I hadn't been planning to make an account here, but I've been working on this for a while and wanted to share my accomplishments and discuss some of the issues I ran across in trying to do it, as I think there might be some broader applications for breeders and competitive theory discussions. Although I'll be mainly talking about what I personally did and how, I'll try to turn this into a more general guide on using and abusing Sketch (for fun and profit) in Uranium.
Smeargle has been one of my favourite canonical pokemon for a while, and I've always loved working out new interesting things to do with them, so it was only natural I'd eventually come round to experimenting with the Uranium equivalent. Let's start with some pictures of the results of my latest projects.
Here's Project Number One:
I'd seen some discussion on the forums about using Sketch to get Atomic Punch, but I hesitated to try it so long as that would have required revisiting the final battle (heh, shows what I know; more on this later). I settled on this moveset after reading some of the threads in Competitive Discussion on here, and also thinking about what I would want to have alongside Atomic Punch (I dithered a bit on choosing Flame Impact versus Infernal Blade, another thing I'll get into later), and Sleep Powder seems like the best sleep move in the current version (Spore would be better, and is what Smeargle tends to use, but it doesn't exist in Uranium right now; I think a case could be made for Glare or Thunder Wave or even Yawn, too, but Sleep Powder's what I decided to go with in the end).
If I was going to put all this effort into getting my moveset, I thought it was only natural to breed perfect IVs first so I didn't waste it on something subpar (I even considered trying for a shiny, but in the end decided that probably wasn't worth the frustration). Thankfully that part wasn't too hard, I had a Jolly Puffluff from the Christmas event that could just breed iteratively with female Ratsy (puts a new interpretation on the phrase 'Father Christmas', doesn't it?) and ended up getting what I wanted after about two boxes' worth. I opted for Insomnia over Quick Feet (or Pickup) because, while none of these abilities are really anything major to write home about, I thought Quick Feet would be redundant at best and useless at worst (Shell Smash takes care of speed and I planned on using Focus Sash to make sure that got off; likewise, because Toxic Orb takes effect at the end of turn I couldn't really think of an effective way to use that without wasting a move slot on Protect).
And Project Number Two:
The other "classic" use for Sketch (and hence Smeargle) is in breeding, to pass on egg moves that are otherwise unavailable. I wanted to see if this was possible with Raffiti in Uranium; more specifically, I chose Synchronoise to test with because it could potentially be really useful for Nucleon (to use against opposing Nuclear types) and the general consensus had been that it was not possible to get on an Eevee (I think because it's only learned by the Comite line through level-up, and they're genderless so can't participate in breeding chains). But as you see, by Sketching it and then using a Raffiti as the father, you can pass it on to an Eevee just fine. And then once you have Eevee parents with it, they should be able to pass it on themselves as well.
I just did this one as proof-of-concept, I haven't decided if I want to seriously breed (i.e. IVs and Hidden Power, natures, etc) Synchronoise Eevees or not. But if anyone does, I would be happy to provide a starter parent (Eevee, not Raffiti) with the move to work with.
I haven't looked into other egg moves that might be possible to access this way, but it might be worth doing. Please do share your ideas if you come up with something.
So how'd I do this? It's actually surprisingly tricky to Sketch moves in Uranium; people love theorising Sketch movesets (and playing them in Pokemon Showdown) but I see a lot less discussion of how to actually get the moves.
My first plan was to use double battles, like I usually do in the canonical games. This is actually trickier in Uranium than official Pokemon games, because there are no repeatable double battles in Uranium. (For example, when I do Sketching in Gen 3 I usually use the journalists Gabby & Ty who can be rematched indefinitely with no wait between fights, and have only one pokemon each so the battle is short.) So I had to prepare a file that had a few unfought double battles on it - specifically, I'd need four, because you can only sketch one move per battle and need to visit the Move Reminder to learn Sketch again in between*.
*Technically this isn't entirely accurate. You could learn two moves in the same battle if you set it up with a Ratsy/Raffiti at level 50 or 80 with enough experience to level up from the next kill, Sketch a move and then knock something out so it levels up and learns Sketch again. But you'd still need at least two battles to Sketch a full complement of four moves, and you still need to visit the Move Reminder at least once if you want four Sketched moves.
I've put together a list of all the double battles I could find in Uranium (let me know if I missed any, I'm not 100% certain I got them all and, e.g., the Uranium wiki only specifies the "couple" type and doesn't label when a single trainer is a double battle or when you can walk between two trainers to fight them together). I'm not counting ones where you have a NPC partner, because those are useless for Sketching purposes.
There are a few more double battles that must be finished to pass through the area, so you can't save them for later:
If I've missed any for either list, please let me know; if I'd been thinking about it, I'd have done a playthrough and taken notes to be sure I got all of them.
Anyway, I raised Urayne to level 80 to get Atomic Punch, and caught and raised a Cararalm to 52 for Shell Smash, and a Cubblfly for Sleep Powder (wild ones in Route 06 know it when caught so that's not too bad). I already had an Inflagetah with Flame Impact and a Beliaddon with Infernal Blade depending on which of those I wanted to go with. I also raised a Cometeor to level 60 to learn Synchronoise.
Then I actually tried to do this, and hit a wall. I hadn't realised previously, but Sketch works differently in Uranium than in the official games. In official Pokemon games, Sketch can target any other pokemon, but in Uranium (I think this might be a bug? if so, I hope the devs will take note for the next patch) it's limited to targeting the enemy team. So that strategy went utterly down the drain. I wish I'd thought to check this before actually raising all the Pokemon and preparing the file; instead, I would need to find opponents who could use all of the moves I wanted, and find a reliable way to bait them into using the move so I could Sketch it.
So here's the method I actually used to get each move:
Atomic Punch: Final boss Gamma Urayne, obviously. This actually wasn't as tough to get as I first expected, after putting some thought into it (and thankfully, the file I'd been playing to find double battles wasn't too far from the end). All I had to do was save just before starting the Urayne fight, stick a Focus Sash on Raffiti, and use Sketch the first turn, then reset if it didn't go right (the AI isn't completely consistent in what move it decides to use, if it has multiple that it thinks will get a KO). As long as Urayne is faster (which it should be), Raffiti will survive the hit and Sketch whatever move Urayne used (this will be true even if Fallout is later fixed; I don't think the periodic damage from it is working because I've never seen it hit anything, but even if so, it will hit at the end of turn after Raffiti has already learned the move).
Flame Impact: I used the Speed Ninja battle for this, which was pretty easy (although again I had to reset a few times, because the AI seems to choose the move at random). This is one of the reasons I ended up choosing Flame Impact over Infernal Blade; I think it's technically possible to get that from Tiko's Beliaddon during the Championship, but there's a lot more randomness/reset-spamming involved in even getting him to show up at all, even before considering trying to bait and Sketch it during the battle. I think I prefer Flame Impact anyway, after some consideration, since it actually has a full animation instead of Infernal Blade using the placeholder strike animation right now, and a priority move is really important to have in this set.
Shell Smash: Ooh boy. This one was tricky. I ended up using Fisherman Luke in Route 05 (he's the one you can rematch via Pokepod). He leads with a pokemon of Cocaran line, which can learn it (Cocaran at L47; Cararalm and Cocancer at L52), which does mean that as long as it's at least level 52 it will know the move. The problem is that it's tricky to get him to actually use it, especially because it can potentially know Guillotine just a few levels later (Cocaran 52, Cararalm/Cocancer 57) and the AI likes to go for the KO whenever possible. Especially if the Raffiti is lower level, it's likely to just hit and take it out. So there are a couple of options: revive spam etc to run out his PP of everything else, or build a special team to try to manipulate his levels and bait the moves etc.
I ended up building a team just for this. It wasn't as hard as it sounds, actually, though I was annoyed to have to do it. Essentially, I took advantage of the fact the rematch trainers' pokemon levels seem to be based on the levels of things in your party (I don't know if it's based on a random pokemon or some kind of level average), and the fact the AI has some predictable tendencies (it seems to like using status moves most when everything else it has will be resisted; it always uses an attacking move if it will KO). I led with a fresh-caught L55 Antarki from the Victory Road volcano area, stuffed the rest of my team with random level 50-55 things I had lying around, and had my Raffiti along. I chose Antarki because it's immune to Guillotine thanks to being Ghost type, and resists Leaf Blade/Wood Hammer (which is the other main attacking move it has). Antarki also has the advantage of being able to easily sweep his team after a successful attempt, with Flamethrower for Cocancer, Psychic for Blubelrog and Shadow Ball for Brainoar (even with no EVs it was fine).
It actually might have been better to have used a higher-level Antarki (and rest of the team), come to think of it, to avoid tempting it into using Dig which would be super-effective on Antarki (Guillotine will replace Dig when learned, leaving it only that and Leaf Blade/Wood Hammer as attacking moves) but I didn't end up running into that problem. Leading with Antarki, it would use Shell Smash first turn more often than not, so I could just swap to Raffiti the first turn and (as long as Raffiti outspeeds Cararalm/Cocancer after Shell Smash, which might require EV training and a Choice Scarf depending on level) sketch Shell Smash immediately at the start of the following turn, or reset and try again if it didn't work.
Sleep Powder: This was the easiest one, actually, thanks to Raffiti's Insomnia. All I had to do was ensure Raffiti was faster than a L15 Cubblfly (the ones in Route 06), fight a wild one, and keep wasting my turn (I spammed Shell Smash and potions) until it tried to use Sleep Powder. Because I was faster I could just Sketch it easily the next turn.
Synchronoise: Another really tricky one. The only source I managed to find was a trainer in the Lanthanite Core postgame area: Acolyte Elwyn has a level 71 Astronite in his party that uses level-up moveset (at that level it's Recover, Stone Edge, Moonblast, Synchronoise). Unfortunately, unlike Shell Smash I couldn't work out a reliable way to tempt it to use this without running out the PP on the rest of its moves, which was tricky but doable (the hardest part was running out Recover, but it's feasible as long as you have something which will reliably do ~1/4 to 1/3 of its health in damage so you aren't risking killing it). Once Synchronoise is all that's left, it's easy to Sketch since it can't hurt Raffiti at all. I recommend using something Synchronoise won't work on while you're doing this, so it doesn't run that out of PP before the other moves. (This took me two attempts because I lost count the first time and got hit with a surprise Recover when I thought it was safe to go for the Sketch; the second time I kept a written tally.)
So this post isn't just me long-windedly bragging about my exploits, let's discuss how Sketch works and how you might take advantage of it. Here's the Bulbapedia page.
Sketch targets a pokemon other than the user (currently, as of Uranium 1.1, it is further limited to target only an enemy, unlike the official games, but I'm hoping this will eventually be fixed). If successful, Sketch will permanently replace itself with the last move used by the target. The learned move will have full PP and will be ready to use immediately. (In official games, learning the move is permanent only if you would receive EXP for the battle, and lasts only for the battle otherwise - so it's temporary in VS battles, battle facilities, and so on; I would assume this is true in Uranium also, though I have not tested Virtual Trainer or the Online Lobby.)
Sketch will fail if the target has not used a move that battle (for example, if you use it on the first turn and go first), if your pokemon already knows the move you're trying to Sketch, and possibly in special circumstances with certain moves (I haven't tested with e.g. moves like Fly/Dig/Bounce/Dive in Uranium, but in official games Sketch fails on the setup turn of those moves). I would also like to see how it behaves with Metronome in Uranium, which is not currently possible to test because (to my knowledge) you cannot fight anything that knows it.
Sketch has 1 PP (it is the only move in any Pokemon game ever not to have base PP in a multiple of 5). This cannot be increased with the use of PP Ups; I did test this, and Uranium will allow you to apply PP Ups to it as normal (and the game remembers how many you've applied, and will stop you at the maximum of three) but the actual number of PP will not change; this is probably a bug, as official Pokemon games do not allow you to use PP Ups on Sketch at all (I've tested this too). This means you have effectively one try to Sketch a move correctly, although this is moot in the case you mess up by hitting the wrong move (in which case Sketch will be overwritten by that move). If Sketch does "fail", however, you can use a PP-restoring item like Ether or Leppa Berry and use it again even in the same battle.
A Pokemon cannot have multiple copies of the same move. Notably: you cannot have multiple copies of Sketch, to try learning multiple moves at once; you need to learn moves one at a time and re-teach Sketch in between (whether you do this via level-up or Move Reminder).
Sketch behaves interestingly if the pokemon using it holds a Choice item. If you use Sketch while holding a choice item, you will not get choice-locked (even if it fails) and can choose any move the following turn (including but not limited to a move learned by Sketch). In the official games, you will become choice-locked upon choosing a move other than Sketch. When I tested this in Uranium, I found it behaved differently, and I never became choice-locked at all for the remainder of the battle (this is probably a bug). I think this is of primarily academic interest, as I can't think of any particular reason you'd want to do this in actual gameplay.
So, a few general tips on Sketching moves from enemies, based on what I learned while putting my Raffiti together.
(1) It's a lot easier to guarantee you hit the right move if your Sketch pokemon is faster than the pokemon with the move you want. As long as you go first, you can just use Sketch the turn after seeing the move you want, provided you survive. Which leads into...
(2) Focus Sashes are your friend, if your Sketch pokemon doesn't have the bulk (or levels) to survive being hit by the move you're after. They're a bit expensive - 14000 each from the shop in Tsukinami - and they get consumed on activation, but they really do make the task a lot easier. They are also a big help if you have no hope of outspeeding your target (e.g., trying to sketch a priority move).
(3) The AI seems to choose its moves based on the pokemon you currently have in, it doesn't ever try to predict a switch. So, if your Sketch pokemon is faster than the target, you can try to bait the AI by sending in a pokemon that will tempt it to use the move you want (e.g. if it's an attacking move, use something weak to it, etc), immediately switching out for your Sketch user, and using Sketch first thing the following turn if it cooperated and used the move you wanted.
(4) It helps if you find a source of the move that you can save before encountering, so you can reset if it goes badly (so if there are multiple places you can encounter a move, it might be worth taking this into account when you're trying to choose one). This is especially true in some cases, like I encountered with Atomic Punch and Synchronoise, where there might be only one source of the move that can only be fought once per save file. Go in with a plan, but make sure you save first or you may not get another attempt.
(5) Not every move is available in the game right now (and a move being available doesn't mean you can Sketch it either: there are some obtainable pokemon that never appear in the wild and no trainers use against you), and some are much harder to find than others. Take this into account when developing your moveset, especially because some moves can take a lot of work and be frustrating to successfully Sketch.
So why use Sketch?
There are two main purposes, as I alluded to earlier.
The first is to put together unique movesets; having Sketch gives Smeargle (and, in Uranium, Raffiti) access to nearly every move in the game, and certainly more moves than any other pokemon can hope to have access to! Whatever cool or broken or quirky interactions between moves you can think of, you can have it. (In theory, anyway; admittedly, right now in Uranium we're limited to moves we encounter opponents using, but hopefully that will eventually change.) The most notable and splashy example right now is probably the Nuclear moves: Raffiti is the only pokemon capable of learning Nuclear-type moves while not being Nuclear-typed itself (and therefore foregoing its weaknesses).
I've shown one such moveset above, and the methods I used to actually learn it, but that's far from exhausting the potential. There are so many more things you can try.
The second purpose is for breeding. If you want an egg move but can't find a parent that can learn it, you might be able to use Sketch. It's easiest for pokemon in the Field egg group, which Ratsy and Raffiti belong to (also Smeargle, in the official games); I know in the official games there are plenty of breeding chains that have to begin with a Smeargle in order to pass the necessary moves, but I haven't exhaustively checked Uranium yet. As long as you use a male pokemon to do the Sketching, you can use it as the father and pass the moves on to children (if you use a female, all the children will be Ratsy, and the Ratsy species has no egg moves so nothing will be inherited). That said, Uranium uses generation 6 breeding rules, which means that both parents can pass on egg moves to the child (unlike earlier games where it was limited to the father).
Further notes on egg moves: Generation 6 also added the ability to relearn egg moves with the Move Reminder. I did a bit of testing and this functionality is in Uranium, but the opportunity is lost on evolution, which I think is different from the official games (I took one of my Synchronoise Eevees, overwrote Synchronoise and tried to relearn it. As an Eevee it could relearn it fine, but after evolving it to Flareon Synchronoise was no longer present in the menu at the Move Reminder). This may or may not be a bug, but regardless that is how it works right now.
I have not made anywhere near an exhaustive examination of Uranium's egg moves and which ones are potentially accessible via Sketch, outside of demonstrating Synchronoise and looking into the possibility of getting Volt Switch on Volchik (which is not currently possible as far as I can tell, and wouldn't be particularly helped by Sketch since Slothohm learns it naturally and is in the same Field group as Ratsy; there's no intermediate pokemon to use for chain breeding).
And if anyone wants one, I have quite a few 5IV Ratsy left over (outside of the ones I dumped into Wonder Trade a week or so ago; OT Aristea if you saw any of them), and some Synchronoise Eevees, and it wouldn't be too difficult for me to make more if I run out. If they'd help with any of your projects I'm sure we can work out a way to get one to you.
Smeargle has been one of my favourite canonical pokemon for a while, and I've always loved working out new interesting things to do with them, so it was only natural I'd eventually come round to experimenting with the Uranium equivalent. Let's start with some pictures of the results of my latest projects.
Here's Project Number One:
I'd seen some discussion on the forums about using Sketch to get Atomic Punch, but I hesitated to try it so long as that would have required revisiting the final battle (heh, shows what I know; more on this later). I settled on this moveset after reading some of the threads in Competitive Discussion on here, and also thinking about what I would want to have alongside Atomic Punch (I dithered a bit on choosing Flame Impact versus Infernal Blade, another thing I'll get into later), and Sleep Powder seems like the best sleep move in the current version (Spore would be better, and is what Smeargle tends to use, but it doesn't exist in Uranium right now; I think a case could be made for Glare or Thunder Wave or even Yawn, too, but Sleep Powder's what I decided to go with in the end).
If I was going to put all this effort into getting my moveset, I thought it was only natural to breed perfect IVs first so I didn't waste it on something subpar (I even considered trying for a shiny, but in the end decided that probably wasn't worth the frustration). Thankfully that part wasn't too hard, I had a Jolly Puffluff from the Christmas event that could just breed iteratively with female Ratsy (puts a new interpretation on the phrase 'Father Christmas', doesn't it?) and ended up getting what I wanted after about two boxes' worth. I opted for Insomnia over Quick Feet (or Pickup) because, while none of these abilities are really anything major to write home about, I thought Quick Feet would be redundant at best and useless at worst (Shell Smash takes care of speed and I planned on using Focus Sash to make sure that got off; likewise, because Toxic Orb takes effect at the end of turn I couldn't really think of an effective way to use that without wasting a move slot on Protect).
And Project Number Two:
The other "classic" use for Sketch (and hence Smeargle) is in breeding, to pass on egg moves that are otherwise unavailable. I wanted to see if this was possible with Raffiti in Uranium; more specifically, I chose Synchronoise to test with because it could potentially be really useful for Nucleon (to use against opposing Nuclear types) and the general consensus had been that it was not possible to get on an Eevee (I think because it's only learned by the Comite line through level-up, and they're genderless so can't participate in breeding chains). But as you see, by Sketching it and then using a Raffiti as the father, you can pass it on to an Eevee just fine. And then once you have Eevee parents with it, they should be able to pass it on themselves as well.
I just did this one as proof-of-concept, I haven't decided if I want to seriously breed (i.e. IVs and Hidden Power, natures, etc) Synchronoise Eevees or not. But if anyone does, I would be happy to provide a starter parent (Eevee, not Raffiti) with the move to work with.
I haven't looked into other egg moves that might be possible to access this way, but it might be worth doing. Please do share your ideas if you come up with something.
So how'd I do this? It's actually surprisingly tricky to Sketch moves in Uranium; people love theorising Sketch movesets (and playing them in Pokemon Showdown) but I see a lot less discussion of how to actually get the moves.
My first plan was to use double battles, like I usually do in the canonical games. This is actually trickier in Uranium than official Pokemon games, because there are no repeatable double battles in Uranium. (For example, when I do Sketching in Gen 3 I usually use the journalists Gabby & Ty who can be rematched indefinitely with no wait between fights, and have only one pokemon each so the battle is short.) So I had to prepare a file that had a few unfought double battles on it - specifically, I'd need four, because you can only sketch one move per battle and need to visit the Move Reminder to learn Sketch again in between*.
*Technically this isn't entirely accurate. You could learn two moves in the same battle if you set it up with a Ratsy/Raffiti at level 50 or 80 with enough experience to level up from the next kill, Sketch a move and then knock something out so it levels up and learns Sketch again. But you'd still need at least two battles to Sketch a full complement of four moves, and you still need to visit the Move Reminder at least once if you want four Sketched moves.
I've put together a list of all the double battles I could find in Uranium (let me know if I missed any, I'm not 100% certain I got them all and, e.g., the Uranium wiki only specifies the "couple" type and doesn't label when a single trainer is a double battle or when you can walk between two trainers to fight them together). I'm not counting ones where you have a NPC partner, because those are useless for Sketching purposes.
There are a few more double battles that must be finished to pass through the area, so you can't save them for later:
If I've missed any for either list, please let me know; if I'd been thinking about it, I'd have done a playthrough and taken notes to be sure I got all of them.
Anyway, I raised Urayne to level 80 to get Atomic Punch, and caught and raised a Cararalm to 52 for Shell Smash, and a Cubblfly for Sleep Powder (wild ones in Route 06 know it when caught so that's not too bad). I already had an Inflagetah with Flame Impact and a Beliaddon with Infernal Blade depending on which of those I wanted to go with. I also raised a Cometeor to level 60 to learn Synchronoise.
Then I actually tried to do this, and hit a wall. I hadn't realised previously, but Sketch works differently in Uranium than in the official games. In official Pokemon games, Sketch can target any other pokemon, but in Uranium (I think this might be a bug? if so, I hope the devs will take note for the next patch) it's limited to targeting the enemy team. So that strategy went utterly down the drain. I wish I'd thought to check this before actually raising all the Pokemon and preparing the file; instead, I would need to find opponents who could use all of the moves I wanted, and find a reliable way to bait them into using the move so I could Sketch it.
So here's the method I actually used to get each move:
Atomic Punch: Final boss Gamma Urayne, obviously. This actually wasn't as tough to get as I first expected, after putting some thought into it (and thankfully, the file I'd been playing to find double battles wasn't too far from the end). All I had to do was save just before starting the Urayne fight, stick a Focus Sash on Raffiti, and use Sketch the first turn, then reset if it didn't go right (the AI isn't completely consistent in what move it decides to use, if it has multiple that it thinks will get a KO). As long as Urayne is faster (which it should be), Raffiti will survive the hit and Sketch whatever move Urayne used (this will be true even if Fallout is later fixed; I don't think the periodic damage from it is working because I've never seen it hit anything, but even if so, it will hit at the end of turn after Raffiti has already learned the move).
Flame Impact: I used the Speed Ninja battle for this, which was pretty easy (although again I had to reset a few times, because the AI seems to choose the move at random). This is one of the reasons I ended up choosing Flame Impact over Infernal Blade; I think it's technically possible to get that from Tiko's Beliaddon during the Championship, but there's a lot more randomness/reset-spamming involved in even getting him to show up at all, even before considering trying to bait and Sketch it during the battle. I think I prefer Flame Impact anyway, after some consideration, since it actually has a full animation instead of Infernal Blade using the placeholder strike animation right now, and a priority move is really important to have in this set.
Shell Smash: Ooh boy. This one was tricky. I ended up using Fisherman Luke in Route 05 (he's the one you can rematch via Pokepod). He leads with a pokemon of Cocaran line, which can learn it (Cocaran at L47; Cararalm and Cocancer at L52), which does mean that as long as it's at least level 52 it will know the move. The problem is that it's tricky to get him to actually use it, especially because it can potentially know Guillotine just a few levels later (Cocaran 52, Cararalm/Cocancer 57) and the AI likes to go for the KO whenever possible. Especially if the Raffiti is lower level, it's likely to just hit and take it out. So there are a couple of options: revive spam etc to run out his PP of everything else, or build a special team to try to manipulate his levels and bait the moves etc.
I ended up building a team just for this. It wasn't as hard as it sounds, actually, though I was annoyed to have to do it. Essentially, I took advantage of the fact the rematch trainers' pokemon levels seem to be based on the levels of things in your party (I don't know if it's based on a random pokemon or some kind of level average), and the fact the AI has some predictable tendencies (it seems to like using status moves most when everything else it has will be resisted; it always uses an attacking move if it will KO). I led with a fresh-caught L55 Antarki from the Victory Road volcano area, stuffed the rest of my team with random level 50-55 things I had lying around, and had my Raffiti along. I chose Antarki because it's immune to Guillotine thanks to being Ghost type, and resists Leaf Blade/Wood Hammer (which is the other main attacking move it has). Antarki also has the advantage of being able to easily sweep his team after a successful attempt, with Flamethrower for Cocancer, Psychic for Blubelrog and Shadow Ball for Brainoar (even with no EVs it was fine).
It actually might have been better to have used a higher-level Antarki (and rest of the team), come to think of it, to avoid tempting it into using Dig which would be super-effective on Antarki (Guillotine will replace Dig when learned, leaving it only that and Leaf Blade/Wood Hammer as attacking moves) but I didn't end up running into that problem. Leading with Antarki, it would use Shell Smash first turn more often than not, so I could just swap to Raffiti the first turn and (as long as Raffiti outspeeds Cararalm/Cocancer after Shell Smash, which might require EV training and a Choice Scarf depending on level) sketch Shell Smash immediately at the start of the following turn, or reset and try again if it didn't work.
Sleep Powder: This was the easiest one, actually, thanks to Raffiti's Insomnia. All I had to do was ensure Raffiti was faster than a L15 Cubblfly (the ones in Route 06), fight a wild one, and keep wasting my turn (I spammed Shell Smash and potions) until it tried to use Sleep Powder. Because I was faster I could just Sketch it easily the next turn.
Synchronoise: Another really tricky one. The only source I managed to find was a trainer in the Lanthanite Core postgame area: Acolyte Elwyn has a level 71 Astronite in his party that uses level-up moveset (at that level it's Recover, Stone Edge, Moonblast, Synchronoise). Unfortunately, unlike Shell Smash I couldn't work out a reliable way to tempt it to use this without running out the PP on the rest of its moves, which was tricky but doable (the hardest part was running out Recover, but it's feasible as long as you have something which will reliably do ~1/4 to 1/3 of its health in damage so you aren't risking killing it). Once Synchronoise is all that's left, it's easy to Sketch since it can't hurt Raffiti at all. I recommend using something Synchronoise won't work on while you're doing this, so it doesn't run that out of PP before the other moves. (This took me two attempts because I lost count the first time and got hit with a surprise Recover when I thought it was safe to go for the Sketch; the second time I kept a written tally.)
So this post isn't just me long-windedly bragging about my exploits, let's discuss how Sketch works and how you might take advantage of it. Here's the Bulbapedia page.
Sketch targets a pokemon other than the user (currently, as of Uranium 1.1, it is further limited to target only an enemy, unlike the official games, but I'm hoping this will eventually be fixed). If successful, Sketch will permanently replace itself with the last move used by the target. The learned move will have full PP and will be ready to use immediately. (In official games, learning the move is permanent only if you would receive EXP for the battle, and lasts only for the battle otherwise - so it's temporary in VS battles, battle facilities, and so on; I would assume this is true in Uranium also, though I have not tested Virtual Trainer or the Online Lobby.)
Sketch will fail if the target has not used a move that battle (for example, if you use it on the first turn and go first), if your pokemon already knows the move you're trying to Sketch, and possibly in special circumstances with certain moves (I haven't tested with e.g. moves like Fly/Dig/Bounce/Dive in Uranium, but in official games Sketch fails on the setup turn of those moves). I would also like to see how it behaves with Metronome in Uranium, which is not currently possible to test because (to my knowledge) you cannot fight anything that knows it.
Sketch has 1 PP (it is the only move in any Pokemon game ever not to have base PP in a multiple of 5). This cannot be increased with the use of PP Ups; I did test this, and Uranium will allow you to apply PP Ups to it as normal (and the game remembers how many you've applied, and will stop you at the maximum of three) but the actual number of PP will not change; this is probably a bug, as official Pokemon games do not allow you to use PP Ups on Sketch at all (I've tested this too). This means you have effectively one try to Sketch a move correctly, although this is moot in the case you mess up by hitting the wrong move (in which case Sketch will be overwritten by that move). If Sketch does "fail", however, you can use a PP-restoring item like Ether or Leppa Berry and use it again even in the same battle.
A Pokemon cannot have multiple copies of the same move. Notably: you cannot have multiple copies of Sketch, to try learning multiple moves at once; you need to learn moves one at a time and re-teach Sketch in between (whether you do this via level-up or Move Reminder).
Sketch behaves interestingly if the pokemon using it holds a Choice item. If you use Sketch while holding a choice item, you will not get choice-locked (even if it fails) and can choose any move the following turn (including but not limited to a move learned by Sketch). In the official games, you will become choice-locked upon choosing a move other than Sketch. When I tested this in Uranium, I found it behaved differently, and I never became choice-locked at all for the remainder of the battle (this is probably a bug). I think this is of primarily academic interest, as I can't think of any particular reason you'd want to do this in actual gameplay.
So, a few general tips on Sketching moves from enemies, based on what I learned while putting my Raffiti together.
(1) It's a lot easier to guarantee you hit the right move if your Sketch pokemon is faster than the pokemon with the move you want. As long as you go first, you can just use Sketch the turn after seeing the move you want, provided you survive. Which leads into...
(2) Focus Sashes are your friend, if your Sketch pokemon doesn't have the bulk (or levels) to survive being hit by the move you're after. They're a bit expensive - 14000 each from the shop in Tsukinami - and they get consumed on activation, but they really do make the task a lot easier. They are also a big help if you have no hope of outspeeding your target (e.g., trying to sketch a priority move).
(3) The AI seems to choose its moves based on the pokemon you currently have in, it doesn't ever try to predict a switch. So, if your Sketch pokemon is faster than the target, you can try to bait the AI by sending in a pokemon that will tempt it to use the move you want (e.g. if it's an attacking move, use something weak to it, etc), immediately switching out for your Sketch user, and using Sketch first thing the following turn if it cooperated and used the move you wanted.
(4) It helps if you find a source of the move that you can save before encountering, so you can reset if it goes badly (so if there are multiple places you can encounter a move, it might be worth taking this into account when you're trying to choose one). This is especially true in some cases, like I encountered with Atomic Punch and Synchronoise, where there might be only one source of the move that can only be fought once per save file. Go in with a plan, but make sure you save first or you may not get another attempt.
(5) Not every move is available in the game right now (and a move being available doesn't mean you can Sketch it either: there are some obtainable pokemon that never appear in the wild and no trainers use against you), and some are much harder to find than others. Take this into account when developing your moveset, especially because some moves can take a lot of work and be frustrating to successfully Sketch.
So why use Sketch?
There are two main purposes, as I alluded to earlier.
The first is to put together unique movesets; having Sketch gives Smeargle (and, in Uranium, Raffiti) access to nearly every move in the game, and certainly more moves than any other pokemon can hope to have access to! Whatever cool or broken or quirky interactions between moves you can think of, you can have it. (In theory, anyway; admittedly, right now in Uranium we're limited to moves we encounter opponents using, but hopefully that will eventually change.) The most notable and splashy example right now is probably the Nuclear moves: Raffiti is the only pokemon capable of learning Nuclear-type moves while not being Nuclear-typed itself (and therefore foregoing its weaknesses).
I've shown one such moveset above, and the methods I used to actually learn it, but that's far from exhausting the potential. There are so many more things you can try.
The second purpose is for breeding. If you want an egg move but can't find a parent that can learn it, you might be able to use Sketch. It's easiest for pokemon in the Field egg group, which Ratsy and Raffiti belong to (also Smeargle, in the official games); I know in the official games there are plenty of breeding chains that have to begin with a Smeargle in order to pass the necessary moves, but I haven't exhaustively checked Uranium yet. As long as you use a male pokemon to do the Sketching, you can use it as the father and pass the moves on to children (if you use a female, all the children will be Ratsy, and the Ratsy species has no egg moves so nothing will be inherited). That said, Uranium uses generation 6 breeding rules, which means that both parents can pass on egg moves to the child (unlike earlier games where it was limited to the father).
Further notes on egg moves: Generation 6 also added the ability to relearn egg moves with the Move Reminder. I did a bit of testing and this functionality is in Uranium, but the opportunity is lost on evolution, which I think is different from the official games (I took one of my Synchronoise Eevees, overwrote Synchronoise and tried to relearn it. As an Eevee it could relearn it fine, but after evolving it to Flareon Synchronoise was no longer present in the menu at the Move Reminder). This may or may not be a bug, but regardless that is how it works right now.
I have not made anywhere near an exhaustive examination of Uranium's egg moves and which ones are potentially accessible via Sketch, outside of demonstrating Synchronoise and looking into the possibility of getting Volt Switch on Volchik (which is not currently possible as far as I can tell, and wouldn't be particularly helped by Sketch since Slothohm learns it naturally and is in the same Field group as Ratsy; there's no intermediate pokemon to use for chain breeding).
And if anyone wants one, I have quite a few 5IV Ratsy left over (outside of the ones I dumped into Wonder Trade a week or so ago; OT Aristea if you saw any of them), and some Synchronoise Eevees, and it wouldn't be too difficult for me to make more if I run out. If they'd help with any of your projects I'm sure we can work out a way to get one to you.