Release tournament new category idea: Sudden Death

Back Room Dart Board
1

Update:
I never saw the original CSDC site, until now. Just found it via wayback machine:
https://web.archive.org/web/20240716081758/https://csclub.uwaterloo.ca/~ebering/crawl/csdc/0.24/index.html

And there’s this very interesting sentence:

The Crawl Sudden Death Challenges is a competition that aims to fill a Crawl niche not currently filled by the biannual version release tournaments.

Tbh, I forgot the CSDC had been hosted by the dev team themselves. And it sounds like they already considered filling this niche in the release tourney? I’d really like to know more about the decision to keep the formats separate. Was it for technical or other reasons?

Original Post:
The 0.34 tournament is over, and I had this idea that a “sudden death” or “one and done” tourney category might be fun. This thread is not about fixing issues with the tourney rules in general. That’s a related topic, but please discuss that in another thread. This thread is about adding one new category, or maybe modifying one of the existing categories.

All the existing categories encourage you to play more. The overall ranking also encourages you to play fast and efficient, and in a way that earns points in multiple categories at the same time. But there is no category that truly reflects crawl’s core philosophy of overcoming RNG, anti-grinding, and permadeath. Not really.

You can play for win rate. If you splat, that’s a setback. Keep playing, and you can make your win rate go up again, significantly. If you break your streak (early enough), you can start a new streak. You can almost always improve your ranking by playing more, by trying again. Over and over. The only limit is the end date of the tourney (and not being among the fastest nine for a nemchoice, I guess).

Someone over on Reddit suggested a Nemelex’ Choice One and Done category. Meaning, only your first game with a nemchoice combo would count for the category. I really like the idea. But what I would like even more would be a category that encouraged you to not just win any run you started, but win it with the best score you can.

Yet, both Best High Score and Combo High Scores encourage start scumming. Not a big problem, even in a tourney category. I just feel it’s odd, and not in line with crawl’s philosophy, that there are two high score categories that allow start scumming and no high score category that does not allow start scumming. I do not mean to offend players who enjoy chasing crazy good high scores, which require very high skill in addition to start scumming. These players are nice and humble people, as far as I can tell. I’ve never seen them pretend it’s all skill, and if everyone is allowed to start scum, it’s perfectly fair. It essentially eliminates part of early game RNG, imho. No more, no less. Finishing the lucky run, with a record turn count is still extremely challenging and impressive.

But maybe a high score category without start scumming would be a fun and interesting addition to the tourney. Some players don’t enjoy start scumming.

I feel Combo High Scores One and Done could be fun. (I’m sure a cooler, shorter name for such a category can be found, perhaps just “Sudden Death”), Say you’d be limited to one game per species and background. (Listing every playable combo would be too much, imho.) Wouldn’t it be more comparable to a sports tournament? If you run a race and stumble, well, better luck next year. Best out of three, like in weightlifting, could also be an option. But one and done appeals to me the most. There are 27 species and 26 backgrounds. You could take 26 shots at the one and done scoreboard, and no need to play your one least favorite species. Would that not be enough and a unique challenge?

There are countless possibilities how points could be calculated in the category. I feel a relative approach would be best. Second or third best with a species or background should be worth something.

Many top players already get their combo high scores first try, but wouldn’t it be more exciting to balance risk and reward without the safety net of starting over again, repeatedly, with the same combo? To be encouraged to finish, despite unlucky RNG or some early mistake?

As mentioned, I don’t insist on modifying an existing category. It could and probably should be an additional, new category.

I’m aware one and done (combo high score, nemchoice, or any other implementation of the principle) isn’t newbie-friendly and probably frustrating for players who like to play fast and o-tab. But, Fastest Real Time Win isn’t newbie-friendly either, and frustrating for slow players, like me. In a one and done category, playing fast (but controlled) would still be beneficial, but not overwhelmingly so. I’m not arguing fast and slow players deserve equal chances to win any category, let alone the tourney. But must every single tournament category for a turn-based game favor the faster players so strongly?

I think a one and done category would appeal especially to players who have jobs and families and cannot or do not want to take off two weeks to play DCSS. I also suspect some people don’t participate in the tourney because they know they would spend too much time playing, despite their best efforts not to. A one and done category would be self-regulating in that regard.

Sorry for the length of this post, I’ve tried to keep it short, but also wanted to bring the idea across fully. It’s still a rough idea, not a detailed proposal. Any thoughts, other suggestions?

Edit: some minor but annoying typos

Edit 2: Sergey pointed out on Reddit that this idea integrates poorly with existing tournament rules. That’s very true. If there was a quick and easy way to flag your games for sudden death, however, and have only these flagged games count in the category? What about that?

Edit 3: I placed a little update above my original post.

5 Thanks
2

So yeah. Turns out thinking about one category isn’t enough. If it’s meant to be part of the release tournament, it has to integrate well with the release tournament’s other categories. A one and done combo high score category would mean you’d have to play it at the very start of the tourney, before all other categories, or there won’t be many combos left.

Maybe the nemchoice variant, then?

Or some way to explicitly register sudden death runs, before you play them. Via rc file, perhaps. But that would be clunky, and not in line with the low entry barrier of the tourney. I really like that you only have to log in and play.

Maybe if everyone had an automatically generated sudden death account that would be merged with your main account, after the tourney? Meh. I bet that’s too complicated to develop and would mess up statistics. And also not really user-friendly.

3

I currently am not participating because I have much to learn. However, I like the idea and was thinking: what if a player could select a seed from their own morgue to play this sudden death tourney? Each player gets 1 chance to try their dungeon. Go back to that nostalgic game or that silly death and get a 2nd chance and have it count.

My apologies in advance if this is the antithesis of DCSS’ spirit.

4

That’s a really creative idea. I just don’t see how you would even begin to rank the results. And the release tournament is all about the release of the newest version.

But I wonder, could there be a way to detect sudden death games via seed? The seeds would have to be random, and different for every game, but also have some property to tell the tourney scripts that it’s a sudden death run. Players would need a comfortable way to start games with these special seeds, on demand. I suspect that would be a whole lot of development work, if it’s possible at all.

Or maybe some other, much simpler way to let users flag certain games could be developed? Maybe via a quick and easy command ingame? Like entering wiz mode, only instead of wiz mode, you enter sudden death mode? Obviously, you should only be able to do that before you spend your first turn. Devs?

5

I placed a little update above my original post.