12-Mar-2013, 4:14 AM
The first submission is in so I thought I'd write something about how I made these levels, as promised.
Hexominos
.
I wanted a simple level to solve yet one that would not be easy to optimize. Collecting chips is simple, and I like pop-up walls so I thought I'd combine the two in some nice pattern. I played around with some pentomino shapes, but there are only twelve of those, for a total of 60 tiles/chips and that didn't provide enough complexity. Going to hexominos wasn't an easy choice as there are 35 of these things for a total of 210 chips. I was afraid that might be a bit too much. However, the longer I fiddled around with how to pack them neatly, I slowly convinced myself that it would not be too bad. So after I settled on an arrangement, I still needed to work out how to restrict movement. I mean I could just have no walls and it would still be a decent level to optimize, but maybe a bit boring. I wanted to find a way to use pop-up walls to force a bit of thinking just to solve the level, and of course not have the optimal route be evident. First I tried to have each hexomino only have three points of entry, so that you couldn't visit any of them more than once. This proved to be too easy to optimize, and too hard to solve casually, as there were only around ten solutions (which didn't use the outside area to go between rooms). I then increased the number of paths between rooms that connected more than two rooms and allowed some of the rooms to have four exits. Hopefully I now found the right balance. As a last minute addition, I added the teleports to a allow entry from more than just near the starting position.
This graph should help you solve the level:
<a class="ipsAttachLink ipsAttachLink_image" href="<fileStore.core_Attachment>/monthly_03_2013/post-19-0-22751800-1363074521.jpg" data-fileid="5">[img]<fileStore.core_Attachment>/monthly_03_2013/post-19-0-22751800-1363074521_thumb.jpg[/img]</a>
Note that you can visit a node more than once, but can't travel through a path more than once. You also have to visit each node, but not each path, and then end up on the outside. Once you find a path like that, you can directly translate it into the actual level.
Nested Replay
One of my first ideas for a TT level was to try a Replay
variant, which essentially means that not all the areas are required to solve the level. I did that here with the added layer of picking which quadrant you visit and in each quadrant you pick which rooms to visit. I first worked out the mechanism for this, and drew out the rooms. I was quite happy with how none of the quadrants are closer to each other or the starting position (with respect to the distance Chip needs to travel). Of course the whole idea still rested on filling in those sixteen rooms with something. I wanted to have some repetition of themes in the rooms, but not too much.
I first worked on the corner sokobans, thinking that block pushing always takes time and has optimization potential. Since I wanted four of these rooms, I set some small restrictions on what I could do in those rooms. Each needed three keys, and they should be in the corners. Each room should be solvable in both MS and Lynx. There should be as many floor tiles as possible without losing the puzzle. There should only be one way to connect the traps and still be able to get all three keys. (This last restriction didn't quite survive in all the rooms.) It took an incredibly long time to come up with the rooms that are in the finished product, and still the trap connections are not easy to guess. (In my defense, I was making a TT level and assumed most people would just look in an editor even before playing.) You can see the trap connections in this map.
<a class="ipsAttachLink ipsAttachLink_image" href="<fileStore.core_Attachment>/monthly_03_2013/post-19-0-49765200-1363076552.gif" data-fileid="7">[img]<fileStore.core_Attachment>/monthly_03_2013/post-19-0-49765200-1363076552_thumb.gif[/img]</a>
I think I next worked on the pink ball and tank rooms. These were meant as the action packed rooms of the level, but I wanted them static in a way that no matter what time you arrived at the room, it would take you as long to complete. I think they turned out quite as I expected, and they are the toughest to solve. I thought this would be ok, since you could avoid any of them by picking different rooms to play, but still some of the other rooms also proved to be tougher than I anticipated. Here I also settled on making all the rooms workable in both MS and Lynx, as I had first thought it wouldn't be bad if some keys were not reachable in Lynx, since you could still solve the level by going elsewhere. I'm now glad I strived to keep the level as equal as possible.
Then I mapped out what types of rooms I still had left in me to build. I wanted each remaining theme to appear twice, but not so that two themes would appear in the same quadrant together twice. I wanted at most a couple of the themes to not be static, since more than that would draw too much focus. So the Bug and Paramecia room really require you to arrive at the door at the right time or you lose moves. I couldn't use keys as part of a puzzle in any of the rooms, as they were already in use for the mechanism of the whole level. So for two rooms I used footwear and some simple block pushing. I also explored teeth leading in two rooms that were shaped the same but one was vertical and the other horizontal, which meant I couldn't use the same bomb arrangement even if I wanted to. The trap rooms with fireballs were the last ones I made, and at that point I was trying to balance the time it took to complete a room.
That topic deserves its own paragraph. If I made all the rooms take exactly the same amount of time, it wouldn't matter which rooms you visited and you could still score the exact same time. So there needed to be variance, and I had to decide how to distribute this variance around the level. I couldn't just make one quadrant with all the slow rooms. I also had to take into account how long it took to pick up only two keys in a room, as this was part of the plan to allow some craftiness in the solution. (See, I didn't allow for practically any interactions between the rooms.) Some rooms could be slow to pick up all three keys, but fast if you only took two, and actually what mattered was the difference between those two times. Hopefully I ended up with something that isn't too obvious which quadrant to avoid (I'm not 100% sure myself), though the rooms in the quadrants you pick might be easier to choose Oddly enough the "time consuming sokobans" I had envisioned, are now some of the quickest rooms to complete
But this is fine, as they are still tricky to figure out and are definitely the easiest to play through.
In the end I'm pretty happy with how the level turned out, though it did prove to be harder than I meant it to be. I thought that being allowed to go practice any room from the start would help, as well as not having to do all sixteen rooms, but still even nine of these rooms are not simple to complete casually. And of course there is always the threat that someone crazy will come and bust the level really badly
Can't wait to see more solutions and maybe hear your experiences with these levels!
-Miika
P.S. If you only manage to solve one of the levels this month, you can still send in your solution and take part.
Hexominos
.
I wanted a simple level to solve yet one that would not be easy to optimize. Collecting chips is simple, and I like pop-up walls so I thought I'd combine the two in some nice pattern. I played around with some pentomino shapes, but there are only twelve of those, for a total of 60 tiles/chips and that didn't provide enough complexity. Going to hexominos wasn't an easy choice as there are 35 of these things for a total of 210 chips. I was afraid that might be a bit too much. However, the longer I fiddled around with how to pack them neatly, I slowly convinced myself that it would not be too bad. So after I settled on an arrangement, I still needed to work out how to restrict movement. I mean I could just have no walls and it would still be a decent level to optimize, but maybe a bit boring. I wanted to find a way to use pop-up walls to force a bit of thinking just to solve the level, and of course not have the optimal route be evident. First I tried to have each hexomino only have three points of entry, so that you couldn't visit any of them more than once. This proved to be too easy to optimize, and too hard to solve casually, as there were only around ten solutions (which didn't use the outside area to go between rooms). I then increased the number of paths between rooms that connected more than two rooms and allowed some of the rooms to have four exits. Hopefully I now found the right balance. As a last minute addition, I added the teleports to a allow entry from more than just near the starting position.
This graph should help you solve the level:
<a class="ipsAttachLink ipsAttachLink_image" href="<fileStore.core_Attachment>/monthly_03_2013/post-19-0-22751800-1363074521.jpg" data-fileid="5">[img]<fileStore.core_Attachment>/monthly_03_2013/post-19-0-22751800-1363074521_thumb.jpg[/img]</a>
Note that you can visit a node more than once, but can't travel through a path more than once. You also have to visit each node, but not each path, and then end up on the outside. Once you find a path like that, you can directly translate it into the actual level.
Nested Replay
One of my first ideas for a TT level was to try a Replay
variant, which essentially means that not all the areas are required to solve the level. I did that here with the added layer of picking which quadrant you visit and in each quadrant you pick which rooms to visit. I first worked out the mechanism for this, and drew out the rooms. I was quite happy with how none of the quadrants are closer to each other or the starting position (with respect to the distance Chip needs to travel). Of course the whole idea still rested on filling in those sixteen rooms with something. I wanted to have some repetition of themes in the rooms, but not too much.
I first worked on the corner sokobans, thinking that block pushing always takes time and has optimization potential. Since I wanted four of these rooms, I set some small restrictions on what I could do in those rooms. Each needed three keys, and they should be in the corners. Each room should be solvable in both MS and Lynx. There should be as many floor tiles as possible without losing the puzzle. There should only be one way to connect the traps and still be able to get all three keys. (This last restriction didn't quite survive in all the rooms.) It took an incredibly long time to come up with the rooms that are in the finished product, and still the trap connections are not easy to guess. (In my defense, I was making a TT level and assumed most people would just look in an editor even before playing.) You can see the trap connections in this map.
<a class="ipsAttachLink ipsAttachLink_image" href="<fileStore.core_Attachment>/monthly_03_2013/post-19-0-49765200-1363076552.gif" data-fileid="7">[img]<fileStore.core_Attachment>/monthly_03_2013/post-19-0-49765200-1363076552_thumb.gif[/img]</a>
I think I next worked on the pink ball and tank rooms. These were meant as the action packed rooms of the level, but I wanted them static in a way that no matter what time you arrived at the room, it would take you as long to complete. I think they turned out quite as I expected, and they are the toughest to solve. I thought this would be ok, since you could avoid any of them by picking different rooms to play, but still some of the other rooms also proved to be tougher than I anticipated. Here I also settled on making all the rooms workable in both MS and Lynx, as I had first thought it wouldn't be bad if some keys were not reachable in Lynx, since you could still solve the level by going elsewhere. I'm now glad I strived to keep the level as equal as possible.
Then I mapped out what types of rooms I still had left in me to build. I wanted each remaining theme to appear twice, but not so that two themes would appear in the same quadrant together twice. I wanted at most a couple of the themes to not be static, since more than that would draw too much focus. So the Bug and Paramecia room really require you to arrive at the door at the right time or you lose moves. I couldn't use keys as part of a puzzle in any of the rooms, as they were already in use for the mechanism of the whole level. So for two rooms I used footwear and some simple block pushing. I also explored teeth leading in two rooms that were shaped the same but one was vertical and the other horizontal, which meant I couldn't use the same bomb arrangement even if I wanted to. The trap rooms with fireballs were the last ones I made, and at that point I was trying to balance the time it took to complete a room.
That topic deserves its own paragraph. If I made all the rooms take exactly the same amount of time, it wouldn't matter which rooms you visited and you could still score the exact same time. So there needed to be variance, and I had to decide how to distribute this variance around the level. I couldn't just make one quadrant with all the slow rooms. I also had to take into account how long it took to pick up only two keys in a room, as this was part of the plan to allow some craftiness in the solution. (See, I didn't allow for practically any interactions between the rooms.) Some rooms could be slow to pick up all three keys, but fast if you only took two, and actually what mattered was the difference between those two times. Hopefully I ended up with something that isn't too obvious which quadrant to avoid (I'm not 100% sure myself), though the rooms in the quadrants you pick might be easier to choose Oddly enough the "time consuming sokobans" I had envisioned, are now some of the quickest rooms to complete
But this is fine, as they are still tricky to figure out and are definitely the easiest to play through.In the end I'm pretty happy with how the level turned out, though it did prove to be harder than I meant it to be. I thought that being allowed to go practice any room from the start would help, as well as not having to do all sixteen rooms, but still even nine of these rooms are not simple to complete casually. And of course there is always the threat that someone crazy will come and bust the level really badly
Can't wait to see more solutions and maybe hear your experiences with these levels!
-Miika
P.S. If you only manage to solve one of the levels this month, you can still send in your solution and take part.