New Release

Hey everyone,

We've just pushed the changes that were in the developer preview into the main game. This may affect (increase) your score in puzzle 652 (http://fold.it/portal/node/993872). We have extended this puzzle by a day to make sure everyone can get the new version to ensure everyone is getting the correct scores. You can find a list of the new features and fixes below:

Residue IE
* Fixed a bug in symmetry puzzles where the dependent residues were being scored with the ResidueIE filter.

General
* Fixed bug where clashing importance was set incorrectly when done through LUA
* Clashing Importance slider now gives more control on low values. The values themselves still correspond to the same clashing importance, but you now have finer grained control closer to 0.

(Wed, 11/21/2012 - 00:12  |  9 comments)
Joined: 06/17/2010
CI slider

Is changed in 2 ways:
- it is kind of logarithmic now, not linear
- CI is shown up to 4 decimals after comma
It is intended?
For me log scale of CI should be hidden somehow. When I put slider in the middle I expect that algo is "watching" for clashes at half power and my score can drop by half. If I slide to 1/10 I expect that my score can go down to 10% of current.
In current scale half CI is about 0.3, and 0.5 is like 70%.

tamirh's picture
User offline. Last seen 6 years 18 weeks ago. Offline
Joined: 05/11/2012
Yes, the CI GUI has changed

Yes, sorry I just noticed that it wasn't introduced in the patch notes. It should have been in the notes for last release.

You have much more control over the lower end because from about 0.0-0.2ish is where it's the most useful (or should I say you can see the most difference). Because of that we increased the number of decimals shown so you can see more of what the actual value is.

The actual values will still do the same thing, but pulling the slider to 1/4 of the total length won't go to 0.25 anymore. In your scripts, setting the CI to 0.1 still does the same thing as before.

I'm not quite sure I understand what you say when you are expecting your score to drop by half when the clashing importance is at 0.5.

Joined: 06/17/2010
CI expectations

I use CI slider/settings to make protein more flexible (like everyone I hope). By setting 0.5 leading to 50% score drop I meant that I expect that CI settings then shake/wiggle will change my score down by close/same amount of points. So 50% CI = 50% starting score after shake/wiggle. I know that is not so liner, maybe this current slider will make behavior close to that.
I remember "feeling" of wiggle/shake about 2yrs back where setting lower than 0.1 was on the edge of implosion, and 0.5 was pretty drastic. At some point (probably on wiggle fix) it has changed - almost no change form 0.5 to 1.0 and 0.02 still not imploding in most cases.
Maybe players that not have so old/odd memories not feeling bad/mad about how it works now :P

marie_s's picture
User offline. Last seen 1 year 50 weeks ago. Offline
Joined: 05/18/2008
Groups: None
btw

can you list updates here
http://fold.it/portal/node/989977

Joined: 04/19/2009
Is this slider accuracy going to be useful in the future?

For now - not only does it seem that no one asked for it, but it is irritating the players and confusing the heck out of the newcomers.

In fact, in either my group room or global, I've only heard complaints - not one positive comment.

Unless there is an immediate need, I'd suggest that you go back to what it was.

Comment from my group chat room: "it's like the strange gift from your aunt myrna that leaves everyone scratching their heads... that thing with the feathers".

Susume's picture
User offline. Last seen 1 week 4 days ago. Offline
Joined: 10/02/2011
Are there really players

Are there really players using the lowest end of the ci scale during hand folding, who have a need for more stops below 0.1? I assume you can tell this from the data collected on player actions. Even at the low end (which for me is 0.07 and 0.2), I have certain levels I use and have never wished for more intermediate levels between them. It is no easier now to land on 0.2043xx (which is what I have to settle for now when I want 0.2) than it was to land on 0.2 before.

If there is cool stuff you can do by varying ci by tiny amounts below 0.1, I somehow never learned it. If it's just for control over clashing, I manage that with zero length bands rather than the CI slider. Anyone out there who can enlighten me what this is for?

jflat06's picture
User offline. Last seen 14 hours 29 min ago. Offline
Joined: 09/29/2010
Groups: Window Group
.

The reasoning is that most of the useful clashing importance values lie within the smaller values. Is this not the case?

Joined: 04/19/2009
.

Anthropic Dreams team members do a lot of work in low CI - I have never heard a wish for more decimal points at any value. I have posted in feedback with a link to this discussion - hopefully more players will weigh in on whether they will find it useful, or would like it rolled back.

jflat06's picture
User offline. Last seen 14 hours 29 min ago. Offline
Joined: 09/29/2010
Groups: Window Group
.

The point isn't necessarily to give more values. It's to make it so that a single pixel does not separate useful values.

Get Started: Download
  Windows    OSX    Linux  
Windows
(7/8/10)
OSX
(10.7 or later)
Linux
(64-bit)

Are you new to Foldit? Click here.

Are you a student? Click here.

Are you an educator? Click here.
Search
Only search fold.it
Recommend Foldit
User login
Soloists
Evolvers
Groups
Topics
Top New Users
Sitemap

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, RosettaCommons