SMAX - The Will to Power - mod

[DrazharLn]

I dunno about that. Is tech trading or stealing even a significant factor in making the game shorter? I also don't see how it makes the game more fun. Stealing techs is fun and creative and trading techs makes the AI feel more real because they proactively contact you and they have well-written dialogue. I don't want to be discouraged to do that by a penalty I'll be paying forever. A new nation-wide maintenance cost would also be difficult to make readable and understandable to the player.

I think the real problem has to be established to try to solve problems like this. What about a short game is bad for the experience of human players and why? Once that has been established then there may be many different approaches worth trying for solving the problem and these don't necessarily need to change player-visible mechanics.

E.g. these are different problems with potentially distinct solutions:

- human players get conquered out of nowhere by AIs that are more advanced than they expected
- human players lose to AI builder or diplomatic victories before they've got into the game
- tech progression feels too fast or easy and that undermines feeling of achievement or time passing, or makes it unsatisfying to play some factions
- players want more time to play with stuff before new things are introduced
- players find conquering the world too quick or easy
- etc

The first two can be fixed solely by changing AI behaviour; the last two may be better fixed by changes to combat, etc.

A kinda relevant story: the Old World designers don't think it's fun if the AI wins a builder-type victory in a 4X game because it's hard for players to see it coming and usually feels unsatisfying. So, in their game the AI cannot win via the main "ambition" system that players are intended to win by. AI can only win by "victory points". AI victory points act as a kind of visible, variable clock on the length of the game. The feature is really easy for players to read and avoids underwhelming conclusions where an AI wins a victory condition you didn't understand or couldn't really influence. Instead, the AI can still win, but it's clear to the player when that might happen and it is easy to understand how to delay an AI victory (because the players also score victory points).

This solution was easier for them to develop and meant they could focus on building a rich "ambition" victory system for players without worrying about how the AI would use it or how that could be visible to the player.

Rambling a bit more, some games go beyond this kind of asymmetry to choose or influence what the NPCs and environment will do based on a director system that tries to model a human game facilitator. The director makes predictions about how the players are feeling about the game and makes changes to the environment, AI behaviour, number and location of AI, equipment, etc. to try and induce emotional changes to match desirable patteens (e.g. tension builds to a climax, then there's a relaxation before the next thing; or, the surprise and elation of succeeding when you thought you would fail, or finding something you need or that would be really fun just as you need it). This is a big shift from the more obvious simulation approach where the environment mostly follows physical rules and the NPCs are modelled as individual agents that try to achieve their own simple objectives.

Obviously game designers try to build systems and agents that focus on player experience with either system, but a director approach is systematic rather than ad-hoc about modeling the player's experience and changing the game to try and induce specific emotions and behaviours from players.

I think it would be cool to know more about how different strategy game designers have experimented with more director-type systems.

[bvanevery]

I guess faked emotional journeys are all very fine and well for people who don't actually wrap their heads around the gory details of 4X games.  But I think overcoming gory details is how intelligent challenge oriented players actually achieve satisfaction in the genre.  The main emotional difficulty for such people is whether the AI's behavior is perceptible, and within a ballpark of being fair / not cheating.

If the game was lying to me about my agency within it, just rubberbanding me if I did well or poorly, I'd be quite annoyed with the game.  It's my prerogative to stomp a game.  I can't stomp something if there's no valid resistance.

[Alpha Centauri Bear]

to slow the game down, we could encumber players that buy and steal tech.

I have already scrutinized AI willingness to trade techs. As for stealing, I think it is already kind of not easy. One need to have a war with other faction (or risk a war by stealing from friend). That makes probes vulnerable. Stealing second time from the same base is more difficult, etc. If needed, I may make it even more difficult.

Some people say lowering probability does not matter as player just build more probes to steal it. Sounds pretty strange to me for the game where economy is everything. There is a significant difference between building 10 probes or 100 probes. One can build more combat units instead. Everything has its price and trade-off. Again, I am ready to discuss means to reduce it. Share ideas.

[bvanevery]

The main actual cost to me as a player in the real world, when stealing techs, is how long it takes me to push a probe team across a map.  I play on Huge maps and there's definitely a lot of times in a game, where it's less effort to build my own research infrastructure than to push the probe teams to other empires.  I often wait until other empires make themselves closer to my own, before bothering to raid them.

[Alpha Centauri Bear]

Is tech trading or stealing even a significant factor in making the game shorter? I also don't see how it makes the game more fun. Stealing techs is fun and creative and trading techs makes the AI feel more real because they proactively contact you and they have well-written dialogue. I don't want to be discouraged to do that by a penalty I'll be paying forever. A new nation-wide maintenance cost would also be difficult to make readable and understandable to the player.

I think the real problem has to be established to try to solve problems like this. What about a short game is bad for the experience of human players and why? Once that has been established then there may be many different approaches worth trying for solving the problem and these don't necessarily need to change player-visible mechanics.

E.g. these are different problems with potentially distinct solutions:

- human players get conquered out of nowhere by AIs that are more advanced than they expected
- human players lose to AI builder or diplomatic victories before they've got into the game
- tech progression feels too fast or easy and that undermines feeling of achievement or time passing, or makes it unsatisfying to play some factions
- players want more time to play with stuff before new things are introduced
- players find conquering the world too quick or easy

Good analysis. I also don't think preventing/slowing technology exchange affects the game duration much. In my personal strong option it is the nature of 4X game accelerated growth. Since it is highly accelerated the smallest speed up or slow down at the beginning have a tremendous effect on future progress and, therefore, on the game duration. Original game designers did a lot of adjustments to keep accelerated growth at bay. Like early yield restrictions, etc. That was an exact reason why early game felt soooo slow. Well, it is either that or super fast growth later. Removing these restrictions + adding formers at start *alone* may account for fast game progress. I tried to tune up tech progression but this is indeed secondary. Faster expansion and economy growth is still there.

Keep also in mind that all AI improvement mods are built with advanced players in mind seeking challenge. With that in mind most of the "AI beats the player" problems you described above are the matter of difficulty level adjustment. When you play chess program you probably not playing the highest difficulty to get beaten all the time. Same here.

- tech progression feels too fast or easy and that undermines feeling of achievement or time passing, or makes it unsatisfying to play some factions
That is not the problem of this mod but the overall number of technologies in game. 85 tech in ~350 turns = one tech per 4 turns. That's the average you should see. Taking that most techs give you some small bonuses this seems to be normal way accumulating small benefits over time. The fact that in vanilla it took 20 turns to discover a tech early game and then 2-3 tech per turn late game is just a skew. I am not even sure if this was on purpose or just lousy balance.

- players want more time to play with stuff before new things are introduced
I agree with that. And that is quite difficult to achieve. Sure things can be more evenly distributed across time. Like 12 different weapons can each occupy about 350/12 = 30 turns, which seems to be adequate enough, providing not all of them even can be used. Plus 20-50 turns after discovering last tech to play with latest inventions.
I would probably accept slightly shorter game of 250-300 turns but anything shorter than 200 creates the sense of not controlling the progress. Things just happen.

- players find conquering the world too quick or easy
Hmm. That should not be a problem anymore in this mod. Or I need to continue working on AI improvement.
😉

[Alpha Centauri Bear]

The main actual cost to me as a player in the real world, when stealing techs, is how long it takes me to push a probe team across a map.  I play on Huge maps and there's definitely a lot of times in a game, where it's less effort to build my own research infrastructure than to push the probe teams to other empires.  I often wait until other empires make themselves closer to my own, before bothering to raid them.

Yep. There are many elements to stealing. All things combined I don't feel it is super easy and game breaking exploit.

[valox]

If the game was lying to me about my agency within it, just rubberbanding me if I did well or poorly, I'd be quite annoyed with the game.

this is a great point.  Negative and Positive feedback loops can be very destructive when they

1) reduce player agency

2) create game-play experiences that contradict the game narrative. (Like when racing games slow you down when you are in first place, but speed you up when in last place)

In my personal strong option it is the nature of 4X game accelerated growth. Since it is highly accelerated the smallest speed up or slow down at the beginning have a tremendous effect on future progress and, therefore, on the game duration. Original game designers did a lot of adjustments to keep accelerated growth at bay.

A while back, I found myself puzzled by a game design choice: the inclusion of a feature called HabComplex. Initially, it felt like an unnecessary obstacle, but upon reflection, it serves as an excellent negative feedback loop. For players with strong starts, HabComplex requirements act as bottlenecks, dampening the effects of an early lead. On the flip side, those lagging behind typically have the required tech in place, thus avoiding population caps. This feature enhances the player's sense of agency, offering an obstacle that can be overcome in various ways.

However, there's room to refine the player experience, particularly in regard to the aligning the game's typical economic curve and the games tech cost curve. As AlphaCentauriBear pointed out, the game's duration is highly dependent on the pacing of the early game. Tech advancement can snowball from mid-to-late game, which we seem to agree is not ideal.  We seem to agree that the early game should be a little more spicy, but we don't want shorter games over all.

We could get the best of both worlds! You know, add additional terraformer at the start or something.  Make the game spicy like that--BUT then re-calibrate the cost of tech advancements. Run 10 or 100 test games and create a table of economic curve.  Also, table the cost curve of the tech.  Then we write a function that uniformly adjusts the tech costs to align with the typical economic curves we have harvested. I hope I'm making my point clear, which is that we can use two or more techniques to get the experience we seem to be aiming for.

[valox]

Oh, and concerning probes, they are 100% the best example of a positive-feedback loop.  Categorically, they allow losing players a come-back mechanism.  They also increase player agency, because the stealer has to make a plan and implement it.  The system also maintains the victim's agency, because they can either do due diligence to protect themselves or not.

By framing it in this way, I think I can carve out the issue that i see (and this may be relatively minor in comparison to other opportunities being discussed).  Its possible for the human players to think quite a bit more abstractly than an AI.  As such, it is possible to pretty much ignore tech on purpose, and focus more heavily on economic and military goals, and use the "come-back" mechanism of the probes and trading to gain an advantage. And there you have it, the core problem that I'm trying to point out with trading/stealing.  It would be fun to find an elegant way to maintain player agency, but not give players a way to leverage those options WHILE in 1st place.

My original suggestion about implementing a penalty for stealing and trading tech too many times... falls short of this goal.  Ironically, the logic above demonstrates that such a change would actually hurt losing players ability to make a "strong comeback" and wouldn't really stop a "winning" player from the gimmick i described. But, maybe i'll think of something truly elegant in the future. Or, maybe i need to demonstrate the issue more clearly if other's don't know what i'm talking about.

[Alpha Centauri Bear]

We could get the best of both worlds! You know, add additional terraformer at the start or something.  Make the game spicy like that--BUT then re-calibrate the cost of tech advancements. Run 10 or 100 test games and create a table of economic curve.  Also, table the cost curve of the tech.  Then we write a function that uniformly adjusts the tech costs to align with the typical economic curves we have harvested. I hope I'm making my point clear, which is that we can use two or more techniques to get the experience we seem to be aiming for.

Yep. That is exactly what I did. I ran many test games and calibrated tech cost so that the tech pace is roughly uniform across the game. Keep in mind that this calibration is specific for the mod and even version of the mod as changes from vanilla economics curve could be quite significant.
Average pace, obviously, does not guarantee same experience between games and poor/great experience, random factors, etc. 50% variation from average is a norm.

This being said, let me reiterate my point. Tech pace definitely needs to be tied to economy pace. Otherwise, one would just either discover everything too early or too late comparing to logical game evolution which creates kind of unpleasant effect. It is economy pace we need to take care of. After that readjusting tech pace is a piece of cake.

So far many modders/users agreed that vanilla early game is excruciatingly slow. Whereas late game is a complete opposite with all the satellite resources and other multipliers. This does not only refer tech pace but the whole development in general. As such, a lot of effort was spent on smoothing it up. Giving players some actions early but tone it down later on. I saw it in many other mods and tried to follow same principle in this one too.
Un-very-fortunately, changing game mechanics is a very dangerous adventure. There is a risk breaking it down beyond repair. All low hanged fruit were picked up already. Like removing yield restrictions and giving formers at the beginning or cutting down satellite resource yield at the end, etc. Some super powerful projects and strategies were nerfed, etc.
I would say at the moment it does look much-much smoother than vanilla. The current discussion is just about possible options to make it even more smoother. However, even if we don't solve it, and game ends in 250 turns, this still would be interesting enough playable experience.
Then, keep in mind, that this is only for highest difficulty. On lower difficulties AI has less economical bonuses and develop slower, so overall game progress is also (could be) slower. Need to check.

[Alpha Centauri Bear]

Oh, and concerning probes, they are 100% the best example of a positive-feedback loop.  Categorically, they allow losing players a come-back mechanism.  They also increase player agency, because the stealer has to make a plan and implement it.  The system also maintains the victim's agency, because they can either do due diligence to protect themselves or not.

By framing it in this way, I think I can carve out the issue that i see (and this may be relatively minor in comparison to other opportunities being discussed).  Its possible for the human players to think quite a bit more abstractly than an AI.  As such, it is possible to pretty much ignore tech on purpose, and focus more heavily on economic and military goals, and use the "come-back" mechanism of the probes and trading to gain an advantage. And there you have it, the core problem that I'm trying to point out with trading/stealing.  It would be fun to find an elegant way to maintain player agency, but not give players a way to leverage those options WHILE in 1st place.

My original suggestion about implementing a penalty for stealing and trading tech too many times... falls short of this goal.  Ironically, the logic above demonstrates that such a change would actually hurt losing players ability to make a "strong comeback" and wouldn't really stop a "winning" player from the gimmick i described. But, maybe i'll think of something truly elegant in the future. Or, maybe i need to demonstrate the issue more clearly if other's don't know what i'm talking about.

I think you are placing carriage in front of the horse even though you described the situation quite well.
You are right that obtaining tech by other means than research diminishes research value. It is quite possible that reducing research allocation in favor of economy + tech stealing/trading may be more effective. Well, go ahead and try it! 😊 Maybe you discover some new super strategy nobody thought of.

As a rule of thumb, any not dumb game rules make player/AI to search for optimal strategy. The one more knowledgeable of the game rules and more capable of deeper choices analysis wins. Changing game rules to tilt balance between different strategies is kind of neutral proposition. It does not make it better or worse (for most) - just different.

Imagine you play chess and dislike playing knights (for some reason). Then you propose reducing knight value since other players use them too effective against you. Thus you invent another game but not many people will be too excited, I guess.

The goal of this mod and similar mods (I am not talking about total conversions) was to keep game as close to original playing expectations as possible. I.e. reduce grinding, any micromanagement that does not actually support player agency in the game, restore to expected balance of different features: economy, research, conquest, exploration, infrastructure, expansion, improvement, diplomacy, intrigues, etc. - making sure none of them shadows others.

[valox]

I think you are placing carriage in front of the horse even though you described the situation quite well.

hehe, i'll own that for sure.  I'm probably a hammer looking for a nail here.

The goal of this mod and similar mods (I am not talking about total conversions) was to keep game as close to original playing expectations as possible. I.e. reduce grinding, any micromanagement that does not actually support player agency in the game, restore to expected balance of different features: economy, research, conquest, exploration, infrastructure, expansion, improvement, diplomacy, intrigues, etc. - making sure none of them shadows others.

And this is the honey pot that compelled me to join the community.  When i heard about this mod, I quickly saw the value of this mods' goals. Very excited to play it this weekend or something for the first time.

As a rule of thumb, any not dumb game rules make player/AI to search for optimal strategy.

yeah, that's one of my favorite topics.  Elegant game design is about increasing the number of meaningful & enjoyable ways to solve a problem with the minimal amount of complexity required to do so.  So a 'dumb rule' is one that leads to "automatic" decisions for decent or expert level players.

also, as long as the 'stakes' are clear for a given game piece, a multiplayer (err multi-player/ai) environment create balance organically through player on player action.  And that is EXACTLY why late game isn't very enjoyable in the vanilla version.  Its hard to clearly predict the balance of powers or do anything fast enough to re balance the game through agency.. turns into a death march to victory or death.

Anywho, i'm rambling. I love this stuff.

[valox]

This being said, let me reiterate my point.

i'm tracking with you much better now.  I see that it is the bending of the avg economic curve that's a challenge, not the tech cost curve; and that solutions need to be as non-invasive as possible.

One idea, is to not be so totally universal in how you bend the tech cost curve.  For instance, maybe you scale the economic-tech costs up differently?  Perhaps, you provide a transformer at the beginning, but explicitly increase the cost of the first few economy techs? 

In StarCraft2, eventually the did the same thing.  Early game was too boring for expert players and for audiences, so they gave you like 3x the number of workers at the beginning of the game, but made your resource supply less durable.... leading players to lots of fun action at the start, without causing the game to end any faster and without skipping the early game entirely. 

So, in terms of this mod, something spiritually similar could be implemented.  Heck, it could even be a matter of giving people a terraformer, but then increase the cost of building terraformer, by 15%, decreasing the efficiency of terraforming by 15%.... leading to a more interesting early game, without shortening the game?

[bvanevery]

I just played a ton of WtP and I don't remember the early game being the problem.  I remember quitting in mid to late game, because I couldn't stand to do something anymore.  I'm not clear on what, because I lost discipline for writing things down / providing feedback.  I'll be honest: I eventually went back to playing my own mod.

In other words, "should there be 1 more terraformer at the beginning?" is barking up the wrong tree, IMO.

I do think that going back to WtP, after playing vanilla or my mod or even just Thinker, always requires a big mental adjustment on the part of the player.  A key thing is how to make use of Recycling Tanks early on.  They aren't just a little bonus, they're a really important early factory.   Pretty much your entire early game strategy should revolve around how / where to build those.

A related adjustment for me personally, is that my usual "forest and forget" strategy is not correct for WtP.  You need to build Mines, early, not late.  The Recycling Tanks need to be fed minerals, and the global warming / mindworm overrun consequences aren't remotely as bad as in the vanilla game, or my mod.  It's inheriting Thinker's lenity / nerfing in that dept., I'm guessing.  So mines early instead of mines way late, like in the era of Hybrid Forests, is way different gameplay.

A noob who doesn't know the intricacies of WtP's different resource progressions, is going to do poorly compared to someone like myself who (belatedly) knows what he's doing.  After several plays, my early game wasn't the problem.

I think I quit mid to late game because there was no point to having a war for some reason.  This happened every game.  Was it the sheer pile of cities on the map?  I played a Huge map, so that I can accurately compare the differences with my own mod, which is designed for Huge maps.  I know WtP wasn't designed to be especially so.  But for me personally, it's the best yardstick for comparing apples to apples.

It wasn't because militarily I was losing.  Generally speaking, I seemed to be winning.  But by midgame, bored out of my mind.

Oh, and I do remember one total non-feature compared to my mod.  You got rid of instant movement mag tubes.  Yeah, it's more realistic.  It's also so incredibly dull, trying to move units on a Huge map when you can't get them anywhere fast.  Push, push, push, push, push.

I'm remembering another thing that really irked me.  Secret Projects are godawful prohibitively expensive.  Yet, the AIs manage to build just about all of 'em.  At least on a Huge map, things are so spread out and it's so hard to move units to interfere with anybody, that AIs can just allocate a few cities to build SPs for quite a long time.  Because they're so spammy with cities, in a way that I'll never be as a human player.  And they've got the Transcend minerals bonus, so it's not quite as painful for them to build long term stuff as it is for me.  So they end up with 90% of the SPs, split between all of them more or less equally, it seemed.  I could get 2 at the beginning if I really tried hard, but otherwise, forget it.  Tech acceleration guaranteed they'd start earlier than me and finish before I even got a chance to try.

[DrazharLn]

@Bear, to be clear, my list of problems was hypothetical.

[Alpha Centauri Bear]

I just played a ton of WtP and I don't remember the early game being the problem.  I remember quitting in mid to late game, because I couldn't stand to do something anymore.  I'm not clear on what, because I lost discipline for writing things down / providing feedback.  I'll be honest: I eventually went back to playing my own mod.

No foul in that, man. I appreciate all you play testing and the feedback. It brought an enormous amount of improvements to this mod!
Feel free to enjoy yourself for a while.
😉

Even with all the negative points you are highlighting below, I dare to say it stepped pretty far from vanilla misbalance. We are just more or less polishing it now to turn into pearl that we dream of. And we sure can continue doing that as long as anybody cares. Thanks again.

In other words, "should there be 1 more terraformer at the beginning?" is barking up the wrong tree, IMO.

Absolutely agree. This is not something changing the generic game feeling.

I do think that going back to WtP, after playing vanilla or my mod or even just Thinker, always requires a big mental adjustment on the part of the player.  A key thing is how to make use of Recycling Tanks early on.  They aren't just a little bonus, they're a really important early factory.   Pretty much your entire early game strategy should revolve around how / where to build those.

You are right. That is the serious change from vanilla. I thought about it long and hard with help of the community and found this to be most logical conversion.
Even though, I don't thing reverting it back to vanilla functionality changes much for this mod. I will just redistribute appearance of other factories and that'll be it. Maybe they will start appearing slightly later resembling vanilla progression more. Which may be more suitable. Let's share ideas.

What to do with RT?

• Revert to vanilla function. Factories will start appearing slightly later in the game. Resulting in total mineral output progression reduction.
• Reduce multiplication coefficient to +25% instead of +50%. Similar reduction can be applied to other factories. Also will result in total mineral output progression reduction.
• Move later in tech tree.
• Make more expensive making it less desirable to build in all bases, only in really high producing ones already.
• Anything else.

A related adjustment for me personally, is that my usual "forest and forget" strategy is not correct for WtP.  You need to build Mines, early, not late.  The Recycling Tanks need to be fed minerals, and the global warming / mindworm overrun consequences aren't remotely as bad as in the vanilla game, or my mod.  It's inheriting Thinker's lenity / nerfing in that dept., I'm guessing.  So mines early instead of mines way late, like in the era of Hybrid Forests, is way different gameplay.

Right. Wet and rocky landscape benefits from farms + mines more. Forest is not the ultimate solution anymore. However, it is still useful in dry and flat lands. I still use it occasionally and AI does too. Should we somehow improve it to be able to compete with mines? Not sure. With forest facilities it becomes powerful enough to be used on regular basis. Probably that should be its place. After all it reduces eco-damage so maybe mid game is its rightful place.

I think I quit mid to late game because there was no point to having a war for some reason.  This happened every game.  Was it the sheer pile of cities on the map?  I played a Huge map, so that I can accurately compare the differences with my own mod, which is designed for Huge maps.  I know WtP wasn't designed to be especially so.  But for me personally, it's the best yardstick for comparing apples to apples.

It wasn't because militarily I was losing.  Generally speaking, I seemed to be winning.  But by midgame, bored out of my mind.

You are absolutely on the point here. I was trying to steer from vanilla total war and indestructible army in mid-late game. I.e. partially shifted from weapon prevalence to war of economies. I admit it may result in too much slow down and grinding for human player. Pretty difficult to find a sweet spot. Some people like to finish it quick, some like more positional game. Please share your suggestions.

Oh, and I do remember one total non-feature compared to my mod.  You got rid of instant movement mag tubes.  Yeah, it's more realistic.  It's also so incredibly dull, trying to move units on a Huge map when you can't get them anywhere fast.  Push, push, push, push, push.

Not sure about push-push thing. Can you send your units to destination or set up automatic relocation? This way you click just one (or none at all) to deliver them to the front line. True, it will take some turn for them to auto move but that does not require any more button pushing.

This is not conceptual change. I can revert it to infinite movement or increase multiplier or even make it proportional to the map size: say 6-9-12-15-18, which will be quite close to instant movement.

I'm remembering another thing that really irked me.  Secret Projects are godawful prohibitively expensive.  Yet, the AIs manage to build just about all of 'em.  At least on a Huge map, things are so spread out and it's so hard to move units to interfere with anybody, that AIs can just allocate a few cities to build SPs for quite a long time.  Because they're so spammy with cities, in a way that I'll never be as a human player.  And they've got the Transcend minerals bonus, so it's not quite as painful for them to build long term stuff as it is for me.  So they end up with 90% of the SPs, split between all of them more or less equally, it seemed.  I could get 2 at the beginning if I really tried hard, but otherwise, forget it.  Tech acceleration guaranteed they'd start earlier than me and finish before I even got a chance to try.

The only reasoning for that is increasing their build time to something more than 20 turns to let technologically slower factions to be able to catch up and compete. From the other side, building them fast gives benefit to research and research factions. Either way is fine. We can proportionally reduce their cost, no biggie.