I have been waiting for answers but I guess we have enough after so long.
Pool Results.
I had to remove the plots because they made the file too large.
Only IDs 8 and 19 had more than 5 mistakes in the LC60® test.
Since they were only 2, I did not remove them from the results.
LC60® had a lot of score value decrease for LDs, so techniques plus an LD would still be worth less than in past LCs.
Still, it could be further improved.
A quick solution that would completely work is to make the techniques non-accumulative.
Some might ask to add a limit higher than 1 for them, however it would require substantial coding on the system if it’s going to be used again (unless the limit is made manually).
In my original plan for LC60® value variation, they would be more aggressive than 0 to 7 points. It was a point of concern (pun intended) since it would be hard to people to calculate their scores, so they would have trust issues with my score keeping.[spoiler]In the original plan, the variable scores would curve down exponentially when someone had too many successes, to effectively force people to vary their techniques/stop hoarding points.
The effect would work like a stock crash.
[/spoiler]
Many people asked for less subtasks. Adding less subtasks is actually easier to do.
In my original plan, there wouldn’t be as many subtasks, but they would still be plenty.
Since many of them were simple, they wouldn’t require much incubation, but more recall.
The idea behind the amount would be to encourage dreamers to improve their recall by checking the open subtasks.
Things that seem to work:
- More than one host.
- Art points.
- Daily updates.
- Technique points.
- Task, subtasks and subsubtasks.
- Weekly deadlines.
- Badges.
- Strict day of week for new tasks.
- Subtask Voting.
- Challenge Length Voting.
- Feedback Surveys.
- Dream of Participant.
Decreasing the value of LDs was positive to balance. But balancing is sometimes hard to notice when you are a participant.
Removing the task performed in LD bonus seems positive. It is already easier to perform tasks when you are lucid. This means that bonus was terrible for balancing.
Length voting had a high rating. So it is nice to give the people a chance to extend what they like or shorten what they don’t like.
People seem to like dreaming of participant subtasks.
I did not expect people to like the deadlines, but it had high rating.
The plan behind deadlines is twofold:1. To calculate the variable score;
- to avoid a huge dump from someone who is away;
Funfact: The badges were not initially planned. They were added after I wrote the signups. I did not expect it to be valuable, but it probably motivates the participants.
It is actually easy to add and if the next challenges use it, they could have fixed badge system to avoid partiality.
It is important to remember that we did not have all the participants voting.
This means that only people who like surveys may have completed it, affecting this result.
Things that do not have much effect and are up to the host:
- IRC bot.
- Subtask owning and sharing.
- More points for late tasks.
- Unlockables.
- Named Subtask categories (colors).
- Similar Scoreboard System.
- Tie-break.
We do not have strong preference/hate for the subtask owning. So it is up to the host.
The participants might not notice that it is important while they are just participating.
The fixed base score increase is effective for balancing.
This is when the new task has more valuable subtasks than previous tasks.
I argue that it is valid since the new task is valid for less time (while the past tasks are still open).
It was already implemented in LC47 and is not a new thing (further thanks to Mew151 for collaborating with our score system).
Qualitative answers praise the tie-breaking because it is exciting.
We do not have strong quantitative preference/hate for the tie-breaking, so it is up to the host.
Giveaways did not work well, still, things were much better than I expected. I expected most participants to just ignore it.
From the host perspective, unlockables and voting are good since the participants have power to add the subtasks they prefer.
The participants might feel motivated because of their contribution and at the same time the host has a feedback on how they like the subtasks before the first success.
Without unlocking, the colors do not add much.
You should not add score for completing a color/set/rainbow because it is not balancing. There could just be more badges.
[spoiler]Without the unlocking, the colors are almost useless. They were added to balance the demand for subtasks.
Spoiler: It is proven to work with Monopoly. The cards were not a coincidence.
[/spoiler]
For some reason, surveys and poll have been high rated but subtask voting did not have the same result.
Some people did not like the scoreboard. Some prefer spreadsheets.
This means the system should be improved to show data like spreadsheets.
Perhaps some day we could have collaboration on the LC system, which was one of my ideas.
Things that did not work:
- Newsroom.
- RSS feed.
- Value variation for subtasks.
- Named Subtask types.
Apparently hardly anyone cared for the news page.
I don’t know if anyone tried using the RSS.
People are apparently afraid of the value variation and it wasn’t welcomed.
The named subtask types (ex: simple task, lucidity power, DC-related) were added to balance the tasks. Tasks from 1 to 4 had the same number of each subtask type.
This is actually good for the task design but participants do not care much about it.
I hope it helps. I could add more trivia about the LC60 that could actually be helpful.
Plus it helps that you can only do the subsubtaks or however they are called once.
GitHub