If you're interested in maps, information about the game mechanics, speedrun techniques etc for Amped 2, go to (https://kb.speeddemosarchive.com/Amped_2.

Introduction

This speedrun has been in the works for a long, long time. Not actively, but it's been stirring in the back of my head ever since I discovered speedrunning, several years ago. I played through this game for 100% completion shortly before then. I remember it as being really difficult, but also that I had a good time with it.

Anyways, while I've been wishing that someone would pick this game up and do the speedrun I was hoping for, it's been pretty clear to me for some time that it probably wouldn't happen. However, it's not until recently that I mustered the courage to do this run.

Segmentation
Several of the challenges and many of the gaps and snowmen locations have never before been documented (at least not on the net). So if you've played this game, I think there are a few areas of particular interest. Here is therefore a list of what each segment contains and their respective start times.

(otm – own the mountain; Sp – Sponsor; PS – Photoshoot; Pr – Pro; L – Legend)

1 (0m05) Milli 1 score, otm
2 (3m25) Event 1
3 (6m35) Brecken score
4 (8m13) Brecken otm1, PS123, Pr1
5 (12m28) Bear 1 score
6 (14m06) Hood score
7 (16m03) Hood otm
8 (19m17) Event 2
9 (22m02) Bear 1 Sp123, PS123, Pr1
10 (28m09) Laax 1 otm1
11 (30m16) Laax 1 score, otm2
12 (34m23) Laax 1 Sp123, PS12
13 (39m32) Milli 1 Sp123, PS123
14 (44m03) Buller score
15 (46m11) Zealand otm1
16 (47m30) Zealand score
17 (49m30) Milli score, Sp12
18 (53m54) Milli Sp3, PS12, Pr1
19 (57m48) Event 3
20 (1h00m04) Milli 1 Pr123
21 (1h03m33) Bear 1 otm1, Pr23
22 (1h07m38) Bear 1 otm2
23 (1h09m36) Zealand Sp123, PS12
24 (1h13m34) Zealand PS3, Pr123
25 (1h18m00) Hood Sp123, PS12, Pr12
26 (1h24m33) Bear score
27 (1h27m15) Bear otm12
28 (1h31m05) Bear PS123
29 (1h33m40) Laax score, otm
30 (1h38m35) Zealand otm2
31 (1h39m56) Buller otm
32 (1h43m05) Brecken otm2, Sp123
33 (1h48m53) Milli otm 123
34 (1h54m30) Milli PS3, Pr2
35 (1h56m43) Event 4
36 (1h59m28) Milli Pr3, L
37 (2h02m12) Brecken Pr23, L
38 (2h05m37) Bear 1 L
39 (2h07m32) Hood PS3, Pr3, L
40 (2h10m18) Laax 1 PS3, Pr123, L
41 (2h14m58) Buller Sp123, PS12
42 (2h19m35) Buller PS3, Pr123
43 (2h24m05) Buller L
44 (2h25m55) Laax PS123
45 (2h28m14) Laax Sp123, Pr12
46 (2h33m40) Laax Pr3, L
47 (2h37m42) Bear Sp123, Pr123
48 (2h45m30) Bear L
49 (2h46m16) Zealand L
50 (2h47m39) Milli 1 L

FAQ

Ok, no one asked me even a single one of these questions, but that's probably not the first time that the term ”faq” has been mis-used. These are just questions that I hope will shed some light on what's going on in this speedrun and some trivia around it and the game.

Q: What's 100% in this game?

A: The game consists of 10 mountain levels and 4 events. Each mountain level is built up the same way. You have a few start locations on the mountain that allow you to ”free style” your way down. The goal of these runs is to fulfill certain score requirements and also to do some side stuff (own the mountain). Each mountain level also contains 3 sponsor, 3 photoshoot, 3 pro and 1 legend challenge. In the event levels, your goal is to finish ahead of the cpu opponents by scoring points in four different categories (total score, media score, single trick score and combo score). To get 100% in the game, you need to complete everything in the game. So all challenges, all the own the mountain goals and place first in the events.

Q: What's ”own the mountain” (otm)?

A: Each mountain level contains 8 snowmen to crash into, 5 gaps to jump and 5 tricks to perform. These can only be done during the ”free-style” runs (so not during any of the sponsor challenges etc).

Q: What's the general outline for the segmentation?

A: There are several rules that determine the segment order and how much it contains (not listed in any particular order):

• Every segment is a hefty time loss (~5-10 secs, depending on which mountain level it is, the ones at the bottom of the menu being the most time consuming to reach). If the segment number can be reduced by cramming more content in another segment, it's always beneficial, even if it means ramping up the difficulty of the segment and risk a few more execution errors.
• There is no point in combining challenges from different mountain levels. Leaving a mountain takes several seconds and is always slower than starting from a new segment. That means you need to make a plan for the whole level in terms of the number of segments you plan to spend on it. Otherwise you risk ending up with one easy challenge being left on a mountain at the end.
• Because of the above rule, I tried to keep segments on one mountain consecutive, all other things equal. In the end, that allowed me to come back and combine certain segments into one, change the order of them or mixing the content.
• Some of the content is unlocked as you progress, so it's impossible to fully complete the mountain levels one at a time.
• Some challenges get easier and some get harder as you upgrade the board stats.
• After each challenge, there is a small animation where the ranking drops one step at a time. Once you reach rank 8, the ranking can only be dropped further by event 4. All challenges completed when at rank 8 or lower will save ~1 second of menu time for each avoided ranking drop. The ”free-style” runs are different in that they don't have this animation, so I made sure to complete them before reaching rank #8 to avoid the most ”rank drop animations” as possible.
• All other things equal, it's about half a second faster to end a segment on a challenge that unlocks a video.

Q: How do we know that you didn't use cheat codes?

A: For some reason, the ”tutorial” option is grayed out on the first menu screen if you have any cheats activated.

Q: How optimized is this speedrun? Can a lot of time be shaved off?

A: Let's divide the question into four parts: routes, ”time-savers”, segmentation and execution.

”Time-savers”
This speedrun is very close to what I know about the game in terms of tricks and minor time-savers. There are a few possible improvements, but they are either small or incredibly hard.

• In segment 10, I have once or twice managed to go clockwise after landing the first gap. That's already a pretty tough segment though, so grinding it out for a rare trick that would save 2-3 seconds is something I opted out of.
• In segment 7, I didn't do the hand plant from a switch stance. Switch tricks give 20% more points, which would have sped up the trick by maybe half a second. It's not something I'm going to redo the segment over, but I'm always annoyed when I realize that I lost time without gaining anything back (like more consistency).
• When leaving the crevice in segment 16 (after the last snowman), I lost half a second by going straight, instead of going down on the right side. It's a bit unintuitive that the right side would be faster and only something I discovered when working on a different segment. Otherwise there are no shortcuts or alternative ways down the mountains that would save time, but were left out because of being too hard or because I found faster ways down the mountain after the segment was done.
• Again in segment 16, you get onto the grinding from below (the start of the crevace) instead of going around and approaching it from above. This would save a few seconds, but was hard to do consistently.
• How grind scoring worked was for the longest time a mystery to me. From early on, I noticed that the score would sometimes get a huge boost, while other times the score would be disappointingly low for no apparent reason. It wasn't until I was done with the speedrun and sat down to study grinding in more detail for the knowledge base article (linked above) that I started to better understand it. When writing these comments, I can't on top of my head point out any particular run or challenge that could be improved with my new knowledge by any significant margin, but I'm pretty sure one could save a few tenths of seconds here and there in several challenges, where grinding is used.

Segmentation
I started this speedrun almost in the blind. I had completed the game several years ago, but then never played it until now. I didn't play through the game again (it's pretty big) to decide on how to divide it into appropriate segments. This was of course a huge gamble as I could easily have found myself in non-recoverable situations (mistakes in the board upgrade strategy, poor choices of segments etc). Overall, it worked out surpringly well and I'm happy with the end result. I did make a few mistakes along the way, but not worse than something I could fix by redoing a few segments. My skill and standards obviously increased the further I came into the run. This meant I had to come back and combine some of the early segments into one, which also worked out really well. When I look back at the number of segments I used for each mountain, I don't see any glaring mistakes that I have any big regrets over. I do however see a few opportunties to reduce the total number even more, but none without a considerable effort.

• I think it would be possible to squeeze one segment out of Millicent 1. Segment 20 (pro123) is pretty light. While the legend challenge is imo the hardest challenge in the game, it's still fairly doable once you start to know the game. I'm thinking segment 20 could be divided into one part going to segment 13 and one part to segment 50. Regardless of it would be possible or not, it's too late now. Doing so would break the segmentation continuity and pretty much the whole run would have to be redone for the 5 seconds that are at stake.
• The second possibility I have in mind is in Bear Mtn. 1. I remembered the legend challenge as being very challenging and some quick tests early on seemed to confirm this, so I planned for that challenge to be in a segment of its own. When I arrived there in the speedrun, I realized that it wasn't as bad as I had first thought. Segment 21 is also on the light side, so it might be possible to do something there. Combining would be no small task though. Like with Millicent 1, it's too late to try it out now, since the segment continuity would break unless redoing a huge part of the speedrun.
• The third place I see the possibility of reducing the number of segments would be on Mt. Buller. I left the legend challenge in its own segment, because of the difficulty. While practicing and doing attempts, I discovered a nice trick that made it a lot easier. The segment therefore became much more doable than I first thought it would be. There aren't really any other simple segments on that mountain, but one idea could be to move pro 1 and 2 from segment 42 to segment 41 and then try to squeeze in the legend challenge in segment 42. Unlike the previous two cases, this one would be possible to do without breaking the segmentation continuity. In the end I decided against it, because of the effort involved.
• The fourth, and last, place is a bit of a longshot. I actually managed to complete segment 46 on my very first attempt. That obviously raises the question if it could have been combined with something else. I shrudder when I think about adding content to the other three Laax segments though. It's also worth mentioning that while it's not the heaviest segment, getting segment 46 done on the first attempt was a bit of fluke and not really representative for my consistency of the two challenges involved.

Execution
As I've already mentioned above, I started this speedrun without almost any preparations. Once I reached the second half of the game, the early segments started to look more and more questionable to me. I have also ended up redoing many of them. Some in order to combine segments and some just to tighten up the execution. Since a new segment is such a huge time penalty, getting as much stuff into each segment has generally taken priority over execution though. Each segment is therefore in general quite difficult to just complete (or at least without any major time losses). Once I defined the content of the segment and routed it, I usually did attempts until I got one successful completion (bad attempts were filtered out by resets). Same thing when I went back to improve on an earlier segment. One completion is of course not exactly a guarantee for a good time, which I'm well aware of. I'm positive that I would beat many of the segments by a few seconds by adding only a handful more completions of each segment. While even small improvements start to add up for a run of 50 segments, the time investment of such an undertaking would be quite overwhelming to me. My wish was to make sure this game got something that at least resembled a speedrun, but I didn't want it to become my life project for the sake of saving the last couple of seconds.

Q: What's up with the pro challenges? They look horribly unoptimized!

A: The pro challenges are unfortunately subject to some absolutely abysmal game design. You're supposed to follow the pro's line and then do tricks at the indicated spots. If you're in front of the pro, the game will warn you, but still accept the tricks as long as they're done in the right spot and with the correct score. The one exception is the last trick. If you're ”too far” ahead of the pro, the game will simply fail you. I'm not sure exactly how the game defines ”too far”, but if you complete the same challenge enough times and compare the completion times, you can determine a range for the minimum allowable completion time. Needless to say, this requires long and boring preparation work, but at least it's not random.
Once the target time has been pinned down with enough accuracy, a route needs to be created to hit as close to it as possible with a good consistency. That's why you'll see me take wide turns, ”reset the boarder” (when the screen goes black) and possibly other weird behavior as well. So the vast majority of unoptimizations in the run are sadly a necessity.

Q: Is this a good speed game?

A: There are some fun aspects from a speed perspective, but overall this is not a good speed game.

Before starting with the complaints, let's say something about what's positive with this game. The part I liked most was easily the score and own the mountain challenges. I had a good time designing the routes through each mountain to get as much stuff done in one go as possible and the execution was also generally challenging for these runs. Unfortunately, they are vastly outnumbered by the other challenges.

The problem with the pro challenges was explained above. Other negative things I can think of:

• The sponsor challenges are generally long, uneventful and with too little variety to make them interesting.
• When you start a run, you can either be in a sitting position or in a standing position (with either a slow or a fast start). If you're sitting down, it will take ~3 seconds to stand up. As far as I can tell, it's entirely random what start you get (I'd say sitting starts happen ~1/3 of the time). I deemed the time loss from sitting starts completely unacceptable throughout the run and always reset if that happened. For segments with multiple challenges (which is true for most segments), that meant a silly amount of failed attempts because of a sitting start somewhere and that was also a limiting factor of how much content could be put into each segment. The pro challenges are an exception to this rule as they will always start from a sitting position (pro legend challenges are not part of this exception). I mentioned that standing starts can be done with either a slow or a fast start. The time difference is maybe half a second between the two, which I chose to ignore for sanity reasons. But there is definitely some fractional seconds that would add up if someone were to be more strict on fast starts.
• This is just a small thing, but something that really became frustrating in the long run. So, the game remembers which challenge you were on when you leave a mountain (if you re-enter the same mountain level, the cursor will be on the same challenge again). This is a problem in a speedrun, because you want to reset all conditions to default (as how they would appear after a hard reset). That means you can't have the cursor suddenly pointing on a challenge somewhere in the middle. The solution is after each failed attempt to go back up and enter the first run before leaving the mountain. This only accounts for a few seconds each time, but the number of attempts needed in a speedrun like this is quite high, so it quickly became very annoying. Especially since it's due to a game mechanic that I don't see would be of much use for anyone.
• My last point is about how terribly laggy and unprecise the menu navigation is. It's also inconsistent. Sometimes the cursor flows through without issues and sometimes it briefly gets stuck on a level. Entering the mountain level you want is therefore a bit of an achievement in itself. Sometimes it felt like navigating to the challenge I wanted to, while also getting a standing start was almost like getting half the job done. Obviously, neither the problems with the menu navigation, nor the sitting starts are apparent from watching this speedrun. I made a small blooper reel though to illustrate what it can look like when doing attempts: https://www.youtube.com/watch?v=I-yaxe4zJH4

Q: When's a single segment run?

A: I guess you should never say something's impossible, but completing this game without including literally 100s of restarts is unthinkable in my eyes. I'm not saying it can't be completed in a single sitting, but having tons of failed attempts doesn't really give me a sense of a speedrun. I hope my speedrun will make you think this game is easy, but the truth is that there are plenty of challenges (mainly photoshoots), where my success rate is down to a few percent. There are also several of the gaps and a few of the snowmen that are hard to get. I haven't kept statistics, but the hardest ones are probably also approaching single digit percentage success rates for me. They aren't on a path or anything and it's pretty obvious that their purpose was to make you explore the game by first finding them and then figure out how to complete them them one by one. I don't think the intent was to do several of them in a row. So yeah, in my opinion, it would require a lot of skill to complete this in something that even remotely looks like an optimized speedrun.

Q: When's any%?

A: The main difference between any% (reach rank #1) and 100% (complete every single challenge and goal in the game) is that you would skip some of the own the mountain stuff and quite a few of the other challenges could also be skipped (there is more than enough challenges to help you reach the #1 rank). Most of the gaps and snowmen would be too far out of the way to be worth the time to collect. However, this would reduce the ”free-style” runs to straight runs to the end, while getting the score requirements done, which would be fairly straight-forward. So basically all the fun routing stuff would be thrown out of the window. A segmented any% run would therefore look mostly the same as a 100% run, minus all the fun stuff. I will however say that an any% single segment doesn't look completely unrealistic without too many failed challenge attempts. Very hard, but probably doable if someone made it their mission to get it done. But I have my doubts about this game ever being speedrun again.

Q: I spotted an inconsistency! You CHEATED!

A: He he, I doubt you did. But there are actually inconsistencies, so let me go through what a diligent observer could find:

• Some high scores have changed between segments for a mountain. This happened because I redid an earlier segment. I think it goes without saying that it's pretty much impossible to get two runs with exactly the same score. However, this has absolutely no consequence because the score itself has no impact on the game. It's only a question whether you scored high enough to complete the score requirements for the mountain or not.
• The character name changes between segments. Sometimes it's ”ktwo” and sometimes ” ktwo”. Let's take an example to understand the reason for this. Assume that the 3rd out of 5 challenges in a segment failed. Because of the auto-save function, that save file is now irreparably lost since two of the challenges will be registered as completed. To remedy this, I actually used many different save files with identical progress and similar names (they are just shown by the game as either ”ktwo” or ” ktwo”). Copying save files back and forth is a bit time consuming, so it was just faster to have a batch of fresh files copied at the same time.
• In segment 5, I completed 4 tricks, for a total of 8 own the mountain goals. However, if you look carefully on the following segments when I sweep over Bear Mtn. 1 in the menu, it will only say 7 own the mountain goals completed. This inconsistency goes on until segment 21. The reason is that I had to redo segment 5. At that time, I had some better knowledge of the game and could use some better strats. Anyways, I was unable to complete it under the same conditions without losing time. This additional trick has no impact whatsoever on the following segments though. If you complete 3 out of the 5 tricks on a mountain, you'll gain a rank. So going from 2 to 3 tricks would not have been acceptable (or would at least have looked pretty bad), but going from 3 to 4 really doesn't change anything until you come back and complete the rest of the own the mountain challenges, which I redid in this case.