Getting more players involved with Lab

Omei’s latest player puzzle/tutorial sparked an idea to get more players involved with lab.  http://eterna.cmu.edu/web/puzzle/6004700/   .  In Omei’s puzzle the suggestion was made to copy their solution and paste it to the lab currently active with the provided link in the comments.  I love the idea so much that i wonder if the player’s solution of the puzzle could be automatically submitted to the lab.  Of course if the constraints were met. If auto-submitted, then the player will get a prompt that they have successfully participated in the active lab. Their “submission” could have a default title with their name in it and they can have a hotlink to the lab to see their solution among the other designs.  Later they could be prompted how their design scored.
On the player puzzle page, the puzzle could look like this: 

As we all know, many new players think when they solve a puzzle their solution is being used in laboratory.  This is not true, but with this idea it could be true.

3 Likes

Additional thoughts:
  1.  granted this would only result in multiple solutions to one design.  However, I think it would only give a few submissions given the number of solvers in that time period.  
  2.  How do we get more than just multiple solutions of one design of the same lab?  There are several unique designs in the lab already, perhaps choose a few (not sure how but would like your thoughts) and post them as puzzles with only the locked bases.  Essentially resulting in mods of that design. 
  3.  Maybe as a player, when we submit a design into lab we have the option to include it as an unsolved puzzle for other players to mod?

2 Likes

Yes, Hyphema!  These are all good ideas.  Now that the experimental pipeline has gotten so large, we need new ways to smooth the transition from in silico puzzles to in vitro experiments.  Keep the ideas coming!  

Hi Hyphema!

On your idea nr. 2, there typically are like 2 or 3 main puzzle solving types that hit through among the high scorers, where most of the puzzle structure stays the same and there just are a lot of variation in the sequence. Picking the high scorer of each major type would be good.

Using things like cluster count and folding error to be more sure to pick a good candidate.

Also set cluster count limit at 20 and folding error at max 1.4, like Omei introduced here:

https://docs.google.com/document/d/1LfCX9pbguzR5OtgYQDakxMmzViRg7WvJKCuj1_kc0s0/edit