Difficulty

[darkgod]

Do I get the feeling greycat does not like the Maze
Dunno about the traps I never had much trouble with them here.
Yes the minotaur is a nasty bastard.
Maybe the maze should be more optional ala elven ruins and be replaced with a "saner" zone ? I have some that I did not implement.
Humm
Maybe I can elevate amon sul to be in the same tier as the lair and the old forest, make he trollshaws the starting zone and create a new second starting zone ?

Oh also remember this is only the first par of the game there is a whole continent full of stuff to be made.

Anyway my real goal was not to prove levelign monsters is neat but rather thsi one point:
To ensure that at any given level the player has a choice of where to go.
This obviously fails a bit since you have to visit most of it anyway.

It isn't that I consider many talents "useless" as they are, it's that they do not synergize in a way that justifies the levelling mechanic you currently have going. To be able to keep up in this kind of arms race you have to specialize by necessity. Diversifying your skillset is punished by the levelling mechanic because you will be that much weaker for trying out something new instead of improving your already strong abilities to be even stronger, in order to keep up with the steadily improving enemies. This results in a boring game with lots of mechanics that are never needed or used, because they don't strictly make you stronger.

Now, how would this differ in a level-per-depth game instead? On the first glance it would seem there is no difference at all, right? After all, the monsters still get stronger. Well, yes, but they get stronger based on how far you've progressed in the game, not how far you've progressed in reaching maximum level. In a regular roguelike, a gimped character can still progress. He'll be at a lower depth that he would normally be, given his level, and he'll have to grind a bit more, but he can still progress. In the current levelling system, he will eventually be unable to progress at all, because the monsters on the next level WILL be as strong as he should be (not as strong as he actually is), and going somewhere else is not an option as all dungeons in their tier are effectively the same.

Again, I do not see the difference, If you are too weak for where you should be you come back to an earlier zone or level no ?
We can all agree the current "steps" are too big, so in the event they are smaller, with zones having a level range of only 5 levels, it works exactly the same no ?

PS: At level 4 in next beta with just 20 cunning you'll get a radius 5 "infravision" for just 1 talent point, should make unlit areas much more bearable

[kaschei]

Since I vehemently dislike the levelled aspect of the ToME module (or rather, the segment that responds to player level -- bad idea is bad idea), but really like the engine in general, anyone can make this change to their copy:

\game\engine\zone.lua:
change
self.base_level = util.bound(plev, self.level_range[1], self.level_range[2])
to
self.base_level = util.bound(plev, self.level_range[1], self.level_range[1])

This may also imply some tweaking to the min ranges of various dungeons. Or I may be reading the code wrong entirely.

Can this sort of engine code be overridden with module-level functions? If so, please drive through. Otherwise, this should be defined at the module level.

[Baker]

To ensure that at any given level the player has a choice of where to go.
This obviously fails a bit since you have to visit most of it anyway.

I get that idea, I just think it doesn't work in practice. Sure, you're free to go wherever you want -- as long as you stay inside your level bracket and minmax your character so the leveled monsters don't insta-murder you. And don't mind that everything is kind of the same anyway.

Having content that requires different levels of strength and ability isn't necessarily a bad thing. I mean, you're still doing it, after all, not everything is leveled.

Again, I do not see the difference, If you are too weak for where you should be you come back to an earlier zone or level no ?
We can all agree the current "steps" are too big, so in the event they are smaller, with zones having a level range of only 5 levels, it works exactly the same no ?

Somewhat, except for the fact that you'd still have to grind for 5 or more levels to stand a chance against the enemies you should already be able to defeat. This just doesn't really solve the problem. How ARE players expected to fix it if their character simply isn't as strong as the leveling algorithm thinks he should be? Remember, leveling up doesn't really help, because the difference between "how strong you are" and "how strong you should be" remains the same and can't be made smaller, only bigger.

Sure, once they've outleveled their current content and enemies stop getting stronger they can FINALLY start moving forward, but as soon as they get to the end of the current dungeon and get to the next one they'll run smack dab into the next wall, because the level range is higher and they have to out-grind the level scaling yet again. And once you're making the steps smaller you're only doing a level-by-depth thing anyway. The result would simply be that people would grind on the first level (like greycat does) until they've outleveled the content and it doesn't scale to their level anymore so they can do the rest of the dungeon at their own pace. Doesn't that strike you as kind of pointless? After all, why have content scaled to your level at all if everyone seems to treat it as an inconvenience and is doing their best to avoid it anyway?

What I'm getting at is really that any XP-based RPG system will always suffer from this kind of problem. In a game that requires you to kill stuff to get stronger you can only get stronger by becoming better at killing stuff, everything else is a waste of points. This is fine in and of itself, but it becomes a problem when you implement a leveling system that doesn't care how strong you are, only how strong you should be. This is why most roguelikes give the player control over the kind of challenge he wants to face, by making stronger enemies a function of dungeon depth or, in other words, optional. The levelling system you've made takes that control away from the player.

I'm at the point with ToME 4 where I'm not as bothered by the leveled monsters as many of the other players. In part, this is because I've adjusted to the expectation that I am going to munchkin-ify myself for combat, by ... munchkin-ifying myself for combat. I pick one method of Killin' Stuff that will work for the class in question, and focus almost exclusively on that. There is almost no deviation in my class builds any more.

I mean, seriously, does this sound like the kind of game you wanted to make?

[Baker]

In Angband as your character advances, the challenge (and reward) level at your current depth decreases. The player is allowed to choose the depth and therefore the challenge level for their character. A new monster must be introduced at certain depths which is superior to what came before, and often times the transition to a new monster from the old is rigid: the old monster is trivial, the new monster is too risky. To compensate, the player must clear the old monsters for longer in order to reduce the risk of the new monster, that is, he must "grind up" trivial content in order to experience new content. Because of this system, I believe that Angband is the heaviest grind of all the roguelikes right now.

When ToME scales the monsters to compensate for the player's strength it keeps the challenge level consistent, so that at no time does the game become completely trivial. Or at least, that is what I believe it is trying to do. As a bonus, it means the content developer can create one set of monsters for each dungeon which will be scaled automatically, rather than having to create many different tiers of the monsters based on the depth. The skill tree provides the player with considerable advantages as well, especially as one levels up and gains access to higher tier abilities.

In my personal opinion I much prefer the scaling approach, as grinding up trivial monsters really isn't my thing; and I like being rewarded for making smart character development choices.

The whole scaling approach would be totally alright with me if it was based on dungeon depth instead of character level. As you said, it offers a consistent challenge level without any jarring jumps in difficulty, AND it means you don't have to create a bunch of redundant (and probably badly balanced) enemy variations.

The problem just lies with the "based on character level" thing. Just imagine Angband if you could walk to the lowest level of the dungeon without problems and defeat Morgoth at level 1. It might just be me, but I don't think this is how it should work.

[Gwai]

By far my favorite thing about the scaling monsters is that it allows me to chose what dungeon I do in a way I never could in Tome2. For instance, my order of dungeons (for any of the classes I play) is Old Forest, trapdoor, maze, then probably amon sul and maybe trollshaws if it didn't annoy me too much. Amon Sul just has too many places for inescapable skeleton archers for me to want to brave it early. Even late I usually only do it for completeness. Similarly, trollshaws is cool looking but way too annoying to clear, so it's not a good place for me to level at all. And I don't want to take on Bill until I'm at least level 4 with most characters. Thus I wait until I can afford to dive.

But if Joe loves amon sul and thinks it excellent for his characters' life expectancies he can do it right off. I really like that.

[Canderel]

The whole scaling approach would be totally alright with me if it was based on dungeon depth instead of character level.

How about both... Make dungeon level count about 70%, and 30% player level to calculate monster level. Or even 50/50...

[darkgod]

Baker I really do not follow, if zones level are choosed only once then all it changes between fixing it and "letting the player" fix it is the order.

Let's say you have 3 places (A B C) with 3 levels each (1 2 3).
If I use fixed levels then I can lets say do this ordering:
* A 1 = 1
* A 2 = 2
* A 3 = 3
* B 1 = 4
* B 2 = 5
* B 3 = 6
* C 1 = 7
* C 2 = 8
* C 3 = 9

Now if I tell the zones that you want them to have a level range (let s say 9 for the ease of example). You can have the player choose this:
* A 1 = 1
* A 2 = 2
* A 3 = 3
* B 1 = 4
* B 2 = 5
* B 3 = 6
* C 1 = 7
* C 2 = 8
* C 3 = 9

or

* B 1 = 1
* B 2 = 2
* B 3 = 3
* A 1 = 4
* A 2 = 5
* A 3 = 6
* C 1 = 7
* C 2 = 8
* C 3 = 9

or

* C 1 = 1
* C 2 = 2
* C 3 = 3
* B 1 = 4
* B 2 = 5
* B 3 = 6
* A 1 = 7
* A 2 = 8
* A 3 = 9

How is that different ? once a level is set it is set, monsters in it will never become stronger.

Sure, once they've outleveled their current content and enemies stop getting stronger they can FINALLY start moving forward, but as soon as they get to the end of the current dungeon and get to the next one they'll run smack dab into the next wall, because the level range is higher and they have to out-grind the level scaling yet again.

It seems to me that people think monsters level with the player ?
It is not the case monsters level with the zone level (which never changes) and the current depth in that zone.

Why do you feel you have to grind the first level forever to outlevel the zone ? If your character build is too weak it's a problem of character build. Fixing the levels of zones would not make you have to grind less.

PS: And yes this is not an engine requirement BTW it's the zone that says it wants to levelup or not

[Baker]

It seems to me that people think monsters level with the player ?

Yeah, actually, I thought that's what they did, mostly because of what you and Nerdanel said on page 1 of the thread. That's also what it felt like for me (I didn't feel really that much more powerful at level 6 than level 1). So I guess I misunderstood because of the bug and enemy level really depends on how many dungeon levels you've progressed through in their difficulty bracket? In that case I was apparently arguing about nothing, sorry. That's actually a pretty nice way of dealing with scaling.

[darkgod]

No are were not, it was constructive
But yes once a zone level is set it should stay this way, bug notwithstanding.
I want the leveling thing to enable players to choose their zones order (but not completly).

But you are probably right that 11 levels of range a too much maybe.
Maybe also the player leveling speed reduces too early ?

[greycat]

Does this mean when beta4 comes out I'll stop dying so damned much? That'd be awesome.

If that's not what this means, then I'm just confused.

/me adds "Annihilation of all healing potions by green worm masses leading to death later" to list of gripes.

[Baker]

But yes once a zone level is set it should stay this way, bug notwithstanding.
I want the leveling thing to enable players to choose their zones order (but not completly).
But you are probably right that 11 levels of range a too much maybe.

I think 5 levels or so would be better, yes, especially if you created a few extra dungeons that overlap a bit level-wise. So you'd maybe have a 1-10 bracket, a 5-15 bracket, etc. This would give players a bit more variety and the chance to take extra risks early on for larger rewards, you know what I mean?

Maybe also the player leveling speed reduces too early ?

I can't really say, because I haven't gotten in very far yet, but it just feels slow, maybe because a single level doesn't mean that much when you've got a level cap of 50. Leveling up doesn't seem to make you that much more durable or better at dishing out damage. In Crawl, for example, where you've only got 27 levels, both you and your enemies get stronger much more quickly. By level 4 you've usually found the Ecumenical Temple and chosen your religion, by level 9 you can cast the highest categories of spells and have probably started raiding a branch dungeon already, by level 15 you can start thinking about assembling your end-game gear. Grinding isn't really an issue either, because skills level up as you use them and can be higher than your character level.

So in comparison, ToME just seems very slow and grind-oriented right now, especially because enemies seem to be very powerful in comparison to your character. Meeting a Rattlesnake on dungeon level 1 that hits you for nearly a quarter of your total health and poisons you to boot is kind of nasty, and having to rest after taking out a pair of crystal rats makes everything slow down to a crawl. I'd prefer it if you progressed a bit (or a lot) faster, because grinding for XP and gear with in a game with permadeath seems kind of pointless. It would help if you made enemies that are a higher level than you more valuable, because right now grinding on weak enemies is much more efficient than taking risks. I killed a skeletal warrior that hit me for nearly half my health once (that took some luck, by the way) and didn't even make half a level or a full piece of gold out of it.

[Canderel]

Seconded... If I manage to kill the elite armoured skeleton, I want a level for it... well, I tend to play skeletons, so maybe I did get one.

But yeah, the tough monsters should be worth more. Especially ranked ones. Anacondas are not worth more exp than snakes (for their level)... or am I understanding this wrong:

Code: Select all

newEntity{ base = "BASE_NPC_SNAKE",
	name = "anaconda", color=colors.YELLOW_GREEN,
	level_range = {20, 50}, exp_worth = 1

vs
newEntity{ base = "BASE_NPC_SNAKE",
	name = "large brown snake", color=colors.UMBER,	
	level_range = {1, 50}, exp_worth = 1,

A level 20 anaconda = level 20 large brown snake in EXP? That lvl 50 brown snake is vanilla, can only bite me... lvl 50 Anaconda can constrict even a skeleton/ghoul to a pulp.

[darkgod]

Skeletons do not need to breath they cannot be constricted to death

I'll make exp affected by the NPC rank, so an elite would get a 3.5 multiplier (against a 2 for normal NPC).

Comparing ToME to crawl is IMO very funny because last time I played it (long time ago I admit) it felt much much worse than T4. I had to dive to find food, I could meet monsters that would simply oneshoot me on a very frequent basis..
If I added a food counter you'd play the same way but I hate food counters, they always feel silly, either too fast or too slow.

[Vanguard]

Oblivion was better in that sense, because everything was levelled.

Oblivion had the worst leveling system of all time.

The sandworm hole is OK if you have the phial of Galadriel, but if you're stuck with a radius 2 lantern, you are toast.

Cast vision or bring scrolls of magic mapping. If you plan ahead a little bit, the only dangerous thing in the sandworm lair is the boss.

And quite often, you can't do the sandworm hole at all due to the stairs being generated wrong.

Never had this happen before.

The maze is completely ridiculous if you aren't maxed in Cunning with trap disarm level 2 or better, because you cannot bypass the traps at all.

Those are annoying, but by the time you're ready to take on the Maze, a single trap isn't going to kill you. Item destruction is the real problem.

Not even mages can Dig those walls (or at least mine couldn't).

The maze is a joke as a mage if you have phase door or teleport (you should have both).

And the boss in the maze, while he has a spiffy hat, is absolutely murder.

His offense is scary, but to a ranged attacker it's a cake walk. For a warrior, yeah, you'll probably want to hold off until you've got enough defense to be able to absorb a few hits. Get the artifact shield first, at least.

Maybe I should just grind on level 2 of the Elven Ruins. That's not so bad. It's the boss on level 3 that tears me into strips

Okay, that guy is a bloody nightmare with his freezing status and 200+ damage supercrits. He's beatable enough as a mage, but a mage won't get much use out of his item. I don't know what a warrior is supposed to do except pray he doesn't hit you.

Really I think all the game needs to fix these problems are about two more dungeons that are at around the same danger level as the old forest and something to protect the player from stunlock kills (skeleton warriors, mummy boss) and instant death ranged attackers (elite archers, vampire boss).

Hey DG, what do you think of the idea of a passive talent for hand-to-hand classes that protects against stunning and freezing, giving 100% immunity to both at level 5? Archers and mages wouldn't need it, because if they're doing their jobs correctly they won't let boss monsters get in close range in the first place.

I also think scroll and potion destruction should either not happen or be very rare, because as was said before, there's nothing you can do to avoid it sometimes, and a lot of classes absolutely need scrolls/potions to succeed.

If I added a food counter you'd play the same way but I hate food counters, they always feel silly, either too fast or too slow.

Don't add a food counter. It has its place in crawl (even though I don't like it), but it wouldn't do anything to improve ToME 4, just like it didn't do anything to improve ToME 2. Same with needing to fuel your lantern, I'm glad that's gone. It was just a hassle that did nothing to improve the challenge or fun.

Personally if I were in charge I would make it so archers didn't need ammo either.

[Canderel]

Skeletons do not need to breath they cannot be constricted to death

I know, that's why I used the example... The lvl 50 anaconda should be (maybe isn't) more dangerous than the lvl 50 brown snake.

Do Skeletons take constriction damage, but stops before dying?