• NeilStevens
• Lessons From 2.3

This is developer discussion and planning. Heavy spoilers may be included without a second thought.

NeilStevens: Well, despite not getting as much testing as I'd have liked, ToME 2.3 seems not to have been a total failure (though some would like to have seen improvements, judging by the sudden agreement I'm getting that more testing would have been nice).

2.3 is done, so now is the time to learn from 2.3, and plan for ToME 3. ToME 2.4 seems unlikely, so we'll just have to tough it out and test the skill improvements in 3.0 with the rest of the game rewrite.

In ToME 2.3, I made a deliberate trade-off. Players lose early-game power in exchange for getting the ability to be as strong as possible in any skill they increase. Generalist classes, such as Mages, Paladins, or most any kind of Loremaster, should benefit greatly. Specialists, such as Warriors and Archers, lose out. This trade-off has been documented thoroughly at GeneralDiscussion/Abilities in 2.3 vs. 2.2, and after reading it I've been chewing on where to go from here.

My first thought is to introduce non-linearity to the skill progressions, but I'm not sure how to do this. Right now, every skill level is worth the same amount. One way to improve the plight of the early game would be to make the early levels worth more. For simplicity I'd probably keep things mostly linear, but make the progression be two lines with two different slopes. The first 5 levels might be worth what 10 or so levels are now.

I have several objections to this idea, though:

• If we're trying to help specialists, this is exactly the wrong thing to do. Specialists will lose the benefits of this change by the time they get their key skills to level 50, but generalists (who don't raise so many skills to 50) will keep the benefits all game long. Changing this would make Mages and Rogues much more powerful, not helping Warriors in the comparative race at all.
• This could create a mid-game hump for players to get past, a point where the player progression slows but the monster progression continues, and make the game less fun. The phase of the game where the player needs to maximize his stats is already enough of a speed bump that we really don't want another one added.
• Thematically, the plan doesn't make sense. Dilettantes who study a little bit of many things should get less, not more, benefit from their studies than someone who focuses on the same few skills. After all, when one goes out and fights the forces of Morgoth, one is not paired against amateurs. These are battles of life-and-death, where someone who didn't study all that well might have the most trouble.
• A minor point: the already-tricky gap between magic skill levels and spell levels will get even trickier.
• Lastly, and most importantly, I'm not even sure there's a problem. I really, really, really don't like the idea of ToME being a slow-starting game like Sangband or Angband, but with 6 points a level now I'm not sure that's really a problem. Sure, the start of ToME 2.3 is harder than the start of ToME 2.2, but is 2.3 really any harder than, say, Pernangband 5 with its deadly, deadly Bree? I remember remarking to the illustrious maintainer over on rec.games.roguelike.angband that Pern was silly because Bree is deadlier than the Barrow-downs. Are we that far back toward things being more difficult? I don't think so, but that's where player feedback is most useful.

Discussion

NeilStevens: I'm wide open to feedback here. What's wrong with ToME 2.3? What's right about it? Is the game too hard at the start now? Are warriors and archers unwinnable now (I know Sorcerors are still winnable)? How about Rogues and Rangers, did they get more hurt or helped by 2.3?

DarkGod: The game being a bit harder in the beginning does not bother me at all, people should suffer and die ! at least in a reasonable measure As for no 2.4, I fully agree, I think(hope ? help ?) that I can rebuild the ToME module to a point where it looks more like a game than a few monsters over an engine in a month or so

RavenRed: A few minor thoughts. Linearity is worth keeping, as it keeps progression logical and predictable. Another approach is make the early game match up to character ability, rather than the reverse. This helps the generalists a bit more in that they don't need relatively high fighting / casting skills in order to survive early. Finally, I think races need to be redone in terms of their starting abilities and ad/disadvantages...

MayLith: I guess this means I need go back on my avowal not to play 2.3.x until I've won 2.2.7, huh? (Either that or get off my duff and win for once...) Having no practical experience with 2.3, I'll say:

• A predictable progression is very desireable. Whether that progression needs to be linear is a separate question.

• One thing I'd like to see (maybe we should file this under your pipe dreams, RavenRed) is an in-game skill points calculator similar to the skill points script on killerbunnies, which I can't use. It would be great to have access to something like that during character generation.

• I agree with RavenRed about character races.

RedNaga: Mmmmm... my points -

• For linearity, i agree with RavenRed - i'd like to keep things simple to understand, without forcing every player to memorize at which point the skill start to become less effective and thus you get less power for skill point. Of course, i assume also that it would just be a matter of getting used to it...

• Starting difficulty - to date, the start is not *that* difficult, even with generalist classes, but this assumes that the player know some effective basic survival tricks - in other words, i think that this will have no effect on even half-expert players, but will greatly discourage new ones, as nethack, crawl and other roguelikes are much more easier to begin with (even if they catch up quickly). However, instead of changing the start difficulty, i would go the other way around - make the characters more powerful at start, also raising their maximal to keep the 'sense of growing', and make the endgame a bit more difficult to compensate.

• I join the chore about the races rework - some of them are *really* inadeguate (and i don't mean yeeks, of course )

NeilStevens: Game-breaking home run (as Kasuga Ayumu might say)! What if we gave the player a certain amount of skill points at the start of the game? Make it 10-20 points, compensating (or not) by making certain character levels only give 5 points instead of 6. That way character development begins immediately, with specialists being able to get a jump start (my sorcerors will LOVE this, but so will my warriors). Too much? Just right?

RedNaga: i don't know at all about balancing (i love chaos, not balance ), but the idea sound right fine. This will give the bootstrap needed to not make the game impossible for new players and still keep the same difficulty in the long run for more advanced player. Maybe that's the right way this time...

RavenRed: Points-based character creation makes a comeback. Looks like a good idea, although 10-20 points is high. You could always compensate by giving no class-based skill points at the start, only skill MODS. This would then roll the starting character class bonus into the creation process for the player to use or not, should they desire. i.e. a class starts out as a true tabula rasa with potential skills rather than actualised ones.

• NeilStevens: It's a good idea for a module, but if we do it in ToME we lose the ability to make easy and hard classes.

  RavenRed: What's wrong with that? It makes all the classes balanced at a stroke! <jk> But seriously speaking, how else would you suggest starting/low-level players get their general (i.e. across abilities) power levels up in the early game? Really hacky things like a town-quest with a guaranteed skill reward begin to play across my mind, but I suppose the real question is how we make the abilities gain a natural part of playing ToME, rather than being something artificially and consciously "tacked on" at a later stage of development. Maybe an across-the-board increase in the starting levels? Something more targetted?

  NeilStevens: I really think that 6 skill points a level should be sufficient to power up our characters, especially if we front-load them as I suggested just above. Making stat gain smoother and more natural would be nice, but that's a different can of worms that I'm not opening yet, heh.

  RavenRed: Thinking of a Steamband-style progression? It's always been a very minor bugbear of mine, too. Thinking of fixing it prior to 3.0.0, or is it more of an Arda goal?

  NeilStevens: I saw that on the newsgroup when the feature was first added, and was infatuated with the idea for a while. Now I lean toward some sort of quest structure with stat gaining rewards.

  PaulMoore: That sounds good - the big tedium of stat gain is the need to linger at about the same level waiting to find potions by basically random chance. In the same way that there are shops which provide "basic resources" like ID, maybe having a set of quests available from somewhere like the Alchemist shop ("Get me the ingredients - an Ancient Red Dragon heart - and I'll distil you a STR potion") would give the player more control over achieving stat gain...

ZizzoTheInfinite: Actually, for what it's replacing, 10-20 points is staggeringly low. My write-up that NeilStevens links to above uses a Wood-Elf Archer as a common example; in 2.3.1, that Wood-Elf Archer would need no fewer than 155 skill points to "buy back" her 2.2.x starting ability levels for even her non-combat-related abilities, and it would take until level 43 (if we leave out Searching and Perception, the two most expensive, it would only cost 104 points and would be achievable by level 20, but the point remains).

Most of the problem appears to come from the resetting of all the starting base ability levels to zero, and I still don't believe that such a drastic change is necessary to achieve the stated goals. Even if you want the initial base level of an ability to be independent of race/subrace/class, it doesn't have to be zero; set it to some non-zero constant that approximates the initial base level of an average character and tweak it from there with positive or negative initial skill levels based on race/subrace/class as desired. It's the huge jump from zero back to where you would have started that's too big to handle easily by tweaking the starting skill level.

NeilStevens: I would be looking in that direction if the illustrious maintainer hadn't already ruled that he's fine with the game being harder. I'd still like to adjust things, but if DG wants things harder now than they were in 2.2, that's his call.

• ZizzoTheInfinite: Fair enough. I've taken up the banner, then, and set up a module for further experimentation along these lines. (After all, you're always saying that this sort of experimentation should be done in modules anyway, yes? )

DanielJoo: Sorry for being offtopic (once again!) but does, what you say above, mean there is absolutely no chanche for another bugfix coming out before Tome3? I read a scary amount of reports of people not being able to win due to the mtdoom bug, with 2.3.0/2.3.1. And that is pretty much the worst kind of bug that there is in a *band, so i was kinda' hoping for a 2.3.2 thingy appearing sometimes soon. But what I've just read here makes me believe it wont happen (You hear some sobbing -more-)

• NeilStevens: Bugfix releases would be in the 2.3 series. We already did release 2.3.1. 2.4, a release with new features, is out of the question. 2.3.2, 2.3.3, and so on are possible.

PeterWiersig: 20? Sound a bit high considering that the Lvl1 Char can't raise a skill above 5.999. And I don't consider the start much harder than 2.2.7.

• RedNaga: Well, i've tried running a troll zombie warrior (one of the favored beginner combinations) and it took me 7 turns killing a large brown snake (with 2 attacks/round). As i said above, it is not much harder *if* you know about tactics, some basic monster knowledge and so on. I think one the best way to hook new players is to let *them* kill some monsters before *monsters* show them what they can do . It surely worked for me at the time...

RedNaga: Relating to the above considerations about the starting difficulty, just a minor point - some suggestion on how to use tactic at the start and how to allocate their starting skill points (if the thing goes on), put on the beginner section, would greatly help them.

TobiasParker: A quick note about the new difficulties of non-specialist classes, My highest level in tome 2 was 35 with a High Elf Swordmaster, In tome 3 i started a hobbit assassin and am now level 40 and reached deeper in every dungeon and have been to many that i have never even visited, so in short, I am not sure if stealth was always this powerful, but now it is downright nasty. (Character Sheet)

Okay, about skills and changes and stuff.

1. Midgame Direction. I always run into a hump at about Clevel 30, not the stat max hump or a "i cant kill monsters at my Dungeon Level" hump, it is a "plot" hump. To help my rambling self make my point let me tell a quick story. I was playing tome on our (Mine and my Girlfriends) laptop the other day and she asked me what the point was. Now this made me think, why was i repeatedly recalling into the Paths of the Dead with 99 identify scrolls and killing the same monsters for the same loot. As i improve my tome-playing abilities i began to understand the need to return repeatedly to the same dungeon for better equipment and also to overcome the "stat-max" hump by fountain scumming. When my Roguelike tactics were not as refined i would hardly ever return to a dungeon that i had completely dived through, i mean why would i? i cleared it and if there were princess quests i beat them, no need to return Eh? I have been playing tome for well over a year and i still do not feel comfortable with repeatedly recalling to the deepest dungeon level i can survive at, hoping to fill my resistances.

But as i tried to explain these essential Tome-tactics to her i could not find the words to describe the process of stat-maxing, resistance filling, and item searching. Could there be a way to let a new player know that these tasks are absolutely essential, because even somewhat-experienced players may have only gotten to this Clevel hump a few times. [reading dumps/wiki, alot more to come]

NeilStevens: (Note that you just need to leave a blank line to mark a new paragraph, and I fixed the link to the character sheet) I agree that we could use more quests. That's not something I changed in 2.3, though, and ideas are welcome for ToME 3. Anyone who comes up with new quest ideas, please just file them in the IdeaArchive.

RavenRed: Maybe revive Blacksmurf's ideas?

TobiasParker: Sorry for the formatting, The wiki and i do not get along.

1. ...Continued - After reading Blacksmurf's ideas i agree with much of it, there should be some direction towards which dungeon or quest should be tackled next. I cannot see a Hero carrousing about the wilderness to find dungeon after dungeon till he/she finds the next one to fit his CLevel. Maybe if there were "dungeon bosses" at the bottom of each dungeon that dropped a parchment that revealed the next point on The Necromancers path of influence. [i really do not have enough time...]

Actius If you want to boost specialist classes at the high end, you can make specialist skills with different tradeoffs than the generalist skills. To some degree you're already doing this, but these characteristics could be made more extreme.

KhymChanur: Instead of giving points to the player at the start of the game, how about giving them more points-per-level at low clevels, and less points as clevel increases, so going from clevel 1 to 2 would get 8 points, but going to clevel 25 to 26 would get only 5 points. Also, the same sort of thing could be done to how high above your clevel you can get at a skill: it could be 8 above at clvl2, so you could pump skill Foo up to 10 (if you have enough points), and decrease to the current 4 above by clvl 25.

NeilStevens: That's basically what I want to do, only wiht an extreme ramp. You get a lot at level 0, and the same amount for every level thereafter. I think the level 0 points are important for ToME not having a slow Sangband start, especially when we make tactics and exploration changes abilities.

KhymChanur: What changes will be made to tactics and exploration?

NeilStevens: The plan is to make them accessible only when you buy the relevant abilities, once the skill points are front-loaded.

KhymChanur: I can understand that for tactics, but why for explore? If you run, you move faster but are nosier and don't notice details as much, so your searching and stealth go down. Hmmm, though I suppose that increasing stealth and searching by going slower could be something you'd have to buy.

NeilStevens: Yeah... I know of somebody who makes excellent use of the large stealth bonus, because the speed cost isn't as big a deal later in the game, especially when combining with other stealth equipment might prevent you from awakening a single monster! And I know you said you agree with the tactics already, but the saving throw bonus for cowardly tactics is significant, too. My sorcerors use it very well.

We do give the player loads of skill points these days, and DG has decreed that making the game harder is OK, so I really want this for ToME 3.

TheFury: So something like giving each new player a potion or two of learning would help offset the lack of class bonuses? I may add that to my module, though players could sell them for ridiculous amounts of gold... Can I give a player something at start, and make it 100% off?

Also, I don't know whether or not anyone else has noticed this, but the void was almost unwinnable with the lack of class bonuses. I couldn't hit anything, so I ended up summoning dragons to beat everything down to the point I could damage them. Here is my dump: http://angband.oook.cz/ladder-show.php?id=4377 .

NeilStevens: Sure, you could give potions, and even set a discount on them, but it'd be easier just to give the points directly with a birth hook.

Post-winning has no impact in my decisions to alter the game. All I care about is whether you can win.

TheFury: Good, once I figure out how, the next version of FuryMod will start each character out with 10 skill points. That should be enough to get going.

KhymChanur: About the Void: the Void and the monsters in it can be tweaked so that meele will work against them without the class bonuses.

NeilStevens: Oh yes, if we wanted to we could reduce the levelling, or reduce the dungeon levels, either way.

HugoTome: I think it is right that a specialist should find the early game harder than a generalist.

I think the really difficult thing for a novice is appropriate use of skill points - I've tended to be pretty random about this, and suffered as a result. Right now I'm playing an unbeliever inspired by ZizzoTheInfinite's excellent DITL in which I copied Zizzo's carefully worked out skill appropriations exactly, and I'm doing far better. I don't know how to fix that, or whether it should be fixed.

I think the specialist would be helped more by RavenRed's suggestion of giving skill mods for the user to assign at will at the start, rather than skill points - maybe half a dozen .1 increments, though it might need some reductions in default starting mods to compensate.

GreyCat: No, archers and warriors are not unwinnable now. I've got an actual living breathing Demonologist Winner in 2.3.1, and I've had some "near misses" with other fighting- and shooting-type characters. Yes, it's harder now. Quite a bit harder, in fact, and it's harder both at the start and in the end. (I used to be able to start a generic fighter and take out a 2-headed hydra as my first kill in the shallow water. Now I'm lucky if I can beat a red naga -- usually I have to settle for a pink frog or a crebain. At the other end, you've taken away our ability to skip the quests in Angband by teleporting to level 99, and you've nuked my random artifact bolts with extra blows.) I've compensated by getting better at the game.

Oh, and of course there's always the simple fact that getting lucky with equipment finds will make the game immensely easier. This is just the nature of the entire Angband family. The character who finds boots of speed +8 on level 3 of the first dungeon is going to find the early game much easier; and the same applies to characters who find, say, rings of +8 dex/speed/attacks....

TheFalcon: Certainly, the game may be slightly harder than it used to be. However, it is still possible to get basically any character straight through the Thieves quest, and to around level 20 in about ten minutes. Until this is no longer possible, I doubt you are making the game too slow starting.

HarryErwin: I'm a generalist in real life http://osiris.sunderland.ac.uk/~cs0her and my experience is that I don't know as much as some specialists, but once I reached a high level in multiple skills, the *interactions* between my skills have had explosive power. Get that flavour, and I think you'll end up with generalists and specialists being very different but equally playable at high levels.