SMAX - The Will to Power - mod

[Alpha Centauri Bear]

Me neither.
😂


Well. Probably "competitive" is not the right word. I'd say not too cooperative.

[bvanevery]

I pioneered the use of r/4Xgaming to get new blood into the SMAC community.  I've brought substantial traffic to this site, and maybe indirectly even a new user or two.  I've mentioned the work of modders I've gotten along well with, i.e. Tim's Will To Power.  I've encouraged the other modders to get out on Reddit and get more users, and they've done so.  I've done plenty to "rise the tide that floats all boats".

But yes, there is still competition.  And I still think that so far, I've done the best job with this game.  As far as what the original game was, and what it should be.  I've contributed substantial playtesting effort to helping WTP "catch up" and polish / iron out its rough edges.  Far more than anyone who was was a strict mercenary about their work, or terribly business minded, should do.  I have that open source background, and ideologically, I still believe in some of that to some extent.

But I sure as heck draw the line on making mod managers for Yitzi's Patch, which helps my work in no way at all, and would just burden me with months of additional boring development.  Someone else can do that if they want it.  That's another thing about open source.  You want something bad?  Then how about you go do it.  I know that lots of people are capable of doing it, with not terribly advanced coding skills.  Do they actually do it?  No, they don't.  So forget 'em.  I'm not here to save 'em.

I mean, I made my own work the best it could be, pouring blood on the table, within the scope of a .txt only mod.

There are a few things I'd actually like, out of some kind of "minimal binary alteration" project.  Fixing probe team rules, is first among them.  But I'm darned if I'm going to get into the difficulties of ASM reverse engineering, and illegal binary distribution, while simultaneously trying to get a commercially viable 4X game going.

I mean, I can be writing a DirectX 12 3D engine.  If only the rest of my life plate would clear up, and I actually got a new laptop in here.

My next "low hanging fruit" would be to look at where OpenSMACX is at, nowadays.  But, you know.  Life.

[DrazharLn]

It's mostly not competitive. Induktio based their work on scient's freely shared work and decompilation database and Alpha Centauri Bear based theirs on Induktio's freely shared work. The various projects don't combine because of creative differences of opinion and because it would be more work to coordinate that.

Bvanery's mod isn't part of that lineage, but they have offered lots of free playtesting to Alpha Centauri Bear.

[bvanevery]

I finally ordered the new, modern laptop, so in about a week, that's going to be focusing my attention rather differently.

It's worth noting that my own work is under Creative Commons Attribution Noncommercial license.  Nothing's stopping anyone from using my work, long as I'm given appropriate credit.  But, with the "creative differences", it seems people want to program things and not mess with .txt files so much.  Or, they figure they can mess with .txt files themselves and don't see value in utilizing my work, despite ~4 calendar years and ~15 person months put into it.  Programs have a way of propagating themselves, in ways that data driven designs do not. 

Some people have done enough of their own .txt modding that they genuinely don't see a need for my work.  They might draw inspiration from some things I've done, similar to how I've drawn inspiration from other people's .txt mods here and there.  But they don't need my work for any heavy lifting.

Whereas, people think they need AI code.  Squeezing every last drop out of existing AI code, via data driven design / environment around the code... I'm not sure why that doesn't occur to more people.

[Alpha Centauri Bear]

Some question I minded to ask everybody long ago. Reminded by tech speed discussion on Thinker forum.

Tuning tech speed to be not too fast and not too slow is an extremely tedious work. It requires a lot of relevant statistics as a basis of all computation. Otherwise, all beautiful formula will just suck.
With that in mind I ask users permission to add tech research collector code that would dump research info to the file that you later can send to me. It should probably accumulate data from multiple games to reduce number of message. I then can share these data with all developer if they are interested in it. Any objections?

[Alpha Centauri Bear]

Some potential solution for problem that was nagging me long time already: reduced efficiency of high end regular units against natives. In short, high end regulars becomes less and less effective against army with natives as their relative cost grows but their psi strength not. This is combined with proposition of increasing unit strength with reactor instead of reducing their cost.

• Reactor adds 10-15% combat bonus. It does not change the price.
• Regular units get small psi bonus proportional to its weapon/armor. Currently, I think something like 5%. I.e. 20 strength item also confers 100% psi bonus.
• Natives cost slightly less than in vanilla in compensation for above.
  

  Mind Worms	4
  Spore Launcher	4
  Sealurk	6
  Isle of the Deep	8
  Locusts of Chiron	8

• Natives do not require maintenance at all and not only they are parked in fungus. Parking them in fungus is too tedious support reducing micromanagement.
• Optionally, natives require regular maintenance. Either way I want to exclude this "in fungus" logic.

Another idea that I don't insist on but that flows along psi bonus above. Empath/Trance ability gives some small flat bonus (25%) plus proportional of of weapon/armor (5%?) and cost proportionally to unit cost, correspondingly.

[bvanevery]

Any objections?

Not personally, but I think it is a misguided use of your effort.  I think you should be playtesting something and making a decision.  Not hoping that by collecting data and throwing it over the fence to someone else, that someone else is actually going to make a decision.  Newsflash: since it's your project, they won't.  Not unless they're motivated enough to fork your project and "put things right".  No reason to believe that's gonna happen at this time.

Here's what I think you should playtest: my mod.  AFAIAC I don't have any problem with tech pacing at all.  Nobody's told me differently either, although I sure wish more people were disposed to chime in about such things.  Anyways, I think I don't have a problem, and you do, because I've played both of our mods plenty.  You should play mine and see if that helps you figure out what's wrong with yours.  I have no doubt your tech pacing is off, as I've filed things in your Issue Tracker about it.

[bvanevery]

Again, play my mod.  Do you even make it to "high end" units, for this to be some big deal?  In the early to midgame, you can't even make your own mindworms anymore.  Last test game I played, it kept me from assaulting Yang's rear line.  He was building all the +100% defense against psi Trance units I told him to build.

Natives do not require maintenance at all and not only they are parked in fungus. Parking them in fungus is too tedious support reducing micromanagement.

That's a bit of a giveaway, increasing the tendency to use the mindworm as the one and only perfect combat unit.  Generally you have to pay for Clean Reactors.  When you capture a unit and it's Independent, that's a somewhat rare resource, so you typically don't want to just throw those things away.

Optionally, natives require regular maintenance. Either way I want to exclude this "in fungus" logic.

This is the better option if you're absolutely determined to simplify.

[Alpha Centauri Bear]

Not personally, but I think it is a misguided use of your effort.  I think you should be playtesting something and making a decision.  Not hoping that by collecting data and throwing it over the fence to someone else, that someone else is actually going to make a decision.  Newsflash: since it's your project, they won't.  Not unless they're motivated enough to fork your project and "put things right".  No reason to believe that's gonna happen at this time.

I don't understand your blame. I did playtest tech speed many times. That's why I picked parameters those are there now. However, I understand that my statistics is not exhaustive just because of my playstyle and sheer number of cases. I just want to expand to many people playstyles and get more data. I don't ask anybody to do any conclusions on this.

Here's what I think you should playtest: my mod.  AFAIAC I don't have any problem with tech pacing at all.  Nobody's told me differently either, although I sure wish more people were disposed to chime in about such things.  Anyways, I think I don't have a problem, and you do, because I've played both of our mods plenty.  You should play mine and see if that helps you figure out what's wrong with yours.  I have no doubt your tech pacing is off, as I've filed things in your Issue Tracker about it.

I understand you mod uses vanilla tech cost formula?

[bvanevery]

I just want to expand to many people playstyles and get more data. I don't ask anybody to do any conclusions on this.

If you like more data, then collect data.  It is of very little use though.  Is the player going to give you an opinion to go with the data?  If not, then all you know is they did X.  You don't know whether they considered it good, bad, or indifferent.  You don't know if it was in harmony with what they thought was their playstyle, or actively working against them.

Garbage in, garbage out.

I understand you mod uses vanilla tech cost formula?

Yes.  And all of my development, was empirically derived, with tons of playtesting, using this as a constant.  When things took too long to get, I put them earlier.  When things were too easy to get, I put them later.  After 4 years it certainly gets results.

Recently I playtested plain Thinker in conjunction with my mod.  Pretty quickly, I turned off the alternate tech costs, because they are not an improvement.

[Hagen0]

Getting data from MP games like that is a reasonable path in principle. I know of a Civ4 modder who did this. The problem here is that you need every player to send you the data by hand which is not that likely to happen. Bvanevery is right about that. He is also right to say that data without interpretation is of limited utility. If there was enough of it to run statistics on it it would be different. However, there are not enough games. Moreover, the rpgcodex games were played with ever-changing mods and setups, the player skill and intent (winning, roleplay) varies wildly. This kind of data is not useful.

As for the tech costs, early on vanilla tech costs work just fine. They are much better than the attempt to rebalance them in Thinker mod. You need certain techs to get going early on. Chiefly Centauri Ecology but also early SE choice (Planned, FM). The Thinker costs make the early game boring and tedious. You also have no shot to research military technology in time if you meet an aggressive neighbour. You just die.

The problem with vanilla tech costs is that they don't increase enough in the late game. However, this may be a lost cause to begin with. The archive of games at http://dos486.com/alpha/ has already been mentioned. In the University game thawk is making 10k research every turn. Any attempt to increase tech cost to meaningfully keep up with this is probably silly.

[Neil]

Getting statistics from people who care enough about your mod to provide it, seems like a good idea to me. If you only measure your own games, you won't capture different play styles. I don't understand the negativity around it. You won't get perfect data, but it's not like your target audience is the mass market anyway.

In my opinion the target should be that "on average", a player reaches the end of the tech tree around the game turn limit, or maybe a bit before. There needs to be enough time to digest and adopt the new stuff from later technologies before even newer stuff comes to replace it or the game ends.

Part of the problem seems to be that the amount of energy you can generate in the late game can vary considerably, depending on what has happened. Perhaps tech costs can also take into account the number of turns since the beginning. If the player is well ahead of the curve (from a theoretical baseline), costs go up a bit. If the player is well behind, costs go down. Obviously you don't want this to be so extreme that it enables bad players. The main objective is to even out tech progression for all factions generally, rather than a "catch up" mechanism for behind factions. I got this idea from Hearts of Iron, which attaches an "estimated" date a faction would be expected to discover a technology, and adjusts the cost based on the current date.

You're not going to please everyone. At some point you have to go with your gut and decide what you think is best.

[Alpha Centauri Bear]

That is correct. There is some average growth pattern and there are adjustments. The latter are there to reduce the fluctuation range between say barren and fertile worlds as needed to make them not too extreme. It should not override average pattern completely. WTP has both average and adjustments components. I am gathering statistics to ensure my basic pattern is not too far off the average gameplay development.

As for early tech cost I agree with you too. Calculations show that one tech should be discovered on average in 350 / 85 = ~4 years. 20 labs is the perfect cost for first technology assuming two bases (= 4 worked tiles) producing 1 energy each tile + 1 from HQ. That gives exactly 4 turn to discover it. Subsequently, following tech should cost somewhere 25-30-35-40-50-60-.... Of course, since Thinker and WTP set tech cost not by research sequence but by level, all first level tech should not all cost 20 as it will make further researches too quick. So the first level tech should be somewhere 35-40 on average defining affordable learning curve.

[DrazharLn]

I like that native units are free when in fungus. I think it's good flavour and encourages you to do thematic things with the worms, like hiding them in the fungus or have them patrol fungus.

[Alpha Centauri Bear]

I like that native units are free when in fungus. I think it's good flavour and encourages you to do thematic things with the worms, like hiding them in the fungus or have them patrol fungus.

There are some interesting aspect playing natives: psi combat ignoring weapon/armor, moving in fungus as on roads, fungus combat bonus, full healing in fungus. These alone is very sufficient to encourage their usage and create a flavor. However, there are "encouraging" and "forcing".

Not paying maintenance only when parked in fungus and only on this turn is too micro and it actually forces player to end turn in fungus or suffer hefty penalty. One can move by road/river but should end turn in fungus. If anything it is discouraging aspect of native management. Think about it. For every native unit on a map every turn player is charged 1 mineral depending on how the hell they completed last turn for these particular unit type! Why anyone should care how do I move my units?! This is pure micro that has nothing to do with strategy at all and quite little with flavor.

Bear with me. I don't absolutely mind any fungus native bonuses. I mind how game penalizes specific parts of unit movement. For example, full healing in fungus is pretty fine feature because 1) player willingly decides to spend a turn to heal unit, 2) player willingly decides where to place unit while healing, 3) healing takes whole turn of immobility. Whereas when human moves unit from A to B they don't control it every freaking turn. They just point to destination and click. If it happens native end their turn not in fungus - player pays. To not pay they have to move these units by hands calculating every move. Ugh.