Rayman 2 Forever speedrunning

[ScalieDan]

This is the main topic to discuss the speedrunning of Rayman 2 Forever for the Game Boy Color. Share the latest news, runs and techniques here, all while competing with the fastest runners!

Links to leaderboards:

• Rayman 2 (GBC)

[hoodlumsworld]

To start the discussion, I'll post the current leaderboard here:

1. Zarcatus: 31m 57s
2. Thextera: 1h 14m 46s

[SpeedBot]

A new run has just been validated on speedrun.com!

Any% in 41m 03s by SkeletKasket - 2nd place

Submitted: 31/07/2024
Platform: Game Boy Color

[hoodlumsworld]

Our first Rayman 2 Forever run in three and a half years! GG SkeletKasket on the 2nd place.

[Zarcatus]

A guide on how to beat the 31:57 and an evaluation of whether a sub 30:00 is realistic

Rayman 2 GBC currently sits at six full game runs by three runners and ten runs of the bonus stages. With just 16 total runs verified on the leaderboards, it shares the 22nd out of 28 places in terms of lifetime activity, only outperforming Rayman 3 (Mobile), Rayman Designer, Rayman Brain Games and Rayman 60 Levels. There has never been serious speedrunning competition for Rayman 2 GBC and my 31:57 is now almost 8.5 years old. With this reply, I'd like to point out where time can easily be saved and then evaluate whether a sub 30:00 is realistic.

First, let's only take old strats into account and look at potential timesaves. Speedrunners usually use Livesplit in order to keep track of their timelosses on each segment of a run. When I ran the game, I used a split for each indivudal screen. My current splits feature 17 attempts, 9 of them being finished runs. I compiled the splits and now I can show you the following comparison:

Segment times of the 31.57.550 are seen on the left / Best Segments are seen on the right
Fairy Glade: 2:09.667/2:04.605
Marshes of Awakening 3:21.425/3:16.900
Whale Bay 2:51.782/2:44.310
Cave of Bad Dreams 3:03.397/2:53.370
The Canopy 2:32.227/2:25.780
Sanctuary of Stone & Fire 2:52.233/2:47.900
Echoing Caves 3:05.615/3:00.030
The Precipice 3:49.154/3:41.350
Sanctuary of Rock & Lava 3:56.750/3:43.750
Tomb of the Ancients 4:15.300/3:57.700

Because I used splits for individual screens rather than full levels back then and now compiled those, potential timesaves are more apparent now. For example, I never got a 2:04.605 Fairy Glade. That time is a sum of best sub-segments itself, a sum of my best individual time on screen 1, 2 and 3. This way, the low number of attempts is compensated to some extent. With that, the biggest timelosses amount to:
1. 17.600 seconds on Tomb of the Ancients
2. 13.000 seconds on Sanctuary of Rock & Lava
3. 10.027 seconds on Cave of Bad Dreams
4. 7.804 seconds on The Precipice
5. 7.472 seconds on Whale Bay
6. 6.447 seconds on The Canopy
7. 5.585 seconds on Echoing Caves
8. 5.062 seconds on Fairy Glade
9. 4.525 seconds on Marshes of Awakening
10. 4.333 seconds on Sanctuary of Stone & Fire

That's 81.855 seconds of timeloss in total or 2.561 seconds of timeloss per minute. The potential timesave alone should be enough to highlight how very improvable this run is. On top of that, there are not a lot of hard strats, with the Sanctuary of Rock & Lava probably being the hardest level. Due to that, most of the timelosses don't stem from hard tricks I messed up but rather from a lot of small mistakes adding up. As a result, investing a bit more time would inevitably lead to a cleaner and ultimately a faster run already. This is particularly obvious on the first and second screen of Tomb of the Ancients where I didn't know which route to take by heart and how exactly the enemies move, leading to a bit of hesistation.

But there are also new strats. Let's take a look at them:

Fairy Glade 3 - Shortcut:
This strat saves about four seconds and it's easier than the intended route as hitting the bomb at the top is not a necessity. Due to that, cycles later in the level should be more consistent too.


Marshes 3 - Slope Skip:
The slope you're sliding down has two reinforcements. Use the second one as a visual queue. Turn right just before passing it and jump right after hitting it. Bumping into the ceiling resets all momentum at once and allows you to hover to a reach a ledge you're not intended to reach. It's very easy and saves 16 whole seconds.


Sanctuary of Stone & Fire 2 - Plum Skip:
Every Rayman game needs a plum skip. This strat certainly looks scary but it seems pretty consistent to me. Hitting the prickly ball allows you to jump again for some reason, preventing a certain death in the lava. Be sure to never release right while performing the trick. But keep in mind that the stakes are high here as messing up the trick amounts to a timeloss of about 28 seconds.


The Precipice 1 - Slopes Optimisation:
This optimisation is a little tricky to pull of optimally but even pulling it off unoptimally will usually still be faster than not going for it at all.


Oh, and there is also the MADMAN SKIP on Echoing Caves 2:
I discovered this strat in 2017. The idea is to take intentional damage and jump while touching the green liquid which would otherwise instantly kill Rayman. The trick can be applied in other places of the game too. However, it is extremly inconsistant, failing it loses 25 seconds and suceeding only saves 9. I don´t think it´s rta-viable. You might wanna consider going for this if a run is pretty much dead or you truly want a sub 30:00.


Speaking about a sub 30:00. It's proven to be possible. The TAS of this game is a 25:35.66. But unsurprisingly, a lot of the tricks are clearly TAS only. On the other hand, the TAS is not very optimised. Most notably, it's missing the shortcut on Fairy Glade 3, doesn't go through the laser on Whale Bay 3 and also doesn't do Madman Skip.

So the question to ask should be: Is a sub 30:00 viable for humans? I believe it is. Sure, the record being almost two minutes off doesn't mean that it'd be easy. But if I had played each individual screen as well as I ever had, I would have gotten a mid 30 in 2016. The new strats (excluding Madman Skip) would allow for another 27ish seconds to be shaved off. That would land you pretty much at a 30:00 flat. However, a flawless run is far from realistic. The reason I believe a sub 30:00 will happen in spite of that is because of sheer potential of undiscovered strats. Keep in mind that the game hasn't been run a lot. In other words, with a few more strats found, an exceptional run would likely be able to break the 30 minute barrier.

Summary: Beating the 31:57 should be more than doable. Some of the new strats are going to make things much easier than they were years ago. A sub 30:00 is possible but probably only realistic if a few more strats are being discovered which is very likely to happen.

[Greengoop]

Nice guide! It’s impressive how you still top the leaderboard after 8 years!

[Torsinn]

Very interesting Zarcatus, I want to play this version casually but speedrunning it looks fun as well so maybe I'll try that sub 30!

[Zarcatus]

I like to hear that. Looking at your success with Rayman 2 speedrunning so far, you definitely have the skill required. But don't expect too much from a casual playthrough of this game

[Hunchman801]

Very interesting breakdown, no doubt there's still a lot of time to be found in speedrunning the game! I suppose it's also nice that the typical run is not that long compared to other Rayman games.

[Zarcatus]

Absolutely! I belive that a good way to find some more timesaves would be looking into the TAS, comparing it to current strats and thinking out of the box. Many speedgames that were thought to be pretty optimised ended up being broken up further (like Ocarina of Time, or even Rayman 1 if you count the new, true Any%) so it'd be even more likely that an unoptimsed game still holds secrets.

The run only being about 30 minutes long is pretty neat, especially if you don't have as much time in your day. Rayman 1 GBC is even better in that aspect tho, as well as a few other games in the series.

[Hunchman801]

It's hard for me to judge which stats from the TAS are possible for humans or not, having never speedrun the game, but after watching both runs, it feels as though yours is more advanced human strat-wise, and the TAS wins because of optimal execution, not strats that you could easily have used. That'll make it a challenge for someone attempting to beat your run!

[hoodlumsworld]

Very cool guide Zarcatus, thanks for the insights.

[Zarcatus]

So as SuperSqank created a new TAS for Rayman 1 GBC, the RTA runners started implementing GDJs (Glitched Double Jumps). They work just the same in Rayman 2 GBC so I looked into how much time can potentially be saved in this game and the answer is 2:20. I didn't expect it to be more like one minute but hey, sub 30 is free now! And we've finally got more hard strats.

(changed the link to an updated version)

I also found this free 5 second timesave (even over TAS):

[SpeedBot]

New runs have just been validated on speedrun.com!

Any% in 28m 53s by Zarcatus - 1st place

Submitted: 28/08/2025
Platform: Game Boy Color

This is the first run I'm genuinely happy about since I came back to this category. It feels like poor GDJs are the main thing to improve upon here.

Any% in 28m 43s by Zarcatus - 1st place

Submitted: 07/09/2025
Platform: Game Boy Color

This run has some of the most unbalanced splits I've seen in speedrunning.

Fairy Glade: Any% in 01m 43s 200ms by Zarcatus - 1st place

Submitted: 07/09/2025
Platform: Game Boy Color

There is a delayed GDJ on the third screen I don't go for in full game runs. You can jump off the second leaf and grab the leaf to the right of it. But considering the strats I go for in runs, this is a good IL.

Start Frame: 235, End Frame: 6427, Framerate: 60

Any% in 27m 52s by Zarcatus - 1st place

Submitted: 08/09/2025
Platform: Game Boy Color

It's probably time for me to switch to Rayman 1 GBC now. This run beats the long-standing 31:57 by over four minutes and is only 2:17 behind the current TAS. It can still be improved upon by a fair amount: The Precipice was atrocious and Tomb of the Ancients was mid, the rest was good. My best possible time is a 27:19 now. Dunno what to do about the ILs I thought would be cool to have on the boards, let's see.

Start frame is 165. End frame is 100491 I believe. So with a framerate of 60 this should be a 52.100.

[Greengoop]

Duuude those are some good runs Zarc!

[Hunchman801]

This is really cool, I looked at the detailed breakdown of time saves in your video and found it interesting to see which ones are really hard and for IL or TAS and which ones are safer to use in a full run:

GDJ stands for Glitched Double Jump. For this video, I compiled potential applications for RTA (real time attacks).

To execute a GDJ, jump while Rayman starts to walk off of a platform. At a desired time mid-air, quickly hover, cancel hover, and jump again. You should get an unintended double jump. This also works with this game's predecessor, Rayman 1 GBC.

My opinions on each GDJ:

Fairy Glade 3 (1) - An outlier: You need to keep your initial hover for a bit before doing the GDJ but not more than five frames or the trick will fail. It's inconsistent and failing it means death. I consider this to be a reset trap and would only recommend it for IL (Individual Level) attempts.

Fairy Glade 3 (2) - It's tricky to determine when exactly you should jump here. But it's worth it if you're willing to implement GDJs at all as it's early into the run and there's no big risk involved.

Marshes 1 (1) - You need to jump off the lily pad rather early. The turnaround introduces an additional layer of difficulty but it's manageable. This is likely only worth it if you're looking for a good time (~ sub 30).

Marshes 1 (2) - It often saves more than 1.56 seconds as it's hard to execute the old strat well. Failing it doesn't lose as much as you can simply let yourself fall down. This trick is probably only worth it if you're looking for a good time.

Whale Bay 1 (1) - Not too consistent and failing equals death here. This is likely only worth it if you're looking for a good time and you don't tend to game over.

Whale Bay 1 (2) - Failing it usually means falling down and losing a lot of time. The timesave is miniscule and the platform is suboptimal for GDJs as it's much shorter in the background. Not worth it, unless you're very consistent at it and seek a top tier time.

Whale Bay 1 (3) - Another outlier: You hover for a bit before doing the GDJ. It's inconsistent but if you fail, you don't lose a lot of time. Personally, I prefer to do the alternative method in runs.

Whale Bay 2 - It should be clear that this is only worth it for top tier runs or ILs.

Whale Bay 3 - If you fail this GDJ, you should slow yourself down before landing on the downslope in order to be able to pull off the old strat as backup. I'd recommend this to people looking for an okay time. Maybe don't learn this GDJ first because of the unique (early) timing for GDJs on trees which might lead to a lower consistency with other GDJs.

The Canopy 1 - It requires a turnaround and two GDJs. The second one should be done rather early. Failing it will land you in a cobweb below. Be sure to keep in mind how much hp you need for damage boosts on the next screen and when you would like to collect the next red lum.

Sanctuary I 1 - Just don't land in the lava. I recommend this GDJ to everyone who intends to implement GDJs.

Sanctuary I 2 - The only tricky part here is to pay attention to the geyser. It saves so much that I believe this is worth it for most runners who seek to go for GDJs.

Echoing Caves 2 (1) - "Madman Skip" incorporates a moving platform and you need to keep your initial hover for very long. Failing it loses 25 seconds. This is the hardest strat in the game and absolutely IL only.

Echoing Caves 2 (2) - The timing for the first GDJ is a bit different (later) because of the "3D" platform. The first GDJ is also more risky. It's probably only worth it if you're looking for a good time. The second GDJ is likely worth it for most runners and you need to jump rather early.

The Precipice 2 - You need to get a lot of height, meaning that you can fail even though you pull off a GDJ. This is certainly not worth it unless you seek to achieve a top tier time.

Sanctuary II 1 (1) - The first GDJ seems odd to me. It's extremely consistent when you first enter the level but inconsistent after. The second one saves more time than any other GDJ in the game.

Sanctuary II 1 (2) - Your positioning needs to be strong, or else you'll hit the portal and get warped to a bonus stage. Make sure to face right before landing on the ground to avoid that. The height you need off of this GDJ is superb so you'll often fail the trick and risk falling into the lava. Don't go for the optimal version of this strat unless you do ILs. The alternative methods may be worth considering.

Sanctuary II 2 - The third GDJ is the one that matters. It's rather consistent and failing it doesn't mean you'll die.

Tomb of the Ancients 1 - It saves quite a lot but it's also very late into the run and incorporates a moving platform. Be sure to learn to not fall all the way down if you fail this strat. You can regrab the platform with spikes and make it across without the moving platform. Also learn the backup GDJ if you intend to go for this in runs.

Of course, at some point you have no choice but attempt more and more of those if you're looking to improve your time, but new runners will certainly find it useful.

Congrats on the new record, and good luck with R1 GBC!

[Zarcatus]

Thank you, Green and Hunch.

I certainly hope for some more runners. It's much more interesting with more competition. But I don't have particularly high hopes for this game.

About R1 GBC: I'm pretty busy with uni at the moment so it might take me a while to want to invest time into speedrunning again. On the other hand, I learnt most tricks until Fiery Depths 1. So I might get into it earlier. Can't really tell yet.