The Defining Features of ToME4

[Grey]

This is mostly an expression of my own thoughts on what's good about ToME4 and what makes it stand out, and a reflection on how dramatically the game has changed since earlier betas. However it may also provide a focus for future development to play on the game's strengths, and serve as an example to other games which wish to have the same features of ToME4. It also includes points of reflection on things that should change.


The Defining Features of ToME4

Procedural content

A staple for roguelikes, ToME4 randomly generates dungeon levels and places enemies and items randomly. This helps improve the replayability of the game.

Exceptions: World map, certain static levels (though this is quite normal in the genre).
Problems: Disconnected levels. Many samey dungeons. More variety of generators may help. Another issue is a linear and predictable feeling to some areas, since there isn't that much randomisation to enemy types and bosses. Enemy egos could help.

Permafailure

Dying makes you lose a life, or ultimately makes you fail the game. There is no reloading from a save game. This combines well with the procedural content to encourage multiple playthroughs. The rigour of punishment for death enforces a much stronger learning mechanic. The variety of starting options lets you try something new each time you play.

Exceptions: Infinite lives is available for donators, though arguably this doesn't impact on the normal game modes.
Problems: Some players may find that the amount of time and thought put into building a character feels too wasted when they lose everything. But this is a minimal issue, since it's a roguelike, and if they want to play something different they can pay for the unlimited lives version...

No regeneration of content

Levels are constant once generated, and do not respawn items or enemies over time. Shops only restock under certain circumstances. You cannot easily "scum" areas to indefinitely improve your character.

Exceptions: World map encounters and farportals, though the latter is limited by gold. The former is not easy to scum, especially as they become very powerful later on. However perhaps to make them even less attractive sources of xp/loot they should give no xp and drop sub-standard equipment.
Problems: None, I'd say. There is sufficient experience and items in the game to be harvested without the need for scumming.

Minimal resource management

There are no food items, no non-unique potions, no other little things to worry about. Between each fight you are pretty much the same, leaving decisions of strategy to be focussed purely on build and eq choices, and decisions in battle based on clever use of replenishable resources that go on cooldown. Timing plays as big a part of battle as worrying about resources. Most class resources are easily to replenish, whilst some abilities are purely cooldown restricted.

Exceptions: Quite a few actually, some of which should be considered closely:
* Stamina - it is not easily restored, even between battles
* Hate - again not easily recovered, but this is a class feature
* Negative/positive energy - awkward to recover between battles since it degrades
* Ammo - it's been suggested that ammo be removed before and extra egos put on bows/slings. This would be good I think, since ammo as it stands feels very out of place.
* Alchemist gems - not played alchemists enough to comment on these
* Wands - as I've said before, these should go.
* Some long cooldowns (maybe they should be toned down?)
* Gold - it has never been balanced well in the game, with players having either too much or too little. It still feels out of place as of b34, and yet it feels like it can't be removed either.
Problems: If you use up your resources mid-battle then you may be left with zero abilities open. This can make protracted fights very difficult for some classes, especially if there is little room to manoeuvre/escape. This is most problematic in world map encounters and vaults.

Individual classes

The classes in ToME4 are not what you'd normally expect in other games. Some are incredibly unique, with their own odd mechanics and weird resource bars. Most classes have a cohesive feel to their skills, with many complimenting abilities. There is no simple fighter class, with even melee classes relying on careful use of sustains and ways to debuff enemies. Each class type tends to have its own unique resource which is thematic with the class and used in interesting ways. There is a small element of multi-classing through the attainment of generic trees, but this is rare.

Exceptions: None. All classes offer something quite unique amongst roguelikes.
Problems: A bit weird for outsiders to get used to perhaps, but also very exciting for those that appreciate this sort of depth.

Talent trees

Abilities are gained through talent trees, which themselves provide the vast majority of complexity to the game. Other roguelikes tend to have complexity through item interactions, but ToME4 builds its variety with talent trees. The trees themselves are shallow with fairly light restrictions, meaning that you're not railroaded down a specific build (other games that use talent trees tend to require a long path to be followed before reaching some of the higher abilities). Though equipment can support talents they ultimately are not necessary. Most classes have enough variety of trees to allow for numerous build types.

Exceptions: None.
Problems: Early game you will have few abilities and more restricted tactics. Later on you can have too many, or feel the need to pick up too many, making the game more complex and potentially harder to play. This is ultimately an essential part of the game, but new players could benefit from having more guides on how to build certain classes to help them advance in the game.

Advancement restricted to discrete points

All advancement comes in the form of stat points and talent points, and there is a maximum number you can have of these in the game, mostly through level progression. Thus you can't do everything - hard decisions must be made on where and when to invest your points. There is also a flat cap on how much you can invest in any one talent or ability, preventing extreme builds.

Exceptions: Equipment provides further boosts and the ability to use certain extra abilities. Certain talents also can be used to boost stats beyond more regular limits, sometimes imbalancing the game.
Problems: Some enemies can get to levels in stats or talents far beyond what the player is capable of, potentially breaking certain game elements. Some classes struggle to invest points well due to a large number of trees or many relevant stats, but this is part of the game.

Races distinguished by racial abilities

The defining feature of each race is its talent tree. The other stat bonuses are almost irrelevant, though the hp per level difference can have a significant effect. These talent trees compete with other generic trees for attention, and can compliment other talent trees in interesting ways.

Exceptions: Cornacs, who get a category point instead of a racial tree.
Problems: Little difference in races early on. Some racial abilities are of little use, and should potentially be looked at. Some racial trees are too focused on niche stats, limiting their usefulness across classes. Initial talents on racial trees don't scale with talent level, discouraging investment.

Player as monster

Whatever a player can do, a monster can do, and vice-versa. Well, for the most part. Stats are the same between them too. The generic nature of the engine makes for easier design of player and monster abilities, whilst also allowing for many interesting combinations. It also gives the player a deeper understanding of how monsters operate without having to learn about abilities unique to certain enemies. This feature is most seen through fights against adventurer style characters, the shade boss, and in controlling other NPCs.

Exceptions: There are certain monster-only skills, and some player skills that the monster AI can't operate.
Problems: Balance can be an issue, as a cool player ability can be game-breaking in the hands of the enemy, especially when they get them at high talent levels.

No insta-kills

There is no medusa gaze or cockatrice petrification or death ray that will kill you instantly. Likewise there's few abilities that kill an enemy instantly. Death comes purely from loss of hit points. There are also no status effects that leave you completely defenseless leading to an unescapable death. This gives a feeling of fairness to the game that is lacking from some other roguelikes. It also lets you venture into unknown areas with some confidence that defensive play can at least let you survive.

Exceptions: Disruption shield and a few other talents can cause crazy amounts of damage and potentially need looking at. Some statuses can also be very restrictive.
Problems: Later in the game high damage attacks become regular, so although you may not be one-shotted there is a significant risk of being two or three-shotted.

Transparency of content

Enemy stats are laid bare, the exact amount of damage exchanged is clearly shown, and items are automatically identified. You can also see exactly what statuses are impairing you and how long until they wear off. This wealth of information lets you make informed tactical decisions.

Exceptions: Combat mechanics are still very obscure, and status effects on enemies cannot be seen clearly. The actual abilities of enemies can also be a mystery.
Problems: Some might complain of information overload and too many stats in the game. I'd also argue that the identification system is a bit obsolete now, especially since you can't even speak to Elisa in her house (and Anorithils get this magic orb too?) Occam's Razor suggests the identification abilities should just be removed.

Aesthetically pleasing

The introduction of Shockbolt tiles has revolutionised the aesthetic goals of the game. There has been a lot of excellent emphasis on improving the general look of the GUI, including fancy borders, talent icons, nice backgrounds to menus, and so on. It is, simply put, the nicest looking roguelike ever made. The game also supports full mouse control, including the binding of talents to mouse buttons which get contextually used. These all help attract new players to the game and the genre. It also gives the great impression of this being a professional quality game, making it more likely for people to donate.

Exceptions: There is still some glitchy tile behaviour here and there, and we're still lacking status effect icons. I think we could do with icons to replace the resource bar labels too (a picture of a heart instead of the words "Life" for instance). We need to look very critically at all aspects of the game to consider where even minor improvements can be made to bring it further up to the level of AAA games. One major improvement needed is new sounds and more music.
Problems: Some tiles are actually impractical, getting in the way of gameplay, which goes against the whole roguelike ethos of gameplay over graphics. Some work needs to be done to ensure graphical elements don't obscure important gameplay features. Also some tiles, especially items, are hard to see against the backgrounds. There are also the odd sdl problems that crop up from time to time for some users.

Retro support

Even with the fancy mouse gestures and the stunning graphics there is keyboard support and ASCII graphics. Keyboard controls are in fact much faster for many playstyles.

Exceptions: No terminal support, which some old-schoolers consider essential in roguelikes. There are also some mouse commands that have no equivalent keyboard command.
Problems: Maintaining two interface types can lead to trouble, such as some things being made the same colour as the background, but thankfully we have many ASCII players willing to point these problems out.

Original setting

Sine the departure from Tolkien the game has been given its own universe, with its own spin on many races, some unique races and a relatively deep background history. Much of the detail behind this is in lore pieces, found randomly or in special places, and these are generally optional items for those that want to just dungeon crawl.

Exceptions: Some little outside references here and there, though generally in theme with the setting. The game doesn't go overboard with jokey elements and pop culture references.
Problems: The setting is perhaps not entirely engaging, with little interaction with actual people in the game apart from for sidequest purposes. There's room for a lot more lore to be delivered through NPCs in town, and through more lore in libraries that players can browse at their leisure. I'm very open to suggestions from others in that regard ;) There are also some issues with consistency of setting - chronomancers and celestials especially jut out a little, but ultimately add more to the game than they detract.


And... that's about it. Comments and discussion welcome of course.

[bricks]

Very solid analysis. I feel the NPC vs. player character skill balance is the most problematic part of the game. It really makes certain encounters, including a great deal of the mid/end-game, very risky. One thing that occurred to me; drowning/suffocating is something of an instant-kill mechanic, or at least it's disproportionally punishing. Instead of outright dying, I think the player should take a certain percent of their health as damage every turn spent at 0 air.

[Grey]

I was thinking just that earlier. I've never had any problems with drowning/suffocation myself, but it does seem strange that you can so easily lose 1k hit points from staying in water too long.

[bricks]

Yeah, I died from drowning once just due to lack of attention, and nearly died another time because I coudn't find any air bubbles near the entrance to the Lake Nur dungeon. (Teleported randomly and found a bubble just within reach.) Oh, there was a second time, back when the air bubbles lasted for a single turn. I think that was the last time I tried to play a Yeek.

[edge2054]

Darkgod and I have discussed where exactly chronomancers fit into things (though I don't want to spoil any of his plans for them). Maybe some lore someplace could make their place in the setting feel more cohesive or more time travel based quests could explore when chronomancers come from (because the when is more important then the where in this regard )

I'd also like to do more temporal monsters, not just horrors but other stuff too. Things that are out of phase with time or a temporal 'wraith' that's the time echo of a powerful chronomancer (maybe even a unique).

Temporal Drakes could be fun too.

[bricks]

Another frustration with the current setup, discussing resource management: Equilibrium management is very meh. You can take Antimagic and have no problem with Equilibrium, or you can try to irk by with the Jelly/Swallow in combat. Toning down these two and adding more options would make Equilibrium a more dynamic resource, as would adding better effects to Equilibrium, for example: change skill failure (which is boring and frustrating, especially when compared to Paradox management) to a more interesting effect on gameplay, such as scaling skill effectiveness inversely with Equilibrium, or granting special effects (Nature's Boons) for low Equilibrium. Meditation is also a very obnoxious skill.

Sorry to keep bringing in new thoughts; I figured an edit of my last post wouldn't be noticed.

[edge2054]

Equilibrium could be reworked to play more like Tome 2's Mimic failures. In other words higher equilibrium could have a chance to lower your attributes or turn you into an abomination for a time.

Or the idea could go of a more Cursed type of way (referring to their talent trees). So you may have a whole generic tree dedicated to how you interact with high equilibrium, possibly even changing higher equilibrium scores into boons.

Just a couple of ideas.

Twisted Form - Replaces the current failure rates. When twisted form triggers you get a random effect based on animal traits (some examples; rooted to the ground as roots grow from your feet, a primal howl escapes your lips alerting all creatures to your presence, twisted limbs reduce your strength and movement speed for a time, barkskin lowers fire resistance drastically but increases armor hardiness, etc.)

Abomination - Turns you into an abomination, reducing all stats by 20%, speed by 20%, and all damage and resists by 20%.

[Mithril]

Interesting analysis!

I would add a few charateristics

No global timer
-No global food timer, as in Crawl

Combat focus -That is, most of the time is spent fighting ("quality time"), rather than boring grinding or trying to identify items or wandering back to town to restock things such as food.
-Auto-id of times
-Universal recall item to avoid boring backtracking
-Quick hp regeneration
-Relatively little grinding
-No food
-No temporary lightning such as torches
-Few puzzles

Here I will also make some advertisement for what I consider to be the most important improvement that can be made currently: Auto-explore which would reduce boring non-combat time enourmously. See:
http://forums.te4.org/viewtopic.php?f=46&t=25608

[darkgod]

Nice analysis, you should post it in many places

Makes me want to play this game .. oh wait ..

[budswell]

...Occam's Razor suggests the identification abilities should just be removed.

I thought about this for a while too, but ended up deciding I quite like the way it is. Forcing you to stand on an item before you know it's abilities is good, means magic users cannot just light up a level (vision) and mouse over to find the good items without risk. And still having to actively pick up and ideintify artifacts is good too, as that still takes a turn so is not a cusual thing in desperate situations. And still gives a nice sense of discovery once you identify a good one.

[reverend]

...Occam's Razor suggests the identification abilities should just be removed.

I thought about this for a while too, but ended up deciding I quite like the way it is. Forcing you to stand on an item before you know it's abilities is good, means magic users cannot just light up a level (vision) and mouse over to find the good items without risk. And still having to actively pick up and ideintify artifacts is good too, as that still takes a turn so is not a cusual thing in desperate situations. And still gives a nice sense of discovery once you identify a good one.

Seconded. I really like the way it currently is. It's a nice compromise between realism (not knowing beforehand if an item is any good without actually being there to examine it) and playability (no micromanagement of identify scrolls or spells).

[Grey]

Interesting analysis!

I would add a few charateristics

No global timer
-No global food timer, as in Crawl

Combat focus -That is, most of the time is spent fighting ("quality time"), rather than boring grinding or trying to identify items or wandering back to town to restock things such as food.
-Auto-id of times
-Universal recall item to avoid boring backtracking
-Quick hp regeneration
-Relatively little grinding
-No food
-No temporary lightning such as torches
-Few puzzles

There tie in with minimal resource management. But yes, you have a point in that the game is entirely focussed on the combat, and everything else can be ignored. Side-quests are just excuses to explore a new dungeon.

One thing that doesn't fit in with that is the alchemist side-quests. These probably need some attention...

Here I will also make some advertisement for what I consider to be the most important improvement that can be made currently: Auto-explore which would reduce boring non-combat time enourmously. See:
http://forums.te4.org/viewtopic.php?f=46&t=25608

A Crawl fan? Some people list auto-explore as a weakness of Crawl, rather than a strength. It doesn't address the real issue of dungeons being boring and low on content, making movement between fights dull. There are other ways of dealing with this, such as:
- more vaults, especially mini vaults that don't have high level content or gear and are just a way of adding in extra enemies to fight in specific designs (Trollmire could do with a few more of these to break up the terrain)
- enemies that seek you out or move about randomly, rather than sitting and waiting for you to find them
- tighter level generators, with less corridors and empty spaces

[Hirumakai]

The Defining Features of ToME4

Procedural content

A staple for roguelikes, ToME4 randomly generates dungeon levels and places enemies and items randomly. This helps improve the replayability of the game.

Exceptions: World map, certain static levels (though this is quite normal in the genre).
Problems: Disconnected levels. Many samey dungeons. More variety of generators may help. Another issue is a linear and predictable feeling to some areas, since there isn't that much randomisation to enemy types and bosses. Enemy egos could help.

No regeneration of content

Levels are constant once generated, and do not respawn items or enemies over time. Shops only restock under certain circumstances. You cannot easily "scum" areas to indefinitely improve your character.

Exceptions: World map encounters and farportals, though the latter is limited by gold. The former is not easy to scum, especially as they become very powerful later on. However perhaps to make them even less attractive sources of xp/loot they should give no xp and drop sub-standard equipment.
Problems: None, I'd say. There is sufficient experience and items in the game to be harvested without the need for scumming.

Just to play devil's advocate, since I don't think saying Problems: None for No regeneration of Content is quite right. The above two combined allows (and in some rare, but noticeable number of situations, require) stair scumming which is generally considered a negative. It can also mean its possible to have map layouts and enemies in some of the required dungeons which cannot be beaten by some classes.

As an example, I've seen the bottom level in Telos' Tower be a single single screen sized, empty circle. No place to teleport and no cover. Filled with high level undead, mages, archers, plus the Telos boss. Watching a ~800 hit point Aegis enchanced shielding rune drop in 2 rounds indicates that the level is not doable without scumming. But you need to kill Telos in order to finish the game. So you stair scum, since unlike in many other rogue-like you can't just go elsewhere.

The fact that you have a limited set of equipment, generated completely at random, means characters will also sometimes not receive equipment that people take for granted in dealing with certain situations. Its very possible for certain types of runes and infusions to not drop, or of sufficient quality to make a difference. Take shielding runes, regeneration infusions, and healing infusions. A large portion of many character's success can be directly attributed to the number and quality of those that drop for them. Basically the difference between having an unlimited stack of light healing potions and an unlimited stack of *HEALING* potions.

Of course, this is a rogue-like, so maybe its fine that some characters get lucky and have an easier time than others, but just be aware that there are some trade offs between "random but limited and permanent generation" and "random and unlimited and non-permanent generation".

[darkgod]

I'm perfectly fine with some chars not doing some bosses (Telos is certainly not needed to finish the game).
Your point about inscriptions is more valid but I have never found it to be a problem, in the end there is so much stuff dropping that you are bound to find something, it might be a super healing infusion instead of a shield, or a regen instead of a heal, .. but you just adapt

[Mithril]

Here I will also make some advertisement for what I consider to be the most important improvement that can be made currently: Auto-explore which would reduce boring non-combat time enourmously. See:
http://forums.te4.org/viewtopic.php?f=46&t=25608

A Crawl fan? Some people list auto-explore as a weakness of Crawl, rather than a strength. It doesn't address the real issue of dungeons being boring and low on content, making movement between fights dull. There are other ways of dealing with this, such as:
- more vaults, especially mini vaults that don't have high level content or gear and are just a way of adding in extra enemies to fight in specific designs (Trollmire could do with a few more of these to break up the terrain)
- enemies that seek you out or move about randomly, rather than sitting and waiting for you to find them
- tighter level generators, with less corridors and empty spaces

Auto-explore would fix boring non-combat traveling without having to change dungeons. Likely much easier to implement. Also, since there are persistent dungeons it is preferable to explore every nook and cranny of a level which if often rather tedious and boring currently. Especially in the end when you squint at the mini-map to see if you have missed anything on the current level. Not quality time! With auto-explore it is very easy and quick.

There are also some dungeons that is just a study in tediousness like the Maze. Auto-explore would be invaluable here.

I cannot see any down-side to auto-explore. If some people want to be masochistic and trudge back and forth manually and tediously, then they can just avoid using the auto-explore button.

Auto-explore would dramatically reduce tedious, boring, repetitive non-combat time and increase the proportion of quality combat time! It would seem to exactly adhere to the defining features of Tome.