Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Aseaheru]

you have to tell them to.
ive had groups spar in my first winter.

[Hyndis]

Multiple sparring matches can happen at the same time, in the same squad, at the same barracks.

I use a danger room briefly to jumpstart the melee core of my military, just to fill out my first squad, then I split them up and spread the trained dwarves amongst multiple squads, and fill in the rest of the squads with newbies.

Once everyone is recruited and is at level 4-5, its quite common to see 6/10 dwarves in the squad actively sparring at any one time. Occasionally all 10 are sparring all at once, though that is rare due to nap times.

What I do for my military is equip my dwarves with the finest weapons and armor I can forge. I assign them to a squad at the entrance to my fortress. Everyone is assigned to active/training at the same barracks right at the entrance. If there is something to stab, they stab it. Otherwise they train amongst themselves. Its a great passive defense. No stealthers can get by it, because the blob of dwarves physically blocks the entire entrance width.

However the experienced gained when on duty but not engaged in combat with goblins is depressingly slow. Even if I jumpstart training by throwing the first ten dwarves into a danger room, and then divide up those ten danger room trained dwarves into 3-4 squads with the hope that the elites will be able to train the newbies, it still doesn't really work out well. Training just seems woefully slow, to the point that experience loss via the occasional dwarf being killed or maimed to the point they can no longer fight can outstrip experience gain.

Of course once you get a critical mass of elite dwarves all in a blob of stabby death, then losses pretty much never happen. Its getting up to that point which is very hard to do.

[Damiac]

Socks, it sounds like your training success mostly comes from the live training.  Combat XP is pretty good, that's how I get my archers to elite pretty quickly.  With melee, you have to find a way to keep the targets from running around too much, (like putting them in a pit).  Otherwise your soldiers just waste all their time chasing things around.

I've considered using live training dummies by capturing creatures in cages, but cage traps just seem too exploity to me.  But that might be the only way to really train up the melee troops effectively.

I usually put them in big squads of 10, but with 5 seperate train 2 orders.  This does result in fair amount of sparring, but skills gains are still pretty slow.

So I guess the conclusion is that training just needs to work better and faster.

[GreatWyrmGold]

For live training, I just send my dwarves out, armed or no, unarmed or no, to attack wild animals.
It gets food, to!

[Damiac]

I tried doing that for a while, and they usually have to chase something back and forth over the entire map 3 or 4 times before they get lucky and corner it, and are able to actually hit it.  So it seemed that it was even less efficient than just normal training.

I suppose if they weren't wearing armor they'd have better luck catching stuff, but then they might get killed by it...

[sockless]

But.... why? If someone wants to use an obvious exploit to train their dwarves extra fast, that's their call.  It's a single player game, who cares if people want to cheat?
If you use legitimate training methods, the experience gain is quite slow.  It's not as if you can accidentally exploit the system as it is now, you really have to go out of your way.

That said, your suggestion doesn't seem entirely unreasonable, it just seems like it wouldn't impact people playing the game normally.  I don't see the point of making a change just to stop training exploits.

This.

If people want to use exploits, then go ahead. That's why we still have Dwarven Atom Smashers. Until we get online leaderboards or something, there's no incentive for Toady to fix exploits when he can use that time to add new features.

[expwnent]

The point is that danger rooms are unbalanced. It's a valid strategy that is too powerful. It is a (small) flaw in the game, and the purpose of the discussion is to come up with a good solution. I agree Toady has better things to work on but that doesn't mean we can't discuss anything else.

[10ebbor10]

Maybe we should scrap the entire skill thing, and replace it by something else. A couple of bitmaps representing complimentary skills for example. A dangerroom would only be able to train some of these(Endurance, evasion, ...) and only up to a certain level, at which point it'll slowly reduce. Same for other things.

This solves the issue where any repetitive task can create a legendary fellow, as you really have to do a variety of tasks to unlock all the subskills.

[GreatWyrmGold]

No, skills are important. Carpentry may be similar to masonry, but they don't have that much in common. Changing the skill system might be useful, but not destroying it altogether. Why should a dwarf need to learn masonry-related skills like engraving or mining or stonecrafting to make masterful statues or thrones?

[expwnent]

Actually, I think that would be neat. To truly master one skill, you also have to know things about related skills.

As long as it's done in a not-annoying way, of course.

[Fluoman]

You mean "pyramidal" skills, where for example, one cannot become legendary woodworker if not already at least competent woodworker (or some other combination of skills)?
That won't affect danger rooms, since they train everything at the same time.

[GreatWyrmGold]

Actually, I think that would be neat. To truly master one skill, you also have to know things about related skills.
As long as it's done in a not-annoying way, of course.

Explain how related the skills have to be.
I think the inverse would work better--say, being a legendary mason makes you a better stonecrafter than a random cheesemaker.

[expwnent]

Hm. I guess that makes more sense. But it might be more efficient to learn how to approach the problem from multiple perspectives.

[GreatWyrmGold]

Hm. I guess that makes more sense. But it might be more efficient to learn how to approach the problem from multiple perspectives.

How does it make sense that you need to learn how to carve pretty pictures in walls or make stone rings to make great stone furniture?

[Maklak]

Personally I have no problem with danger rooms or coin tossers. I used that once and the skill gains were great, but stat gains not so much. Some DR trained soldiers still got killed too. Now I struggle to make normal training work and, for the most part, fail at it. After about 3 years I can get to weapon skill and shield user 10, but few soldiers live that long. I even had 8 squads of two recruits train for a year with almost no gains whatsoever.

If you want to "fix" DRs anyway, it would be enough to limit experience gains from all types of traps, be they weapon or upright spear.

As a side note, I once had a single Unicorn with a ripper (dagger) and 4 shields train up to 18 Knife User, 20 Fighter, 12 Shield User and 1 Armour User in a short fight with a Steel Ranger after 20+ pages of combat reports. He was in a caste with 240 skill rate for Dagger and Fighter, but he still made it from recruit to hero in record time.

A few problems with normal training are outlined here: http://www.bay12games.com/dwarves/mantisbt/view.php?id=3855#c23741 I typically have very poor learning rates for dodge and armour user. I even tried squads of 2 without shields and it still didn't work too good.

As for the system where training one skill helps train another, you misunderstand. Currently we have skills and stats. Similar skills train similar stats, making training more skills just a bit easier. A while ago there was a proposal for a lot more complex system.
* On the first level you have stats, same as now. These slowly increase with abilities.
* On the second level you have abilities, such as smoothing, carving, cutting and so on. These train when a skill is used.
* Finally you have skills that are weighted averages of abilities. Stone-crafting uses some of the same abilities that bone-carving. While you couldn't train a skill directly, using it would increase it's base abilities and therefore the skill itself. Also, a legendary stonecrafter would have mastered some of the skills needed for bonecarving and vice-versa, giving us skill synergies. There was even an idea that goes even further, to skip skills and directly use weighted abilities for each reaction and job.
While this system is interesting, is would be difficult to decide what abilities there should be and how to choose their weights to determine skill level for a job, such as "make rock throne".