max is given a very high number to ensure it does just that.
This is a follow on from the work I did a few months back on prioritising the Population modifiers (before I did this virtually all modifiers happened before the species effect making Good Population stupid-good and Bad Population actively debilitating. I'm going to be working on Research priorities over the next couple weeks (for which a separate thread will be started when I've got my head around the basics).

So, as you can see from my brief commit message above, effects are split into two main groups, before and after the species modifier.

I have two questions for the other testers (ie anyone playing with the new Test builds onwards).

1) Have I set the priorities for all/most techs in the right places, are there any you think should/should not be modified by the species traits?

2) Does this make any particular species substantially weaker or stronger in a bad way?

For 2) I am specifically worried about Gysache (because I'm always worried about them) and Egassem (because, seriously, get a large world full of these guys with end game level of bonuses and see how good Ultimate Industry now is).

Due to the problems Bad Population gave them, Egassem have over the last couple years been given huge numbers of other bonuses, Great Supply and Great Attack Troops being the most obviously powerful bonuses. I need feedback from some players starting games with these new numbers and playing through to see if, when you play to those specific strengths, we need to tone down some of their bonuses in some way.

I genuinely think we almost certainly need to here. There's a good chance that for the next week at least the Egassem are the most powerful Playable race.

(NB: they'll lose a bit of power regardless when the Research priorities are set as they'll almost completely lose the ability to research things effectively on their own, probably, but even so the early power they have will be scary I think)

For Q 1) There are a few I was unsure about, I think setting Pure Energy Metabolism in the Early group is potentially too much (both ways, Bad traits lose a fair bit as well—on the other hand PEM is possibly an overpowered tech regardless of what we do with it), on the other hand having Collective Thought Network in the late group makes it a bit weaker for Good/Great species.

NB: the idea of setting some priorities to strengthen these traits in the early, mid and late game is definitely happening. The specifics of which effects are set where and just how good the difference is/ought to be is definitely negotiable. Whether we need to then rebalance other effects to reduce or strengthen species that win/lose as a result of these changes is also negotiable.

Re: Feedback needed: Production Priorities

Posted: Wed May 18, 2016 10:20 pm
by wobbly
Was there a discussion of this on the forum (if so link), so I can avoid rehasing old ground? Anyway I'm up to turn 36 of an Egassem start, default settings (including seed). I don't have many comments yet, but I'll give some raw figures in case your interested.

Planetary Ecology (Turn 3) -> Sub Hab (T8) -> Algorithmic Elegance (T13) -> Nascent (T19) -> Robot Prod (T21) -> Mil Robot (T25) -> Sym Biology (T26) -> Planet Bunker/Fusion Generation (T28) -> Planet Defence Network (T31) -> Laser/Zortrium (T32) -> Domestic Mega Fauna (T35) -> Orbital Construction (T36) -> Laser 2 (Expected T37) -> Organic (Exp T40)

Note: 1st Run w/new figures, obviously shifting robot production to 1st or 2nd in that order would of been better

Invade Scyllior (T12) -> Placed Starting Colony on inferno (T12) -> Invade Abbaddoni (T14) -> 2nd Scyllior Established (T25)

Starting World: Production 44.5, Research 7, Pop 24

Empire: Production 65.4, Research 28.9, Pop 79.9 (32 Egassem, 26 Scyllior, 21 Abaddoni)
Ships: 3 scouts, 2 destroyers, 1 frigate, 2 corvetes, 1 robotic (single laser, 3 zortium) - the robotic is new this round.
Production queue: 2 scyllior colonies, 1 roboship
1 outpost(for egassem) waiting on a scyllior colony to conect supply
MatGB wrote:With [url=https://github.com/freeorion/freeorion/ ... e65b7c9e18]
1) Have I set the priorities for all/most techs in the right places, are there any you think should/should not be modified by the species traits?
I question robotic production. Both for aesthetics & balance. It makes sense for a high-tech race that's bad at industry to replace it's workers with robots. Egassem simply don't need the bonus. Something that's bad at industry does need the full bonus if it wants to improve production early. Personal opinion is this is taking away an option of mitigating an early weakness. I'd wait for fusion generation to apply the bonus, otherwise good production races can crank early robotic ships very fast. I'd say similar for algorithmic elegance when it comes to research.

Here's my homeworld no.s: Expected Industry 60, Egassem industry +4.8, Fusion +4.8, Robotic +2.39, Ultimate Industry +36, Cultural Archives +12
MatGB wrote:With [url=https://github.com/freeorion/freeorion/ ... e65b7c9e18]
I genuinely think we almost certainly need to here. There's a good chance that for the next week at least the Egassem are the most powerful Playable race.

(NB: they'll lose a bit of power regardless when the Research priorities are set as they'll almost completely lose the ability to research things effectively on their own, probably, but even so the early power they have will be scary I think)
Not so sure Egassem actually have a research issue. Hard to tell this play through as I had Scyllior's early, but besides invading that good production & supply rate sets them up for exploring almost as quickly as laenfa. They do have the option of churning out colonies (fast production & good supply out way the narrow planet range on base settings) & invading fast then getting nascent. Long term they'll need a better research race but they're in a pretty good position to grab 1 well before they need it.

Re: Feedback needed: Production Priorities

Posted: Thu May 19, 2016 5:00 pm
by MatGB
This is really nice feedback, thank you.

Of course, playing as Egassem and having Scylior as your first native is as close to dream start that you could hope for, they're the perfect complement, so yeah, you had an advantage.

I see your point about Robotic Production and Algorithmic Elegance, I shall ponder this further—I was partially looking to make sure that species with an advantage were able to take advantage of it relatively quickly, and also that a disadvantage (like Poor Industry) is actually a disadvantage, but your counterpoint is sound.

Re: Feedback needed: Production Priorities

Posted: Fri May 20, 2016 8:14 am
by wobbly
Another 1 to look at is xenaphobics. I captured a trith world & if I set it to production I get 31.6 prod, -14.8 Xenaphobic Frenzy. Though maybe this is a good thing, the penalty used to be pretty weak.

Re: Feedback needed: Production Priorities

Posted: Fri May 20, 2016 8:34 pm
by Piwoslaw
I am 170 turns into my first game with this week's build (2016-05-16.b0c0de8). I almost always play Gysache, so have I noticed any difference in this game, compared to previous builds? Not really. (Maybe I don't know where to look?)

I was lucky to be close to a planet with Gis Guf Gthrim, so I could produce regular strength pilots from early in the game. My strategy is to invest as much as possible into research before cranking out just enough ships to fend off enemies. I did not notice my production nor research differing from previous games.

Re: Feedback needed: Production Priorities

Posted: Sat May 21, 2016 12:29 am
by MatGB
OK, wasn't as clear, to explain better.

Open a system containing your Gyisache set to Industry, hover the mouse over their industrial output and you'll get the breakdown of how you're getting the total number.

Before, the first number would be Gysache Industry Focus, the second number would be Gyisache Great Industry, then you'd get everything else.

Now, before the Great Industry modifier is applied, you get some of the other bonuses, including both Robotic Production and Fusion Generation, making the Great Industry modifier (which is a percentage multiplier) bigger. For large worlds in the later game that can be quite a significant difference, for smaller worlds before you've got that many techs you probably won't actually notice it much but it is going to be there.

As a contrast, put some of the Gis Guff Ghrim on Industry, they get the Bad Industry malus, this now will also be bigger, so the Gysache will now produce a bit more and the GGG a bit less than before. I haven't run the full numbers to see just how big the difference would be but it is there and in some cases (Ultimate Industry as a specific) it has the potential to be massive.

I haven't done Research yet, that's the next project, don't worry about that (yet).

The task doing this is big for two reasons: 1) I have to work out where all the potential bonuses to output are, from all the techs, buildings, specials, species macros, etc. I keep missing them or overlooking some obcure thing that rarely turns up or only applies in those weird circumstances 2) I then, after identifying and assigning priorities, have to play several test games switching between versions and reloading to see if I've made the right calls and if I've got the balance right. After I've done all that I then push it for the next Test version and ask you lot to give feedback, and potentially modify things). The actual coding is really really simple, but a little bit dull—it is, literally, copy/pasting one of 4 lines of code into every effect. I have the potential to screw things up in a very subtle, but potentially massive way doing this. Very very unlikely, but you could end up with some species being weirdly overpowered for no obvious reason at all (or, alternatively, completely and utterly useless, I suspect the Silexians will be even worse when I'm done research—ooo! Raargh, someone grab a Raargh homeworld and tell me whether they're worth the effort at all, ever, they've got Bad Industry and other weird maluses).

Re: Feedback needed: Production Priorities

Posted: Sat May 21, 2016 3:48 am
by Piwoslaw
OK, how's this? I loaded my game at turn 150, set GGG to industry and waited ~40 turns for it to stabilize. So here is a breakdown of the output:

1) Gis Guf Gthrim, large toxic planet with red star, population 40.0:
- Target Industry 63.0
- GGG Industry Focus +8.00
- Tech Fusion Generation +8.00
- Tech Robotic Production +4.00
- GGG Bad Industry -10.00
- Special Ferric Minerals +8.00
- Tech Adaptive Automation +5.00
- Building Industrial Center +24.00
- Building Solar Orbital Generator +16.00

2) Gysache Homeworld, large swamp planet with orange star, population 44.0:
- Target Industry 114
- Gysache Industry Focus +8.80
- Tech Fusion Generation +8.80
- Tech Robotic Production +4.39
- Gysache Good Industry +11.0
- Tech Adaptive Automation +5.0
- Building Cultural Archives +22.0
- Building Gas Giant Generator +10.0
- Building Industrial Center +26.4
- Building Solar Orbital Generator +17.5

3) Raaagh, small terran planet with orange star, population 22.0:
- Target Industry 42.5
- Raaagh Industry Focus +4.40
- Tech Fusion Generation +4.40
- Tech Robotic Production +2.19
- GGG Bad Industry -5.50
- Tech Adaptive Automation +5.00
- Building Gas Giant Generator +10.0
- Building Industrial Center +13.2
- Building Solar Orbital Generator +8.79

Re: Feedback needed: Production Priorities

Posted: Sun May 22, 2016 12:21 pm
by wobbly
Ok here's a weird one from today. Cray blocked in by maintenance ship (attack 3, struct 425, shield 3). So I can send a colony ship through but can't connect the supply line. With Planetary Ecology, Subterranean Habitat, Robotic Production & Fusion Generation:

Pop 25 (when fully grown)

Target Industry: 16.25 (after waiting a long time from switching focus)
Cray Industry Focus +5
Robotic Production +2.5
Fusion Generation +5
Cray Bad Industry -6.25
Cultural Archives +10

So some possible solutions to my issue:
Rage-quit & regen (most sensible but dull/unideal)
robo w/lasers (easy to research but built at above rate then takes time to actually clear maintenance ship)
orbital dry dock for repairs on other side (needs 30pp on other side)
using self-healing organics (this one probably works & cheap enough in pp)
sending ships across in clumps (kinda interesting, kinda annoying)
industrial centre (needs another 60pp)
etc. etc.

Basically Cray are going to need a reasonable way to get enough pp to deal with a sentry or maintenance ship if you don't want the solution to be quit & regen or hit end turn lots & wait.

Side note: Interestingly enough losing 2 production boosters pushes cray away from the robotic line for me. Robotics are expensive to build. Industrial Centres need to be built. Adaptive Auto needs colonies. Gas Giants need an outpost+build. Organics are the cheapest troop carriers, easiest way to grab a better production race. They can fairly quickly research the robotic techs after they've grabbed something that can build.

Re: Feedback needed: Production Priorities

Posted: Sun May 22, 2016 1:34 pm
by MatGB
Hmm, yes, that's useful. It pushes me more in favour of moving Robotic Production out of being affected and putting something later in instead.

(alternate suggesting: a single ship with defense grid sat in the system with the maintenence ship will open supply up for you)

Re: Feedback needed: Production Priorities

Posted: Sun May 22, 2016 6:53 pm
by wobbly
So decided to have a try at proving myself wrong before calling it enough testing for me. Cray, default settings.

Turn 4: +1 scout, algorithmic elegance, planetary ecology
Turn 5: +1 scout
Turn 6: Subterranean Habitat, Becrux colonized - medium radiated, 2xgas giants
T10 - Nascent, Large Hull single gun scout
T12 - Robot Prod. , switch homeworld to prod
T15 - Fusion Generation
T16 - Robot Hulls, 1st outpost ship
T20 - Orbital Construction
T21 - Kalas II Outpost (Medium Barren)
T22 - Mass Driver 2, Robotic anti-sentry 2 gun/2 armour (37 structure), switch homeworld to research
T24 - Mass Driver 3
T27 - Kalas II Colony established
T31 - Orbital Generation + Outpost on Becrux Gas Giant
T32 - Sentry Dead - Steredenn (Medium Barren Silmaline Crystals)
T34 - Zortrium
T35 - Becrux Gas Giant Generator
T37 - 2nd Outpost Ship
T38 - Industrial Centres
T40 - Planetary Bunkers
T41 - Lasers, Steredenn Outpost
T43 - Industrial Centre built, all switched to production
T45 - Active Radar
T47 - Planetary Defence Network
T48 - Steredenn Colony

So:

Homeworld: Pop 25(full), 20.5 ind (exp 23.7), 8 research (exp 7)
- Cray Industry +5, Fusion Generation +5, Robot Prod +2.5, Cray Bad -6.25, Cultural Archives +12, Ind Centre +5

Steredenn: Pop 1.12(13 max), 0 ind (exp 0.73), 0 research (exp 2)
- Cray Industry +0.23, Fusion +0.23, Robot Prod +0.11, Cray Bad -0.28, Silmaline +0.23, Ind Centre +0.23

Becrux: Pop 3.40 (4 full), 11.5 prod, 2 research
- Cray Industry +0.68, Fusion +0.68, Robot Prod +0.34, Cray Bad -0.85, Gas Giant +10, Ind Centre +0.68

Kalas: Pop 8.46 (13 full), 3.59 ind, 2 research
- Cray Industry +1.69, Fusion +1.69, Robot Prod +0.85, Cray Bad - 2.11, Ind Centre +1.69

or 4 turns/robocrusier (L) with all set to production focus. 1 industrial centre, 1 gas giant generator, 1 silmaline crystal
2 turns to laser 2
1 robotic mass-driver, 3 scouts, 1 combat scout, 2 turns to 2nd robotic mass-driver

Just looking at those numbers, I'm thinking I'll try making a beeline for that industrial centre. Then home world produces, colonies research, try & get a gas giant generator (or micro gravity industry) up quick smart.

Edit: Rushing Industrial Centre works: Algorithmic Elegance/Planet Ecology (T4), Sub Hab(T6), Robot Production(T9), Nascent(T12), Industrial Centres(T16), 3rd Colony(T18), Industrial Centre Produced(T22), Orbital Construction+Next Outpost(expected T24), 3 gas giant spots.

Re: Feedback needed: Production Priorities

Posted: Sun May 22, 2016 7:32 pm
by MatGB
wobbly wrote: Homeworld: Pop 25(full), 20.5 ind (exp 23.7), 8 research (exp 7)
- Cray Industry +5, Fusion Generation +5, Robot Prod +2.5, Cray Bad -6.25, Cultural Archives +12, Ind Centre +5

Steredenn: Pop 1.12(13 max), 0 ind (exp 0.73), 0 research (exp 2)
- Cray Industry +0.23, Fusion +0.23, Robot Prod +0.11, Cray Bad -0.28, Silmaline +0.23, Ind Centre +0.23

Becrux: Pop 3.40 (4 full), 11.5 prod, 2 research
- Cray Industry +0.68, Fusion +0.68, Robot Prod +0.34, Cray Bad -0.85, Gas Giant +10, Ind Centre +0.68

Kalas: Pop 8.46 (13 full), 3.59 ind, 2 research
- Cray Industry +1.69, Fusion +1.69, Robot Prod +0.85, Cray Bad - 2.11, Ind Centre +1.69
This is excellent, thank you.

I'm not happy that it's knocking out that much production to be honest, losing 50% of the first three industrial sources is over the top. It might be need to look at the 50% number, it was fine when it barely affected anything, now not so much.

So it's not debilitating, but it does make things harder and force you to do things you wouldn't necessarily want to do—looking at the Growth priorities, we did deliberately have one of the early techs not affected and the other affected, so I think that's the way to go, but not sure.

Does anyone have a strong opinion on this? Bad Industry is, with the current priority setting, particularly debilitating in the early game—is it better to reduce the affect, remove a tech from those subject to it, or accept that the malus is meant to be a malus and it's not excessive?

Re: Feedback needed: Production Priorities

Posted: Mon May 23, 2016 8:16 am
by Vezzra
MatGB wrote:Does anyone have a strong opinion on this? Bad Industry is, with the current priority setting, particularly debilitating in the early game—is it better to reduce the affect, remove a tech from those subject to it, or accept that the malus is meant to be a malus and it's not excessive?
Hm, if the good/bad boni/mali of species traits turn out to be too powerful/weakening I'd rather go down the reduce-the-effect road than have fewer boni/mali be affected by the trait. I'd prefer the majority of boni/mali to be affected by species traits, and have the percentage factor of the good/bad traits reduced.

Just my 2c. :)

Re: Feedback needed: Production Priorities

Posted: Thu May 26, 2016 12:13 am
by MatGB
Vezzra wrote:
MatGB wrote:Does anyone have a strong opinion on this? Bad Industry is, with the current priority setting, particularly debilitating in the early game—is it better to reduce the affect, remove a tech from those subject to it, or accept that the malus is meant to be a malus and it's not excessive?
Hm, if the good/bad boni/mali of species traits turn out to be too powerful/weakening I'd rather go down the reduce-the-effect road than have fewer boni/mali be affected by the trait. I'd prefer the majority of boni/mali to be affected by species traits, and have the percentage factor of the good/bad traits reduced.

Just my 2c. :)
Pretty much agree, having finally figured out how to use Grep to do what I needed, the BAD_MULTIPLIER is only used for bad industry and research, so when I do research I'll knock it down to 0.75 so that the bonus for Good is the same as the malus for Bad, and throw in some more effects—at some point in the future we might decide introducing VERY_BAD for some species to counterbalance Great & Ultimate might be necessary, but I'm not even going to think about that in the immediate future.

Aside: I'd been worried about Gyisache, but of course they have Good Industry and Great Research, so the bonus isn't going to be massive for them the way it is for Egassem.

Re: Feedback needed: Production Priorities

Posted: Fri May 27, 2016 1:27 pm
by LGM-Doyle
Wheel barrow, forklift, cyborg.

With a wheelbarrow a person's initial strength bonus scales linearly.

With a forklift the bonus no longer matters.

With a cyborg body the bonus no longer exists.

Mostly, I'm with Vezzra on this issue the species benefits/penalties should apply to everything. It is more flavorful, less arbitrary and species always matters. Everything is wheelbarrows.

From a 'realism' position one can always argue that any technology falls in the forklift category. However, from a game mechanics perspective that is the worst category, because it makes the choice of colonizing species less important. The player is forced to make a large number of pointless colonization decisions, because the benefits/penalties will be wiped out by forklift style techs anyway. So avoid forklift and make species benefits affect everything.

Except for Transcendence. Transcendence is the cyborg body. Break the Transcendence win tech into a series of techs that modify all of the species in your empire into a common proto-transcendent species. When an empire transcends all of their citizens disappear, leaving abandoned planets. Having the only techs that change species solely on the path to transcendence is an acceptable complexity increase.

Make the pre-transcendent techs that change species not uniformly good. This balances these techs, makes the Transcendence win progress visible to other civilizations and forces interaction of the different win conditions.

For example.

Pre Transcendence: Withdraw ++ Research -- Supply
Pre-Transcedence: Reflect ++ Research -- Industry

Lastly, make Transcendence require a Transcendence Beacon and a contiguous empire. It allows empires with a conquest goal to interact with empires trying to transcend, by isolating planets from the Transendence Beacon. Of course, you could just exterminate your own population, but who would do that.

Re: Feedback needed: Production Priorities

Posted: Sat Jun 04, 2016 10:43 am
by MatGB
LGM-Doyle wrote:Wheel barrow, forklift, cyborg.

With a wheelbarrow a person's initial strength bonus scales linearly.

With a forklift the bonus no longer matters.

With a cyborg body the bonus no longer exists.

Mostly, I'm with Vezzra on this issue the species benefits/penalties should apply to everything. It is more flavorful, less arbitrary and species always matters. Everything is wheelbarrows.

From a 'realism' position one can always argue that any technology falls in the forklift category. However, from a game mechanics perspective that is the worst category, because it makes the choice of colonizing species less important. The player is forced to make a large number of pointless colonization decisions, because the benefits/penalties will be wiped out by forklift style techs anyway. So avoid forklift and make species benefits affect everything.
You have persuaded me on this, almost completely.

However, doing it, assigning virtually everything to be modifiable and then balancing the modifier numbers is almost certainly going to be a fairly big task (the actual scripting a lot less than getting the numbers right after much gameplay). In addition, currently I don't think the AI pays much attention to good/bad industry/research when for example picking colonies or researching techs, if we rush this in then the AI will be making really bad decision on colonising fairly regularly which would degrade the play experience.

Thus, unless the AI team says getting it updated would be a trivial task, it's not one I want to do close to a Release (because even if that is delayed for real world reasons I don't want to force an even bigger delay because the balance is screwed up).

So, I'm going to do a simple balance pass for research priorities now so we've got at least something in there but in a way that it doesn't severely hamper the AI, then during the next cycle we can redo the whole paradigm and make sure the AI can handle the more complex colonising choices that it implies. Make sense?
Except for Transcendence.
This requires a whole new thread.

Rethinking Transcendence