New Class of Puzzles: No Scripts allowed.

Case number:671071-988901
Topic:Game: Other
Opened by:TheGUmmer
Opened on:Tuesday, November 16, 2010 - 16:45
Last modified:Monday, October 1, 2018 - 14:57

Would it be possible to have a class of puzzles where scripts are not allowed. Don't get me wrong I like scripts and owe the majority of my ranking to them, but it seems that Foldit has become more about the scripts and less about the human ability to see what a computer can't. Just a thought.

(Tue, 11/16/2010 - 16:45  |  24 comments)

Joined: 08/06/2010
Groups: Contenders

Scriptless puzzles would certainly change the complextion of the game.
Scriptless puzzles would take lots of time to work, especially for most of us.

beta_helix's picture
User offline. Last seen 4 hours 13 min ago. Offline
Joined: 05/09/2008
Groups: None
Status: Open » Open

This is an interesting idea.
We have been discussing similar ideas (disabling scripts for a certain period in a puzzle for example) and we'll keep you posted if we decide to try that out!

Joined: 06/17/2010

Run a contest for a try :)
I`m sure, I will end in last 10% :P

Mark-'s picture
User offline. Last seen 1 year 19 weeks ago. Offline
Joined: 08/05/2009
Groups: Contenders

Although I would love to see some of these puzzles, I do feel that it may disadvantage players that have little free time to spend endless hours manually rebuilding and tweaking a puzzle. On the other hand, if everyone were to just run endless scripts on puzzles, where does the 'human element' come into the folding equation? Why are we here..? Is the answer really 42..? Sorry, I digresss.

I would happily participate in some 'script-free' contests, but that would disadvantge the people who play them? After all, they would have to spend those endless hours on point, and ranking, free contests.

Maybe the answer would be a different class of puzzles, running alongside the normal puzzles, that would have a different ranking and points table, separate from the others. They would also also provide some useful comparisons between the two groups of solutions.


mat747's picture
User offline. Last seen 1 year 32 weeks ago. Offline
Joined: 10/15/2008
Groups: Void Crushers

I agree with rav3n_pl and Mark try it as a Contest first.
The idea of having to play without scripts AGAIN very unappealing, with the endless endless hours of manual play.

Joined: 09/18/2009
Groups: SETI.Germany

How about releasing a puzzle with no scripts allowed first, and when it is over, releasing it again with puzzles allowed, playable for the same time.

This would give a possibility to compare how far recipes pull a solution further.

Joined: 09/18/2009
Groups: SETI.Germany

Of course, I wouldn't try this on big puzzles, where manual manipulation takes a lot of time (puzzles with many mutable sidechains, for example).

Joined: 08/06/2010
Groups: Contenders

Many of the scripts don't work well on small proteins. Still, that makes small proteins a good class to work on for comparison. It could help develop better scripts as well. Getting baseline comparisons would have to involve the same puzzles with and without scripts. It seems to me this is being done all the time right now. The best players tend to the top on all puzzles. I suspect they use a mix of personal actions and scripts. The scripts they use are an extention of their personal actions and the human input. The humans are to the scripts as drivers are to cars. Sure different people have different cars and this gives them an advantage in a race but so does driving skill.

Joined: 06/17/2010

Maybe it should be time or move limit on that type of puzzle? ie max 1`000 moves per day?

Joined: 09/16/2010
Groups: None

When I first started, my first protein, I tried to do manual only, got a fairly good score too top 10 in a <15 puzzle but it would not have make the top 30 in the all hands of the same protein.

To off set this, please make manual only puzzles with a very long deadline, so there is not pressure to spend so many hours each day (allow us to spread our effort over weeks or months)

Joined: 05/09/2008
Groups: Void Crushers

Shameless Bump.

infjamc's picture
User offline. Last seen 30 min 18 sec ago. Offline
Joined: 02/20/2009
Groups: Contenders

Re: Crashguard303

"How about releasing a puzzle with no scripts allowed first, and when it is over, releasing it again with puzzles allowed, playable for the same time. This would give a possibility to compare how far recipes pull a solution further."

This option has one major technical limitation, though: memory from the first attempt. A possible alternative would be doing a "randomized control trial" (namely, using a random number generators to conduct the coin flip that would determine whether scripts would be allowed for a puzzle) and then compare the results between the scripts-allowed puzzles and the scripts-disallowed puzzles. The tricky part is ensuring that the difficulty between the two groups are roughly equal.

Joined: 08/24/2010

Manual method folders usually develop great intuition and imagination. You'll find their top scoring solutions varying greatly. They should wear this like a badge of honor.

Joined: 11/10/2007
Groups: Window Group

I agree this would be interesting, especially if we could compare puzzles that were run with and withou scripts in some way.

Joined: 05/09/2008
Groups: Void Crushers

bump - since it came up in the chat today

Joined: 11/21/2008
Groups: None

I agree with Seagat. In the early days, before scripts and with only basic in-game tools, I feel there was much more diversity in the scores between different puzzles, as players had to rely on intuition and experimentation, and hone their skills according to their strengths. This must have helped to expand the game into its present form. In addition, it would probably help new players master the basic tools more quickly.

Joined: 09/21/2011
Groups: Void Crushers

First of all, I like the idea. But maybe just a standard Blue Fuze script allowed, no others.
I think that would make it more playable without killing the idea of increasing hand folding capabilities.

mat747's picture
User offline. Last seen 1 year 32 weeks ago. Offline
Joined: 10/15/2008
Groups: Void Crushers

I post a comment on "Wed, 11/17/2010 - 22:49" I haven`t change my views and have less time to play now then before.

But I think we could have a Non Global point scoring category with no scripts allowed.

steveB's picture
User offline. Last seen 1 year 27 weeks ago. Offline
Joined: 01/08/2009
Groups: Void Crushers

A question worth asking is how were all the scripts first devised ?

The answer is that people used their experiences from manual folding to design the scripts, converting the manual techniques into automated ones. If foldit does not run manual only puzzles, then the knowledge base of manual techniques will soon dissapear from the community, and consequently new script ideas will not materialise, stagnating the foldit project. This effect can already be seen in the puzzles which require manual techniques such as flu design and beginner QTTN puzzles, where the older generation of players tend to dominate the top 10 results in these puzzles.

A second consideration is that from the players point of view, foldit is a game and not a scientific research project. There used to be a lot of fun in the game knowing that when you passed someone on the scoreboard that they were watching, whereas there is now no pleasure going past a script running player on the scoreboard as they are most probably at the shops. There used to be some excellent close finishes to the puzzles, with some hillarious comments being made in general chat as the clock counted down to the finish. If the human interaction and fun is removed from the game the player base will shrink.

Another consideration is deversity of results, as the scripts are not able to make intuative major rebuilds which involve taking whole sections of the proteins into new positions - the very reason that foldit was started.

Here's an idea......why don't you do manual only puzzles in a knockout cup format? Put all the team names into a hat, draw out pairs of teams to play each other on manual only puzzles, and have the 'foldit cup' awarded every 6 months ? That would help keep the manual techniques going, and add some seriously good fun to the game. Semi final: Contenders vs Anthropic would be worth buying a ticket for !

Joined: 04/19/2009

@steveB -

I completely, 100% agree with your comment - thank you for putting into words exactly my own thoughts and concerns (and doing it better than I have been able to).

And your idea of a 'foldit cup' team competition for manual only puzzles is brilliant... I'd certainly buy a ticket for that!

mimi's picture
User offline. Last seen 2 years 43 weeks ago. Offline
Joined: 11/17/2008
Groups: Contenders

Only problem is that some of the "features" of the latest versions of the program make the old hand folding techniques difficult. For example Mark did a recent puzzle completely by hand and it still didn't respond to being walked - so its not just that all the walk room is taken out by scripts.

Joined: 12/06/2008
Groups: Contenders

Please allow me the liberty of using an analogy.

Using scripts while folding is like doing the dishes in a dishwasher, rather than washing them by hand. The dishwasher does a marvelous job of cleanup, more quickly and thoroughly than if you'd washed them by hand. But... if you don't take the time to scrape off the excess food and to line the dishes in their proper places, no number of cycles through the dishwasher will get them more clean than if you did them by yourself.

In this dishwasher analogy, the best results come from the optimum usage of time and effort: doing some manual work preparing the dishes, then letting the machine blast away. Translating this into game terms, the best-folded proteins will be the ones where the players manually work them to remove the major problems, and then allow scripts to make the proteins pretty.

Giving us more de-novo puzzles will satisfy the needs of both camps.

Joined: 06/17/2010

Scissors tool and we will be happily fold ever after :)

bertro's picture
User offline. Last seen 16 weeks 4 days ago. Offline
Joined: 05/02/2011
Groups: Beta Folders
Type: Bug » Suggestion

not a bug, returned to open / suggestion


Developed by: UW Center for Game Science, UW Institute for Protein Design, Northeastern University, Vanderbilt University Meiler Lab, UC Davis
Supported by: DARPA, NSF, NIH, HHMI, Amazon, Microsoft, Adobe, Boehringer Ingelheim, RosettaCommons