Alternative character experience - Potential / Affinity

[Wraith_Magus]

In the current game, experience/stat growth is fairly gamey in nature. You load up weight to train strength. I keep a list of characters to check off, making sure everyone gets their turn to have a reaction shot before I go for the kill. It's especially bad with the Light Scouts, which are farmable practically only for their experience.

To a degree, this is what the "donating"/"airstrike" idea was for, but I want to propose a different scheme for expeience overall, so that players aren't so tempted to just go into missions for character leveling sake, alone.

Potential is a scheme for growth-through-use experience that aims to control player's instinct to just spam a skill to train it that I first came across in the roguelike game Elona. Potential is fairly simple in concept - it's just an experience point multiplier that goes down every time whatever the potential is tied to ranks up. At 100% potential, you gain experience normally. At 150% or 50% potential, you gain experience at 1.5 or 0.5 times normal. Each time you rank up a skill, your potential goes down.

For brevity's sake, I am spoiling the details. Those interested in its technical workings click:

In Elona, specifically, the potential is multiplied by 0.9 every rank up. Potential goes up in various ways for different character stats, but for the main ones, attributes and spells, you have to sleep (which requires passing many turns before you can do again) to get potential back, and skills, which requires you go to trainers and pay special coins you only get from completing random quests or random side dungeons. Basically, by cutting the rate at which you gain experience, it dissuades players from spamming skills to boost them, since the more you spam a skill to boost it, the lower its experience multiplier drops. If you instead do the (more interesting) side quests, you can boost your potential, even as you actually train your skills through use, rather than just spamming abilities from a safe location. When potential goes up, however, how much it goes up is based upon how high it already is - pushing the multiplier into a median range. (For skills, this is 15 - current potential/15 (minimum 1). Because of integer math, it means that, at 59 potential, you gain 12 potential for training, and only gain 11 when you have 60 potential. Hence, since you drop 10% of your potential every rank-up, if you want to only spend one training session per rank up of a skill, you wind up averaging being stuck between 90 and 99 potential, since that's where you will both go down and up 10 potential points each rank-up or training.)

Now, in the current way we use stat-ups, you can just gain a stat point for every mission in everything reliably but bravery. Using potential, however, you can reverse the way in which characters gain stats.

Characters can have invisible progress bars that fill up over hours/days of down-time. That is, they are gaining stats any time they are not in combat, but are instead honing their skills. As they gain stats, however, their potential goes down, and the rate at which they increase their skills drops, as well. By going into battle, their actions will increase their potentials for their different stats, rather than the stats, themselves. This can follow a similar model to what was used in Elona - a reaction shot, instead of raising reflexes by one, could instead increase reflexes potential by 15 - current potential/150 (minimum 1) for every reaction shot fired. (It could count more than once per mission, although it would make balancing a little more tricky. This proposed balance presumes about 10 reaction shots would be made between each stat gain, and would keep players at around 1000 potential, so how much time it takes to gain a stat would have to be balanced against that target number of 1000.)

The short of it is, you gain stats over time as you sit in base, but your potential to gain more stats goes down - making it take longer, and giving you less and less actual stat growth as you sit in base, doing nothing but drills. You get a sudden boost to your potential when you actually use your character in battle, giving them expanded potential, and letting them train faster. But that practical experience isn't going to give you actual stats, it's just making stats gained from training come faster.

Hence, the way to keep soldiers going at peak efficiency isn't so much to keep them constantly in battle, but to find the proper balance between being in downtime and being in battle.

This, in turn, can help with the problem of players hunting down every UFO - since taking on a UFO means taking time away from character training, then the player will eventually have enough potential that they gain no real benefit from putting characters in battle over and over, while they are actively losing their training time.

Another way to keep overpowered player characters from being a problem is to make the amount of experience (since we aren't tied to a strict "one reaction shot = one point of reflexes" model, but an experience bar that fills progressively with downtime) that it takes to gain a stat increase as the stats themselves are higher values. (A rookie with 50 in Strength might take half as long to train in strength as a veteran with 100 strength... or four times as long, if you want a harsher quadratic curve.)

If you make soldiers passively train, (rather than spend money on training, unless you want to make spending money be an "intensive training" option) you can also include some research options involving training - for example, after battles with specific types of aliens, you can have "sebilian noncombatant tactics" research or the like, which would also provide bonuses to the rate of experience gain, making late-game recruits train up faster than early-game recruits. So could a training yard base extension.

---

Affinity is a different animal, but it relates to the problem I have with all the soldiers in my squad being basically the same for any given amount of training, with only their initial randomization of between 50 and 60 in their scores making them slightly different.

Affinity is a natural predisposition towards certain stats - it means that a character with a high affinity for strength, for example, would gain strength points faster than a character with a low affinity, with or without potential.

Without potential also being in play, to take hit points as an example, you gain 1 HP per every 5 other stats that go up currently, but a character with high affinity for HP might go up every 3 other stats going up, while one with low affinity for HP would go up an HP for every 8 other stats going up. You simply change the requirements to make one character capable of going up in stats more quickly than another, meaning that stats aren't going to uniformly go up for everyone, just so long as you grind them enough. Characters have randomly selected specializations, with a natural talent for one attribute, but a weakness in another that isn't so easily covered up with time.

WITH potential being used in conjunction, it changes the way that potential goes up. A character with high affinity, for example, would have a different formula for how much potential goes up. Instead of gaining 15 - current affinity/150 (minimum 1) for a reaction shot, it would be gaining 20 - current affinity/200 or 10 - current affinity/100 for very high or very low affinities in reflexes.

Either way, this would mean that, while starting character strength is still largely ignorable, affinity would be a major difference in what characters you recruited.

Edited September 30, 2013 by Wraith_Magus
Spoiling the math for brevity

[Gorlom]

Interesting mechanics and could probably be very fun to play with. Although I wonder if they aren't overly complicated for a very minor issue that may or may not even be a problem?

It might be too late in development to change this now?

[Wraith_Magus]

While it's complicated to explain, it's not too complicated to put into practice. We already have many things that occur over time (a functional "progress bar") in the geoscape, and this would just be changing the variables related to stat gain to have higher integer values that fill up over time on the geoscape. All the math being used is still quite simple. (Basically everything is just four-function math.) Making a training yard base upgrade would take a little more effort.

It's the ramifications that are quite different - it means that combat is merely there to keep your growth rate high, not to actually give you stats. Especially with some of the extra research paths, it also means that recruits are more easily trained up, since you can bench them for long periods of time, and only occasionally give them a swing at the enemy to up their potentials, and let them grow at sub-optimal rates while the elites aren't growing that much faster than them (although still fast enough to be a dozen or so points ahead).

It therefore also helps keep soldiers expendable.

The most complicated portion of it would just be figuring out how fast you want characters to grow in stats, and how many battles you want them to average to achieve that rate of growth.

However, it seems like, judging from how Chris posted about players taking on multiple crash sites, some sort of mechanics that helps bring power-players more in line with someone who only does just barely enough to get by is a serious concern. The problem is ultimately caused by linear rewards for effort, as far as I see it. Making diminishing rewards for greater effort brings back a balancing effect that was the goal of the "airstrike"/"donate" option.

[Gorlom]

I think some of the advocates of versimilitude (or realism if you prefere) would complain against such a system though. It doesn't reallty seem to make sense other as a gamemechanic. And only thne if you know how it works. Playing with no knowledge of the system would likely be confusing and annoying imo

At least growth by doing is somewhat intuitive.

I wonder if having benched troops passivly grow in the background and making soldiers more expendable is desirable. I would think making the soldiers more expendible is counter productive. The (naturally flowing, as opposed to forced) attachment to the soldiers is one of the selling points of X-com and also Xenonauts.

Hampering player progress and making the player feel like someone is reining him in usually leaves a bad taste in most peoples mouthes. Making so that the player feelss his/her choices doesn't matter is lets the player down and makes the game less fun.

I'm not saying it's not a good idea/mechanic, but some of your arguments worries me.

Edited September 30, 2013 by Gorlom

[Amaror]

I have to admit i haven't read the entire thing, but how is it that gh seems to get tons and tons of suggestions that involve changing a major gameplay mechanic with the argument: "I currently can do x and that's unrealistic/immersion breaking/gamey..."

There really have been a lot of these:

I can carry so much ammo, please don't let me carry ammo in the backpack..

I can carry many shields and make my soldier very hard to kill, please don't let me keep shields in the backpack...

I mean if you don't want to play gamey, just don't play gamey.

Sorry if i come across as aggressive, i've been just a bit annoyed.

I really don't think they will or should for that matter, overhaul any major gameplay mechanics. The Game is in beta, that means bug fixing and fine tuning but not major overhauls of the gameplay.

The game is fine as it is in this matter. It still need fine tuning, absolutly, but not major overhauls.

[Wraith_Magus]

I'm not making an argument based on realism, Amaror. (Well, partially upon realism, but not mostly upon realism.) I'm making an argument based upon the fact that the current system is rather exploitable, and takes away a bit of the fun.

*Sigh* However, this isn't the Dwarf Fortress forums, so I guess I should truncate my argument, some.

It doesn't just feel gamey, it makes me change the way I play the game to make sure I stuff enough grenades in everyone's pockets to weigh them down for their daily strength boost. I have to make sure everyone on the checklist has "had their turn" with a reflexive shot to gain reflexes. That occasionally means taking a few turns sitting behind some wall, letting sebs take pot shots at me, burning through flashbangs, waiting for the riflemen to finally get a good roll on their REF check. It's not just unrealistic, it takes away from the fun of the game to have this checklist mentality.

Beyond that, there is the problem that the devs at least seemed to think was important enough to warrant a pretty big thread to cover finding ways to dissuade players from wanting to fight every single mission - this is a way to solve that problem, while also solving the problem of necessitating that players sit there make sure that every soldier gets a reflex point every fight.

The system I am suggesting is, in basic terms, that they train in their off-time, and gain stats between missions, rather than on missions. While that is more realistic, it also means that it takes away the compulsion to go fighting every fight just for experience.

You don't need to fully understand the mechanics to use it - you gain more experience the longer you leave your troops in the training range, and they get experience point bonuses that make them train faster when they complete a mission. If they don't go on missions long enough, the rate at which they get better slows down.

As for Gorlom finding a problem with making soldiers "too expendable", then I'd just remind you that the difference isn't that you're losing soldiers with stats in the 80s, and picking up a soldier with stats in the 80s, you're replacing them with soldiers with stats in the upper 60s. That's still a sting, but not quite the sting that going all the way down to a raw recruit would be. You still don't want to lose an elite soldier with massive stats. Of course, it all depends on how you actually fiddle with the numbers.

And as for a more overarching argument, I honestly just like these more abstracted forms of game mechanics, because unlike the more brute-forcey types of mechanics that are most obvious, the abstractions tend to be less breakable and exploitable.

Consider the differences between the earlier Total War games, and the newer ones - you improved your nation in the long run, generally, with simply buying better farms or sanitation, which improved the rate of population growth. It was just simple +0.5% per turn stuff, with -0.5% per turn penalties. More advanced nations get their top-level cities first. It takes seconds to just punch the button to make the next form of farm when you see your cities have completed their last project on a new turn, and you're back to playing the game. By the time of Empire, you're sitting there, trying to game having Gentlemen gain the right types of stats, put into the proper colleges, and working on the most advantageous aspects of the tech tree. It's taking much longer, fiddling with more complicated things, but achieving nothing the abstract, elegant solution didn't accomplish, while taking away from playing the game you wanted to play to worry over things unrelated to why you want to play the game.

I considered suggesting it be simply like Elona's potential system, where you still gain skill directly through use, but suffer what are functionally experience penalties until you spend several days recovering potential. However, the more I thought about it, the more this "backwards" system seemed to solve more problems more thoroughly. If practical application merely makes training faster (being an experience growth bonus), then it's the sitting in the base training that you never get enough of, and the practical application that is just the booster. (Hence, putting a player pressure on keeping from taking a swing at absolutely everything - at least, not with the same team.)