I see some people at this tread voted for Dwelp for game of the week and that's great, maybe it can win, hope so.
Guys, you're too kind! Big big thanks! @y2kmp3 Thanks for providing those hints! I'm going to prepare the list for the X puzzles soon. In the meantime, I have just submitted an update for review. All of the issues discussed here have been solved. Let's hope it gets through soon.
Congrats, y2kmp3! You're the first to reach a perfect 100/100/100. I'm "only" at 96/100/100. I'm the no.1 on GC with 426, but we know the real champ is you! I still need to purple/master 90, 95, 99, 100 (and 19 of course). No X-levels yet, as I'm focusing on the normal levels. Also, they look scaringly brutal! Unfortunately, week days I don't have so much time to spend on playing. I envy your work, Roundabyte! Tip for Testopi (58): Spoiler solve green and purple first before even starting with yellow and blue . I hope it narrows down the possibilities enough to make it manageable, else just ask for another tip 85: incredible elegant solution! And thanks for the tip, Roger. It limited enough the heuristic solution space to get to the solution and still allowed me to feel proud to have solved it And 77 is so easy once you see it! These are the puzzles I like most. Once you know it, they are easy to remember and repeat. I have the 15-moves achievement too, but that is less easy to repeat, actually. What is the link to the Game-of-the-Week poll? I cannot find it and I want, no, need, to vote Dwelp!
What is the link to the Game-of-the-Week poll? I cannot find it and I want, no, need, to vote Dwelp![/QUOTE] You can find it on the TA website under General discussion.
Thank you for the kind words, saosijs. Although I managed to complete the main campaign, like you, I too have found the X levels to be unreachable (I only stumbled upon the solution for X3).
As promised, here are the hints with the first move for each of the X puzzles. I have no idea if these are going to help. If you would like me to prepare some other kind of hints (not necessarily moves) for these puzzles, let me know. The board is numbered like in chess. Letters across, numbers go up. So bottom left is A1, bottom right G1, top left A7, top right G7 on a 7x7. X1 : Spoiler B1 -> C1 X2 : Spoiler D7 -> E3 X3 : Spoiler D1 -> F2 X4 : Spoiler A4 -> G3 X5 : Spoiler F6 -> E2 X6 : Spoiler C6 -> D4 These are the first moves in order to get the optimal number of moves (the solution I have found, but there are probably different ways to achieve it). Have fun and looking forward to hearing your progress!
Hi, Roundabyte, Please go ahead and prepare whatever hints you feel are more appropriate to these levels. Because these levels (all except X3 which I solved already) have significant amount of symmetry, I entirely agree with you that my initial suggestion for showing first step is not likely to be very helpful.
@y2kmp3 Would you like to prepare a smart hint for X3, since you have already solved that one? If you post it here, I'll include it in my list. Will try to do the other ones now.
Here are somewhat "smarter" hints for the X puzzles. These should give you a nudge in the right direction. X1 : Spoiler The first move merges two yellow dots. The second move is with yellow as well. X2 : Spoiler At some point, you should have a shape of 5 purple dots in a row. X3 : Spoiler In the first move, connect two purple drops. In the second move, connect three yellow dots. X4 : Spoiler Connect yellow completely first. X5 : The sequence of colors and how many dots get attached to the current shape (or to the first dot) on each move: Spoiler G2 - G1 - G3 - Y2 - Y2 - G2 - Y3 - Y2 - G1 X6 : Spoiler Purple ends up as┣ and green as┏
I have now completed 96 levels and 85 purpele'd. The hard levels are really hard, so I focus for now on the 100 normal levels first. Stumbled now on level 97. Wow just did it. 3 more to go! Make that 2.
Can someone take a look at X-6 and tell me the NUMBER shown as the optimal step for this level BEFORE YOU start playing the level? Humor me on this... There is a reason... will explain in a later post.
Completed the game (see screenshot)! All 100 Levels, 9 Bonus, 6 Extra. All optimal solutions found (notwithstanding bug in Level 19). Roundabyte, thank you so much for your "smarter" hints. I agree that these hints are much better than just showing the first steps. Very thoughtful hints too. They were very helpful but without too spoiler-y. Though I solved X3 without your hint, I like your hint suggestion for that level and would not change it otherwise. Two very interesting tidbits: X5: I found an alternate solution that entirely differs from the one suggested by your hint (the ending configuration is completely different). Very happy to find this, as it sorta allows me to deem myself by not using your hinted solution. Following the same hint format, here is my solution: Spoiler G2 - G1 - G3 - Y2 - Y1 - Y2 - G3 - Y2 - Y2 X6: I will wait for someone else to confirm the puzzle's initial state before I report my finding. Found a completely unexpected behavior of the game. First time I've seen this in the entire game. The X-levels are truly, truly difficult. In some ways, I think it is even harder than the late levels in the bonus pack of Trainyard, which I have long considered to be the gold standard for a "difficult" logic puzzle game with a pathfinding mechanic.
A huge congrats to the first player (to my knowledge) that has fully completed Dwelp! You are the best! Your solving skills and perseverance are amazing and one of the main reasons why I will definitely try to come up with more and more puzzles in the updates. Re. X6: It feels so good to beat the machine, doesn't it? =) So, what happened, you may ask... You know I have mentioned that I ran all the puzzles through an algorithm, to find the true optimal solution. Well, X6 was the only one which the algorithm wasn't able to solve in less than 13 moves in a timely fashion. Being so complex (not necessarily complicated - but with 4 colors and more than a billion possible moves), it ran through the solver for more than three days and I had to stop it and assume that the solution was 13. The funny thing is that I only found the 12 steps solution myself yesterday, when solving it again so I can provide the hints and was very eagerly waiting to see if anyone will find it and notice that. So, X6 is the only puzzle which you can solve in less moves than stated in the current version (I have changed the number in the update submitted yesterday). Well, congratulations again! There is one more challenge for now that I'm not sure you are aware of (though saosijs has mentioned it briefly). In Game Center there is an achievement called "No. 77 in 15 moves". Indeed, while the optimal number for 77 is 5 moves, it can be solved in 15 moves (and I believe it's much harder to do than in 5 moves). You could give it a try if you want. PS. Comparing Dwelp to Trainyard is another great compliment! I was a big fan of Trainyard too back then. Thanks a lot!
Yep, that's why I wanted to confirm the target goal before I disclose my finding publicly. As you correctly guessed, my solution used only 12 steps (and not 13 steps hinted by the game). Knowing that you used an algorithm to check for all possible solutions, I was therefore very surprised to find this. I actually think you should not alter this target in the update. I think it makes for a nice easter egg to reward players who manage to reach this stage. This level is also an excellent example to demonstrate the point I made earlier about what makes good game design. It is always true that a good game teaches and leverages heuristics in its solutions rather than algorithms. As I said before, humans solve problems (this game or other games such as chess) using heuristics. Machines solve, to most degree, using algorithms. Heuristics, despite its flaws, is generalizable. Algorithms do not. An example of this differences in approach is in Chess or Go. For example, to date, no machine-learning that is algorithm-based can play Go well or beat an expert Go player. For casual games, the same principle applies. The mere fact that both you and I found a better solution and beat the machine is testament that heuristic-based is superior to algorithmic-based problem solving. Over-engineering a puzzle to be too complicated by generating it using an algorithm does not necessarily make a good puzzle. Rather, it just makes a tedious puzzle that is difficult to solve but not fun to play because it does not lend itself to heuristic attack. Thank you again for your kind words. This is among the best logic puzzle games I have ever played. Some truly incredibly difficult puzzles. Looking back, I am just so glad that you are here to guide us. I am not sure I would persevere to finally beat the game without your ongoing guidance and encouragement.
Here you go (screenshot without showing the entire solution). Level 77 in 15 moves. Agree with you that this challenge is much harder but has a neat solution. I presumed your algorithm checked that this is the solution with the highest possible number of moves. Aside from rotational symmetry, are there more than one truly distinct solution?
Level 79 join bug Discovered a bug on Level 79 that displayed my top row as being disconnected, then moving that row as one, disjointed unit. I hit undo and performed the moves again and ended up with two separate connections. iPhone 5, iOS 8.1.1 http://imgur.com/hBCrLMx.jpg http://imgur.com/8r8w0DY.jpg http://imgur.com/4F0jSJI.jpg http://imgur.com/mEZJZus.jpg
Way to go, Roger! You're pretty close and I'm sure you'll be able to solve the X ones as well. They do look somewhat scary, but they are manageable. If you want, it might be a good idea to start with X3, the one that y2kmp3 has solved first and without any hints. Great job! Indeed, 15 moves is the highest solution and I know there are at least two of these that are not symmetrical, but haven't counted how many. Thanks, Hotcreme! Wow! I haven't seen one of those since the early stages of development. Will look into it. Are you able to reproduce it, or it just happened once and you don't know how?
Not positive how it happened, but I believe it was letting go of the dots while they weren't perfectly located at their destination. It's happened once before, so I'll take note of what I'm doing if it happens again.
@Hotcreme Thanks a lot for the feedback! I appreciate it! If anyone else has encountered this behaviour, please let me know.
