Skill training overhaul

[facepalm]

Since Cata's current experience system is a bit of a mess, I thought I'd try to unmess it a bit.

Significant changes
Experience needed to level up is now a simple exponential: 100x level^2, instead of 100x level
This has the effect of slowing down skill progression considerably - you're not going to be a melee god by the time you leave town, and disassembling a single van down to the rims should no longer take you to Mechanics 10.

Books are adjusted to match the new level scaling, so their performance is unchanged.

Skill rust works with the new scaling, but the amount of experience lost due to rust is unchanged. This should effectively decrease its influence at higher levels, with the help of uncapped skills (see below) making it easier to counteract.

Crafting is no longer hard-capped
Everything you do will contribute to your skill, regardless of level. Since higher levels now require much more xp to advance, this isn't as problematic as it was. Crafting experience scaling has also been tweaked a bit to scale more with crafting difficulty. It's still a linear relationship, though, so you'll still need more high-level jobs than you used to.

Crafting is tied to craft duration
Previously, a 6 hour armor crafting job offered the same experience potential as a 1 minute slap-together. This meant that quick, simple crafts were ideal grinding opportunities, while the things you actually wanted to use gave essentially no xp for their creation. I've tied crafting experience into crafting duration (works for batch crafting too), normalizing xp gain to a 10-minute craft time. The aforementioned crafts would now have 36x and 0.1x experience modifiers, respectively.

[PropaneSoup]

Crafting is tied to craft duration

This is the biggest change that I immediately like. The others...I'd need to playtest to see how they feel. Because my first impression is that it sounds like just more grinding.

[facepalm]

The present need for grinding is mostly what motivated me to dig into the code for the change. As it stands, skills can only be increased by crafting at your current level... which is always a fairly limited selection, and usually not anything I wanted to use. This capping is currently necessary, because without a cap any source of xp will get you to ridiculous skills in no time (examples: mechanics, melee). Which ironically promotes grinding, because it's easier to grind up to Level Awesome in the beginning than suffer through the lower levels in real gameplay.

At any rate, I found myself grinding wooden spears to get Fab to 3, despite being an archery-only build, because all of the arrow options I wanted to use were Fab 0 and 1. Meanwhile I was already melee 7 just from whacking the occasional zed who wandered over to me, and I had no real need for archery at all because I could gib anything that approached. I wanted a skill system which I could more or less ignore, not be forced to game in order to advance or would force me down a particular path.

But if your goal is "grind to Mechanics 6 as soon as possible," then yes, you will have to grind more to get there. I follow the school of thought which argues that grinding is its own punishment.

Regardless, note that the performance of book reading has been updated to match the scaling. Books are currently one of the best ways to gain crafting skills, and they will be no less so given this change.

[Zireael07]

Nice!

[kevingranade]

Points one and three are OK, point two is a big no. We've gone around and around with this, and we've settled on caps. What I'd like to address the issue of grinding useless items for levels is to add training/practice activities, so if there's nothing to raise your level right now, you can still make progress doing something sensical. The more targeted fix for this is to add usefull crafting chains for various character types, so that archers, melee characters, builders etc always have something worthwhile to craft.

[Coolthulhu]

I follow the school of thought which argues that grinding is its own punishment.

This can very easily lead to really, really bad design. When the optimal choices in the game require you to play in the most boring way possible, being good at the game requires you to have no fun.

I wanted a skill system which I could more or less ignore, not be forced to game in order to advance or would force me down a particular path.

Your solution heavily nerfs skill gain for ranged weapons with limited ammo. Even archery (with nearly free ammo) would probably become unbearably tedious to level, forcing the player to rely on melee and firearms fired at point-blank with no skill.

That said, having a world option that would change leveling speed is recurring feature request on the forums. Having something like this as an option could be a good idea.

[facepalm]

kevingranade:

Points one and three are OK, point two is a big no. We've gone around and around with this, and we've settled on caps.

Fine, but at the very least you should consider softcapping. Reduced xp for (level-1) crafts would add a lot of options to what one could use to usefully train skills without crafting just for the experience alone. As it is, I don't much bother training at all unless I can find books to do it for me. Would you mind if I replaced point 2 with a simple soft cap for you to play around with instead?

I think you'll find, though, that exponential levels greatly reduces the need for capping. Right now, going from level 9 to 10 (which is supposed to be a massive undertaking, right?) only requires ten times the experience of going from 0 to 1 (which is supposed to be simple usage). It's quadratic fluff with linear crunch: any constant xp source will always train high levels too quickly or low levels too slowly. Fix that, either by requiring exponential xp to level (as I've done) or decaying overleveled xp gain, and most of the problems just disappear.

coolthulhu:

This can very easily lead to really, really bad design. When the optimal choices in the game require you to play in the most boring way possible, being good at the game requires you to have no fun.

Can you define "optimal" and "being good at the game?" Do you mean "squirreling yourself away to hone your skills in perfect safety, then stomping forth from your shelter as an unstoppable badass?" If so, then I'm afraid I haven't much sympathy for extending the lack of fun you'll be having. If that's how someone wants to play the game, the optimal route should be to open the debug menu and increase their skills there.

Your solution heavily nerfs skill gain for ranged weapons with limited ammo. Even archery (with nearly free ammo) would probably become unbearably tedious to level, forcing the player to rely on melee and firearms fired at point-blank with no skill.

Good point. Those are a little slow even for the vanilla game, and could certainly use a tweak (maybe with a modifier based on steadiness? i.e. "I know I was aiming right, so...). I'll play with possibilities tonight.

Also, you won't have "no skill." Level 2-3 isn't much harder to reach than it is currently. It's the extremely high levels that are most affected. But then, how many arrows or rounds do you think a world-class archer or marksman has to fire to reach that level of achievement?

[ETA] Ah, I think I misread your final point. Fact is, you need to do that even in the current game, especially with hordes on. Ranged combat just lags far behind melee in general utility.

[Coolthulhu]

Can you define "optimal" and "being good at the game?" Do you mean "squirreling yourself away to hone your skills in perfect safety, then stomping forth from your shelter as an unstoppable badass?" If so, then I'm afraid I haven't much sympathy for extending the lack of fun you'll be having.

Lifting all caps promotes this kind of gameplay. Caps are there so that you have to do different things to level, not just sit in one spot and craft same stuff.
Plus, by nerfing skill gain, you'd not just promote grinding, but REQUIRE it to gain some skills. For example, tailoring - even with current skill gain, you won't gain high tailoring level just by producing what you need. Your nerf would make it necessary for players to either find a book to level it or grind away the differences between levels. Except instead of grinding it from 20% to 100%, players would have to grind it from 20% to 500%.

Having fun and "winning" shouldn't be mutually exclusive. Forcing the player to cripple themselves and play badly in order to enjoy the game is one of the worst things you can do to a game.
In this example you'd be forcing the players to either make the correct decision and stay in the cave crafting armor or charge out naked, YOLO at zeds and die.

Don't punish those who chose not to grind.

[facepalm]

In this example you'd be forcing the players to either make the correct decision and stay in the cave crafting armor or charge out naked, YOLO at zeds and die.

I'm curious how you'd come to this false dilemma, considering that zombies are the primary source for armor crafting materials. The player is already strongly encouraged to gank a few lone zeds for the tattered rags and bits of armor they wear, and most early-game supplies are to be found in town.

For example, tailoring - even with current skill gain, you won't gain high tailoring level just by producing what you need.

Tailoring's fine, actually. If you'll notice how long most tailoring crafts are, it benefits the most from point 3, tying it in to duration. Nearly all tailoring crafts have an implicit 6-10x multiplier, and that only gets bigger as you go up in level and crafts take longer (and due to another change, scale better). Playing with it last night, it took a little longer to get up to tailoring 2 via reinforcement, at which point I could actually trust myself with repairing real stuff, but beyond that it was smooth sailing.

Cooking was similar. Quite a lot of level 2+ food takes most of an hour, and the slight boosts from boiling water helped.

For Mechanics, I debugged my skill up to 3 (i.e. finished Under The Hood) and snuck into town to disassemble an entire car, brought myself to just above 4. You only really need, what, 5 or 6 to go full deathmobile? Yeah, it's doable.

What am I leaving out - fabrication? Construction? Fabrication seemed fine, nothing out of the ordinary for early-game activity, at least. Construction is a grindfest even in the current game, but everything takes quite a bit of time to build, so it may be a wash.

Electronics, computers and first aid are all book larnin' right now anyway, so that's not much affected.

Anything else need looking into?

[Coolthulhu]

Did you test it up to lvl 6 in tailoring? You need about this much to get the survivor gear.
If it's fine around that, then I don't really have much to complain about.

Cooking was similar. Quite a lot of level 2+ food takes most of an hour, and the slight boosts from boiling water helped.

Problem with cooking is that low level cooking and high level cooking are barely related. That's why the cap is important. Making ton of jerky won't help with mutagen.
The cap helps to separate those two categories.

Mechanics

You may want to boost the skill gain from mechanics depending on time. I recall it being rather low, making removal of cheap parts a better way to grind than taking the valuables. This would also allow actually leveling mechanics without consciously choosing to grind by dismantling entire cars.

Fabrication

That's the one skill that I actually see getting buffed by the changes - higher level fabrication recipes take ages (3-9 hours).

Construction

Construction should level no slower than it does now. Most recipes don't take much time (20-40 minutes). Though in this case I'd actually support the full cap removal. This skill simply doesn't have enough uses for cap to work well.

You may want to do something about driving. It is incredibly anti-fun to level it without a book even right now and you need it at lvl 4 to actually benefit from having it at all.
Leveling swimming is also pure grind, but I don't think anyone cares about swimming.

[Zireael07]

I like the tying to duration thing.

EDIT: Driving and swimming are terrible to level up, agreed.

[facepalm]

Did you test it up to lvl 6 in tailoring? You need about this much to get the survivor gear.
If it's fine around that, then I don't really have much to complain about.

I brought it to about halfway past three just from repairing and reinforcing my gear, without taking a serious look at crafting at all, crafted some long underwear and receive a ~20% skill gain, and figured it was fine. I'll spawn a crapton of rags and leather and go to town, then post my "grinding pile" tomorrow when I get to 6.

Problem with cooking is that low level cooking and high level cooking are barely related. That's why the cap is important. Making ton of jerky won't help with mutagen.
The cap helps to separate those two categories.

If they're that unrelated, they ought to be separate skills. They're already separate crafting tabs. With the addition of more low-lvl chemical recipes recentlyish, why not spin that into a Chemistry skill?

[Coolthulhu]

It would be a rather bare skill right now, especially at low levels. Plus the dedicated cooking would have no recipes at higher levels. They complement each other very well.

It's not like those skills are totally unconnected, it's more like what you're doing at low levels is different to what you're doing at high levels. Learning chemistry from cooking soups would be like becoming better at installing CBMs after dismantling few cars (mechanics grinding is uncapped and that's the result).

Why do you want to remove the caps? You said the exp gain from boiling water would be minimal at higher levels. Caps solve the unrealistic part of grinding.

[KA101]

Hard cap or this doesn't get merged: soft cap just makes the grinding take longer.

Shame, really, as the duration is a good idea; exponential XP does screw over ranged fighters, so I'd want that revisited. Possibly some bonus for tougher shots, perhaps?

[facepalm]

Caps solve the unrealistic part of grinding.

They CAN, sure. But Cata's caps as currently implemented go too far, and add back in unrealistic grinding. Since crafts have to be exactly our level to count for xp, else nothing, that often results in forcing the players to repetitively craft stuff they don't need or want in order to make progress. That's the very definition of grinding. Like I was saying to Kevin, at the very least we ought to have softcaps instead.

But! Tailoring tests!

On review, it appears that tying crafting xp to crafting time buffs much more than exponential levels nerfed it. I spawned a naked survivor in an LMOE shelter with all the rags I could eat, and got to crafting.

Here's what I did for each level:

• Level 0: blanket, socks, reinforced socks, tried to reinforce blanket
• Level 1: long patchwork scarf, light gloves, stockings, keffiyeh, pouch, reinforced ~half
• Level 2: duffel bag, long underwear top & bottom, cotton hat, boots, reinforce all going forward
• Level 3: backpack, boonie hat, trenchcoat, leather armor gauntlets, short rope (rag) x2
• Level 4: leather body armor, sleeping bag, cargo pants x2, drop leg pouches, duster, knit cowl (failed)
• Level 5: pouch, chest rig, nomad gear, hunting cap, safety glasses, nomad cowl, (52%) drop leg pouches, backpack, survivor backpack (fail, then success) (95%), rain coat, drop leg pouches (x2)
• Level 6: survivor suit (48%), survivor helmet (58%), survivor boots (83%), survivor cargo pants (97%), survivor gloves
• Level 7: stopping here, but heavy survivor gear is now craftable

I've bolded the things that I crafted because I ran out of other ideas for that level. Basically, zero grind, because average crafting time keeps up with the exponential xp.

Which is actually a problem.

To break the system, I made another survivor who'd grind their way to the top instead with basic items. Since it was apparent by then that long duration crafts were the way to go, Wool Gloves (level 1, 4 felt, 90 minutes, so 9x xp) gave the best rate of return. Dude loved him some wool gloves, tell you what.

The results were actually kind of disappointing:

• Level 1: 2 wool gloves
• Level 2: 5 wool gloves
• Level 3: 8 wool gloves
• Level 4: 12 wool gloves
• Level 5: 17 wool gloves

I think we can agree that this would be a bad thing as-is. It would mean that making anything but wool gloves would become somewhat pointless.

So, since I promised Kevin I'd add in a soft cap instead, let's do that and see what happens. We'll divide crafting xp by 2^N, where N is the positive difference between our level and the craft's level. So being level+1 means you get half xp, level+2 is 1/4, etc.

Also, since it's kind of silly for crafting two of anything to raise your level, enormous crafting time and low level notwithstanding, I tweaked the normalization to be base_xp times (1+time/half_hour). So crafting anything still gets base xp, and longer crafting still gets a bonus. But in the case of these wool gloves, it effectively halves the xp they give.

With these changes combined, the gloves-to-level becomes:

• Level 1: 4 wool gloves
• Level 2: 20 wool gloves
• Level 3: 64 wool gloves
• Level 4: 192 wool gloves
• Level 5: 544 wool gloves

I'd say that sufficiently tapers off in effectiveness, wouldn't you?

Archery

Also, I looked into ranged skill gain. Yeah, they're crap for a reason. At most you get 5 xp per arrow/bullet, when even a level 0 craft gives you 20. Needs something of a boost.

[ETA] Sorry, that's 5 for a normal hit, 4 for a miss, and a whopping 9 for a headshot. Be still, my heart.

KA101:

Hard cap or this doesn't get merged.

If need be, but first I'd like to ask you to at least try out the current build with the soft cap. I think it works pretty darn well.

[KA101]

No means no. I don't care to have the forums whine about grinding fishhooks to make swords again, gagh.

[facepalm]

I don't care to have the forums whine about grinding fishhooks to make swords again, gagh.

Now THAT'S a good reason, you should lead with that next time. The reduction of internet whining is truly the noblest of causes.

Any chance I could argue for a level+1 hard cap, then, with the current xp reduction? Even having one extra level's worth of possible crafts would add a lot of flexibility to what is currently a limited and very grindy affair. Or is there some specific scenario where people were griping about a level 0 craft helping you reach the cloud-scraping pinnacle of level 2?

[KA101]

Kevin made the point a bit more obliquely in his post with the round & round note. ;-)

Or is there some specific scenario

Yeah, the fishhook one, pretty much. Grinders are apparently very very good at finding optimal ways to grind. -_-

[facepalm]

Ah, I assumed he meant "round and round [trying to improve the game]" which meant that other changes might alter the dynamic enough for soft caps to be revisited, not "round and round [listening to people piss and moan]."

[facepalm]

Okay, switched back to hard cap (boo), and took a first pass at ranged improvement (yay). In quick testing, ranged level 2 was quick and easy to get to. I'll make a few dedicated ranged characters tomorrow and see how far I can push it - it's likely to need more scaling.

[Podesta]

Am I playing cata in some terribly wrong way? Because honestly, grinding is very much still needed. Definitely, instead of grinding based on the same item over and over, you instead grind over one specific item for each level... I totally understand the history behind this decision, but maybe add an option for people to test it out and see how it goes, just like with rust? This way nobody is pissed, and if the grinding gets worse we scrap that away...
Maybe as a suggestion add the softcap and also make the lower level recipes require less time, giving even less xp?
I don't know, I just mean that it's worth a try, because saying that the hardcap have solved the grinding problem is just denial.. It's definitely better than it was before, but I think we should test new things to see if we can get a better solution..

[KA101]

It solved the problem of people complaining that they could grind on easy/useful things and then go directly to endgame crafting.

Kevin's idea is the way to go, if we could free up the time to implement it.

[DeNarr]

Throwing in my 2 cents, I find crafting to have been incredibly grindy. My typical playthrough involves trying to find the books I need, so that I don't have to deal with that crap. With the possible exception of grinding on one specific recipe if I need to get my skill up to read the next book.

Cooking is the biggest issue here. I don't have the inventory space to create all the weird dishes I need for higher levels, and for some reason I need to create a bunch of jerky if I want to learn chemistry properly. I find myself starting as a cannibal just to help skip some of that process.

[KA101]

I don't have the inventory space to create all the weird dishes I need for higher levels

Was worse with skill rust, ugh. I remember making all sorts of weird shit just to get from 7 to 8 cooking before the skill rusted away. Cooking-practice would be tough to justify as a task--measuring out precision mixed drinks/prescriptions or something? but would be very welcomed.

[facepalm]

What about changing the hard cap from level based to recipe based ( along with the other changes mentioned )?

I think that would just result in people complaining that instead of having to make lots of one thing, they have to make two of everything.

Fundamentally, we're trying to compress all of human knowledge and experience into about a dozen integers that go up to about 10. No matter how we try to approach it, there's going to be edge cases where it's silly.

Having thought more about it as well, if I were to refactor the skills system entirely, I'd do a system which is on its surface like Llamageddon proposed, and split skills into knowledge and ability scores. Ability would function basically as skills do now, but capped by knowledge. Knowledge (not ability) would be gained by reading books, and if knowledge and ability were equal, would slowly rise over time to reflect one's internalizing of practical lessons, until knowledge = ability+1, and then ability could level again. Recipes would be learned/auto-learned based on knowledge, not ability, so a player with a lot of book-smarts can try harder things, fail, and learn something from it.

But it'd be a lot of work for something that's not all that different from what we have already.

[KA101]

I think that would just result in people complaining that instead of having to make lots of one thing, they have to make two of everything.

Yeah, the necromancy-leveling system in Solid & Shade. Make one of everything at this tier to go up to the next one; functionally the first two tiers are more or less a resource sink, as are most of the third. Interesting mod but I can't say I approve of the design.

[ejseto]

Little late to the conversation

Me too. Didn't even realize this was happening until it was in. It's impossible to learn anything now. Square the required experience? So, what, are we all expected to run around with sub 3 in everything now? There goes my stint at not cheating. I'd just revert it in my local fork but why bother restraining myself then?

Is this just for testing/feedback or is this seriously staying in?

all of human knowledge and experience into about a dozen integers that go up to about 10

The hard skillcap is 100 and there are definitely things that need above 10. You can't even add a second engine to your car without 12 mechanics.

[Coolthulhu]

Mechanics needed 12 because it was just so easy to get there.
Now it should be lowered hard.
And mechanics gain should be tied to difficulty of the part being removed (much more than it is now, at least).

[facepalm]

So, what, are we all expected to run around with sub 3 in everything now?

Crafting isn't greatly impacted, since craft duration serves as an xp multplier now. Books aren't affected - they scale with the new level xp. Ranged xp sucked as it was, it should scale okay now. The only things that I didn't touch were melee and mechanics, mostly because they were already so ridiculously easy to gain that any further changes would have been made in a vaccuum: I don't know what further tweaking they'll need. In playtesting, they both did okay.

I've noticed Melee is now significantly helped by getting a good weapon. Previously you could assault a horde of zeds with a pocket knife and come out the other side covered in gore and with a ridiculous Melee skill.

Mechanics is slower than before, but fine. Coolthulhu's right though, requiring 12 (or was it 16?) Mech to add in a second engine was entirely because it was so easy to increase. That should be lowered, sure, but to what? No idea yet.

The hard skillcap is 100 and there are definitely things that need above 10.

Yeah, but 10 is generally considered "godlike," as evidenced by all of the crafts topping out at 8 or 9 being the pinnacles of pre-Cataclysm civilization.

[DeNarr]

Hmm, I really liked how Mechanics made use of high skill level. I was actually hoping other crafts would get some end-game use as well. Maybe making "high quality" versions of the same items. Like food that taste better. (A steak cooked by a really good chef tastes much better than one cooked by a fry cook).

[ejseto]

Crafting isn't greatly impacted

Yeah, I posted that before I tried crafting anything. After blowing a stack of 12 first aid kits to heal after a fight, I didn't get even 1% at skill 4. I also didn't get even 1% to my unarmed after taking out half a dozen Z's, though that was at skill 8. Still, why are you nerfing combat skills? It's not like I was grinding unarmed. The Z's were in the way of supplies I need to grind other skills. Getting points for that is nice, but not an end in itself.

Why is it better to slow advancement to a crawl than just cap skill improvement? Instead of, "Well, now I have to pummel a 10000 zombies instead of just 100," wouldn't it be better to have something like, "You're already an expert at pummeling things, you're going to need to learn a real fighting technique to improve," the same way crafting is currently capped?

[Coolthulhu]

wouldn't it be better to have something like, "You're already an expert at pummeling things, you're going to need to learn a real fighting technique to improve," the same way crafting is currently capped?

It would be great to have a system like that too, but how would you balance it? What would be "a real technique"?
Just requiring stronger enemies could easily promote leaving stronger enemies for later. Requiring better weapons would defeat the point. Requiring more martial arts would be big RNG and a hard nerf for low item spawn rate worlds.

Still, why are you nerfing combat skills?

Melee combat skills grind themselves. Before, they were so easy to gain that you could easily get lvl 8 or so before getting any other skill at 4, without even trying.
Didn't yet test if this nerf is good, but I'm certain that a nerf was needed.
The only really bad thing from the melee skill rate nerf I can think of is that now dexterity is even more important than before. Though that should be addressed by nerfing dex rather than buffing skills.

I agree about the first aid. This skill needs a skill rate buff - would be cool if healing self actually gave respectable xp.

[facepalm]

First aid could probably use some love. I lumped it in with electronics, computers and construction as "skills you already need books to train, so I'm not going to worry about them."

ejesto

I also didn't get even 1% to my unarmed after taking out half a dozen Z's, though that was at skill 8. Still, why are you nerfing combat skills?

followed by

Why is it better to slow advancement to a crawl than just cap skill improvement? Instead of, "Well, now I have to pummel a 10000 zombies instead of just 100," wouldn't it be better to have something like, "You're already an expert at pummeling things, you're going to need to learn a real fighting technique to improve," the same way crafting is currently capped?

Why is it better to get no skill increase than <1%? You're complaining that melee skills got nerfed, and then saying they should be nerfed even harder. At skill 8, you are already an expert at pummeling things. Why wouldn't you expect further progress to be slow?

[ejseto]

Because you get a notification that there is something impeding your progress, which makes sense. No matter how much you master the art of smashing things with your fist, you can't progress beyond that without real training. No matter how much you master the art of roasting squirrels, you still won't be able to make mutagen. Part of the problem is the abstraction of placing all these essentially dissimilar things into the same categories. I'd rather be notified that something, which I can somewhat control, i.e. by learning the appropriate technique/recipe, is holding me back, than have interminably long bouts with no apparent progress.

Requiring more martial arts would be big RNG and a hard nerf for low item spawn rate worlds.

Well there's supposed to be NPCs that can teach you. There used to be, though all I remember them doing is causing segfaults, but still. Considering there's still placeholder text that implies that's what they'll do, once they're working, there should eventually be another way to learn martial arts.

would be cool if healing self actually gave respectable xp.

Isn't this what should logically be giving them most (outside of skill books) xp? How else would you improve it, by crafting bandages? Does that even make sense? Gee, I figured out that alcohol kills germs, medical school here I come!

but I'm certain that a nerf was needed.

Perhaps, but this is too much. Everything was harshly nerfed, then crafting and reading were buffed to compensate, while skills for actually doing stuff outside of crafting were left to rot. Crafting recipes are already skill capped, why not combat? Why can't it be something like "you can't advance beyond 3*(number of martial arts techniques known + 1) skill" or something (or I suppose you could assign values to them but that seems arbitrary)?

[KA101]

Is this just for testing/feedback or is this seriously staying in?

Your feedback is greatly appreciated and I assumed that we'd need to tweak this based on playtesting. My testing focused on ranged combat, where the .22LR provided pretty decent growth at low levels.

MA skills are currently restricted by access to opportunities to gain them; I suppose they could be capped by levels in Unarmed, though that'd just enforce optimization harder.

[ejseto]

Archery is actually really easy to level up, it seems. Starting at 2 which I got from a book, I crafted a slingshot and tons of pebbles and just hold down 'f'. Z's can't touch me through my armor and every shot is a headshot for ~2.5% per shot. I think it's down to ~0.4% at skill 4, but still pretty ridiculous for a weapon made of twigs and rocks.

[tyrael93]

You think that's bad? Blowguns can get you to archery 10 in a moment. Apparently someone forgot about capping XP from blowguns and since they're craftable from the start you can easily become a master archer in two days tops. Now, I haven't checked if shooting against a wall still works, but if it does...
But it probably wasn't a bug, right? I guess breathing control really does play a big role when firing an arrow!

[Coolthulhu]

How about scaling with (base or scaled) attack speed?

• Pistols have low move cost, so you can fire often
• Shotguns and rifles have higher cost, especially heavy rifles
• Launchers higher than that
• Bows and blowguns have moderate base cost, but it drops really quickly with xp
• Axes and heavy hammers are quite bad as weapons due to their high cost (makes it hard to kite and dodge), they could use a buff
• Firsts and combat knives are amazing weapons that exp fast and drop focus to 20 after few minutes of combat

Explanation: IRL training doesn't strongly depend on number of repetitions. Otherwise combat training would be about wild swinging. Plus, swinging a heavy axe gives more "feedback" from the body (balancing, pain from effort etc.) than thrusting a punch. Not that there is no "punch thrusting theory", it's just that just punching mindlessly probably won't help you develop it.

As for wall shooting: maybe cap the missed_by value when it doesn't actually hit anything? That would be a simple "no grinding" switch, but it makes sense to require live targets to train well.
Even better would be also scaling the xp with target size (headshotting a fly is harder than headshotting a hulk).

[Eliijahh]

Yeah I agree. The fastest weapon should train the skill at the same rate as the slowest weapon.

Returning to the main problem, grinding vs not making you a god too fast with any skill:
Why don't you make skills cap? You do your sword fighting, then you get to 4 max. If you want to improve, you must find someone that can train you. Then your skill cap gets to 6, and you can train ,till six. In real life people were trained by other people, and the knowledge was passed generation to generation, each generation adding a bit and improving the art. And before you are really capable of adding up, you need to have a huge knowledge, that is impossible to learn by yourself.

More radically, have you ever thought to add limitations tied to the players stats? Like, if you choose strength 10, you will never be able to get more than 10 in all strength related skills? (Like melee could be a medium between strength, agility, perception and intelligence, cos you need all of them) I mean, it is very radical, and a sort of limitation of the game, but maybe with the npcs being fleshed out, it would make sense for each of the characters to not be perfect jack of all trades. In today's world nobody can be a MASTER electrician/mechanic/swordsman/chemist.
I have no idea how to implement it, but just a thought. This way the grinding is not really required, because you would have a sphere of expertise, (that you could improve of course) but that's it. Then, in really late game, you could find ways to modify your brain or body with implants and mutations, so that you could become not a master but still good in all the skills.
In this way character creation is much more stressful, each time you choose a set of skills, stats, perks you have a different playthrough.

[Shoes01]

I find it very interesting to lock the skill cap... books or trainers could raise the cap, but then you still have to grind/train/fight to level up your skill.

[Eliijahh]

Yeah it is true. It is also hard to simulate something that in real life is a grind, but since it is made interactively it feels less grindy, and with big limitations: look at learning how to play the piano, I think someone used the example here, you do countless of hours of the same basic repetitions to teach your body-memory how to use those fingers.
So, grinding is a part of learning, of course. So, either you find ways to make the grind, actually feel like it is a fun part of the game, but it is very complex. Or, you try to remove it from the game.

One way would be to tie a set % of success to make a item. The more you try to make the item, the higher the % goes up. (with some items having the same %, or bonuses, so if cooking meat as a 95% to be made than cooking a more complex meat dish has a 50% (the base) + 10% of the 95% of the normal meat dish.) So you start by not knowing how to make anything, you learn recipes from books, and then you start with a set % that increases each time you try to make it. So knowing how to make a arrow perfectly will not help you with building a metal axe. This way, at least for crafting, you could also not use levels.

[facepalm]

Coolthulhu:

How about scaling with (base or scaled) attack speed?

Not a bad idea.

But I think the simplest solution for slings/blowguns is to add back in the damage scaling I had earlier just for the archery category, which would cause those two to provide much less experience, since they deal so little damage. That kind of scaling was explicitly vetoed by the powers that be though, so no, it's not a bug, it's a current feature.

Plus adding some more dispersion on the blowgun, because that is some sniper-level blowdarting now that I look at the statistics.

[ejseto]

Why should blowguns provide less experience? Why should someone who has spent a year learning to use a blowgun be less proficient at it than someone who spent a year learning to shoot a bow? The real problem is that blowguns and bows have almost nothing in common. The problem is that I can make 1000 darts out of driftwood, hold down f, and get to level 7, just because I have good armor, and then translate that to better weapons that really have nothing to do with blowguns. And I'll probably go further, just cuz.

add limitations tied to the players stats

This is a good idea. There has to be something that prevents people from just becoming masters of everything for all time. Skill rust does something but not really anything you can't work around. Oh, no, my archery is rusting? Good thing I kept my trusty blowgun on me. It's an easy-to-game system, and it's not even fun, because it encourages that kind of boring play.

[KA101]

Plus adding some more dispersion on the blowgun, because that is some sniper-level blowdarting

By all means; perhaps have it strictly generi-ranged Marksmanship rather than archery.

Hard caps are out until there's some (probably NPCs) way to otherwise access that content: no making content mutually exclusive (because you developed the skill to do foo, you may not develop the skill to do bar with the same character) w/o workarounds.

I've worked* through Arcanum and GearHead2, both of which prominently featured that sort of skill lock, and they were rather unrewarding, to put it nicely.

• I did not play those. They were tedious and non-fun, thanks to having to pick a tight focus from the start and mortgage character development to that focus.

[facepalm]

Ah, Arcanum. Such great ideas, with such terrible design. I remember being especially impressed when I realized that literally every tech consumable involved at least one thing that could only be found in garbage cans, only one at a time, refreshed once a day.

[kevingranade]

I'm totally behind scaling skill gain with time spent in general (See here
http:// http://smf.cataclysmdda.com/index.php?topic=3728.msg53912#msg53912
smf.cataclysmdda.com
http://smf.cataclysmdda.com/index.php?topic=3728.msg53912#msg53912/
http://smf.cataclysmdda.com/index.php?topic=3728.msg53912#msg53912
index.php
http://smf.cataclysmdda.com/index.php?topic=3728.msg53912#msg53912
?topic=3728.msg53912#msg53912
http://smf.cataclysmdda.com/index.php?topic=3728.msg53912#msg53912 for my
proposal along the same lines).
As for the blowgun, reducing its accuracy and making it not use the archery
skill sound fine to me.

[DeNarr]

Hmm, if Blowguns didn't use Archery, what would they use? On a similar topic, I find it very odd that Crossbows use Archery. They have much more in common with a handgun (or rifle for large crossbows) than they do with a bow.

[ejseto]

As for the blowgun, reducing its accuracy

That's not really the problem. It doesn't really do any damage anyway, which is actually a good thing, since it allows you to just sit there and spam fire without even looking for new targets. Even the weakest Z's will take ~5-10 shots at 6+ skill to go down, at point blank range.

making it not use the archery skill

This is what needs to happen. The problem is the blowgun isn't a weapon unto itself, it's just a training tool for real archery weapons. Whoever made the blowgun, I wonder what they were really thinking. Were they thinking, "Gee, it would be really cool if I could be a ninja and shoot people with poison darts!" or were they thinking, "Archery is too hard to level, it needs a stepping stone." The problem is the stepping stone is a whole staircase.

[Snaaty]

Well, I think making the blowgun useful for injecting poison/chemicals into your enemies would be really awesome. I imagine a more advanced poison-making process, and at some point even chemicals/poison that you can use to inject zombies and slow them/kill them.

[ejseto]

more advanced poison-making process

This is another alternative. One of the reasons blowguns are so good as a stepping stone is that when archery was overhauled, it only included bows. So making arrows is a multi-step process, whereas I can dismantle a dresser and make hundreds of darts out of the wood. Blowguns just plain aren't good weapons currently, and their only use it to make it easier to use weapons that realistically have nothing to do with them. And they're REALLY good at that.