I agree on entertainment taking the forefront of this TAS. I think unorthodox and humorous, yet quick solutions is the best path to take. The U-Boat for rescuing the kitty is a great example of that. The wave for cleanup is quick and hilarious. More exotic animals for the farmer rather than three cats. The airvent on the flowers was great though because it requires TAS precision to complete.
Be obscure and diverse in your choices. Empanada is going in that direction. Take advantage of the huge dictionary available.
I remember reading something that said it was possible to beat this game without using the same item more than once throughout the whole game (excluding puzzle mode). Maybe you could make a TAS based on that.
Sorry to necro this but I thought you guys would be interested in seeing that the Japanese version of the game has a pretty entertaining WIP for puzzle mode
http://www.nicovideo.jp/watch/sm16129010
Note that this uses SRAM to skip the tutorial.
Sorry to necro this but I thought you guys would be interested in seeing that the Japanese version of the game has a pretty entertaining WIP for puzzle mode
http://www.nicovideo.jp/watch/sm16129010
Note that this uses SRAM to skip the tutorial.
necroed again then. Not very optimized, but this game is really hard to optimize.
Interesting that a Japanese run of the game, that did it for speed would use completely different strategies, beacuse the words have a different number of characters.
Joined: 4/2/2017
Posts: 13
Location: TASVideos.org, of course...
I have made my own WIP of the puzzle portion of the first world, although I used the Scribblenauts Collection release of the game, as it allows for button input, which makes for a much better TASing experience. Button controls also allow for a very easy glitch: the double jump. It is very self-explanatory, but I will say that it allows for several levels to require 1 less object when there is something that is just barely out of reach otherwise, and it may allow for small optimizations in other levels.
Movie file:
User movie #75167431570737622
Encode (using Lua text to substitute for the top screen, for a more console-like screen resolution):
Link to video
I don't want to make this post too long, but I will say that I feel like I typically want to go for more technically impressive or interesting/entertaining solutions (I'm particularly proud of what I came up with for P1-5, and also P1-9). This means that I don't really want to do a "cop out" solution for a level like P1-8 where you just use an "apocalypse object" to get rid of everything. Another goal of mine is to never repeat objects, so I'll have to decide where it is most important to do vehicle clips, spring clips and the like.
Here is a list of objects I've used so far: https://docs.google.com/spreadsheets/d/1TDRPkPElwoH2oEZqTlXZcc0Jq-catRBaCrTTogD5Mo8/#gid=793384863
Feedback is welcomed/appreciated. Let me know if I should take a different approach!
Looks good! I liked 1-7 and 1-9 the best.
You may want to Ask a Judge about the acceptability of this release compared to the original release, especially if using button inputs instead of touchscreen inputs is slower, or if there are other notable differences.
SMWAgent09AF wrote:
Another goal of mine is to never repeat objects, so I'll have to decide where it is most important to do vehicle clips, spring clips and the like.
Is this goal equivalent to getting the "All New" Merit in every stage? I think the speed/entertainment tradeoff to use only unique objects makes sense for the game, and if it can be neatly defined based on that Merit then that's handy (in the case that someone wants to obsolete the movie while following the same goals, for example).
Joined: 4/2/2017
Posts: 13
Location: TASVideos.org, of course...
Really_Tall wrote:
You may want to Ask a Judge about the acceptability of this release compared to the original release, especially if using button inputs instead of touchscreen inputs is slower, or if there are other notable differences.
I don't see any reason why an updated, official release of the same game should present any problems with its ability to be accepted, as the only differences that affect TASing make it better. (Well, besides the few seconds in selecting the game at the beginning, but it approximately evens out because you can skip the opening credits afterwards, which you couldn't do before)
Either way, here are the key changes introduced in the Scribblenauts Collection rerelease:
Button input can now be used to control Maxwell. This allows for much more precise and immediate control of Maxwell. With button input, Maxwell can jump whenever a jump button is pressed, and he won't auto-jump next to a ledge unless the player tells him to. Button input also allows for a double jump glitch that enables more interesting platforming than just summoning a flying machine to get over every 2 foot high ledge. In the unlikely event that touchscreen input is preferred for something, it still can be used (it's a hybrid layout).
The camera doesn't try to lock on Maxwell every time you are manipulating something off-screen. In general, it seems to let the player control it just fine.
An oversight in word recognition has been corrected. Formerly, to bypass the restrictions of Advanced Mode (you must beat the level 3 times in a row, but each time you must use a different set of items), you could literally enter a single letter and a space before a word and the game wouldn't recognize it as an item you had already used. In the rerelease, this no longer works, so you can't cheese Advanced Mode. For our purposes, this isn't that important, since I don't plan to do advanced mode for every level (that would take upward of 6 hours), but it's worth mentioning.
An oversight/bug allowing the player to move the Starite in an action level wherever they want has been fixed. Formerly, there was some setup involving attaching the Starite to handcuffs, attaching those to a bag, basket, etc., then putting the Starite inside the bag. This made it possible to move the bag with the Starite in it, but also cheapens the effect of playing through all the levels.
Per the above, I feel that there is no reason to use the old version when this one is more conducive to creative TASing.
Really_Tall wrote:
SMWAgent09AF wrote:
Another goal of mine is to never repeat objects, so I'll have to decide where it is most important to do vehicle clips, spring clips and the like.
Is this goal equivalent to getting the "All New" Merit in every stage? I think the speed/entertainment tradeoff to use only unique objects makes sense for the game, and if it can be neatly defined based on that Merit then that's handy (in the case that someone wants to obsolete the movie while following the same goals, for example).
This almost works, but unfortunately there are a couple cases where this won't be an accurate measure of if the goal was met. For some reason, if Maxwell holds or wears an object in a level, this will count as an object that the player summoned, even if it was already a part of the level. (This can even be seen in a couple levels in the WIP above.) If you hold an object that you have summoned in a previous level, you will fail the "All New" merit, regardless of if you wrote it yourself. Also, if you beat a level without summoning anything, you won't get the merit. (This is less important, since you will notice that nothing was summoned anyway, but it does mean it isn't that neatly defined.)
Now that I think about it, I wonder if I should go for all merits? There is one specific merit that effectively requires me to summon a bunch of stale objects, and then sandbag in a level for about a minute, but it could be worth it since few people have seen it (it is glitched and unobtainable in the original release). I don't want to make this decision on my own and then find that I've done quite a bit of unnecessary work, so I would appreciate other people giving their thoughts here.
Joined: 4/2/2017
Posts: 13
Location: TASVideos.org, of course...
Okay, I actually have some new thoughts on this. Or rather, I'm rereading other people's thoughts in this thread, and I think using entirely unique objects for all 220 levels might be too much challenge on my part without enough reward. I particularly like Chef Stef's suggestion to aim for unique and unexpected solutions in puzzle levels, but aim mostly for speed in action levels.
A TAS of this game could potentially take two directions: a demonstration of the quirks of the game concept (i.e. using unusual synonyms, finding bizarrely creative ways to solve a basic challenge), or a demonstration of the quirks of the game engine (mainly aiming for speed, using various clipping glitches and speed tricks which are each only possible with a select set of items). I just feel like stretching all the more broken items in the game across some arbitrary spread out levels will make lots of other sections of such a long TAS feel quite bland, and having more varied goals will make for a better demonstration of both. If a viewer doesn't like one approach, they can just skip half of each world, although I imagine both approaches should have their appeal if I can execute them well. In any case, performing the speed tricks reliably and quickly is a good example of superhuman play, since they usually involve precise touchscreen positioning and attention to detail that can't be accomplished as fast in an RTA context.
I don't want to get too far in WIP without hearing some current thoughts regarding goals from other people interested in a TAS of this game, since there are a lot of potential courses of action.
