Twilight Forums

I didn't want to derail the Time/Day discussion. But now that we're a month into Retcon activity, I thought we might discuss competition brackets. This could either be implemented by Ryme via leaderboards, or via something like KoLDB.

I think it's pretty clear that not all mettle is created equally. Thus, my 15mettle run with 5 pulls/day is not equal to someone else's 15mettle run w/o pulls but with a hand full of skills. So i think a leaderboard based on mettle doesn't so much say that the runs are of equal difficulty so much as they encourage people to find the optimal restrictions to hit that board and be fastest. This is quite different from KoL, and adds another level of strategy to run-planning. I think it's a wonderful enhancement.

The trouble with the current structure is that there isn't a way to compete between people with similar restrictions. The notable exception being Maso (which has no leaderboard).

The main difficulty here is the influence of pull runs vs non-pull runs. Items make such a big difference it biases all competition. That is, at all competitive levels so far, pulls top the list. Maybe once people have 60 permed skills the balance will tip. That would be interesting.

So, one idea is to post leaderboards for the named restriction sets "Grimdark" etc. That's a lot of leaderboards -- and might serve to dilute competition too much for a small playerbase. Another is to divide between pull and no-pull runs. And/or class-perms vs. non-class-perms.

Perhaps this will add to the ideas Ryme already has for his enhanced Retcon boards.

_________________

Navpane Rebuff: http://userscripts.org/scripts/show/87404

Heh. Not to steal your thunder too much, but I have said literally everything that you just said, at one point or another in the last few months.

First, there *will* be a KoLDB equivalent. It's phase II of the leaderboards. It may not come out immediately, but it will before we've got too much history. (More than a month, likely less than 6 months.)

During the "how should retcon work" discussion, the roll your own retcon idea was very popular. It wasn't even my idea, but something that a lot of players clamored for, and I ran with it as the best idea out there. I pointed out then that the only sensible method was to do brackets, which would 100% guarantee other people would be outraged by the placing of those brackets, and argue vociferously that run X worth Y mettle isn't really the same as run Z worth Y mettle, and it's not fair (or not balanced) that they're in the same bracket. I said then I'd do an advanced, searchable system at some point, but for the base brackets people were just going to have to accept that inherent inequalities exist, and if they care to be more specific they'll have to tune their searches instead of me tuning the leaderboards. In my memory this statement was met with apathy and/or silence for the most part, but I could be wrong. I've got a small chip on my shoulder (not directed against you, Nardo, just in general) that despite having already acknowledged all these limitations ahead of time, since then a swarm of players have "discovered" this same issue and told me the various ways in which the system they wanted is broken. (Sorry if I sound grumpy. I feel like I've been fielding a lot of this in various ways over the last few weeks. I expected some, but there's been a lot more, some of it about things that were obvious enough I pointed them out more than a quarter of a year ago, but nobody considered them a problem until now; and of course now it's urgent to fix, despite the relative youth of retcon.)

I don't think it's a good idea to have much greater bracket sprawl (excluding perhaps some class-based divisions) -- definitely think it would be confusing overkill to try one for each named type, and it'd only cause outrage for anyone who has a favorite run that's not one of the named ones. I'd rather leave it to the KoLDB-style outcome, where if you want to search runs with less than 5 skills using restrictions P, Q, R, and S, you can do so.

In a general sense I see the leaderboards as a mini game within the greater game of TH. As such they have their own rules and ways to optimize them, as predefined mini competitions. They will reward certain behaviors, but aren't built to reward *every* style in which people will do runs. For instance, if you want to compete in the highest bracket, doing Maso simply is NOT going to be competitive. This doesn't mean Maso needs its own board. It means if you're doing Maso, it's because you want to do Maso. If you want to be on the leaderboard, you pick something in that bracket that lets you go faster.

By providing four brackets I've tried to present some flexibility for players who prefer different play styles, but going beyond the system to cater to *every* play style just isn't going to make a lot of sense, except via the KoLDB-style "roll your own" leaderboard, to match the "roll your own" retcon system.

I'm beginning to think I'm the only one who doesn't care about speed runs.

_________________

I've won the 100k DD bet so many times, I should have the title "Mr. Luck"

Well, I wasn't around for the retcon pre-release discussion, so don't blame me for having asked for anything.

But, if the mettle value of a restriction is supposed to reflect its impact on speed (which admittedly I'm not sure is the correct assumption), then I would suggest the modification of some mettle values. "All items" should be something like -10 or -15 mettle instead of just -4. If the diet restriction gives +2 mettle, then decaf should be worth +6. (...unless good caffeine gets harder to get, but right now the black market makes it so easy to get decent quality drinks without even using pulls.) These are just some initial thoughts.

Except that makes playable runs useless... since negative-mettle runs wouldn't be able to perm any skills.

... with a minimum of 1.

You'd get the benefit of perming, a single mettle, one foil, and a line in your history. That's still very much to get from a zero brains run. And I'm saying this as someone which only does these no brain runs.

An alternative way to look at it:
Having full items means you don't need a permanent sidekick, as you can always use the best non-permanent ones. So why do I get the bonus for them?
It means I don't need skills, since I have items which do better than the skills. So why do I get a bonus for not using skills?

I guess what I'm saying is, don't do "-10 or -15". Instead, having a full item run does not give you the bonus for sidekick and skills. This means the most you can get out of it is 5 mettle, for diet, which frankly makes sense, it is a real limitation even with full items.

What temporary sidekick gives -time? Which temporary sidekick heals? Which temporary sidekick does +items and blocking? Or damage and XP? Or +PP and -FT and +chips?

_________________

The churches are empty / The priest has gone home / And we are left standing / Together alone
--October Project: "Dark Time"

Count me in on the vote for making All Items run a greater negative. -10 would be good, since you'd still have positive mettle with just a few restrictions in the other categories.

I'd say the reason this issue has been met with apathy until now is that it's not really an urgent one. From what I can see in chat, people do look at the leaderboards, but they're not really focusing on them, everyone's mostly just having fun exploring the possibilities of retcon for now.

Not to say that I personally wouldn't love KoLDB-like boards. I can very well picture a situation that would in a way be similar to KoL's, with the bracket boards to show off and the more precise boards for the actual competition. The precise boards would be the reference for the speed retconners community, while getting to the top of the bracket boards would require playing specific run types.

That said, I too think that 1) different bracket boards should represent different ways of playing and 2) the current point system is too kind with pulls (compared to skills, mostly).

-4 is already a big penalty. With all item runs, it's already impossible to unlock everything in Nocturne. I think it's a big enough penalty. Yes, they're easier runs and should get less rewards, but I don't think we should be punishing people who want to use their items.

_________________

The churches are empty / The priest has gone home / And we are left standing / Together alone
--October Project: "Dark Time"

Im with Cristiona. Why should I or anyone who has donation items be punished for using them? Why pay $10 for an item, and then be penalized by getting no rewards? I can't see the logic.

Ryme said that more advanced brackets would come, so i'll take his word and wait for that.
Untill then, just enjoy the game.

_________________

Buying Stainless steelbreaker, n-barrelled shotgun and Seppia's cephalopod skullcap. Pm me here or ingame!

What, it's changed now? What happened to 'skills can only be permed on runs of >1 mettle'?

And there's no way to avoid needing the sidekick... there's too little -time at early levels to give up a source. Giving up sugar, yeah... not a problem. Skills are mostly pointless until you have a half dozen or more permed. (And even then, without a stockpile of preprepared skill points, they're still mostly useless.) Caffeine... well, if you really wanna spend twice as many days, you can go decaf. (Now, if IOTM-only were a -2 mettle option, and all-item were -7 mettle, I could see that as fairer. It still allows playable runs (all-item, class skills, all-caf, diet, full sidekick) to perm skills. Maybe -9. and going no-skill. Anything beyond -9 is just insane.)

I missed this thread before I posted over in the retcon time thread, but I did think of a thing or two to add. To be honest Ryme, I think that bracketed leaderboards don't make a lot of sense as things stand with retcon, and that there are too many issues with them for them to be meaningful. A 9-mettle run and a 10-mettle run are a much closer comparison than a 10-mettle run and a 14-mettle run, but the arbitrary break point compares the runs that are further apart. You're right that there is no good place to put the brackets that won't result in poor comparisons like this, and that KoLDB-style boards are the only real comparison method. With that said, I think it would have been better to avoid implementing leaderboards at all until you could get the KoLDB-style ones up and running. I doubt that people who are interested in speed competition will be satisfied until they can get more accurate comparisons. Whether or not that should be a high priority for you is totally your call.

I see what you're saying, but if I lived by "do it right or don't do it at all" this game simply wouldn't exist. I almost always have to do everything in stages, and the intermediate stage tends to be pretty important in terms of helping me figure out the next stage.

I'd much rather start with "usable enough" so that there's *something* for players to enjoy, while getting enough feedback to find out what I need to do to make it "pretty good." I can guarantee that even with a community this size there's no such thing as doing something in a way that will make everyone happy.

I know at least some people get some satisfaction out of the current boards, and that's justification enough to keep them in instead of just deleting them entirely.