This is the story of Magnus, the valiant hero of Palantius. He leads his army, the gallant Blue Knights, in a war of revolution. Along the way, he ignores the plight of the downtrodden, spurns the help of local leaders, and generally makes an ass of himself. His forces eventually kill the goddess of fertility, and the entire world goes straight to hell a few years later.

But who are we kidding? This is actually the story of Bruce, a lucky dude who gets hyped up on steroids, grows a huge ass grey beard in the span of a couple of minutes, then nukes his way through a horde of people and monsters using what can only be described as the apocalypse in book form. Along the way, he gets a bit of help from his sidekick Parker, who spurns the overuse of steroids in favour of watching Bruce mow down poor sods, but pays for it dearly when he gets in the way of the world's most deadly Venus fly trap.

Hello everyone, and thanks for taking the time to read these comments about my OB64 speed run.

OB64 is a game that once I thought would be impossible to speedrun. I was certain that there would be no way to get powerful enough characters to beat the entire game in a reasonable time. I thought that the minimum time would be somewhere close to 20 hours, let alone less than 5.

Then, I read Amazing Ampharos' topic on the forum. The combination of the item dupe trick plus abusing Odiron's gifts from the beginning of the game made a reasonable speed-run possible. It may be a bit boring to watch, but that's why there are audio commentaries, right?

The speed run is very straightforward, even if the game is not. I'm not going to write a whole essay on the game mechanics (check CyricZ's Walkthrough at GameFAQs for anything you need to know), but instead I'm just going to explain some of the choices I made. I'll try to make an audio commentary which details these things, as well as an explanation of the game's plot which I so compulsively fly through.

One very important note: during the run, you may notice that the number of Angel Fruits, Champion Statuettes, and Altars of Resurrection in my inventory jumps around sporadically. This is due to the use of the item duplication glitch, not because I was entering cheat codes in between segments.

Segment 1

This was the second hardest segment to do. It has the most random factors involved, and although it's only four scenes long, there are random elements in each stage that can instantly end a run. I've actually lost to each and every boss in this segment at least once while recording. No word of a lie.

Choosing not to rename Magnus was an eleventh hour decision. It probably makes a difference of a couple of milliseconds either way, if at all. Other than that, I zip through the opening questions as quickly as I can. What I want out of this part is to get the Magicians Robe, which is one of the pieces of equipment needed to make Archmages ? and this is the only place to get it this early. Luckily, the first answer to every question not only gets the Magician's Robe, but also nets me some good characters to use in the time before I set up my end-game Wizards.

You can see that I screw up in the action menus for the first couple of tries. The menus in this game are weird ? some of them you can blitz through in the blink of an eye, some take ages to navigate. Luckily, all my mistakes happened in menus where it cost me virtually no time. The only thing really hurt by these mistakes is my pride.

The idea to use the Doll Master unit in Scenes 2 and 3 came to me late as well. If you take Magnus' unit to Ishro, you get an extra long cutscene, which, while making more sense (where did that old man come from?) wastes time, so I saved time there. It also only has 7 attacks for the entire unit, unlike Leia's (whom I had originally planned on using to defend my base), which has 10, so using it to defend my HQ in Scene 3 saved some more time.

In Scene 4, something strange happened: I killed an extra enemy character, allowing my Phalanx and Sorceresses to level up before fighting Debonair. This is likely what gave me the power I needed to beat Debonair in one fight, despite lacking combo magic.

Speaking of Debonair, he's just a major pain. His class has very good growths, so he towers above everyone else at this low level, and I'm not sure, but he may actually have high end-equipment at this point as well. It is very hard to take him down in one battle, but I was determined to do it. It just looks better, plus it saves a decent amount of time.

I likely could have gained another minute, give or take, had I only had to fight the Dragon Tamer and Wizard in Scene 3 once each. This is, however, even more uncommon than a one-encounter Debonair fight. Besides, the extra experience gained by those fights helps out for the Efeminette/Debonair fights. I could also shave a second by using the R button to cancel some of the commands in the Organize screen (most notably, "Remove Character"), but I didn't know this until after the segment was completed.

Other than those few things, I'd have had to get really lucky with critical hits and combo magic in order to gain any more time.

Segment 2:

This segment took the most number of retries, because I wanted to pull off the item duping part as flawlessly as possible. I end up with a pretty good result ? not perfect, but after 120+ attempts, I was getting mighty sick of it (R, A, R, Right, A, Left, A, A...).

Here we have the infamous item duping glitch. In a nutshell, I abuse the game's auto-organize function in conjunction with bad programming to increase the number of certain items in my inventory by several hundred fold.

In detail, this is what happens: Magnus' unit has 8 slots open to carry items. The first slot contains the unused Power Fruit, and the rest are empty. I want to dupe Champion Statuettes (raises a character's level by one), Angel Fruits (a better Power Fruit) and Altars of Resurrection (for selling purposes). I place these items in the bottom three slots of Magnus' unit, leaving four empty slots in the middle of the list. Now, certain events cause the game to run a check on a unit's inventory, and if it finds that there are empty spaces it shifts all the items up so that the empty slots are at the bottom ? and one such place that this check is performed is when the members of a unit change. Now, when I remove the Phalanx and Sorceresses from Magnus' unit, I'm also taking away those four empty inventory slots, so the other items shift up. What happens next is anyone's guess ? if you then remove the items that were shifted, the game glitches out and displays the total as 55/01 (or 55/02, in the case of Altars), which doesn't make a lot of sense. If I were to sell one more of a glitched item than I started with, the display changes to 55/55. I can actually sell as many as I want, all the way up until I max my funds, but I don't need to do that. I, however, sell a bunch of Altars to give me the funds I need for some important items a bit later. Upon exiting and re-entering the Organize Screen, the number has changed to 0/55. That 55 is actually 255, but the game isn't able to display that. Because of this display error, and because you're never supposed to have more than 99 of any one specific item, every time I gain one of the duped items by defeating an enemy, the number of items displayed messes up again, and it looks like my inventory jumps around. Again, I repeat: no cheat codes used to alter my inventory, just abuse of a glitch.

Booting Leia and Dio saves some time in extra text, as their stand-ins have fewer lines 'n such. This is particularly evident after the execution scene, as neither Leia nor Dio have to make their entrance.

I save a couple of seconds from my practice runs because my Phalanx is able to become a Wizard, meaning I don't have to level up a second Fighter. In my mind, this makes up for the time lost by mistakes in the first segment.

Big thanks to Amazing Ampharos here ? he suggested level 23 for Scene 5, and as it turns out it's probably the ideal level to get the Wizards chaotic enough to class change. I tried for so long to do it at 20 before going back to 23, whereupon I got the segment done in a few more tries. I think in one of my practice runs I got it to go at 20, but I couldn't get it in a recording. In the end, it's just a few points of damage lost, which may or may not save a few seconds very late in the game.

Speaking of the class change, you'll notice that I mess it up a bit, as well as screw up in a couple of other places. This was not altogether unexpected, as the later parts of this segment are not near as well-practiced as the earlier bits.

If there's any part you would want to skip, it's the several minutes long section where I level up Magnus and an Archmage. Unfortunately, this is the fastest way I know of to make your characters strong enough to blow through the game. Training would likely take 10 times as long, and I doubt I could get strong enough just by fighting enemy units.

I got pretty unlucky with combo magic; it didn't really start to take off until Scene 8, and it's the wrong combination to kill all those Golems, so I have to let them attack me. In hindsight, I probably should have picked a different fighter to save that bit of time.

Segment 3:

The segments from here on in were much easier to make. Very little randomness can occur (though it does rear its ugly head now and again). Every battle is pretty much the same now ? deploy Archmages, beeline to enemy headquarters. I skip all the optional stages along the way, obviously.

My Archmages decided it would be a good idea to hip-check Rhade, so I have to endure a couple of extra attacks. No biggie, just a few seconds lost. THEN they do the exact same thing on the Witch/Siren unit in the next level. This costs me quite a bit, as my level 70 Archmage gets incapacitated and I have to sit through all the attacks from their unit.

Note the on-the-fly adjustments to Magnus' unit ? I actually made the decision to do that a split-second before the encounter. If this were a TAS, I would be losing time. But because this is a console run, those adjustments were necessary to ensure that I didn't get in any extra fights.

Saradin joins, which guarantees that I won't get the secret Caliber mission at the end of the game. Choosing to fight the Western Division saves time, no doubt there. On an interesting (but largely irrelevant) note, Mount Keryloth 1 is actually impossible to do in a run like this, as a Diana-led Pumpkin Head unit attacks your headquarters, and the chances of this unit killing Magnus are very high, not to mention that it probably out-damages him as well.

Segment 4:

This segment is probably the easiest of all of to make, as there are only two places where the run can outright end (and I circumvent one of them). That said, many of the smaller things went right throughout ? Magnus attacked the correct character 95% of the time ? so by the end I was actually worried about losing something so stellar, hence slowing down and ensuring that my one guy didn't die.

I did another small bit of levelling here. In theory, it's possible for a level 50 Magnus to get through all of Chapter 3, but in practice he has to fight enemy units far too many times to take them out of commission. While I would have liked to have waited until Chapter 4 to do this and take advantage of Magnus' extremely nice General growths, it's just faster to do it this way.

Going to Capitrium is about a minute faster than going to Tremos by my testing, and also nets a nice weapon for Magnus ? the Evil Blade. This sword not only has more attack power than most of the swords in the game that I collect, it also has the special ability to put enemies to Sleep. As you'll see in the next segment, this means that if Magnus doesn't end up killing an enemy, he'll at least incapacitate them.

We have another on-the-fly strategy adjustment here. I'd say that in at least 2/3rds of my practice runs, my Archmages side-attack that last Templar unit, and because Parker is already very weak (and slow), he can't survive the attacks coming his way in this fight if he's in the front row, barring an extremely lucky dodge (or two). Because this segment had gone so well up to this point, I decided to quickly switch the destination to ensure a head-on battle. In hindsight, I maybe could have equipped a Bloodstained Robe, which might have given me enough defense to survive (not to mention that I end up equipping the Robe anyhow). As it stands, I feel justified in my decision.

Segment 5:

Probably my sloppiest segment in terms of mistakes and randomness, but most of the mistakes made don't cost an appreciable amount of time ? again, the only thing hurt is my pride. You'll notice that I deck out my Archmages in the best gear I have. The Bloodstained Robe for Parker allows him to survive ? there are two points (the 2nd and 3rd siege missions) where he'll die if he's wearing his normal robe. The other stuff for Bruce allows him to one-shot Amrius, Procus, and Yumil nearly 100% of the time, not to mention take out a couple of random Ogres that have a habit of eating more than one attack.

In the third-to-last scene of the segment, I get into a random encounter that costs me a bit of time. I was going to reset when I thought, "what the hell, I'll just go with it." Well, the next two stages went very well, and I DIDN'T get murdered by Richard (though it was mighty close). I had in mind to save in the 1st slot and try again to see if I could get a better time. In my haste, I saved in the 2nd slot, thus forcing me to keep this segment unless I wanted to do segments 2, 3, and 4 again.

In the end, it's not a big deal. Though the actual time I could have saved is measurable, the only real regret I have is that I cut it really close to the 4:30 mark.

Segment 6:

I made the final stage its own segment so that you all could see the in-game clock one last time. The actual scene is straightforward, and I think there's more cutscene here than actual gameplay.

This segment is almost perfect; I only mess up in the menu once, and I get the best attack angle twice (!).

You'll notice that my weaker Archmage dies in the fight against Danika. I planned it this way; it's not an error. He's used throughout the run to haul around an extra Angel fruit (saves trips to the Organize Screen), he's needed in a few places to mop up, and he's required in this fight to eat a couple of hits. RIP, Parker.

The game does NOT end when the final hit on Danika comes down, as it's still possible to fast-scroll through the conversation between Magnus and Danika. After that the alst dialogue box disappears and the conversation ends, the timer stops and the run is done.

Because I got Saradin, I don't have to do the secret Caliber mission (good thing too; I don't know if my unit could handle the double Sphinx unit). As such, I get the absolute worst possible ending: no Caliber + low Chaos Frame. In my practice runs, I've also gotten the middle Chaos Frame ending. I'm 99% sure which ending I would get hinges on whether or not the enemy captures that town all the way back in Scene 5 (you know, the one with Yumil running around?).

Possible improvements?:

- Better execution of commands. I bobble the menus sometimes and don't skip scenes quite as fast as I could. I also don't input commands ASAP sometimes, particularly in the long levelling segments. In addition, I don't use the fastest menu speed, as it is actually too fast most of the time and causes way too many errors. If someone wants to master it and beat my time by a few seconds, I would love to see it (not being sarcastic here).

- Tweaking the levels of the characters. I didn't test the levels much, I just ran off of what Amazing Ampharos had planned for the most part. It may be possible to beat the game at a much lower level for both the main Archmage and Magnus. In fact, it may be possible, with a bit more tweaking, to manage the game with only a single Archmage, and just adding a Soldier group or something for the two stages that require an extra Angel Fruit

- More complex route planning, like I did in scene 3. I just sent my unit straight at the enemy headquarters, and let it find its own way there. Units in this game always take the path that uses up the least Fatigue (which can get frustrating when playing normally). However, time might be saved if you micromanaged the unit around enemy units.

- More segments. Doing 1-3 scenes per segment would allow for a lot of time to be shaved off, as you could keep resetting until Magnus attacked the correct guy every time. This would also cut down on errors heavily.

- I used Drakonite books. Amazing Ampharos had planned to use combo magic. I'm pretty sure that magic is the best way to do it, as physical attacks are always single-target, and you can't get any good physical units until later, not to mention that you'd need to get elemental weapons for them. However, if someone comes up with some other crazy strategy that saves time, all the power to them.

Thanks and credits:

• Amazing Ampharos, for bringing up the idea and pioneering running this game
• CyricZ on GameFAQs, for his amazing guide to this game
• The SDA staff for maintaining this site
• Atlus and Quest for this wonderful game. Make another one PLZKTHX
• PNP Games for selling me this game cheap
• You guys for putting up with the run and commentary

Future plans:
- There aren't that many games I've considered speedrunning (I generally suck at execution-heavy games), but I'm thinking of tackling Eternal Darkness next, then maybe Paper Mario.

- I do not plan on doing runs for alternate endings. Using the fate of Magnus as the basis, there are 5 different endings in the game. High Chaos Frame is probably the only other one worth doing, but would likely be extremely similar to this run, the only differences being an extra unit or two liberating strongholds, as well as choosing to make peace with the Western Division. Middle Chaos Frame ending could be achieved simply be resetting at Scene 5 until the enemy does not take the town. Low Chaos Frame + Caliber mission can end two ways, depending on whether you win or lose Caliber - and making two separate runs for those would be pointless. The only other type of run would be "100%", aka 100 Chaos Frame ? that would be a real challenge. If someone wants to see these, someone else will have to do the run.

- I think I would do another run of this game at some point. If my time gets beaten, or I get bored again, I might shoot for an improvement. But right now, I'm awfully turned off of this game; I can barely look at the cartridge.