First a quick note: you can teleport out of getting mirror fragments, but it just softlocks the game for Zonplas and Ud-Meiyu, and for Orjugan even though you can duplicate the fragment (and get infinite Capla from Olha as a result), can make it so you can pause during its battle (teleport anywhere that isn’t Lake and enter the battle from the normal route), it still seems useless. Teleporting after it dies skips the death cutscene if you do it late, and respawns Orjugan if you do it early so infinite Orjugan% is also a thing.

What I really wanted to mention was the bat farming for Emel. If we consider each extra room visit takes 17″ (rough estimate that I really should try and get more accurate Edit: 14″ is more accurate I think, no magic use), and each visit yields 1000 Emel, getting Brillante to lv10 from lv9 is an extra 3′07″ (11 visits, maybe 10 with good RNG). Does the extra level really save all that time against Orjugan and Lanaluna (and Romun fleet / Kishgal door bosses)? This needs investigation. On a similar note going for lv10 Ericcil instead would be + 51″ (3 extra visits), so I really don’t think this is a viable route. Even if Ericcil saved that much on Lanaluna (I don’t think so), it would still lose time at Orjugan and Kishgal, albeit maybe not very much. Edit: Just checked my run and it took me 2′20″ to farm for lv10 Brillante, I suppose 17″ for each visit was a bad estimate. // It takes the JP runner only 1′40″ to go from 4k Emel to 14k… damn maybe my farming is just slow, lv10 Brillante might be worth it if you’re fast.

I just tested fighting Orjugan lv37 (267HP) and lv9 Brillante… it’s very much doable. Still only takes 3 magics + a few normal hits to kill. I have no idea why I ever went for level 10. 2022 Edit: You need to go for lv10 because it takes forever to get lv37 anyway, so it’s not worth it to leave before grabbing a few more Emel. Also confirmed that while Lanaluna is more annoying and takes longer, is still very much killable with lv9 Brillante (only used 2 Herbs) and at most lv10 might save like, 1 minute maybe? I don’t even think the entire fight took me 2 minutes. Edit: Got 1′30″ with all Luck equip lv9 Brillante lv38 (273 HP). In my run it takes me ≈ 53″. Edit: 1′08″ all STR equipment, even without coin. Edit: 1′05″ using the coin instead of Wooden Bracelet with Galba Shield. Btw all these times used the Emel Mallet, so there was no Luck/STR mixing, I just tested full Luck vs. full STR (except for the last one). So: maybe Lucky Silver Coin is slightly better than Wooden Bracelet against Lanaluna, but how long does it take to get it…? Not getting it might be faster in the end depending on how much it costs, it doesn’t seem to save a ton overall. Getting the Lucky Silver Coin costs ≈ 40″ (without considering equipping it) and I honestly kinda doubt it saves all that time back. Who knows.

Kishgal grind: let’s say 6″ to kill 3 bombs and go back in/out (JP run time, he does it well) 1004 EXP × 3 3012 × 6″ Piana-Pullus is 29030 29030 ÷ 3012 ≈ 9.6 6 × 9.6 = 57.6″ teleporting to Port and killing Piana-Pullus is ≈ 40″ if done first try (lv44, Brillante + Ericcil magic and a couple attacks) 20″90 Gaposasura kill (lv41 Brillante lv10) Edit: 17″55 on a good run 9″16 getting back on track 6″57 going to it 36″63 total not grabbing the chest 27858 EXP 1205 bomb × 3 3615 27858 ÷ 3615 = 7.7 × 6 = 46″2 2022 Edit: I had the wrong Gaposasura EXP value. Correct: 26098 ÷ 3615 ≈ 7.2 × 6 ≈ 43″3 so Gaposasura is still faster, it took me ≈ 40–41″ in my WR with item both Gaposasura and Piana-Pullus are faster EXP than the Kishgal bombs it seems. There could be something wrong with my math, though… Also math’d that there’s no need to have high lv41 EXP before Gaposasura, you get all the levels you need without even killing a bomb, except for lv50 at Ark but as shown by JP runner that’s pretty irrelevant (this is with Starlight Medal). lv49 good enough, and depending on how much lv41 EXP you had you reach lv50 anyway. It seems the Ark quickkill only requires 3 Capla (lv49), so you can use 1 on Ernst I guess. Not fantastic… but better than nothing I guess. 2 Capla are almost enough actually, only takes 1 Brillante after the Ericcil spam. At level 50 2 might be enough.

…I honestly don’t know. There used to be an official site but I guess that’s been replaced with DX now.

Well there is a new run but it’s not from me!

I don’t think I saw anything new in it other than the skip near the end which saves ≈ 10–11 seconds only. Edit: at 56′35″ he also skips a cutscene where Ord is wounded by teleporting out of it! 31″ save there. Rest of the run has a whole bunch of differences compared to mine but I don’t think there’s anything too relevant that I didn’t already know. Oh yeah he gets double hits on Demi-Galba which I thought were too inconsistent to go for but apparently it’s possible. Edit: Oh and he annihilates the Ark in like 3 seconds with Capla even being 1 level lower than me. That’s probably faster than using it on Ernst, but makes the run much harder as Ernst is difficult to fight without it… Edit: Also he only kills 2 bees at the start and goes straight to Grana-Vallis from there. Considering he does NM route that makes sense as you don’t need the 5 Emel at the start of the game. I suppose killing 2 bees should be faster than 3 plants for the level, but killing plants is also pretty quick so who knows… and 3 plants are slightly more EXP I believe but probably doesn’t matter. Hmm it might be possible to kill 2 bees at the same time when they’re in a group though, and that might be really fast. Edit: You start the game 150 NEXT, plants are 50 EXP, bees 77 which × 2 = 154 more than 150 from 3 plants. Edit: Timed first plant room as ≈ 9″ without kill, + 1″62–1″78 with kill, and killing 2 bees is taking me long… 7″50–7″92 no kill and 14″19 2 kills even with a crit, I’m sure a TAS would go for bees but for a human 3 plants look faster to me. Just got 13″08 with a crit and hitting 2 bees at once while moving towards the correct exit (upright), super lucky. Even this might not be faster than 3 plants, lol.

Edit: Looking at vids I calculated the strat where you take damage before Geis loses ≈ 2″10 in the room, but thanks to RNG gains it back at Geis. 2 extra hits from Geis can take just ≈ 1″90 with perfect RNG, but that almost never happens. Much more often it’ll take 3+ seconds.

Going by IGT sounds like a complete mess. It would require a ton of math for each run and it would preclude players from using the strategy of quitting out of a mission right before it ends.

What categories are we talking about here? I don’t think it makes any sense to go by RTA for this game anyway except for Missions where I guess bullshit loads would still affect it…

“Why don't the rules just say "early exiting is not allowed"?” It’s a completely legitimate technique that cuts off minutes from the run if done perfectly. If anyone is serious about running the category I don’t see why they’d prefer disallowing the technique and be slower. The only inconvenience with it is that you don’t get to see your individual Mission times unless you exit out late enough.

I think the rules for this category need amending. The fastest way of completing missions is to exit out right before the end (the game keeps playing in the background for a few seconds and allows you to beat the mission anyway). The only way to know whether you’ve exited out correctly is to have a clean save so you can tell whether you’ve gotten a rating on the mission or not (exited too soon).

Here are my proposed rules: if you’re gonna use the exit-out trick, you need a clean save. Otherwise, it doesn’t matter (I think a completed save that doesn’t exit out missions is slower than one that does). I think the only way to get a clean Missions slate is to delete all of your data, which does suck balls. There’s probably an AR code to do it too, and people with flashcarts can always backup their save, but to force people to run on a clean save seems a bit much.

I don’t know if cyghfer reads this but I don’t know how to get to him so I’ll just say it here, has anyone tried letting Sturm capture the central Airport? I’m not sure if Andy’s Missiles are OHKO on Sturm Bomber/Fighter (whatever it builds idk) but I remember that being part of the fastest win strategy (5 days I think?). You let Sturm capture it so it wastes all its money on a Bomber or something and with Missiles positioned there you just destroy them as they’re built.

Also to cyghfer specifically: when you’re down to moving your last unit, you should press L to select it faster instead of manually moving the cursor all the way to it. Generally it’s down to routing to figure out when L is faster but when you need to select your last unit it always is and it’s easy to see even impromptu.

Just do a run, then ask for the category to be created so you can submit it. I guess it won’t necessarily be created even if you have a run to submit, but I’d say that’s the first step. An empty category is not gonna be created so you need to show a run for it first.

I just noticed that all the obsoleted runs were set to “w/o loads” too, when that isn’t true. Could that be fixed?

In case a mod could fix my obsoleted runs’ video links: 15′35″ https://www.speedrun.com/ykm/run/9mekpo2m 15′41″ https://www.speedrun.com/ykm/run/7z0elvjm 15′53″ https://www.speedrun.com/ykm/run/2yw16v0z 16′00″ https://www.speedrun.com/ykm/run/8yvg03ey 16′05″ https://www.speedrun.com/ykm/run/8y8qex1y

How were all the runs retimed? That must’ve been a lot of work? I was kinda expecting my run to still be WR after load removal but I guess I was wrong. Darned lake skip.

This game was in really bad need of a load remover so thanks for that. Maybe I’ll run it again in the future.

Edit: Is there like somewhere the community for this game resides? This sudden takeover by unknown people surprised me.

Trying to test a new route which is basically NM route but improved thanks to the weaker Normal enemies. Skip spiders, kill flowers until 52 Emel (you’ll get to level 10 or almost, 111HP doing this, can destroy bombs). The flowers themselves take ≈ 1′30″ but it’s an RNG hellhole. Have had 1′38″, 1′27″24, and 1′37″, 1′18″20 with Galba Shield. Without it’s literal hell, almost impossible to do without dying, but with good RNG farming on the first runthrough should be a lot faster than NM route. After the farm you grab the Wing, teleport to Port, get lv4 Livart, grab Galba equips, teleport to mid Grana-Vallis. You now have lv4 Livart for grinding/farming, can do grinding in the Wing room if needed. I don’t know if this is faster or not, 1′30″ is kinda a long time to farm when you’re only saving ≈ 1′10″ from skipping spiders and lv2 Livart. But having lv4 Livart from lv10 for grinding/farming should save some nice time (and should make grabbing the Grana armor unnecessary, even though it might already be with good play). At lv10 with Galba Shield and lv4 Livart you 2-shot bombs, lv11 (117HP) OHKO. Already at lv14 (134HP) you kill moles very quickly (5 hits). At lv15 (140HP) it’s 4 hits. So I tried going through Grana while killing bombs after the flowers farm, by the Wing I’m already lv16 (146HP). I think this route is faster but it’s insanely hard and too complicated to time without doing an actual run.

Edit: timing from flower room to 51 Emel (rip) 1′52″92, finally did it without cheating I think. Don’t feel like that 1′27″24 was legit. lol Took me like 2 hours of attempts, good luck doing it in a run. I’m lv9 106HP with 1097 Next. Takes 1 bomb to level, requires a deep hit to kill I think. After that I guess you can keep destroying as many bombs as you can, I got up to level 13 129HP now (missed some). Got carried away grinding bombs and leveled to 20 before Zonplas, rip. Forgot there was still farming for 140 Emel to get Brillante lv5… that leveled me. Also retimed early dive strat, getting pretty consistently close to 1′30″ at the flowers. It seems slower by maybe 10 seconds: early dive loss: LC entrance to tp spot ≈ 1″, going back after Galba Armor ≈ 45″ (− ″94), going to mid after slimes ≈ 3″ = 48″06 no early dive loss: farming flowers = 20″?, grinding a few bombs maybe? = idk (probably very little), extra tp in LC ≈ 10–11″ = 31″, maybe some seconds more

also straight to Wing would lose 10–11″ from extra Grana entrance tp, like at least 10″? (got 12″21 in test but was bad and stopped a bit late) going from entrance through the flowers path, save 20″ farming flowers… wait a second, is it fucking faster? May not be faster for TAS, but it might be in human play wow… so there is no NM route improvement, NM route is perfect already. lol Low level flowers farming is just too slow/hard…

Also I really don’t know whether this route is faster than the old one I use in my PB. There are many improvements that could be made on my PB’s route like triggering LC tp, leveling on flowers instead of spiders (≈ 14″ vs 20″), not grabbing Banded Mail and probably more stuff.

Edit: OK so following NM route, you should be lv9 when teleporting to mid Mountain, you kill the bombs on the way and grind at Wing room until lv17 (152HP). You then go for the moles and 140 Emel, doing this I got to lv19 with 157 Emel. I used the Rehdan Shield. /// Eh tbh Wing room is SUCH good EXP that lv18 (157HP) might be worth it, fighting Zonplas at lv20. The extra level saves ≈ 20″ at Zonplas I think, but then you also have just overall more EXP and are higher level going to Ud-Meiyu etc. /// It took me 9″42 to go from lv17 to lv18 in Wing room. LMAO lv19 takes 19″22 though and you still end up lv20 for Zonplas so it seems optimal to get lv18 only. 2023 Edit: In my PB I barely get lv20 while grinding to almost lv19 at bombs.

No idea why I didn’t have Zonplas’s EXP in my calculator but after looking into it, turns out my routing is overleveled. lv EXP 18 8404 19 7004 20 5836 21 4864

My run fights Zonplas at level 20, so it misses out on 2568 EXP. My run prioritizes EXP to get to lv20 to fight Zonplas, but it might be better to instead prioritize Emel and fight Zonplas at lv18 with maybe Livart lv5 instead of lv4… testing needs to be done… Tests: 50″24 lv18 lv4 cheating Zonplas in place 32″54 lv19 lv4 but I think I had 2 crits 20″60 lv20 lv4 again 2 crits 50″68 lv18 lv4 no crits lmao this luck 30″00 4 crits lv19 31″02 huh weird only 1 crit :/ 24″16 lv20 no crits but not perfect execution if you remove RNG it doesn’t seem worth it to level to 19, since it took me 37″24 on bombs. (Edit: 33″08 even.) Haven’t looked into how the EXP ends up though. 1′42″79 15 crits lv18 using Rehdan Shield, clearly inferior. lol Even deal 0 damage often. 45″20 lv18 lv4 with 2 crits. Killing at lv18 leaves you at lv19 with half or a little over half EXP. Killing at lv19 leaves you at lv20 with 1/4 EXP. Killing at lv20 leaves you at like, 1 enemy kill from leveling up.

Everything at Grana-Vallis seems to give 5 Emel which with Mallet is ×1.5 so 7, but flowers stay at 2 (1 + 1) due to flooring. Fireflies being the easiest to kill once you’re level enough, though moles may still be best due to combination of how many there are and the EXP… Edit: Moles seem to have a higher drop rate than fireflies too. Tested with full Luck, moles drop 100%, fireflies not quite. Even with Rehdan Shield the fireflies drop basically no Emel, and the EXP is crap compared to moles. Moles still seem best for Emel (bombs are probably better EXP, so they’d be good for Nightmare).

Tempering: lv Emel 2 5 3 13 4 34 5 88 6 230 7 600 8 1,540 9 4,010 10 10,900 11 100,000

I have no idea why I ever fought Zonplas at level 20, it seems totally overleveled. 19 might be optimal, or maybe 18. Edit: Might be leftover from when I didn’t grab Galba Shield, at that point level 20 might actually be good.

Thinking of a skip spiders, straight to Galba Shield, level twice on flowers and go as usual route. Timings:

15″ from Port tele to Limewater entrance 25–26″ from Limewater entrance to Grana-Vallis entrance 4″03 from Port tele to wolf screen 7″90 first wolf screen, up 6″69 first wolf screen, left 4″44 from there to Limewater entrance 6″18 first wolf screen from Limewater side, up 7″67 second wolf screen, up 6″87 third screen to Grana-Vallis entrance 1′33″58 foot diving for Galba Shield only 46″ from Limewater entrance to Galba Shield (right after dialogue box) 47″58 from Galba Shield to exiting Limewater (makes sense…) 33″ going from Galba Armor to midpoint 37″40 going from first room grind to midpoint (Livart) 35″63 same thing Ericcil 30″06 from Galba Shield to midpoint Livart takes 15″90 to grab 2 levels from Flowers (to lv7) with godlike execution and it’s enough to destroy bombs 2019 Edit: normal way: lose 4″ from Port tp to wolves, lose 1″72 wolves room going up, lose ″94 or whatever from LC entrance to tp (probably actually 1″ or so) = 6″66 LC tp trigger early: lose ″94 going out of LC, lose 4″14 going to wolves = 5″08 I believe triggering the LC tp early is faster.

Also in my WR it took me 25″ to go from lv6 to lv7 on the spiders, so I could shave off some seconds if I disregarded them and went for the flowers instead. Requires really good Emel luck though, you still need 5 by Port Rimorge for lv2 Livart in the normal route.

Edit: I tried doing Grana-Vallis with the skip spiders, get Galba Shield route, and it took me 6′25″ from the flowers room to getting the Emel/level requirement (187Emel/169HP) kinda, it was all messy and I actually died on the last wave of moles. You do slightly less damage with Galba Shield than lv2 Livart, and also can’t use Rehdan Shield for the extra Emel drop chance so you need to get lucky and hope you have enough Emel before you level (I cut it real close in my attempt). No idea what this means yet but it’s written down now. Edit: Welp the same segment in my WR takes like 6′10″. :/ But wait a second, if I’m saving 1′14″ from skipping spiders, does that mean this is faster by like a minute? I think this route loses like a minute before Limewater Cave in WR: 1′14″ saved from spiders but in my WR I could’ve saved ≈ 5″ leveling with flowers from 6 to 7; 1′55″ lost to Limewater diving (like 45″ behind), then Grana-Vallis is like 15″ slower so 1′ behind but the WR still has to Limewater dive. So it would lose 14″ going from forging Livart to wolf room, + 6″69 going left, + 4″44 entering Limewater Cave, then 46″ for Galba Shield, then 28″76 to get to the midpoint (this saves time as it isn’t done in the other route)…other route would take 35″63 to get to the midpoint right after the farm/grind, so WR loses like 1′10″ meaning new route is 10″ ahead, but then it loses 7″ going to midpoint… jesus, this stuff is close. I need to grind Grana-Vallis some more with the new route to see what exactly the time is. Well damn… 6′34″44 new route Grana-Vallis with 147 Emel, but I noticed that for some reason I don’t have the Emel Mallet on this save so probably had enough Emel with it. But still, so slow… damn. :/ I question the grinding of moles in the last room though, there must be a better way. 7 moles killed in 20″46 = 1638 EXP Wing room good, should be able to get 7 bombs in 7″ or better = 3465 EXP in 21″ no Emel though 37″24 going from lv18 in this save (almost no extra EXP) to lv19 from bombs in the normal room (not Wing) much much slower with moles, 47″ and still no level Edit: I even got a 33″08 with decent execution at bombs. Can’t find a strat for Wing room. The one the JP WR uses requires Livart lv4…

Clearly my WR is a billion years old and tons of new tricks and such have been found since. I just tried a route where you forgo killing spiders after Demi-Galba and just go for the flowers until you’re high enough to kill bombs, turns out getting to that level (9) takes 1′05″ on a good run which is technically slightly faster than grinding/farming spiders, but the lv1 Livart is ass for killing the moles so it probably loses time in the end, also it’s harder than pulling off Nightmare 100% deathless so I sure hope it’s not faster, I don’t wanna do it.

I looked at my Normal/Hard/NM runs and calculated more or less if the starting NM route would save time for Normal too, it seems to be very slightly worse but it’s hard to say. Stuff: 11′14″ v. 14′51″ Geis black screen 14′02″ v. 16′25″ exit Port = 2′48″ v. 1′34″ = −1′14″ 23′49″ v. 27′36″ exit peak = 9′47″ v. 11′11″ ≈ +13″ 13′40″ hard 25′25″ with 50″ death = 10′55″ so skipping spiders and lv2 Livart saves ≈ 1′14″ but then you lose too much going for the Wing and teleporting I guess… Note that my Normal run gets the Limewater mid teleport and picks up both the armor and shield at Grana-Vallis I think, and the other two runs don’t.

Night63 and I found out like yesterday that the Emelas Crown has a secret + 4 Luck bonus, and an all-Luck equipment might be best for the speedruns, you get a lot of crits. Rehdan Shield +3, Emelas Crown +4, Lucky Silver Coin +6. For Zonplas Galba Shield is a lot faster though, not sure how crit dmg calculation works. All-Luck seems good for the endgame though.

Lots of things to investigate and test, I’ll try to fill up my EXP calculation page with more minibosses and stuff to see if any grinding could be done better.

I assume you’re talking about 32-track speedrun since you mention Keverage.

Charaters affect weight, which affects offroad speed. Dry Bones is the lightest character and as such is the preferred choice, his speed offroad is the fastest. You get pushed around by CPU much more with light weight though, but unless that’s a really big deal for you you should just pick Dry Bones. If you’re PRBing then offroad won’t affect your speed so a heavier character can still keep up if you’re PRBing.

The best kart depends on your skill level. Entry-level is Dry Bomber, if you’re better you should use Egg 1. Polt and BLS might be better but they’re very hard to use so not recommended.

For Time Trials DB/E1/Polt/BLS are still the main karts but there can be some variety, you can check WR history at https://mkwrs.com/mkds/

@JustinBailey I’m not sure if you were talking about Wi-Fi or TTs, but for TTs Egg 1 is almost always better than Dry Bomber. Speed is the most important stat, even without keeping boost all the time the BLS (kart with top speed) is the best TTs kart by far, higher speed is just better when you’re good enough to handle the more awkward handling and shorter MTs. For Wi-Fi some people still use Dry Bomber, and it does have a lot more advantages compared to in TTs but Egg 1 is still widely considered superior if you’re good enough to handle it.

A lot of TTs TASes have been made and they almost always use BLS. Some mission TASes have also been published (much more haven’t been published yet). You can check them out here: https://docs.google.com/document/d/1UGioxOY-ronVo_XPU9rH2AA7PO-bkUdgXzMoE4t06Us

For human TTs WRs, you can check the history here (though it’s a little incomplete): https://mkwrs.com/mkds/

For overall TTs leaderboards there’s the Players’ Page here: http://www.mariokart64.com/mkds/

War Room records aren’t speedruns, you can find those on Cyberscore: https://www.cyberscore.me.uk/game/56

That map just looks like a clusterfuck to me. Just from looking at it I don’t feel like you can blockade the bridges…

Yeah I’d prefer that, I don’t even know where I put my VBA and idk if I even have the AW ROM. It’s also just a lot easier to analyze videos than input files on an emulator.