SpaceChem

SpaceChem

View Stats:
SpaceChem > General Discussions > Topic Details
ToughThought Jun 21, 2014 @ 9:00pm
2014 SpaceChem Tournament
Welcome to the 2014 SpaceChem Tournament thread! The tournament is open to everyone.

EDIT December 12, 2014: The time for posting and solving puzzles has ended. Everyone is still welcome to participate in the discussion.

The puzzles (updated January 9, 2015):
Prime Time -- active Sat., July 5, 2014, 13:55 GMT; deadline Sat., July 19, 13:55 GMT. (Prime Time results.)
Blockheads -- active Sun., July 6, 2014, 14:10 GMT; deadline Sun., July 20, 14:10 GMT. (Blockheads results.)
Budget Reactors is having a laugh -- active Mon., July 7, 2014, 18:11 GMT; deadline Mon., July 21, 18:11 GMT. (Budget Reactors is having a laugh results.)
Martian Respiration -- active Tue., July 8, 2014, 18:21 GMT; deadline Tue., July 22, 18:21 GMT. (Martian Respiration results.)
Prussian Blue -- active Sat., July 19, 2014, 17:12 GMT; deadline Sat., Aug. 9, 17:12 GMT. (Prussian Blue results.)
Multiplier -- active Sun., July 20, 2014, 20:05 GMT; deadline Sun., Aug. 10, 20:05 GMT. (Multiplier results here and starting here and here.)
Greed -- active Thurs., July 24, 2014, 04:05 GMT; deadline Thurs., Aug. 7, 04:05 GMT. (Greed results.)
MeitnHassium -- active Sat., July 26, 2014, 13:57 GMT; deadline Sat., Aug. 9, 13:57 GMT. (MeitnHassium results.)
Ferrous Wheel -- active Wed., August 13, 2014, 06:22 GMT; deadline Wed., Aug. 27, 06:22 GMT (or 2:22 a.m. EDT). (Ferrous Wheel results.)
74th Rule of Acquisition -- active Thurs., August 14, 2014, 06:22 GMT; deadline Thurs., Sept. 4, 06:22 GMT (or 2:22 a.m. EDT). (74th Rule of Acquisition results.)
Comet Chemistry -- active Sun., August 17, 2014, 14:43 GMT; deadline Sun., Aug. 31, 14:43 GMT (or 10:43 a.m. EDT). (Comet Chemistry results here and here.)
Mitosis -- active Mon., August 18, 2014, 17:39 GMT; deadline Mon., Sept. 8, 17:39 GMT (or 1:39 p.m. EDT). (Mitosis results here and starting here.)
Water not WAtEr -- active Fri., September 5, 2014, 01:45 GMT; deadline Fri., Sept. 26, 01:45 GMT (or Thurs., September 25, 9:45 p.m. EDT). (Water not WAtEr results here and here.)
Jadeite -- active Sat., September 27, 2014, 18:00 GMT; deadline Sat., Oct. 18, 18:00 GMT (or 2:00 p.m. EDT). (Jadeite results here and starting here.)
Carbon Products -- active Fri., October 17, 2014, 01:22 GMT; deadline Fri., Nov. 7, 01:22 GMT (or Thurs., Nov. 6, 8:22 p.m. EST). (Carbon Products results.)
Barbituric Acid -- active Fri., October 24, 2014, 18:48 GMT; deadline Fri., Nov. 7, 18:48 GMT (or 1:48 p.m. EST). (Barbituric Acid results.)
Beads -- active Sat., October 25, 2014, 23:51 GMT; deadline Sat., Nov. 8, 23:51 GMT (or 6:51 p.m. EST). (Beads results.)
Back to Counting -- active Thurs., November 13, 2014, 15:58 GMT; deadline Thurs., Dec. 4, 15:58 GMT (or 10:58 a.m. EST). (Back to Counting Results here and starting here.)
Stereo Chemistry -- active Fri., November 14, 2014, 15:58 GMT; deadline Fri., Dec. 5, 15:58 GMT (or 10:58 a.m. EST). (Stereo Chemistry results.)
Final creator ranking is here.
Final standings are here.

GuavaMoment in 2012 and WildM in 2013 hosted wonderful tournaments. This year, we have a decentralized tournament. There is no host, so we all share responsibility to make the tournament go well.

I encourage you to take a look at the previous tournaments. You can find the 2012 tournament here[lparchive.org]. You can find the 2013 tournament here[lparchive.org] or here. If you are new to SpaceChem or could use some pointers (and who couldn't?), you will find valuable tutorials there. In fact, every part of both tournaments presents you with a great learning experience, as well as great fun.

As in past years, the tournament puzzles will be "custom assignments" that you can import into the ResearchNet area of SpaceChem. To import a puzzle:
1. Copy the puzzle code to the clipboard.
2. Enter ResearchNet from SpaceChem's start screen by clicking "Sign On."
3. If you haven't yet progressed far enough into the game to have been offered access to ResearchNet, you will see a button enabling you to "Hack Into the System." Click it. Nothing bad will happen.
4. Click "Create or Import an Assignment" at the bottom of the screen.
5. Click "Import."

This year, puzzle creation will be a part of the tournament itself. I am confident that we will see some outstanding puzzles. Contestants who elect to create puzzles will be the ones collecting, verifying, and tabulating solutions. The showing-off of a puzzle's results (either in a video or in writing) can be done by the creator or can be delegated to someone else willing to help. I am sure that any offers to help others would be appreciated.

To create a puzzle, go to the "Create or Import an Assignment" area and click "New Research." (You may instead choose "New Production" or "New Sandbox," but you must be prepared for the additional difficulty of verifying and showing off solutions to a production or sandbox puzzle.) With your new unnamed puzzle selected, click "Edit." You will be able to give it a name and specify inputs, etc. You can exit the "Edit" screen and come back to it as many times as you like. You can try out a new variant without losing the original, by selecting your puzzle, clicking "Duplicate," and selecting the new duplicate, before clicking "Edit." When you are ready to share your puzzle, with your puzzle selected in the "Create or Import an Assignment" area, click "Export." The puzzle code (which does not include any solution you may have made) will be saved to the clipboard and can then be pasted into a post in this thread.

GuavaMoment has remarked that a good tournament puzzle is simple enough for the average player to be competitive but complex enough for a few to create something really interesting. The scoring formula for created puzzles is based on this concept.

In order to give people time to discover this thread, there will be no active puzzles until July 5, 2014. If you wish, you can use the intervening time to work on creating (and solving) your own puzzle.

My next post will state the rules for this year's tournament. Although there is no host this year, I suppose that it falls upon me, as the rules' author, to resolve issues that may arise about them. (If an issue affects my own participation too directly, I will ask someone else to resolve it.) Issues that may arise about instructions for a particular puzzle would need to be resolved by the puzzle's creator.

Good luck to everyone!

EDIT July 1, 2014: corrected misstatement in step 2, "Enter ResearchNet."
Last edited by ToughThought; Jan 9, 2015 @ 2:34pm
< >
Showing 1-15 of 957 comments
ToughThought Jun 21, 2014 @ 9:04pm 
2014 SPACECHEM TOURNAMENT RULES

ELIGIBILITY

1. The tournament is open to everyone.

CREATION OF PUZZLES

1. Each contestant is encouraged to create a puzzle, but it is not required. Each creator may create at most one puzzle. A creator cannot submit a solution to his or her own puzzle. Instead, the creator can earn points for creating the puzzle, based on maximizing disparity among the top solutions while minimizing disparity between the average and the top, as detailed below. The puzzle should not be too easy, too difficult, previously known, or too similar to a previously known puzzle. The creator should not post a puzzle without first solving it (to better understand its difficulty). The creator should also be prepared to post a video or written discussion of the results, as discussed below.

2. Because random inputs are different on different platforms (e.g., Mac vs. PC), the use of random input puzzles is forbidden.

3. Although puzzles that reward exploitation of bugs in SpaceChem (e.g., "particle smashing," "telekinesis," etc.--see Carlbunk's guide) are not forbidden, please be warned that some players find such puzzles distasteful.

4. The use of out-of-game aids (such as cearn's editor[www.coranac.com]) to create puzzles is permitted. (In consenting to this link, cearn said: "Apologies to those who don't find it editory enough (with proper forms and all).")

5. Each creator posts here in this thread: the puzzle's name; a brief description; SpaceChem code for the puzzle; priorities (e.g., "fewest cycles, then fewest symbols"); any special instructions; and an e-mail address where solvers can submit solutions. (It need not be an e-mail address used for other purposes.) Time of submission is not permitted as a priority, even to break ties.

6. A puzzle should not be posted before 09:00 GMT on Saturday, July 5, 2014, should not be posted until at least 24 hours after the preceding puzzle became active, and should not be posted while 4 puzzles are already active. Please wait. A puzzle is considered to become active when it is posted (or, if this rule is disregarded, at the earliest time after that that it could properly have been posted).

7. The submission deadline for solutions is exactly 2 weeks after the puzzle became active for research puzzles, or exactly 3 weeks after the puzzle became active for production or sandbox puzzles. A puzzle ceases to be active when the deadline occurs.

Example: If research puzzles are posted July 6 at 8:00 a.m., July 7 at 8:30 a.m., July 8 at 9:00 a.m., July 9 at 8:00 a.m., July 15 at 8:00 a.m., and July 15 at 11:00 a.m., the first three of them become active when posted, the fourth one becomes active at 9:00 a.m. on July 9 [24 hours after the previous one became active], the fifth one becomes active at 8:00 a.m. on July 20 [when the first one ceases to be active], and the sixth one becomes active at 8:30 a.m. on July 21 [when the second one ceases to be active].

SUBMISSION OF SOLUTIONS

1. Solvers submit solutions to the creator. Submissions are recommended to be in the form of a screenshot as an e-mail attachment. (For production or sandbox puzzles, screenshots of each reactor are necessary, as well as a screenshot of the entire layout showing pipes and reactors while the solution is running.) The subject line of the e-mail should take the form: "[solver's screen name]: [puzzle name] [cycles]-[reactors]-[symbols]."

2. The following are forbidden: solutions that require action by the user while the solution runs; solutions that depend on out-of-game conditions; solutions that are not reproducible; solutions that depend on a hacked or non-standard version of SpaceChem; solutions whose implementation requires the use of out-of-game aids; tricks that cause SpaceChem to misrepresent the cycle, reactor, symbol, and/or output count; exploitation of bugs that make puzzles trivial (e.g., producing outputs with no symbols). Otherwise, the exploitation of bugs that cause unexpected SpaceChem behavior (such as those described in Carlbunk's guide) are permitted.

3. There is no limit to the number of solutions that may be submitted by a solver. However, since time of submission cannot be a solution priority, there is probably no reason to submit more than one solution (except possibly to make an interesting observation, to make a correction, or to make an unanticipated improvement close to the deadline). Please do not flood the creator with solutions; instead, please wait as long as necessary and submit just the best solution.

4. The creator should verify that the submitted solutions actually do solve the puzzle and comply with any special instructions.

PUBLICATION OF A PUZZLE'S RESULTS

1. Soon after the puzzle's deadline has passed, the creator should post a results list, showing each solver and the statistics (cycles, symbols, etc.) for each solver's solution.

2. The creator should also link either to a video showing off the top solutions or to a page containing a written discussion and SpaceChem videos of the top solutions. Other interesting solutions may also be shown, at the creator's discretion. Creators are encouraged to mention their own solving statistics, to give some solvers the enjoyment of having beaten the creator at the creator's own puzzle. (There is no problem if the creator delegates the making of the video or write-up to someone else willing to do it, even a contestant in the tournament, so long as no results are shared until after the puzzle's deadline.)

3. Solvers earn points on each puzzle solved as follows, based on the puzzle's priorities: 1st place--10 points; 2nd place--9 points; 3rd place--8 points; 4th place--7 points; 5th place--6 points; 6th place--5 points; 7th place--4 points; 8th place--3 points; other valid solution--2 points.

4. Ties among a puzzle's solvers are not broken. If two or more solutions have the same statistics, the solvers each receive the average number of points associated with the places they tied for. Example: If 3 solvers tie for 8th, 9th and 10th places, they each receive 2.33 points [(3+2+2) / 3].

END OF TOURNAMENT

1. The tournament ends when a week elapses without any active puzzles.

2. Creators of eligible puzzles earn points, based on maximizing disparity among the best solutions while minimizing disparity between the average and the top. To be considered an "eligible puzzle," at least 8 solvers must have solved it, and the top score among the solvers (taking into account all statistics) must have been obtained by fewer than 8 solvers. The creators' scores are determined by ranking, in comparison with all other puzzles, the difference in each puzzle's top-priority statistic (e.g., cycles, in the case of a fewest-cycles puzzle) between the 90th percentile result and the 1st place result, divided by the difference between the 50th percentile result and the 1st place result. If the score is undefined because the 50th percentile and 1st place results are the same, use the next-priority statistic, instead. (Example, using numbers from 2013 puzzle Crossover Filter: In a fewest-cycles puzzle with 28 solutions, if the top 4 results are 220, 228, 228, and 229 cycles, and the 14th and 15th place results are 266 and 287 cycles, the 90th percentile result would be 228.7, the 50th percentile result would be 276.5, and the creator's score would be [228.7 - 220] / [276.5 - 220] = 0.154.) The larger, the better. The scale of points is the same as above, i.e., 1st place--10 points, down to 8th place--3 points, other eligible puzzle--2 points. Ties are not broken; tied creators receive the average number of points associated with the places they tied for.

3. At the end of the tournament, in order to determine the final standings, each contestant's points from all puzzles solved are added together. Creator points are computed and added in.

4. If there is a tie in the standings, it is resolved by head-to-head differential among the tied contestants, counting both solved and created puzzles. Contestant A's "head-to-head differential" against contestants B and C equals the number of contests between A and B in which A scored better than B, minus the number of them in which B scored better than A, plus the number of contests between A and C in which A scored better than C, minus the number of them in which C scored better than A. (A "contest between A and B" is either a puzzle for which both A and B submitted solutions, or puzzle creation by both players.) If some, but not all, of the tied contestants are tied in head-to-head differential, the process (including this sentence) is repeated among just those contestants. Any remaining ties are not broken.

5. The winner receives satisfaction and admiration. (There are no tangible prizes.)

EDIT September 1, 2014: Modified paragraph 2 under "End of Tournament" to change, from 12 to 8, the minimum number of solvers necessary for a puzzle to be eligible for creator points.
Last edited by ToughThought; Sep 1, 2014 @ 5:15pm
ToughThought Jun 22, 2014 @ 5:21pm 
Here are my answers to a couple of questions asked by jfb.

Q. Is fewest reactors allowed as a priority criteria?
A. Yes. Take caution, though, about using a production or sandbox puzzle, since their solutions are more difficult to show off.

Q. Is it just one puzzle per person? Because I have a few puzzles I would like to show off. And it might make the competition a bit short if each participant can only make 1 puzzle each.
A. Yes, just one per person. I believe that there will be enough. I'd be more worried about the competition running too long, if more than one per person were allowed.
ToughThought Jun 25, 2014 @ 3:37pm 
A note to newcomers: you don't have to have competed in a previous SpaceChem tournament to do well. If I'm not mistaken, last year's winner, ecco2, did not compete in 2012. I myself fared pretty well in the 2013 tournament despite just starting to play SpaceChem less than 6 months before. In any event, no matter how well you do, you'll enjoy it. Please join us.
Waypoint Jun 25, 2014 @ 6:50pm 
Holy crap, is it that time of year again?

2013 was tons of fun, even if I had to bow out early when classes blew up. I look forward to taking part in this one with a more stable schedule.
ToughThought Jun 28, 2014 @ 6:50am 
Puzzle activity will start one week from today!

Although Waypoint and I are, so far, the only ones to post in this thread, I have seen indications posted elsewhere from a number of players that they intend to participate. These include carlbunk, gggol, Peffern, jfb1337, Krackor, Blueeyedrat, ProRaptor1, ContentContext, and MrBlarney. I imagine that there are quite a few others who haven't posted or whose posts I've missed.

I don't know whether anyone has posted about this tournament in the Something Awful forums. It would be nice if someone mentions it there.
Sue Jun 28, 2014 @ 6:58am 
Bão! (by GV)
rexkix Jun 28, 2014 @ 10:38am 
I've only recently discovered the 2012 and 2013 tournaments. I'm excited to try my hand here. Nobody need fear me; I'm a truly average SpaceChemist. The tournament rules appear to be well conceived. I see you've already answered it, but I do wonder if the tournament will run long enough with only 1 puzzle per person. I imagine the number of people who are able to submit satisfactory puzzles is small. I know I'm not smart enough, that's for sure.

You might ask people how many expect they'll be submitting puzzles I suspect changing the number of puzzles a person can create from one to two wouldn't make the tournament run too long.

Thanks for organizing this thing.
DariusOne Jun 28, 2014 @ 11:48am 
I'm very much looking forward to this as well, and happy to see you figured out a good format for some competetive play that doesn´t require a massive time investment from a single person to make work. Now to derust myself.

And of course good luck to everyone, and may your solutions function as planned every time you run them :)
Last edited by DariusOne; Jun 28, 2014 @ 11:49am
ToughThought Jun 28, 2014 @ 12:26pm 
Originally posted by rexkix:
I see you've already answered it, but I do wonder if the tournament will run long enough with only 1 puzzle per person. I imagine the number of people who are able to submit satisfactory puzzles is small. I know I'm not smart enough, that's for sure.

You might ask people how many expect they'll be submitting puzzles I suspect changing the number of puzzles a person can create from one to two wouldn't make the tournament run too long.

Okay, who expects to submit a puzzle? I plan to post one. I've also seen that gggol, Peffern, jfb1337, Blueeyedrat, ContentContext, and MrBlarney expect to do so. Who else?

If each contestant focuses on producing one puzzle, I believe that we will see good quality, as well as variety. rexkix, I encourage you to play around with puzzle creation, even if you don't think you'll submit one. It's something different about SpaceChem to enjoy. And maybe you'll find inspiration.
Bashy McFetus Jun 28, 2014 @ 5:25pm 
I've always regretted not getting in on the earlier tounaments, even though my♥♥♥♥♥would've been handed to me far before the finals. Count me in, and I'll try to come up with something interesting.
Waypoint Jun 28, 2014 @ 10:00pm 
Originally posted by rexkix:
Nobody need fear me; I'm a truly average SpaceChemist.

That's what I thought last year until I placed at least once just about every round. Good solutions can come to anyone.
Last edited by Waypoint; Jun 28, 2014 @ 10:01pm
spagchem Jun 29, 2014 @ 5:46am 
I've got a puzzle ready. Now just need to find a less embarrassing solution so beating the creator is a bit of a challenge.
GWen Jun 30, 2014 @ 1:20am 
I've got a puzzle ready too, my solution isn't very elegant but I'm looking forward to see how fast people can do this one.
Waypoint Jun 30, 2014 @ 9:00am 
I've got one with an "easy" but tedious solution and an obvious but difficult path to making it faster. I'll probably save it for later in the tournament though as it's on the tougher side of things.
Carlbunk Jun 30, 2014 @ 3:57pm 
I'll most likely have something to submit but I'm going away for about a week near the end of the month so I probably won't submit any puzzles until then when I have time to collect and review the results.

Speaking of which, is GuavaMoment or WildM around? It would be really great if you could post some details about the software/tools/process you used to create the result videos as a reference for everyone else. Not that we have to make videos, but it would be nice, I think.
< >
Showing 1-15 of 957 comments
Per page: 15 30 50

SpaceChem > General Discussions > Topic Details