Question about new Sprint modules

[Cyclone]

According to the coding wiki, building new Dungeon Sprint modules is relatively simple. This interests me because I've always felt that Crawl's mechanics would work just as well for prebuilt levels as randomly-assembled ones.

However, while I can appreciate the extremely fast-paced gameplay that Sprint offers, I don't really want to design stuff that way. I prefer to build slower and less frantic levels, like Crawl's vaults.

I imagine you could balance this by slashing the XP rewards of everything to nerf Sprint's rapid leveling.

My question is twofold: am I right to think that these types are levels are feasible, and would there be any interest in them (from players as well as developers)?

[KennySheep]

I can only really speak from a player's perspective, but I think it's possible to have both. The Pits (from how far I've penetrated into it, at least) seems to be exactly what you're talking about, the map has a much slower buildup in difficulty than the others and feels much more like completing a series of vaults than one large challenge run.

I don't think the problem is with how much experience sprint gives, I think the problem is with making the player fight massively out-of-depth monsters for the level their at with every fight.

[mumra]

We've discussed some ideas to add more flexibility for designing sprints;

• Changing the experience multiplier
• Enabling multi-floor sprint maps
• Restricting the selection of races/backgrounds
• Customising the starting conditions in other ways

This kind of customisation will probably involve some Lua code (but you'll likely need some of that to do anything with a sprint map anyway).

No real work has begun on any of these possibilities yet, but you can see st's full wishlist here: https://crawl.develz.org/wiki/doku.php?id=dcss:brainstorm:variants:sprint#sprint_wishlist

[Cyclone]

We've discussed some ideas to add more flexibility for designing sprints;

• Changing the experience multiplier
• Enabling multi-floor sprint maps
• Restricting the selection of races/backgrounds
• Customising the starting conditions in other ways

This kind of customisation will probably involve some Lua code (but you'll likely need some of that to do anything with a sprint map anyway).

No real work has begun on any of these possibilities yet, but you can see st's full wishlist here: https://crawl.develz.org/wiki/doku.php?id=dcss:brainstorm:variants:sprint#sprint_wishlist

I hope that stuff gets added some day. It sounds like it could make Sprint just as interesting as the main game.

I'll play around with the files when I get home and see what I can put together. Would the devs include a module in the official releases if it was good enough?

[mumra]

I hope that stuff gets added some day. It sounds like it could make Sprint just as interesting as the main game.

Sure, well I'm planning to implement at least a couple of things on that list myself if nobody else does first, to improve a sprint idea I'm working on myself.

I'll play around with the files when I get home and see what I can put together. Would the devs include a module in the official releases if it was good enough?

Absolutely - that's how most of the existing sprints came to be in the game. As and when you have something you want to submit someone will certainly take a look at it - Mantis is the usual place to post these kind of contributions. Even if wasn't quite up to standard we'd let you know what things needed changing or improving to make it in. So it can be good to post your ideas and works in progress as you go, then improvements can be recommended early on rather than you having to go back and redesign lots of stuff. Everybody's first few vaults and maps will probably fall into a few pitfalls, so it can be a good idea to design a few smaller simpler vaults to go in the game before taking on something as big as a Sprint. Anyway thanks in advance for any contributions, they are always welcome!

[evilmike]

Multilevel sprints would be cool, and I could see letting sprints define their own XP and piety multipliers. However, I would still advise you to keep it fairly short. I think a multi-level sprint should be a series of small-medium levels, rather than one big one. It's ok if some players can take their time, but I think it's important for this mode to be playable in a single sitting too.

So, for a slower paced sprint where you don't gain XP as fast, I think it would be a good idea to design it for a certain level range (either low levels, or give out free xp and equipment at the start). That way you can make an interesting sprint map without it having to be 5x longer than normal.

[Cyclone]

Absolutely - that's how most of the existing sprints came to be in the game. As and when you have something you want to submit someone will certainly take a look at it - Mantis is the usual place to post these kind of contributions. Even if wasn't quite up to standard we'd let you know what things needed changing or improving to make it in. So it can be good to post your ideas and works in progress as you go, then improvements can be recommended early on rather than you having to go back and redesign lots of stuff. Everybody's first few vaults and maps will probably fall into a few pitfalls, so it can be a good idea to design a few smaller simpler vaults to go in the game before taking on something as big as a Sprint. Anyway thanks in advance for any contributions, they are always welcome!

I can draw some vault plans right now and build them when I get back to my computer in the States.

What kind of pitfalls do you mean?

Multilevel sprints would be cool, and I could see letting sprints define their own XP and piety multipliers. However, I would still advise you to keep it fairly short. I think a multi-level sprint should be a series of small-medium levels, rather than one big one. It's ok if some players can take their time, but I think it's important for this mode to be playable in a single sitting too.

So, for a slower paced sprint where you don't gain XP as fast, I think it would be a good idea to design it for a certain level range (either low levels, or give out free xp and equipment at the start). That way you can make an interesting sprint map without it having to be 5x longer than normal.

I think it depends on what you want to make. Sprint was designed for short levels, but there's nothing wrong with having some longer ones.

[Galefury]

What kind of pitfalls do you mean?

Some common ones:
1. Wrong level of challenge (and possibly reward).
2. Too much desire to add completely new things. Especially vault-defined monsters and explicitly specified items.
3. Reliance on gimmicks, such as lua triggers.
These all come from new vaultmakers wanting their vaults to be very special, and different from what's already in crawl. Not necessarily a bad thing, but combined with inexperience the resulting vaults are frequently terrible. Some of Grunt's vaults are perfect examples of this.

If you want to avoid these problems, make a few vaults where you only use random monster glyphs, not too many out of depth monsters, and no lua. Vaults (the branch) rooms might be a nice place to start. Once you know how to make these fun, you can more safely move on to fancier stuff. If you want some strong opinions on this (and good advice on vault making) you should talk to tenofswords or st in ##crawl-dev, or you can read tenofswords in progress vault making guidelines. I can dig those out if you want.

I think it depends on what you want to make. Sprint was designed for short levels, but there's nothing wrong with having some longer ones.

The sprint game mode is called sprint. As in fast and short. It's okay to have some variety, but it would be weird to have some completely atypical sprint maps. When a player chooses to play a sprint map, the player should get what it says on the label, not something completely different. However, if what you design is fun, I'm sure space will somehow be made for it.

[mumra]

2. Too much desire to add completely new things. Especially vault-defined monsters and explicitly specified items.
3. Reliance on gimmicks, such as lua triggers.

Note: these two particular points matter far less when designing with a Sprint, you have a lot more freedom to flaunt the conventions. Although if your Sprint relied only on these things without having something interesting as a core gameplay concept then it would be bad. The other problem that Lua tends to cause is buggy vaults. There are so many unpredictable variable in a game like Crawl, it's very easy to code things in a way that will break due to some unexpected combination of factors, and with Lua you actually have access to a lot of low-level stuff that can create issues if you're not careful. The solution there is to keep testing things thoroughly (with a variety of combinations of characters, gods, transmuted forms...) and also post your work-in-progress at intervals for others to review and test. The different problems that can happen are far too many to list effectively, and that is only the ones we've already discovered

[KennySheep]

What kind of pitfalls do you mean?

If you want some strong opinions on this (and good advice on vault making) you should talk to tenofswords or st in ##crawl-dev, or you can read tenofswords in progress vault making guidelines. I can dig those out if you want.

I would want this.

[Galefury]

I found this version and this version on the wiki. They are both very rough, but have some good content. It's just not very well readable.

Note: hangedman = tenofswords = claws.