Jump to content

Soldiers gaining experience, increasing stats


Gazz

Recommended Posts

There have been many (and extensive =) threads on the old forum about how training in general could work.

Pretty much everyone agrees that while some in-base training is reasonable, experience through missions should be paramount.

So I've been thinking about how exactly the game should decide how much experience / stats a soldier gains from which actions during a mission.

The stats we have are

action points
resilience
strength
accuracy
reflexes
bravery
missions
kills
rank
'rank' experience points as a hidden stat (assumed)

For some stats it's obvious which actions to assign. Firing guns increases accuracy. Duh.

For other stats it's not so clear cut. While it is the obvious thing to attach reflexes to reaction fire during the alien turn, it leaves the issue of overall progression.

Firing guns happens a lot in the player turn but also in combination with reaction fire. That leads to a lot more accuracy skillups than there are for reflexes.

If important stats just won't go up, that is frustrating and often leads to...

Gaming the system / exploits.

A soldier should not gain accuracy by silly actions like firing 20 magazines worth of aimed shots at a ground tile in front of him.

In JA2, the easiest way to gain strength is to overload your soldiers to several hundred percent of their carrying capacity, ordering them to travel to an adjacent tile (ideally of "difficult" terrain), and immediately canceling the order while the game is still paused. Repeat.

Some considerations will certainly apply to all stats, others to particular stats or situations. I don't know what everyone will come up with or in which order so... just gunna have to play it by ear. =P

Link to comment
Share on other sites

Stats to be floating point values

Any system we could come up with would be immensely easier to balance if stats were saved as floating point values... internally.

If a stat can go from 1-100, then the last 5 percent pose a serious problem because it is expected that every point gained at this point requires an extraordinary amount of effort.

This is very hard to do if a single mission is enough to add 1-2 points flat.

With floating point values, fractions can be awarded and several missions may be necessary to increase accuracy from 95 to 96.

Stat caps

Personally I'm against stats having a hard cap of 100 (or whatever).

Even if advancement beyond 100 is excruciatingly slow, it greatly adds to the satisfaction of seeing our soldiers grow "bigger" than the average guy on the streets. =P

Link to comment
Share on other sites

Hmm, good point about the stat caps. Now that I think about it, why are there stat caps? Is Xenonauts likely to have the overflow problems of X-com 1/2? They added some caps as a result of it.

As for stat gains, well, there is a limit to how much you can gain per mission.

Depends on the criteria for raising. If reflexes gain XP for any reaction response, including the turning or whatever they do when they begin to react, that would take out some of the difficulty.

Still wondering how the game knows you're firing at a valid target for Accuracy XP gain. (from something Chris said on the old forum). Does it just not register accidental hits?

Bravery could be increased partially with rank and/or missions. Also not panicking when you should.

Good point about strength though. Carrying heavy stuff apparently increases it, but not sure how much you need to carry. Overloading? Half?

There was also that "linking" stuff I brought up that AD helpfully copied over: How experience is awarded

Link to comment
Share on other sites

Evaluating "skill use" during missions

I'm pathologically in favour of systems with dimnishing returns. It's how our world works. There is always a point where adding 10 % of effort starts adding less than 10 % result...

Let's say that every shot a soldier fires adds ( 100 / accuracy ) counters.

The higher the soldier's personally accuracy the fewer counters added per shot.

On the mission debriefing screen, the game decides how that is going to translate into accuracy advancement.

A soldier with ACC 60, firing 15 shots, will have 25 counters.

ACC 80, 10 shots, 12 counters.

ACC 95, 20 shots, 21 counters.

This system results in a relatively high and constant amount of counters towards the high end with only a moderate advantage at the low end.

The reason is that ( 100 / accuracy ) will be nearing 1... and stay there.

Such a system can be useful if you want to promote "advancement" even at high levels because the game is very unforgiving on soldier deaths. Otherwise everyone ends up as a super soldier before long.

There is no way to get such a simple system to produce low advancement at the high end.

Counters += 10,5 - SQR ( accuracy )

results in a more believeable degression of

30 (acc) : 5

50 : 3,4

70 : 2,1

90 : 1

94 : 0,8

96 : 0,7

alas, about that point it starts to fail like many formulas do at particular ranges.

So I would use an actual cap at this point, meaning that for values of 98+, another formula is used.

Counters += 2,9 - SQR ( accuracy -50 ) / 3

98 : 0,6

102 : 0,5

110 : 0,3

120 : 0,11

125 : 0,01

That would finally be the hard cap but since the chances of anyone ever reaching such a value would be astronomical, I wouldn't worry about it.

Still, a value of 125 looks plenty of "elite" (because most everyone will imagine 100 to be "normal human) without going into the realm of Marvel superheroes.

With this system, the example I used above would look like...

A soldier with ACC 60, firing 15 shots, will have 41 counters.

ACC 80, 10 shots, 12 counters.

ACC 95, 20 shots, 15 counters.

ACC 114, 20 shots, 4,6 counters.

Those are some values that the game can deal with.

For every "full ten" counters, the game could give a 50 % chance of a 1 point increase.

For remaining "fractions" < 10, the game could give a chance eaqual to (fractions / 100).

The 15 counters would then result in 50% chance for 1 point plus 5% chance for another.

The last line with 4,6 counters would only be a 4,6% chance for one point.

In the last example with 50 % chance of a 1 point increase.

The total advancement from a single mission may have to be capped to 2-3 points but that would only ever take effect at the low end of attribute levels.

This way there is always a chance to increase a soldier's skill (up to 125) but it's massively easier to do so at lower skill levels.

Link to comment
Share on other sites

Yeah, there's a few ways for doing stuff like that (skimmed the maths, too tired).

-Higher stat = more xp used per point

-X-com's method for secondary stats (maximum increase scaled based on current stats)

I'm curious as to how actions translate into stat gains at all. Is it xp levels per stat point? Actions add to counter for rolls? What?

Link to comment
Share on other sites

Hmm, good point about the stat caps. Now that I think about it, why are there stat caps? Is Xenonauts likely to have the overflow problems of X-com 1/2? They added some caps as a result of it.

caps...

While I'm completely contradicting myself in posts 2 and 4, there is this annoying issue of common sense. =P

I could buy that some "elite soldier" with years of training has an accuracy of 115. Considerably above a regular, trained shooter.

With no cap whatsoever, he could get up to 180 or 220... and believability starts to fray at the corners. =P

Like it or not, 100 will always be seen as some kind of "normal maximum".

So I'd let soldiers increase their stats above 100 but not so much as to look completely ridiculous.

They are still human...

Depends on the criteria for raising. If reflexes gain XP for any reaction response, including the turning or whatever they do when they begin to react, that would take out some of the difficulty.

I haven't delved into individual attributes, yet, but yes, reflexes are a tricky bit.

One way could be to count "experience" everytime the soldier or an alien has a chance for reaction fire.

That would broaden the playfield, so to speak, and also give soldiers a chance to increase the skill if the player does not "game the system" by parking them with AP reserved for just that.

It could be reasoned that by "potentially getting shot at" they'd use their reflexes by ducking...

There was also that "linking" stuff I brought up that AD helpfully copied over:

I don't like it much.

If you implement a system that arbitrarily grants increases for stats that have not been used, you may as well grant completely random stats at the end of the mission.

Using the formulas I suggested above, a soldier could be granted 1-3 "uses" of every attribute at the end of every mission.

That would do diddly at high skill levels would brush up the "slacker" skills over time.

Of course it's completely arbitrary but the end result would pretty much be the same... with a lot less hassle.

Still wondering how the game knows you're firing at a valid target for Accuracy XP gain. (from something Chris said on the old forum). Does it just not register accidental hits?

That's a key issue for accuracy gain and it's no an easy one.

The "easy but dumb" system would be to grant no increase at all unless a hostile is directly targeted.

Now take a flamer or rocket launcher.

They are AOE weapons and should count even if not directly shooting at an alien - even if no alien is visible on the screen at all!

So how could a useful rule look?

Could be that it "counts" as long as any hostile is visible.

Not perfect but might be a start.

Would only be exploitable when a melee-only alien can not reach your soldier.

Let's say there's a chrysalid and you use your fancy jump armour to hop onto the Chinook, which has no ladder.

The hostile would be in view while you empty 5 magazines into the ground to get the skill-ups.

Needs moar brains.

Bravery could be increased partially with rank and/or missions. Also not panicking when you should.

Depends on whether rank is modifier or not.

If the event of promotion adds bravery that's one thing.

Personally I think it won't but we'll have to see.

I think that the chance to increase bravery should be triggered by anything that does alter morale - whatever events that may be in the game.

Also, it should depends on it being a positive or negative event!

Seeing your comrades massacred left and right may trigger a lot of morale changes but it's not very likely to make the soldier much braver...

Negative events could have 1/6 or 1/10 of the effect of positive events.

Good point about strength though. Carrying heavy stuff apparently increases it, but not sure how much you need to carry. Overloading? Half?

Just carrying (standing around) shouldn't do much... but a little.

How far a soldier moves (carrying x weight) shouldn't be a consideration because the AP are already limited by carried weight.

I think that any weight above "half capacity" should have a chance to train strength. The exact "training" shoulddepend on how much... but it shouldn't be exploitable by overloading a rookie by 300%. =P

I'm curious as to how actions translate into stat gains at all. Is it xp levels per stat point? Actions add to counter for rolls? What?

That's the primary reason for this thread.

The math is actually secondary. A necessary evil but a lot easier to whip up than it is to figure out what exact action should contribute to shich stat, when, and how much.

Math isn't everone's idea of fun, either, so I also included the results of it "in plain". =P

Edited by Gazz
Link to comment
Share on other sites

Yeah, that's what I meant about the reaction stuff. Tried not to mention anything from EU though. Basically, what I meant was something like any time a soldier gets listed in something like EU's reaction priority list, they get REA xp or whatever is used.

Link to comment
Share on other sites

There could be a difference between active and passive use of reflexes.

Active when the player reserves AP and they are actually used for reaction fire.

Passive if the player moves the soldier within an alien's field of view.

The latter should be worth less because it would be too easy to move 1 tile around that wall, step back, repeat.

If that's worth fewer "points", it reduces exploitability while still giving soldiers the chance to improve "naturally" through offensive movement.

Another possibility would be to grant a low chance for everytime AP are reserved.

That could get silly pretty quickly, though, when the rookie you parked back in the Chinook increases his reflexes more than your frontline troops. =P

Still... it's worth a thought.

Edited by Gazz
Link to comment
Share on other sites

Maybe you should put in a limit of 5000 characters and 2 posts per day and user.

Would make it a lot easier to keep up with your forum. =P

I'll have to come back to this one...

You shouldn't say things like that.

That is just going to generate even more posts "where Chris will read them"!

As for resilience, I'm not sure what could be logical events to increase it.

Getting hit with plasma weapons and grenades makes you healthier? Umm...

For action Points, the most bare-bones way would be akin to reflexes.

If you spend AP in sight of aliens, you have a chance to "train" AP.

Sounds simple but it might just be servicable - unless something has a way smarter idea for this...

Edited by Gazz
Link to comment
Share on other sites

You shouldn't say things like that.

That is just going to generate even more posts "where Chris will read them"!

Oh god, the kneeling thread/s.

I'm curious, do strategy game forums usually generate large posts compared to say...arcade or simulator games? Or whatever the other genres.

If so, might be a result of the people the games attract.

Aargh why are there so many massive posts on the forum?

*points at Gazz while whistling innocently*

On Topic: Hmm, never really considered having stuff during the player turn affect reactions. It sounded somehow wrong.

I suggested that bravery go up with rank as well as whatever else due to rank being = experience. Besides, having officers (who increase or harden morale) more likely to panic that their grunts alot of the time does end up being a bit strange.

Bravery going up from not panicking when you should is just stolen wholesale from EU. Its was the only way to gain Bravery, basically any time you don't panic when below the panic threshold, you gain bravery XP. Getting jaded is my guess.

Link to comment
Share on other sites

Teeheehee!

I'm pathologically verbose. Can't help it.

If I didn't break up such things into multiple posts it would be even worse. A non-navigable wall of text helps noone.

never really considered having stuff during the player turn affect reactions. It sounded somehow wrong.

I'm not entirely sold on the idea, either, but it would mean that reflexes isn't such a one-trick-pony where advancement is concerned.

Since there is no actual "ducking" an alien's reaction fire it's a bit of a stretch, but with a bit of artistic license, it could improve gameplay.

Your stormtroopers, who routinely are the first through the door, should increase their reaction as a result of that. Since that cannot be tied to classic reaction fire, tieing it into the aliens' reaction fire seemed the next best thing.

Of course it might end up exploitable but there is always the risk of getting shot dead while doing so.

Also, the player would have to know when exactly the alien reaction fire events are triggered.

Given that, I wouldn't think that it's extremely exploitable. IMO, the good outweighs the bad.

What other approaches could there be?

Link to comment
Share on other sites

Dunno. It would make sense if there's a suppression mechanic though. Being better able to respond quickly under stress and whatnot.

Its just that without it, it sounds wrong.

But yeah, having shocks increase reactions (somehow) as part of their duties would make sense, and be very useful. You can't reaction move all the time (wait, is reaction move still going to be included?).

Huh, now that I think about it, if troops gained reaction XP from being in a local priority list, they would gain XP from charging through doors. They'd be in the list as part of being a target. The things you can do when your opponents aren't expected to survive or re-appear.

Hmm, we might need a third voice.

Link to comment
Share on other sites

You only need a third voice that's useful.

So, let's see...

I'm against accuracy increasing just by firing weapons. I'd tie accuracy specifically to aimed shots, and tie snap fire and burst fire to different stats. Actually, strike that, firing guns does give accuracy increase, but at different returns, aimed shots giving a much higher return. Snap fire could increase reflexes, and burst fire strength, with a higher improvement being granted if the target is hit.

Action points, and to a lesser degree strength, could be improved by movement, rather than just ap expended. Strength would also improve through carrying heavier loads, but that should be balanced by an ap penalty, say, a progressive decrease in ap to zero as a soldiers load increases through his top 20% of capacity.

I prefer having one major way of improving a stat, and then other minor ways just so soldiers don't effectively pigeon hole and require you to train up a new rookie with a role in mind because none of your other soldiers can take over when your launcher guy dies.

Having different fire modes actually hit could be based on something other than a pure accuracy equation as well: have aimed be based on accuracy, burst involve strength and accuracy and snap involve reflexes and accuracy.

Bravery should increase with missions/experience primarily, with rank also granting a boost perhaps, but I'd have thought resilience would be tied to rank more than anything else. Otherwise the only way I could see resilience increasing would be through improvement of other stats, have it's progression tied to average percentage increases of all other stats maybe?

But yeah, so rather than having emptying a clip into the square in front of you giving you no improvement at all, have it give a minor improvement. Repetition does help with skill improvement. But have it massively outgained by actual combat.

Link to comment
Share on other sites

Ah, my bad. Yeah, aimed I consider above half, snap anything below half max AP. Oops. With the graduated nature of the aiming system, it's a little less easy to determining what's responsible for what. Still, having reflexes/accuracy min/max at each appropriate end of the aiming scale, and even out in the middle maybe...

Urg, my head is not in it today at all... I'll come back once I've straightened out a bit.

Link to comment
Share on other sites

It gets even more hazy when you consider that not all weapons have the same number of aiming levels. I think shotguns only have one or two.

That said...burst increasing strength...not bad. There seems to be some kind of recoil stat somewhere.

Link to comment
Share on other sites

I'd tie accuracy specifically to aimed shots, and tie snap fire and burst fire to different stats. Actually, strike that, firing guns does give accuracy increase, but at different returns, aimed shots giving a much higher return. Snap fire could increase reflexes, and burst fire strength, with a higher improvement being granted if the target is hit.

That has potential but in that case I'd prefer to split the gain, like burst having a 65% strength and 35% accuracy component.

For single shot, there would be similiar weighing between accuracy and reflexes, depending on the aiming ticks used.

The obvious drawback is that it's a considerably more complex mechanism.

That, too, would have the effect of stormtroopers, who use 'snapshots' at short range more often, improving their reflexes.

It gets even more hazy when you consider that not all weapons have the same number of aiming levels. I think shotguns only have one or two.

I think that's perfectly okay because someone using a shotgun is not going to be much of a sniper.

Actually, shotguns are used in disciplines like skeet shooting which has a considerable "reflexes" component.

The more you take careful aim, the less your reflexes play a role.

The word-highlighting may look a bit weird but in a thread with such a broad range of topics, how else are you going to find something specific?

Link to comment
Share on other sites

The word-highlighting may look a bit weird but in a thread with such a broad range of topics, how else are you going to find something specific?

Ferocious memory?

That, too, would have the effect of stormtroopers, who use 'snapshots' at short range more often, improving their reflexes.

True about shocks using single-shot. Though they would use burst quite a bit too (t2 CQB weapon being smg/carbine).

I think that's perfectly okay because someone using a shotgun is not going to be much of a sniper.

Rocket launchers also only have one or two ticks.

Edited by Sathra
Link to comment
Share on other sites

Rocket launchers also only have one or two ticks.

Hmmkay, so it's not an ideal first draft.

Making it weapon-specific would get really messy, IMO.

How about using the weight of the weapon to determine if and how much of a strength component is involved?

Using heavy weapons (literally) would give the soldier a work-out, no matter what exactly he does with them...

Link to comment
Share on other sites

That would indeed work. There does seem to be a Heavy Weapon distinction.

Hmm, I wonder if the sniper rifles are considered Heavy. Not exactly relevant, more just general interest.

There is no "heavy weapon" flag in the game files, which suits me just fine. =)

name			weightweapon.pistol		1,5weapon.rifle		5weapon.shotgun		5weapon.sniper		6,5weapon.machinegun	12,5weapon.flamethrower	8weapon.rocketlauncher	12,5

The sniper rifle is a bit on the heavy side but that's not unrealistic.

If I were to make a rule out of this, I'd assign a "strength component" of

INT ( weight - 6 ) * 7

The MG and launcher would end up at 42%, flamer at 14%.

Negative numbers are discarded, of course.

If you wanted to make the flamer more strength-heavy, you'd only increase the weight. Not an unreasonable move, that, since flamers aren't lightweight things and require a lot of fuel.

Personally, I wouldn't allow a flamer to work at all without a 4x4 backpack item - it's fuel tank.

Would require additional coding but I'd have it fueled directly from an inventory item. An item that is too bulky to keep any reloads on hand without damn near using up someone else's backpack, too.

A functionality like that would be a big bonus to make some "heavy weapons" more believeable.

I'm thinking of fun stuff like the XM214.

Edited by Gazz
Link to comment
Share on other sites

I would prefer to tie stat gains into shots that actually damage the enemy.

If you fire and someone takes damage then you have a chance to gain from it.

Firing into the ground or whatever will accomplish nothing.

Even indirect fire weapons will get the benefit from that as long as something is in its area of effect when it lands.

They may fire less often than something like an assault rifle but they could potentially hit more targets with each of those shots.

That may not be ideal so you may limit any potential gain to a single chance per shot rather than per enemy damaged.

Link to comment
Share on other sites

Ok some interesting thoughts here. Rather than directly comment on what has been said, I'll suggest some thoughts on what may cause gains for the stats listed:

The stats we have are

action points
resilience
strength
accuracy
reflexes
bravery
missions
kills
rank

For action points I would think that *any* action taken would attribute some experience towards the stat. The more points a particular action takes (i.e shots at higher aim levels) would attribute more points than actions taking lesser points (i.e turning, walking etc). As a marker of a genuine level of experience the number of missions completed by the soldier should add additional experience, and also for each actual rank increase (Rookie - Private - Corporal - Sergeant etc) should also add a bonus.

Resilience is a tough one, as was said it seems odd to think getting injured makes one healthier, but in this instance I think the utilisation of "what doesn't kill me makes me stronger" is probably a safe bet. So any injury sustained that does not kill the trooper would give experience to the resilience stat. Also gains could be added for rank level and completed missions etc. A further option could be the more missions completed without taking injury could create a multiplication factor for the resilience experience.

Strength again is not an easy one. The simple thing is obviously the more weight a trooper carries (and the longer it goes on for) the more strength experience would accumulate. A different multiplication factor could be applied based on whether the weight is carried in the backpack or in the arms - it should be a question of balance not realism to determine which way it should go. Realistically overloading a trooper could increase the gains (at the expense of other stats due to the lesser AP's available for other things), but weightlifting is a matter of repetitions more so than it is actually weight. So carrying an 80% load for longer and more consistently and contributing with other actions would be better than struggling along with a 130% load and doing nothing.

Accuracy of course, firing weapons, hitting targets and throwing grenades etc. Though hitting your chosen target is obviously best, I would think that within a certain arc of your target, say 30degrees either side, should also give gains to accuracy. Allowing this would also mean you could get reasonable gains for shots specifically towards non-alien static targets.

Reflexes is another tough one. Actions that occur through reaction fire etc, absolutely should led to reflex experience. As had been said burst fire too could give gains, whereas aimed fire may not, or at least would give much less. Another option that could assist with reflexes are having shots fired at the trooper being near misses. Reflex experience could be gained as a nod towards the trooper evading the shot etc, so the closer the shot the more experience could be attributed.

Bravery would have to have gains over time, so for rank increases and again based on the number of missions completed by the trooper. Not panicking when below the threshold makes sense as a bonus too, what I would suggest also would be that for every hit and kill the trooper makes he scores a bonus to bravery experience could be added. If you treat bravery more as confidence, then any encounter face to face with an alien that sees the trooper come out alive (and hopefully uninjured) could give a bonus as well - I'm thinking here more like breaching a UFO or entering a room to be confronted by a group of aliens - soldier goes EEP! - soldier kills or incapacitates aliens and moves on - soldier goes Heck Yeah I'm good! (Bravery experience goes skyrocketing).

Missions, as I have said in some of the above, the number of completed missions should affect the experience levels of some of the stats,but for the others completed missions could generate a multiplication factor, so actions gain more experience (this does not need to be a large number, I would think it should start from say 10 missions and begin at X0.1 - at most, this could increase to as much as X0.25 but I would not go beyond that, otherwise it would be too much)

Kills, as above for bravery, but kills could also have a general value for action points and possibly resilience also.

Rank - assuming rank experience is required for promotions it need not be directly tied to actions taken during a mission, but rather should be gained based on the total experience gain of the primary stats. Say Private GI Joe, received a total of 50 experience across his primary stats in one mission, he accumulates 10 rank experience points towards Corporal. Obviously the actually ratio would need to be fudged based on how much experience is required for each rank and how many missions would be reasonable for someone to achieve the next rank. Gaining of ranks as with many games probably should involve a bonus to all stats, but it does not need to be significant.

I'm sure there are other options for all of these, but to me these seem reasonable

Link to comment
Share on other sites

There is no "heavy weapon" flag in the game files, which suits me just fine. =)

There is something about a "HeavyWeaponMovePenalty" though, which is where I'm getting this from.

@MickeyC: Lots of stuff there. Not much to say though, but it does seem pretty useful.

Poor, poor Chris.

Edited by Sathra
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...