October 1998 marked the end of the triumphant era of platforming bandicoots. Many critics agreed that this game was the pinnacle of what the bandicoot has starred in so far. So, what does Sony do about this? Buy out Naughty Dog as a first party developer doomed to make non-Crash PlayStation-exclusive titles forever, and force them to sell the Crash Bandicoot property to crappy developers who have struggled to produce a decent Crash game since.
Best 100% time with deaths (game says 105%): 2:41:28 by Mychal 'trihex' Jefferson on 2007-01-01 in 18 segments.
Get Flash to see this player.
« t r i . h e x » presents ...
In association with Speed Demos Archive,
Crash Bandicoot 3: Warped
All 105% of the game completed in 18 segments, totaling 2:39:00.
» » » I n t r o
In my perspective, I find that a sequel should only be produced if there warrants a great purpose and an audience to satisfy from whatever the first offering featured. My previous project, Crash Bandicoot 2 [100%; 22 seg.; 1:39:59], contained many sequence breaks, skipping entire levels and still maintaining a 100% in-game accomplishments. I view this speedrun as a sequel to the Crash 2 one, which greatly contributes to my attitude of not being satisfied with this run, mainly because it just can't be as entertaining as Crash 2 to watch because it's so linear. My Crash 3 run doesn't have the flashiness of the previous run, but that allows it to shine in a different aspect of endurance and execution versus perfection. That's because the routes that I found the fastest in most of the levels in Crash 3 required take a major risk, such as a huge ass jump made tricky due to camera angles, that proved very difficult in which merely executing it yielded success. This new play style required to produce a speedrun of the game made it far less fun to do such.
Marked the beginning of any work on a Crash 3 speedrun. This was back in my regretted novice days of course, so I made plenty of mistakes. The main one being very impatient in producing a quality speedrun. That run was started very shortly after completing the rejected 2-hour plus Crash 2 speedrun. The rejected 3:08:xx Crash 3 run is a very important run in the development of me, t r i . h e x, as a speedrunner. Upon Radix receiving the excuse of a speedrun, he replied in the tread I started with the rude-awakening that forever changed my attitude towards producing speedruns:
"You may have practiced before hand, but taking one or two days to do a 3+ hour run doesn't make a good one.
Especially the fact that you segmented the run to 12
segments and died 23 times is laughable. (20 times in 7 segment crash 2) A
few deaths in a long Single-segment run aren't too bad.
Dozens of deaths in a segmented run? I don't think so."
- Radix (Sept. 7, 2005).
Needless to say, I was quite embarrassed about the situation, but I finally saw the need for a revolution. I rewatched all the speedruns I had published on SDA prior to both Crash Bandicoot games being rejected. I began to see that they lacked quality --the deaths, missed jumps, and failed shortcuts in the [obsolete, 1:47:xx SS] Sonic Adventure DX run as Sonic; to the death in 6-4 that cost almost 5 minutes in the [obsolete, 2:19:xx SS] Yoshi's Island 'run. Then I began to get determined to improve my reputation as a speedrunner, and improve all my records drastically. Every day, of every week from then, I practiced. Continually seeing more ways to improve the things I had just improved, I developed my now-attitude of never being satisfied with anything I ever publish. Getting the feeling of "settling for just this due to time restraints", it's a good habit, but probably lessens the overall satisfaction of any completed speedrun.
I managed to improve my Yoshi's Island SS time by nearly 22 minutes. Once the run was published, immediately overnight, my reputation was reestablished, and the YI enthusiasts who saw the run were eager to watch whatever I had planned next. Which, if you don't know, was the 100% table. I'll skip the details about the 100% table, go to the Super Mario World 2:Yoshi's Island page and see the comments on the table for details about it.
July 6th, 2006
Hot after just completing the extremely-satisfactory Crash Bandicoot 2 speedrun, I was anxious to improve my last evidence of suckage --Crash Bandicoot 3 [105% in 3:08:xx; 12 segments, 23 deaths; rejected]. Initially, the rejected speedrun had 12 segments, essentially a segment for each world (levels 26-30 counted as a world), then a segment for each world again to get their relics after beating Cortex the first time, totaling 12 segments. Problem is, each segment in this way easily totaled over 12-15 minutes each. Obviously mistakes are going to happen with the segments being that long. That, and at the time I had it as in intention to get all platinum relics to show off since there was very little else I could do to show off. So I reworked all the routes to include some of the time-savers I discovered from the Crash 2 speedrun, and found a successful pattern in the segmenting. Doing a segment for all the platforming (Crash) levels in a world, then a segment for all the vehicle-based levels (Cocoa/Crash) including those levels' relics proved to work.
July 27th, 2006
I complete the speedrun. I posted in glee the official title of the speedrun "Best 105% time with death-abuse: 2:29:55 in 17 segments," foolishly miscounting the segments, it's actually 18. At this time, this was when SDA was in drastic need of additional help with Radix becoming busier by the day. So, the still-unpublished Crash 2 run was in queue-hell, while I was attempting a Gex 2: Enter the Gecko speedrun to send along with Crash 3. The Gex 2 'run ended up being cancelled because I was growing uninterested because it was far less fun than I originally remembered. DJGrenola and others become close staff of SDA and runs were literally flying off that verification thread. Eventually my Crash 2 run saw the light of day.
Summer was officially ending, school was coming up in less than 2 weeks, and my free time was about to be demolished for no return. During my off-time I was able to improve some of my Yoshi's Island individual 100% level times, but I didn't have the time to start a new speedrun or anything like that. I also never sent that Crash 3 run after officially announcing I cancelled the Gex 2 run, mainly because my comments for Crash 3 were lost, and so were my segment times. I kept procrastinating on doing anything about the Crash 3 run, since I didn't feel like watching and observing the speedrun in-depth a third time. So the weeks kept going by without an update. This gap of nothing probably wouldn't have been so bad if I had a game to speedrun, but with Yoshi's Island DS still months away, and a 100% SS of Yoshi's Island requiring oodles of practice time and preparation which can only be provided during summer, I was blank. Throughout the time, several followers of the Crash 2 speedrun kept bumping the Crash 3 thread daily about when am I ever going to send the damn Crash 3 'run. I promised them during Christmas break '06 I'd have the time to sit through and produce these comments along with acquiring the segment times. The promise was to have Crash 3 ready for sending by New Year's Day.
December 31st, 2006
I decided to dedicate the entire day to progressing on these comments for the 'run. I progressed through each segment until segment 8, and noticed a percentage complete gap, soon realizing I don't have segment 8 on the final disc. I checked all the other nonfinalized discs, hoping I accidentally skipped it in the final recordings. I had no hope in finding it. I eventually concluded I had to replay Crash 3, save at 49%, and redo this segment, saving at 54%, and let the rest of the run proceed with the remaining.
January 1st, 2007
To start off a great year, a great speedrun finally sees its final segment complete. Into the wee hours of the morning, approx. 3:30 AM, the segment was complete. The previous finishing day for this 'run was July 31st, 2006, but is now 01/01/07 due to that lost segment.
» » » T h a n k s
Of course I have people to thank that would've never made this accomplishment possible without:
Radix, for rejecting every single run he ever has that I've ever submitted. He is the one who sparked the revolution within and made me the proud speedrunner I am today. Keep owning!
Nate, for putting up with everyone's constant bitching about Crash 2 never getting published. I was never worried about it not getting done. Also for the usual processing and whatnot of Crash 3.
The Verifiers, whoever they are, thank you. I wasn't worried about needing verifiers after seeing how popular the Crash 2 run was, but you're appreciated regardless.
» » » M i n i - F A Q
There are a few things I want to cover before I proceed into the segment-specific comments.
Q: WTF is SS Jumping (Slide-Spin Jumping)
A: It's the technique System Error revealed during the progress of the Crash 2 speedrun. If you slide jump and spin attack all simultaneously, it results in doing a jump higher than what Naughty Dog thought. The reason it's so useful is because there are many obstacles that can be skipped by doing this because they made them just long enough to be out of range with slide jumping, thus, flirting with your ability to jump; so I dubbed them "flirt gaps." You can do a SS jump and then a double jump to jump over things that shouldn't be possible. A key example is the 2nd boss, defeated in less than 30 seconds when it's supposed to take at least a minute longer.
Q: Crappy Time-Trail Times
A: I'm very aware I skip time boxes throughout the trails whenever I'm trying to get relics. Here's the situation, this game has no "overall" in-game timer. So the game must be timed manually. It would seem logical to manually time everything except the time trails, but getting a lower time in the trails takes more "real" time, so it goes against the timing purposes of manual timing. To keep things universal, I manually time everything in the run, which means, in trails, run in a straight path, do as little as possible, and ignore the timer because any time I pause it, I don't count it. This hardly affects vehicle-based level time trails, but for Crash's levels it does. Also, if I were to use the "time trail time + everything else real timed = SDA time" formula, to achieve the best possible level times, you would need to get as far off-screen as possible from the clock, and use your cannon to shoot it from far away, thus saving 1-2 seconds on trail time, but doing this costs about 5 seconds in real time ... slightly contradictory, huh?
Q: In the forums you promised us all platinum relics!
A: The real time versus trail time argument I mentioned previously covers why in almost all of Crash's levels I have to settle for gold.
» » » S e g m e n t - S p e c i f i c C o m m e n t s
Segment 1 in 0:07:17 (6% complete)
What a segment to make an opening statement for a run! During Level 3, I do stop in about 4 spots, but that's because the collision detection isn't too great when you run in those areas. To hit all the boxes, you will have to slow down in specific spots, this principle carries on with other Cocoa/Tiger levels. During Level 3's Relic, I wasn't planning on glitching through the wall like that. It doesn't slow me down at all, but it does make me skip 2 boxes and loss the platinum relic, which isn't required to begin with (I wanted it to show off). Towards the end of the same trail, I jump atop a false ground near the ceiling of the final platform, this results in Tiger jumping further than usual, closer to the warp, and is faster. Another side effect of this is the timer triggers to stop whenever Tiger lands on the same ground as the warp, so it stops later, another factor into the "crappy" time of 0:20:90.
I hate how I missed a box so early in Level 5, but the fact that everything else went perfect. I decided to settle with it and hate myself.
Segment 2 in 0:10:29 (14% complete)
Man, I can't believe I settled for missing 2 boxes so early into the segment ... what the hell was I thinking?! Aside from that, Level 1 goes fine.
In levels involving the Triceratops, it will always destroy every box you skip, except checkpoint boxes. I save time, not a lot at all, by not needing to spin attack while walking. Goddamnit! I miss the 2nd checkpoint box ... twice! Ugh, what a bad taste that leaves in my mouth!
I was debating to try to include taking the shortcut of swimming past the 2nd electro-circular thingy in Level 2, especially since it's so risky for so late in the segment. I wanted a great speedrun, so I decided to include it in both the first trip, and the trail.
Segment 3 in 0:06:08 (20% complete)
The later I proceeded into completing this speedrun, the more I hated these segments (segments that involved the vehicle-based levels). Level 8's first trip doesn't quite go perfect, as it's notable after the first zapper I hit the edge, but it's close enough. What's freaky about the trail run is how I get the same exact trail time that I did in the previous segment attempt ...
In Level 10's trail, I could get a better time of at least 0.5 seconds, but I need the 2nd mask for the next segment, as I save a lot more time by losing half a second here, to gain 20+ in the next segment through time saved thanks to invincibility masks.
Segment 4 in 0:08:15 (26% complete)
This segment is all about raping masks. I achieve the invincibility mask 5 times total in this segment. Doing Level 7's bonus level is particularly tricky. Doing it fast is very hard, the last 4 boxes require some precision, and exiting off of the 3 boxes in the air without getting hit by the TNT box is easier said than done. I learned through dozens of failed attempts to just wait for the swordsman to turn around as there's no way to get past him going at you without losing a mask.
Level 9 marks the first time you see me abuse ice/oil physics. While achieving invincibility mask, you'll notice I keep stepping atop the boxes while the apples come to me off-screen, I do that to show off and hopefully give a feel of me going fast. I made a slight mess-up during the final oil spill area.
Level 6 is the definition of mask raping. Typically, as I exit to the bonus level, the dumbass frog would own itself and hit the nitro box, it's a miracle that he didn't do it this time.
Boss 02, what a n00b.
Segment 5 in 0:11:27 (31% complete)
Level 14, warp to Level 31
Level 14, warp to Level 31
Just be aware that this is NOT a fun segment AT ALL. I hate Level 14, and I hate the bastard who designed Level 31. There were so many times I wanted to say fuck it and settle for a Gold Relic on Level 14, but thankfully I didn't. There were so many failed attempts where I'd finally get platinum on the Relic for Level 14, only to have 1 of the 17 bajillion possible things possible go wrong on Level 31's Box Gem. There were times I wanted to cut both levels to their own segment. Know that I hate this segment with a passion, and that I was stuck on it for quite some time.
Notable, Level 31 has got to have the easiest Platinum Relic of all the levels. You can skip numerous time boxes and still get Platinum. The better times you see on the 1st and 2nd rank are times I get when I actually try to hit all the time boxes.
Segment 6 in 0:13:52 (41% complete)
Surprisingly, a segment as long as this one proved to be very easy. Maybe because there's not a lot of time-saving shortcuts that are possible in the Crash platforming levels, but there's not much to be able to do other than complete the levels.
Segment 7 in 0:09:13 (49% complete)
This the first world where I chose to do the Crash platforming levels before the vehicle-based ones, that's because I already have 2 masks, which is usually what I hope to obtain from the trails in the vehicle-based levels. I knew that getting all the gems in Level 20 in one trip was certainly do-able, but I also realized it was going to be a total bitch. So, I chose to make it the first level in the segment.
In Level 19, I learned the hard way that no matter how fast you go, after death-warping back to the checkpoint, there's no way to make it past the door without running shoes.
Also notable, this is the only segment where I hybrid a vehicle and platforming level in the same segment. I chose to do Level 17 because Segment 8 is a total bitch, and doing this easy part here was going to make the next segment that much easier (though it still is hard as hell).
Segment 8 in 0:09:35 (54% complete)
A lot of history is involved with this, otherwise, eventless segment. It's the only reason the publish date of this run touches 2007. I pretty much cover everything I care to mention about this run in the final paragraph of the introductory comments.
Segment 9 in 0:14:17 (63% complete)
I'm guessing in the original route plan I had segment 8 and 9 as one segment. But that was probably proving too stressful to do and shortened it to two. In Level 23, you'll notice I try to include the cannon in getting far away boxes faster.
In Level 25, I originally planned to get the box gem from the colored gem path, but I forgot that I didn't have the yellow gem yet and had to put that on hold until the final segment.
Typically, in my segment routing, I place the hardest levels in the beginning so I fail in the beginning of segment attempts to save time. I shuffled the level order for this segment around for a bit. At first I was having trouble doing Level 21, so I placed it in the beginning. Then it became Level 23 that was the problem. Oh, and for some reason I thought getting hit while having a mask would kill the final target enemy, what a dumb assumption to make at the end of a segment.
Segment 10 in 0:10:12 (68% complete)
This is a fine time to share my love of vehicle-based levels late in the game. Hence the sarcasm, I hate them, all of them. In Level 22, it's almost unfair how in the double ramp part, if you go too fast, the collision detection fails, and you "miss" the second box. You have to slow down to hit it, and even then it's still not guaranteed.
I could write a 45-page essay on how Level 24 is NOT fun, and NEVER will be. And if a level is not fun to play, then to speedrun it would be taking the boredom multiplied by infinity squared.
Also, I'm considering writing to Naughty Dog one of these days and tell them about their bosses and how easy they are to bea ... FINALLY, the running shoes!
Segment 11 in 0:04:27 (70% complete)
Goddamn this segment to hell. Getting platinum relics had best been worth it! I was stuck on this segment for 3 days. What I hated the most about this segment was the fact it was the only thing left to do before I could finally get to the fun part of this entire speedrun, RUNNING SHOE TIME TRAILS!
Segment 12 in 0:05:36 (73% complete)
Whew! What a blast! Obviously, I get a Gold Relic on Level 16 because it's faster in real time to skip the four 1 second time boxes in the beginning.
Segment 13 in 0:06:03 (76% complete)
I'm just as surprised as you are about how much I skip to save real time and still achieve platinum on Level 19.
I got stuck on the walls, walking on them too much and prevented me from getting a Platinum relic on Level 15.
Segment 14 in 0:07:37 (80% complete)
Yuk, I can't believe after the carpet glitching mess-up that my time on Level 13 was my best time overall, I really should practice more.
For Level 11's and 6's Relic, it's faster in real time to skip the time boxes for a Gold Relic.
Segment 15 in 0:07:49 (85% complete)
The contrasting effect is very strong in this segment. I say that because you'll see a fully-equiped Crash run through Level 9's very basic bonus level. I skip the 2nd mask to abuse it in the other half of Level 9. That level's box gem proved to be very hard, memorizing the times of each hidden spear in the ground on the purple gem path was very hard.
Level 7's relic is the only one where you'll see me affect level conditions before hitting the timer. Common logic should explain why I did what I did in the beginning. I also skip numerous things to save real time for the Gold Relic.
Segment 16 in 0:06:40 (90% complete)
We return to those vehicle-based levels late in the game. I will once again rant about how unfun they are, reveal that I hated playing this segment, and proceed to brag about how uber the next level is.
I can go on about how many hours I wasted trying to see if it was possible to get that yellow gem from Level 7's entrance. It's possible to make it to the platform with it, but I think it's only placed there from the Level 27 entrance.
Now for the finale of the segment, getting a Gold Relic and the Red Gem Path Gem in the same trip! This proved to be difficult, because in the Gold relic time is 1:41:xx, so I don't have room for ANY mistake.
Segment 17 in 0:09:10 (98% complete)
Allow me to remind everyone that it's not easy, at all, to get Platinum, 1st place, and the Box Gem for Area 51.
As much as I've tried, I still can't get a Platinum relic and the Box Gem in Level 30, I have to slow down for the final 3 boxes.
You'll notice I play a little more "Grandma-like" in Level 29/19 because I'm very afraid to ruin this segment after getting Level 28 done right (which only happened once every 12 tries).
Segment 18 in 0:11:23 (105% complete)
Obtain the 45th gem, achieve 105% complete, and the run is over.
The final segment. Of course I have some sort of finale planned for making it this far into the run. For Level 32, for those unfamiliar to Crash Bandicoot 3, controlling the Eggapus-rex looks far easier than it actually is. Especially whenever I'm halfway into 32 and have to make those tight jumps and turns.
For Level 11's second trip, for the Box Gem and such, I get hit on the Eggapus-rex on purpose to show that it runs away into nothing. :D
The Finale, I complete as much as Level 25 as I can without light. I do this in both the Colored Gem Path Gem and for the Relic. I could've done the entire level without light, but judging when the alligator jumps and a few other objects prevented me from getting a Gold Relic. That, and it's late into the segment, which means I can't take too many risks.
» » » O u t r o
Thank you for waiting for the 6 month plus gap between finishing this run, and having it on SDA. I want to apologize to those who continually supported me throughout the progress of Crash Bandicoot 2 and 3 for having to wait this long to see the sequel speedrun.
So, now that I have reclaimed and fixed every speedrun from my regrettable n00b days (The Sonic Adventure one was obsoleted, so same difference), what's next? I'm not quite sure what my next project of interest will be. I know I'm interested in the following as of January 2007 (In order of what will likely be accomplished):
Thank you for reading my comments on my speedruns. For more « t r i . h e x » coverage, check out my latest batch of Yoshi's Island 100% individual level improvements.
Return to the Game List, the FAQ, or the Home Page.