The Grand Memory Clobber is a glitch in Dragon View resulting from uncleared memory that is set after killing the first boss of the game. If the player doesn’t reset the game at some point after killing the first boss, the Grand Memory Clobber will do all kinds of crazy nonsense when you reach the dragon cutscene in Ortah. The outcome of the crazy nonsense is dependent on the values of every single byte in RAM, and usually crashes the game; in super rare circumstances, it will call the credits. This run was done in real time, single-segment, on console.
There are no other real-time runners of this game, but I need to give a special shoutout to Omnigamer for looking into the glitch at my request and providing me with loads of fun info, and to Khaz, the TASer. Khaz dumped a perfect map from the game’s code in addition to sharing some optimizations, glitches, and route considerations and making an incredible GMC TAS of this game. For additional reading, I would recommend reading the TAS submission notes and TAS discussion thread.
This run uses the Rotayo skip to skip directly to Toronto, skipping the entirety of the first two dungeons. This category also allows resets, which are used twice in this route to manipulate the RNG (for stat-up items and Droog). I believe I am the only one that uses this skip in runs, and it is an enormous skip, so it should likely be split into a different category (the old/other runs should be an “all dungeons” category or something similar).
I will eventually add more info about this game and a detailed breakdown of the route like I did for Robotrek. Please be patient with me in the meantime!
Date to be determined
Esuna had some pretty awful stat growth up to R.Pison, so we had to gain a few extra levels. I save scummed Olvan’s stats at recruitment, and he got really good levels for the entire game, so things were smooth. We captured Perole and brought him back to his throne in the past, as per tradition, and softlocked on Gariso on our first attempt. 1 death at Romus at level 6, 1 death to an Undead crit at level 9, then deathless besides 11 deaths in Saro’s cave. This was a really smooth and enjoyable run, though it felt like I grinded a bit too much. The levels were appreciated at the end, though.
Mario 4: Space Odyssey (Марио 4: Космическая Одиссея) is an original Russian bootleg for the Sega Genesis. I was sent multiple copies of this game by my good friend SireAlligator and did the run on a model 1 Genesis.
I am playing using soundbank 2 (the music for the levels is random; it picks a song randomly from a sound bank, and we can choose 1 of 2 of them) because soundbank 1 hardlocks the game if you break a block or fall in a pit. There is a final boss and ending programmed into the game, but it can’t actually be accessed on console. Gens (emulator) is able to play it, but Bizhawk and console cannot. When you complete the final level, it cuts to the intro again.
This run has no Twitch VOD because I was ashamed of it at the time. I finally encoded and recorded commentary for the local recording, but it’s a pretty terrible run. This is basically my only completed run after the major route overhaul I did, and it missed almost all of the skips and quick kills. I’d guess that an easy 90 seconds can come off of this if someone cares, but I won’t be running it again.
The girl is much slower than the guy in this game (approximately a minute), so they are tracked separately. I did a run of the guy as well, but didn’t even bother saving that VOD. I have no record of his time.
Dagger run. After spending close to 2 years routing this game with every weapon from scratch, and spending several months running this game with the torch without finishing a single run, I decided to spend a week using the dagger just to improve the existing runs. This run is the result of that very, very short grind. Even by 2014 standards, this run was 2 minutes slower than “good”.
I will not be returning to runs of this game, due to ongoing harassment stemming from rumormongers. For your daily dose of SGnG content, I highly recommend the two top Japanese runners, mmmmalcome and dice_ta! They have been grinding this game daily for YEARS. The current best time is a whopping 32:29 by mmmmalcome, but dice_ta is very close behind!
There are several other runners, but I am not terribly involved with this community anymore so I don’t know who is active. I believe the SRC page is updated regularly, so I would recommend checking there.
The Mega Man community maintains their own extremely comprehensive leaderboards, which covers every game in the Mega Man series. This is your one-stop shop for finding the fastest times for these games or new runners to follow. The current best time for MM3 is a 33:42 by ColonelFatso.
I play on the US version because that is the version I own and grew up with; the US and JP versions are tracked together since the time only differs by about a second (in favor of JP).
See the Mega Man 3 leaderboards
The Mega Man community maintains their own extremely comprehensive leaderboards, which covers every game in the Mega Man series. This is your one-stop shop for finding the fastest times for these games or new runners to follow. The current best time for MM6 is a 33:34 by ppotdot1.
I use the “Knight first” route for this right now (still obviously tracked with the other routes in any%). This is a route that was first used by DarkTerrex. While it is a few seconds slower than Wind first, it is much more fun and consistent for me to play.
See the Mega Man 6 leaderboards
This run was played via Nestopia netplay. This uses the “ghost turtle glitch” which allows Mecha Richter to play without the game realizing he is alive. We are able to force single player spawns and boss health for the entire game while still using two players, as long as he doesn’t die.
This run finally beats the best single-player time of 35:38 by omqwat! Omqwat has put in a tremendous amount of work on this game, and was definitely difficult to surpass. I think we’re aiming for a low 34 before we quit.
This Robotrek category is technically single-segment; we do not allow resets due to the impact it has on RNG. I spent several months in late 2018 reverse-engineering the game’s mechanics and did a super aggressive reroute to cut out the crab grab. The listed PB is the result of this. Sub-3:20 is definitely possible (probably even a 3:15 or so, given enough time and good enough luck), but I’m happy with this run. This is the most representative demonstration of Robotrek speedrunning I can imagine.
This is the best time on the NA version, and my 3:14:20 is the best time on the JP version right now. There is no known conversion between the two and the JP version is faster due to text speed. Besides this difference (around 7-8 minutes by my estimates), the two versions are identical.
Big shoutouts to liquid_phase for having the record on the JP version before me, and for being such a fun speedrunner to collaborate with! He and the other JP runners have put a TON of time into this and I expect he’ll take the record back once he learns the new route.
The Mega Man community maintains their own extremely comprehensive leaderboards, which covers every game in the Mega Man series. This is your one-stop shop for finding the fastest times for these games or new runners to follow. My PB is currently the WR for this category.
The JP version of MM7 has a random choice between 3 cutscenes after each of the robot master fights, but the North American version does not. Each cutscene is a different length, so there is no strict conversion between the two versions (cutscene timings can be found here). Gameplay is identical between versions. I use the NA version because I do not want to deal with the RNG, but all the other top runners use JP because it has potential for much faster times.
“Buster Only” is a pretty interesting category. As the name implies, you must use the Mega Buster for the entire game unless you are strictly required to use Rush Coil/Jet to progress. You may use Rush Coil 3 times in this game (once in each of Wily 1, 2, and 3), and cannot circumvent or damage any enemies while it is equipped.
See the Mega Man 7 leaderboards
This combo looks really good on paper, except for the lack of late-game healing. We did really well for the most part, but both characters had consistently terrible stat growth, which ended up catching up to us by the end. Deathless until crash site in the past, 1st try recruitment of Lejes (L14), 12 deaths in Saro’s cave, 2nd try Gorsia. 15 minute long slugfest against Gorsia to win it. Also, Valsu was hiding in the same town as Sky Rune but didn’t have it, so we got a bonus fight against him!
The Mega Man community maintains their own extremely comprehensive leaderboards, which covers every game in the Mega Man series. This is your one-stop shop for finding the fastest times for these games or new runners to follow. My PB is currently the WR for this category.
The JP version of MM7 has a random choice between 3 cutscenes after each of the robot master fights, but the North American version does not. Each cutscene is a different length, so there is no strict conversion between the two versions (cutscene timings can be found here). Gameplay is identical between versions. I use the NA version because I do not want to deal with the RNG, but all the other top runners use JP because it has potential for much faster times.
The following is a list of additional items required for the 100% category: Rush Jet, Rush Search, Super Adapter (via the 4 RUSH plates), Hyper Rush Punch upgrade for S.Adapter, Exit module, Energy Balancer, Proto Shield, Beat, and Auto’s Hyperbolt.
See the Mega Man 7 leaderboards
The Mega Man community maintains their own extremely comprehensive leaderboards, which covers every game in the Mega Man series. This is your one-stop shop for finding the fastest times for these games or new runners to follow. The current best time for MM7 (JP) is a 40:53 by almondcity.
The JP version of MM7 has a random choice between 3 cutscenes after each of the robot master fights, but the North American version does not. Each cutscene is a different length, so there is no strict conversion between the two versions (cutscene timings can be found here). Gameplay is identical between versions. I use the NA version because I do not want to deal with the RNG, but all the other top runners use JP because it has potential for much faster times.
See the Mega Man 7 leaderboards
ActRaiser 2 has a bugged password system which allows me to access difficulties above and beyond “hard”. Full enemy details can be found in the ActRaiser 2 compendium, but in short, difficulty 202 is the hardest difficulty that is possible to complete. In this mode, nearly every enemy and boss instantly kills me on contact and has upwards of 5 times the normal max health for a boss.
This PB has only 1 death! I plan to revisit this to get a deathless 59:XX, but I’m satisfied for now. I got every shadow sprite glitch kill, and executed extremely well throughout the run. This is the world record, mostly because I think I am the only person who can actually complete this difficulty.
Difficulty 202 can be accessed by using the password SWZX LKCF XHLW
100% in Batman Forever is defined as: 100% enemy kills, 100% secrets, 100% blueprints, all possible pickups. The in-game totals will read 100% in all fields appropriate except for pickups. In the SNES version, there are 5 pickups missing from stages 2, 3, 4, and 6, and 3 pickups missing in stage 5. Some of the missing pickups are present in the Genesis version, and some are present in the DOS version. It is not possible to reach 100% pickup collection for these stages on SNES.
In order to achieve 100% enemy kills, it is necessary to deal the finishing blow to an enemy yourself. If an enemy dies by falling off a ledge, taking environmental damage, or taking friendly fire, you will not get credit for the kill and the run is failed.
For more details, please listen to the audio commentary.
Shoutouts to Kandoken for the stage-by-stage guides that I referenced for many areas of my route.
This was a pretty solid run. Only 6 deaths total, one to an early level 12 B.Demon crit, and 5 deaths to the first walk in the past. Kamil joined at 14 at average stats and I didn’t bother to reset for a level 15 recruitment. Monmo was super slow because Fire 2 was only doing 11-20 damage. Entered the past at levels 27/26 and with no mirrors.
My PB in NES Battletoads is unrecorded because I could not get my VCR to work at the time. This was recorded back in the summer of 2010, but I do not remember the exact date. At the time, this was the fastest time, but it was overcome by Icthus’ 16:31 about 1.5 years later. I played using Zitz (2nd player port), and did not use the Turbo Tunnels skip. After this PB, I stopped running Battletoads seriously; I still did runs for GDQs, but did not have any interest in doing serious runs due to my tendency to attract run-ending glitches.
Since then, a load of new runners have picked up Battletoads and pushed the time lower. TheMexicanRunner and JC have been consistently the top two runners among them, continuously pushing the time lower over the past 4 years or so. The current best time is TMR’s 12:42.4, just frames ahead of JC’s 12:42.7.
This run has been absolutely demolished by UnicornSid! Not only did he work in both double ship upgrades, but he managed to find some more quick kills on the bosses and clean up the execution. My eventual goal time for this was around a 22:45, and he pushed out a 22:13! His record is well beyond what I imagined was possible for this game.
All of the NES BioCom times listed are world records. Both US runs use the new barrier skips in area 12, but the JP run has not been updated since that discovery. This game is more active than it has ever been, and I highly recommend checking out the other runners! For those interested in learning how to run this game, I’ll be making tutorials shortly.
While he is not active anymore, I need to give major shoutouts to feasel for getting me interested in this game and teaching me almost everything I know about this game. I also want to credit the legendary Hotarubi for his 15:16 on the Japanese version, which gave me the motivation to grind that category down. Tom Votava and Scott Kessler, other early NES runners, also contributed quite a few strategies. It is because of the tremendous legwork of the early runners and their collaborations that we are able to achieve such fast times now. This should be the goal of every speedrunner.
Pakuliittus (any%)
Koston (100%)
MetroidMcFly (100%)
I absolutely love this game and would happily return to running it again. We run on hard difficulty because the timer is shorter, which allows us to take advantage of some double KO strats at the end of certain stages. Our health also drains faster from attacks, so we are able to abuse the door breaker glitch sooner.
The only reason this category is nearly this fast is thanks to the many numerous strategies that __sdfg found. The current best time is his 15:51.
This is a spectacular game with a lot of potential for speedrunning. This run was the first deathless run of Brandish and the record for about 3 hours before Osse101 murdered it. Hahaha. He has pushed this game far beyond what I even dreamed of. I highly recommend watching his runs. The current best time is his 56:49.
Shoutouts to Breakdown for teaching both of us the route that he developed years ago. Without him, we would’ve struggled to get started.
Elnard is the Japanese version of The 7th Saga, and is significantly easier. This is yet another run included for historical purposes. After concluding my 7th Saga run for Twin Galaxies back in 2009 or so, I spent a few days churning out a route of Elnard, and then like 2 days doing attempts.
In this run, I played as solo Wilme and fought Gorsia at level 33. Most of the grinding was done against Undeeds in Monmo’s cave, and I didn’t do any seed manipulation. This run was on pace for a 4:50 or so, and then lost 40 minutes to the final boss. There is no recording because I rage-deleted it.
I do not think anybody runs this version. I don’t remember if Zakky did runs of this or not. If someone were to seriously run this, I don’t think a 4:40 would be a problem.
This was a fun little casual run that I used to do once in awhile years ago. I never did serious attempts of this, so my PB isn’t recorded. I just listed this run for historical purposes. This run was also done before Dragondarch found the insane experience glitch, which is the greatest glitch of all time. The current best time is an INCREDIBLE 34:32 by Dragondarch.
This category does not use the barrier skip in area 3. I did my run at 1280×720 resolution; the resolution does slightly change your spawn points in the levels, and I believe 1080p would be slightly faster because of advantageous spawns. I could only capture in 720p so I ran that.
Between each run, I restarted the game twice. The in-game timer adds your previous save file’s time to your current run’s time, so a double reset ensures that 0 is being added to our time. The IGT is mostly accurate aside from that, so it could technically be compared between runs. For the 100% category, I had to close the game, manually delete the save data, and then do a double restart because of the way that save data is handled.
While he is not running this game anymore, I need to give special thanks to BagtheTea for all of the help he gave with routing/motivation for this game.
This is a really awful category that nobody runs anymore because it is hardware-dependent. The physics in BCR depend on framerate and resolution of the game, and you can change these settings in the options menu. If you run on very specific settings with the right graphics card, some collision detection doesn’t work right. This run was done on my old PC, which had a low-level AMD card. When played at 800×600 and 56Hz with vsync on, I was able to climb over the barrier in stage 3. This skips areas 5 and 6 entirely.
I do not recommend trying to compete on this category because it causes extreme instability issues and causes the game to crash more than usual. Anybody interested in running this should consider standard any% instead.
This is yet another Twin Galaxies special from 2009 or so. I was not permitted to use several of the beneficial glitches in this run, and I never returned to serious runs after many years of rerouting. No turbo/autofire was used. The best time is currently an incredible 21:42 by Koh1fds!
Koh1fds
Xray662
PacnyTHuk (used autofire)
HalfMoon_Voodoo (used autofire)
This is a run using incredibly archaic rules. This was done originally for Twin Galaxies back in 2009, and was therefore not allowed to reset the game at any point. This was intended to be a solo run using Lejes, but attempted to recruit Olvan at around level 40 because he happened to be in town. He forced a fight and destroyed me, and I spent 50 minutes attempting to recover the lost runes. I also lost approximately 40 minutes to the Sky Rune fight. Without being able to reset, dying to an apprentice is incredibly crippling because they will flee town with your teleport item. I had to buy consumable teleport items (Winball) and try to find them again after every ~5 attempts. Hahaha.
While I do still have the original recording of this, I do not plan to upload it. It is very long and super low quality. This, like some other entries here, is just for historical purposes.
The other runners of this game have split the categories based on whether or not you use an OoB glitch to skip the brain race. My PB is currently the best “no OoB” run, but the current best time with OoB is a 10:43 by WizWar100. It is worth mentioning that the other runners allow continues and I did not in my PB. Death abuse allows some more reckless movement and damage abuse.
Significant improvements were found for this category since my run in 2013 and I have incredible respect for the work these runners put in! I have no plans to return to this game. The current best time with Pimple is a 19:35 by Skinomarsh, and the best time with Rash is a 19:17 by RBMACHOK!
This run was done on console without using autofire. I believe that this is currently the fastest console run, but there were several runs that were faster than ours that were removed from SRC for using turbo. While turbo does have a large impact on the run, they were significantly faster in other areas and we acknowledge them as the better players. The best time (with autofire) that I know of is 21:33 by PacnyTHuk and HalfMoon_Voodoo.
Koh1fds (single player, no turbo)
Xray662 (no turbo)
PacnyTHuk
HalfMoon_Voodoo
All Battletoads Arcade runs were done on an original Battletoads Arcade pcb. The current runs were played on a Smallcab Deluxe supergun (see equipment page for more info) and using a Qanba Q2 stick. I played on maximum difficulty (6/6) and required a 1-credit clear (no continues permitted).
The Rare Replay port cannot access difficulty 6, and many others that I have seen attempt a run of this game do not care about a 1cc. Credit feeding drastically changes the strategies requires to clear stages 4 and 6.
This category does not use the level skip exploit. I don’t know if other runners use any desync bugs to warp across the map, but we did not for this run. Mecha Richter and I currently have no plans to return to this. The current best time is a 1:19:49 by HaosEdge and SeductiveSpatula.
This run is, I believe, the only speedrun of 7th Saga that I did with recruiting a partner. I don’t recall if I used seed manipulation or not, but I did allow myself to reset for recruitment luck. Either way, this was intended to be a test run and then I never ended up doing more runs. Twitch lost the recording of this, so this is really only listed for historical purposes.
The fastest time for this category is currently a 5:26:27 by Zakky!
This run was done on the Rare Replay collection, as opposed to the original release on ZX Spectrum. The lag is similar between the two, but the controls are entirely different. The ZX Spectrum version uses tank controls while this uses more traditional isometric controls.
My notes for this run on SDA are extremely detailed and go more in-depth with the route.
For more info about the game itself, please visit the Knightlore StrategyWiki. As a quick reference, my route manipulates item seed 1 from the chart listed on that page.
This run is extremely outdated, having been completed before the “JP Special mode” TAS, and has been beaten. The current best time is a 23:58 by LeHulk.
Ohon has a 23:06 on the Japanese version (no public video) and claims that another Japanese runner has a time over a minute faster with no video. I have seen Ohon’s run and can vouch for it, and Ohon saw the other runs live although there is no VOD. The Japanese version is slightly more difficult, so the US times should easily be able to cut the 23.
This was an incredibly unlucky run but very fun run with a team that can’t do damage to things. I died twice at level 1 to hermit-bird encounters. When I recruited Esuna, she was unable to get any good power/guard/speed levels. The final few battles were rough, but I was never in danger of dying. They were just slow. The fight with Monmo was baffling, though. I had to reset twice because I ran out of B.Fires before Monmo died. I had 152 magic and was dealing 69-85 damage per cast. On the third attempt, every B.Fire was dealing over 100 damage. Bizarre luck.
This was an interesting run. 19 deaths on Esuna, trying to get the Sky Rune, and 7 deaths trying to get to the final town from the crash site. Saro’s cave and Gorsia only took one try each, though the fight with Gorsia lasted a whopping 23 minutes! Kamil had a 50% chance to hit and was dealing 120 damage.
Another really good team. Grinded in the present until Lejes was over 110 magic (level 34 this run), and then moved on. Extremely good deathless run until Saro’s cave, and then shit hit the fan. I got into an unwinnable battle shortly after Griffan and died. We died 21 times in Saro’s cave, 14 of which were after killing Goron/Griffan. I died 4 more times on the way to Gorsia, and then proceeded to miss him 13 times in the winning battle despite having a 79% chance to hit.
This is the worst combination in the game. These are the two slow non-Lux characters and they are both incredibly expensive to equip in the end-game. I leveled on S.Brains until Kamil’s magic stat was above 110, and then proceeded. Kamil was OHKO’d by every spell Gorsia used, and could barely survive Goron/Griffan’s Blizzard 2. I brought FortSw and 9 diamonds, emeralds, sapphires, and rubies into the past and could still not fully arm both characters. This run had 22 deaths, 20 of which were in the past; 14 of those were in Saro’s cave which is completely impossible to survive with this team. You simply need to make it through without a single encounter.
Lux had absolutely terrible stat rolls until tune-up, but Olvan carried the team to that point without issue. This run was deathless but had to reset once because I got stuck in the wall once during the tune-up! We also kidnapped a child and delivered him 5000 years into the past to rule the world after we are gone.
This run was infuriating. Cart 5 was incredibly cruel to Lejes, giving consistently worst possible luck in every situation. Somewhere on the order of 40 S.Brains ran away during the level grind. We still made it to the final save in 9:26, and then spent 1.5 hours attempting Gorsia. Gorsia set an all-time record for strongest hit, clobbering Lejes for a whopping 1204 damage! The final attempt at Gorsia fully debuffed him and was 1-2 hits away from killing him, but ran out of potions. This is the one run that did not finish this week.
This is the angriest performance I’ve ever seen from a 7th Saga character. Olvan was not having any of it. He hit several enemies for over 1000 damage, and just cruised through the game without difficulty. This run ended with a first-try Gorsia fight (with three 2% dodges!) in just under 11 hours real time.
Esuna has a fine time through the early to mid game, but really struggles killing enemies when magic is no longer effective. She made it to Gorsia and debuffed him without issue, but the fight was super intense afterwards. Gorsia was potentially able to OHKO Esuna with a high damage roll on a critical hit, so I needed to stay at max HP at all times. There were several close calls with single-digit HP!
This team had a pretty easy time for the most part. Olvan and Kamil are trivial for Esuna to deal with. The biggest issue was Monmo, surprisingly. I advanced to Monmo at around level 29, but was struggling to finish the job. I softlocked on Gariso the first time and quit for the night. Second attempt at Monmo/Gariso was better and the past was fine.