Released in 1986 for the Famicom, 1989 for the NES and a bunch of other release dates for various ports by Chunsoft, Dragon Warrior (also known as Dragon Quest) is the game that spawned the JRPG franchise. The player controls a young hero with the aim of defeating a being known as the Dragonlord. Along the way the hero must collect the stolen 'Balls of Light', and the cover of the game box features a dragon that looks suspiciously like the one from Dragonball. Hmm. I wonder if you get a wish at the end.
Best time: 4:47:36 by Eric 'Lhexa' D'Avignon on 2013-09-21, done in 10 segments.
Get Flash to see this player.
Author's comments:
This was my first speedrun, and a pleasure to finish, even if some segments did push me to the limits of my endurance. I find it amusing that the first Dragon Warrior, an exceedingly simple game by JRPG standards, has a speedrun almost as long as its much more complicated sequels already on the site. When people say that this game is all about grinding, they aren't kidding: about 4:10 of the total 4:47 is grinding. Incidentally, here are the experience-grinding rates that I managed to achieve:
Level 1-2: 3.8 XP/min
Level 2-3: 3.6 XP/min
Level 3-4: 6.9 XP/min
Level 4-5: 9.5 XP/min
Level 5-6: 16 XP/min
Level 6-7: 17 XP/min
Level 7-8: 26 XP/min
Level 8-9: 44 XP/min
Level 9-10: 50 XP/min
Level 10-11: 72 XP/min
Level 11-12: 79 XP/min
Level 12-13: 86 XP/min
Level 13-14: 106 XP/min
Level 14-15: 115 XP/min
Level 15-16: 111 XP/min
Level 16-17: 128 XP/min
Level 17-18: 112 XP/min
Segments two, five, and seven were at such long odds that I was willing to keep any success. However, for the other segments, I would complete enough to figure out a mean and standard deviation for their distribution, and then only accept a run that was at least two deviations below the mean. Since the distributions tended to have significant positive skew, this was a stricter standard than it might seem.
Before I move on to individual segments, here are some explanations of the finer points of my run.
Name choice: There are a total of sixteen growth types, depending on name. Two bits of this four-bit number determine a flat bonus from 0 to 3 that you receive from the beginning of the game, so the best choice there is plain. The other two bits determine whether you have strong Str/Agi, Str/HP, HP/MP, or Agi/MP growth. Strength is preferred over MP in order to beat the Dragonlord at level eighteen. So the real contest is between a Str/Agi name, "L", or a Str/HP name, "J". With the equipment I use, and the levels I reach, there are only a handful of differences. J gets Red Slimes into the "safe zone" (a better monster damage formula that takes over when your defense is high enough) one level earlier than L, due to the early agility bonus mentioned earlier, and L gets Rogue Scorpions and Green Dragons into the safe zone one level earlier than J. It's not safe to grind in the Rogue Scorpion zone until I have the Silver Shield, so that advantage does not matter, and the Metal Slime trick requires that I use zones 10 and 12 at level seventeen, so the Green Dragon advantage is also moot. Finally, J has a much easier time beating the Dragonlord, as his extra HP allows the HEALMORE spell to be more effective. So "J" has an easier time in segments one and ten, and "L" would have had an easier time in segment nine had I used a different strategy. J wins. Also, the one-letter name saves a bit of time whenever the name is displayed, which happens a couple thousand times during the run.
Gold glitch: The game can only track up to eight opened chests, and this number does not reset upon death. So if you open some chests in the Mountain Cave, die, and reach eight in Tantegel (the throne room chests, plus the treasury, will get you there), the ninth chest will never disappear. This allows you to pull infinite gold from the castle treasury, albeit at a very slow rate. The best single-segment strategy uses this trick to acquire a Broad Sword, Full Plate, and Large Shield at level four, and a segmented run could do so at level one. However, the bulk of the benefit of this gear comes from the Broad Sword, which I can get via a Death Necklace. When testing, I found the gold glitch to be a few minutes slower than my route, due to the large amount of time you need to spend setting it up. I was comparing two unoptimized runs, however, rather than two runs obeying my "two standard deviations" rule, so it's possible the gold glitch run would have been slightly faster in that latter case.
Metal Slime Manipulation: This is my lone contribution to the lore of Dragon Warrior. By a quirk of the RNG, whenever you go an unusually long time between encounters (assuming you're moving continuously over the most dangerous terrain, i.e. desert or hills), the next fight will be with a specific monsters, namely the fifth one of that area's group. (See Ryan8bit's FAQ for the monster lists.) I use the overworld theme as reference -- when it reaches the sixth bar, I'm going to hit one of these fights. The specific monster is a Demon Knight in Zone 10, and a Metal Slime in Zone 11. If you just hold "A" through the fight, they will play out the same way every time after a fresh reset, so by charting them out beforehand I can get a handful of guaranteed Metal Slime wins. These fights occur about four minutes apart, and repeat with a period of eight fights, the first, second, sixth, and eight being wins. The specific pattern might depend on your save file -- I only tested this trick with a single save.
Segment One: 18:56
My initial equipment purchase is a Club, Clothes, and Dragon Scale. The Dragon's Scale is equipped and then sold, exploiting a bug that keeps the defense bonus from ever being lost. The Club is an obvious choice, since fast levelling requires that I do as much damage as possible. Meanwhile, the Clothes and Dragon's Scale are enough to ensure that Red Slimes virtually always do just 1 HP of damage. Since I kill them in two hits, that means that for the first two levels (reached in Zone 1, with Slimes, Red Slimes, and Drakees) I know that I must always keep my HP above the XP needed for the next level. This consistent damage from Red Slimes also allows me to estimate how many Drakees I can safely kill.
Level four is reached via ghosts in Zone 2, northwest of Tantegel. Finally, level five requires lots of spell-casting against the critters in the Hills of Hubris, a strip of difficult monsters southwest of Tantegel. Until level five, I only visit the inn just after gaining a level, a feat which is just barely possible. The inn is always quicker than the old man in the castle, and 6 GP is not a big enough number to matter. I made a small mistake when purchasing at the item shop, costing a few seconds, but it otherwise went great.
Segment Two: 9:32
For quicker levelling I need a better weapon, and to afford that I need a certain obscure item called the "Death Necklace", which is utterly worthless beyond the fact that it can be sold for a hefty 1200 GP. I use torches because it's slightly faster than bump-walking: you lose a fraction of a second not only with each bump, but with the slight delay before movement from a stop. This was actually my second Death Necklace -- the first try was sloppy and unlucky, so I elected to try again. I kill one Drakeema in a lucky fight (downing it in two shots) because they offer unusually good XP for their HP. This one fight saves about thirty seconds, given the next segment's rate of experience gain. Two trips to the chest are needed to gather enough gold, namely 1500 GP. I almost equipped the Death Necklace by accident on my second trip, since it now occupied a spot normally held by torches.
Segment Three: 35:30
I dash to Rimuldar to buy a shiny new Broad Sword. The monsters there are still too tough, though, so I grind around Kol to level seven, remembering to pick up the Fairy Flute in the meantime. The SLEEP spell, gained at level seven, makes the Rimuldar monsters killable, though by no means safe. They bring me to level eight. In the kept segment I had extraordinary luck towards the end, sleeplocking five enemies in a row. I finish by buying six keys and using Wings to return to the king, rather than death-warping, as gold will continue to be tight all the way into segment six.
Segment Four: 4:57
I need a Large Shield for segment six, so it's off to Garinham. Since I don't want to visit this town more than once, I then head into the Grave, which turned out to be relatively easy even at such a low level. I need the Cursed Belt and 240 gold to fill up on Herbs and Fairy Waters in the next segment, and I reach my target on the very last step thanks to a fortuitous excellent hit.
Segment Five: 5:24
I kill the Axe Knight at level eight, using a lucky SLEEP spell. If you only watch one segment of this speedrun, make it this one. With my shiny new Erdrick's Armor I am as powerful as any level eight character has ever been, but must nonetheless die to a wolf to death warp back to the castle.
Segment Six: 26:47
Level eight, Erdrick's Armor, a Large Shield and the Dragon's Scale are exactly enough to push Wolflords into the safe zone, so I use them to grind to level ten. The regenerative property of Erdrick's armor means that I no longer need inns, and can grind in areas far away from towns. The huge agility boost at level ten pushes the next segment into the realm of the plausible. I also take the opportunity to pick up the Stones of Sunlight and Erdrick's Token.
Segment Seven: 9:05
The various Axe Knights of Charlock Castle sure did exact revenge for their fallen comrade. Anyhow, on this hundred-and-fifth attempt J manages to retrieve Erdrick's Sword, having first finished the remaining quests. While I now have most of the best equipment in the game, I still don't have the OUTSIDE spell, so another death warp takes me back. By the way, although I'm less than two hours in and have completed every quest besides killing the Dragonlord, the run isn't even halfway done.
Segment Eight: 49:14
This segment is for grinding three more levels and obtaining the best shield. It turned out that Zone 9 (Wyverns etc.) is the best zone all the way up to level thirteen. I need thirteen Goldman kills to raise enough money for the Silver Shield, but sometimes twelve will do, so I ended up putting off the thirteenth kill until I was sure I needed it. I was amused by how close the margin for affording the shield ended up being. The segment ends with a quite lucky run to Cantlin.
Segment Nine: 2:00:06
This one is all about grinding to level eighteen, the minimum where I can beat the Dragonlord. After testing, I found that Zone 10 (south of Hauksness) remains the best for grinding all the way up to level seventeen -- it's a big step above Zone 9 (north of Hauksness), and the subsequent ones have time-wasters like Magiwyverns or Starwyverns. For level eighteen, Zones 10, 12 (southeast of Hauksness), and 13 (inside Hauksness) were very close (from 112 to 118 XP/min), with 12 barely winning, but they were so close that I wouldn't be surprised to be wrong. As it happens, in the kept segment I ended up being trounced at level seventeen, losing a minute or two to bad luck. I spent some of level seventeen in a convenient section of Zone 12 hills, resting in 10, but after running out of spare MP I had to spend the remainder in the old Zone. I ruled out Zone 13 because effective grinding there requires multiple HEALMORE casts, and thus a return to the castle, and I can save a minute and a half by making my MP pool last the entire segment.
Using the Metal Slime trick described above, I was able to get eight (out of a possible fourteen) Metal Slime kills. For reference, the Metal Slime fights occur around the following times: 4:30, 8:20, 23:40, 40:10, 56:40, 1:05:10, 1:09:20, 1:13:40. Not included are a few times when I thought I would get a Metal Slime and got a Magiwyvern or Wraith Knight instead. There were also familiar Demon Knight fights around 19:40, 44:00, 52:30, 1:25:40, and 1:34:00, whose corresponding Metal Slime fights would have been losses. I also successfully avoid a nasty Demon Knight at 12:10 by pausing for a moment, letting the RNG proceed past the fight. This one would have been a preemptive strike for maximum damage, so I was happy to be able to miss it. The extra experience from Metal Slimes saves me five or six minutes.
Segment Ten: 7:37
This segment went about as well as one could hope. I got very few encounters in the most dangerous zone, and killed the first form of the Dragonlord in two hits, which is very rare at this level. Usually one needs fifteen hits to kill the second form, but a normal fight will only guarantee thirteen hits (one between each cast of HEALMORE, with two hits at the start and two after you run out of MP), so because I was behind toward the end of the fight I took a swing on 45 HP. A fire attack at that point would have had a 1/2 chance of killing me. But I got lucky there, then got lucky again with all MP depleted, and brought down the big dude.
Total Time: 4:47:08
As a bonus, I used the extra key I bought back in segment three to rescue Gwaelin and finish the game with the full ending. But since this is a speedrun, I didn't have time to drop her off back at Tantegel Castle, and instead she accompanied me to the Dragonlord. Oh well, she doesn't seem worse the wear for all the fiery breath attacks. This version of segment ten took 9:48, for a total time of 4:49:19. If you want to see it, look in the SDA forums.
In closing, here are some things that an improved Dragon Warrior run could try: killing four Metal Slimes to reach level seven, one segment each (yes, this is possible, just very hard), saving around forty minutes; getting both pieces of Erdrick's gear at level seven (the armor saves only a little time, the sword a fair amount more), saving about twenty-five minutes; and killing the Dragonlord at level seventeen using SLEEP (which might not be humanly possible), saving twenty-five minutes. In all, you could bring the time down to about 3:30. But my own strategies were sanity-consuming, and the new ones would be even more so.
Thanks go to Ryan8bit (for his wonderful Formulas FAQ), Darkwing Duck and ShinerCCC (the other prospective runners for this game -- good luck on the single segment!), and the various people who offered encouragement on GameFAQs and the SDA forums.