Listed in order of preference.

Game Name: Persona 4 Golden
System: Playstation TV
Category: Hard Mode, Golden (Truest/Canon) Ending
Gameplay Estimate: 11:00
Ending Estimate: 0. Final boss kill is 15 minutes before the end, but the traditional end timing is on New Game Save
Pros/Cons:
+ Most popular Vita/Persona game casually, and the series is quickly gaining popularity. Limit Break will be shortly after Persona 5 English release as well, so relevant timing
+ Storied history of speedrunning and modern relevance. Initial P4 route had names like Neviutz and Romscout working on it. Vanilla P4 has had two new runners in the last year (Cereth and Oxwas), while P4G has had an explosion of runners from 2 to 5 in the last year alone, with others contributing strats and ideas.
+ Although SMT and Persona are known for their difficult battle system, P4G is completely marathon safe. The player is favored in every fight with speedrun strats, and death is not very punishing.
+ Incredibly deep combat system combines a deep weakness system, unique and interesting mechanics and a demon fusion system allowing us to tailor our character to best handle every scenario he’s in. Particular highlights include: Shadow Yukiko, Shadow Kanji, Shadow Mitsuo, Kusumi no Okami and the Final Boss.
+ Nearly endless incentive and prize potential.
+ I’d like to put an emphasis on the show during the game’s downtime. If it gets accepted, I’d like to reach out to Atlus, voice actors, people involved with the games music etc as potential 30 minute skype interviews to fill gaps in the action.
= Due to the game’s length, splitting the run into two parts may be required or preferred. There are 4 potential places that I’ve noted might be smart to segment the run. Incentives can be balanced across both of them. The places are on Shadow Mitsuo boss kill (ends on the best boss of the first half), between the two (still ends close to a great boss while splitting one of the between dungeon segments), on Shadow Naoto kill (Would end on a boss and would mean that the second half would have zero grinding commitments) and between Naoto and Heaven ( Similar to on Naoto kill, but splits the largest segment of cutscenes into two pieces).
_ While the ratio of downtime to dungeon crawling and combat is actually quite decent for an RPG (approximately 60/40 dungeon crawling to cutscene ratio), the downtime is very segmented from the rest of the game, which means that there will be gaps that need to be filled with strong commentary, donations, interviews etc.
_ Skipping cutscenes involves a VCR fastforward sound that on a vanilla copy of the game, cannot be turned off. I have reached out to the Vita hacking community about potentially helping me to create a hack that either removes it or replaces it with a song in the game. Unfortunately, there are quite a few sound files in the game, so isolating it may not be possible.
_ Some of the grinding has a fair amount of variance in the time taken to reach the necessary leveling milestones. The estimate takes this into account.
Donation Incentive Offers:
[hidden]Naming the Main Character (8 characters for each first and last name) – Self explanatory. Costs no time, other than inputting it.
MC’s Girlfriend/Best Friend (Can choose from any party member) – While it hasn’t been fully tested, the worst case should add about 15 minutes. As a bonus, I can take them into the final fight without affecting the strats used.
Date or don’t date Marie – Marie is a very polarizing character as the major addition to the Vita port. Costs no time and doesn’t interfere with the above incentive
Costume – Rolling  incentive. See the GameFAQs thread here: https://www.gamefaqs.com/vita/641695-persona-4-golden/faqs/65828 - most of the costumes that are not purchased can be obtained with minimal time/money investment.
Ending – We could start as a Bad End run and push up through Golden ending as certain milestones get reached (Bad > Normal > True > Golden). Each ending is about an hour added (8 – 9 – 10 – 11 in the estimate for each). This incentive likely works best if the run were to close the marathon (i.e. X donation dollars for the total marathon to extend the run)
Difficulty – similar to ending, but must be locked in at the start of the run. Very Easy > Easy > Normal > Hard. I can’t quantify the difference in time, other than that in theory, Very Easy would be a much shorter but less interesting run.
Dub (English vs JP) – with the undub now being created, it’s possible to have a bidding war between English and Japanese dub. This would cost no time but would not be a rolling incentive.
Remove the bzzt – This is still being researched, but using similar tech to the undub, it may be possible to remove the bzzt effect from the TV fastforwarding to be replaced with something else (two working ideas are a single song from the game or muting so that external music like remixes of the game’s soundtrack could be played, using a footpedal or keypad to toggle on and off.)
Dancing All Night – This would take 10 minutes max – would involve booting up a copy of Persona 4 Dancing All Night and either playing a high difficulty song or having the whole couch do a dance along. Or both. Could also be used to fill time if the first half ends ahead of schedule. https://www.twitch.tv/ghoul02/v/40006132
MC Cosplay – Self explanatory.
Gone Skiing – Can choose the party member that we go skiing with. Adds negligible time.
Sing the Junes theme– Every time the Junes theme plays, the runner and the couch has to sing along.
Attempt the Rainy Day Beef Bowl – A classic trip to the meat dimension. Normally not done in a speedrun, would maybe add a minute?
Play some animated cutscenes – Quite a variety to choose from. Normally skipped in a run, but some of them like the Concert would have an audience for sure. Add approximately 2-3 min per, depending on which. One of them is a song that the couch could sing along to. (https://www.youtube.com/watch?v=6R0vUAkEeZQ)
Yosuke’s weapon – Yosuke is supposed to use a pair of daggers, however, due to a few necessary quests for the run, we can have him choose between a pair of daggers, a pair of wrenches, a pair of fish, and two ears of corn. None of these would cost time to get or use, and he’s used for quite a few of the end game bosses.[/hidden]
Sample Run Video:


Game name: Ni No Kuni: Wrath of the White Witch
System: Playstation 3
Gameplay Estimate 7:15 (Will be adjusted soon)
Ending Estimate: 9 minutes of good music and art. I’d prefer to not skip it.
Pros/Cons
+ Popular game that combined Studio Ghibli (My Neighbor Totoro, Spirited Away) with Level 5 (Dragon Quest 8, Yo-kai Watch, Dark Cloud). This game was a spiritual predecessor to Yo-kai Watch and the similarities are plain to see.
+ Speedrun has an assortment of fun tricks and glitches, including menu storage, abusing spell text and jumping to avoid unnecessary encounters.
+ The game is seriously stunning. Great music, great art.
+ Combat plays out like a cross between Pokemon, Tales of, and an MMO, taking qualities from each. The strategies are decently varied and most bosses are very plainly different in how they’re tackled. There’s a counterhit system, which encourages kiting attacks and hitting right after they come out, which adds a very cool skill barrier to the run.
= While the story is relevant to the charity I think (Deals with loss and depression), most cutscenes are completely skipped. This does very much help the pace of the run.
_ Some of the bosses are true damage sponges. We’re very underleveled at times, and party AI is terrible leading to one boss in particular being run away and spamming ranged spells.
_ Gold Glim variance means that not getting ultimate attacks in fights may cost a few minutes. This estimate should be safe with bad luck.
_I’ll need to have a backup save for one gambling segment. We need large amounts of money to clear the endgame, which means gambling and abusing bad card AI. However, with the worst luck, we can lose both of our attempts which would force reloading a save either created before (could waste time if still unlucky) or loading a premade save (which would lose naming incentives)
Donation Incentive Offers:
Wizards Book name, naming Mitey, the monkey and the penguin.
2 players, 1 controller for the test of courage
English vs Japanese voices (Can be rolling incentive but I believe requires going to menu each time, so only switchable on saves)
Sample Run Video: I’m still learning this (Been slowly learning it over the last 6 months) and will have a finished run before AGDQ ends

Game Name: Final Fantasy Crystal Chronicles
System: Nintendo Gamecube (but please let me play on a Wii)
Category: Any% 2:45 (English)/2:38 (Japanese)
Ending/Estimate: 8 minutes, but the music is my single favorite part about running this game.
Pros/Cons:
+ Amazing music, and a decently fun Gauntlet style spinoff of the Final Fantasy series
+ The boss strats are very deep, including quick kills, abusing AI etc while being completely underarmored for the portion of the game we’re in. We’re a glass cannon, yet the game is relatively marathon safe, thanks to a forgiving continue system and Phoenix Downs working as a mid-fight revival item
+ 3 out of the 4 races are completely speedrun viable, each with a set of pros and cons. I’ve been competitive with two of the 3 and completed a run with the third for fun.
= English works without a glitch that makes the run much less safe from the early to end game, while Japanese abuses one of the odder glitches, requiring us to play in solo multiplayer mode for the first area and duplicating money using the GBA. This means Japanese does have less variance
_ Much of the time is down to luck for a few specific item spawns and world map “encounters”. The marathon estimate is very safe and takes this into account.
_ Done at RPGLB15, and did a very good showcase with the only exception being a disk read error that caused the final boss to be uncompleteable.
Donation Incentive Offers:
Singing Sound of the Wind/Moonless Starry Night – Probably two of the best vocalized themes for any video game. I used to boot the game up as a kid and roll the opening just to hear it. And the ending is just as good.
Race – Clavat/Lilty/Selkie are all viable, with about a 3 minute difference at top level, and each has unique pros and cons.
Character and town name – 6 character limit I believe.
Sample Run Video: https://www.youtube.com/watch?v=Phkd815kW8w (I promise, I can beat the final boss)

Game Name: Final Fantasy Theatrythm Curtain Call
System: Nintendo 3DS
Category: Best of 5 Tournament finals/showcase
Estimate: 15 minutes, but can be lengthened or shortened to fit
Ending/Bonus Estimate: None
Pros/Cons
= Only submitting as a bonus/filler run. Would be a great, flexible and fun way to fill time should the schedule get way ahead.
= If accepted as bonus, I’ll schedule a time and date to get who will compete on stream. I’ll also recruit some e-sports style commentary to go along with it to make it just silly fun.
+ Where else are you going to have a potential bidding war between Dancing Mad, One Winged Angel, and Swords of Fury?
+This game is seriously fun, and Final Fantasy music is always amazing.
_ Not a true speedrun (And the RPG elements will be less present as well)
_ 3DS setup can be relatively involved, which might limit the appeal of using it as a quick stopgap between runs. Also due to the way save files are managed, would have to be done on two level 1 accounts for parity.
Donation Incentive Offers:
First song bidding war. Can be any Final Fantasy Song in the game.
Sample run video: https://www.youtube.com/watch?v=dgLdXLarppI (obviously not mine. Note this does work with a single screen and audio capture.