Oh, that's awesome. Do you think you can find resterror's WR video?

If there's no videos, then the only solution is to make new ones.

It's a shame that resterror's WR has been lost though. I think I could get down to a low 20:0X with good execution, but 19:36 is impossible with my route. I would love to know what he did to get that time.

Well if anyone does run this game with the death glitch, I'd suggest renaming this category to "Any% Glitchless" and the new category to "Any%".

At some one we should add more details in the rules to make sure when a run is done on emulator we can see the inputs, the rom is valid and it's not a TAS (no movie).

I worry about discouraging runners from getting into this game by making the rules too strict. The last thing I want to do is reject a new runner's first run because they weren't showing their inputs on emulator or they were using Fusion. Requirements like that are best limited to top times (e.g. sub 23) IMO.

No, you can't allow L+R inputs in runs. It's not doable on original hardware and can only be done on emulators with key binding or apparently third-party hardware with terrible controllers. Essentially banning original hardware from serious runs of this game is not a direction that I want to go down.

I guess you could go with a separate category, but I don't like categories based on emulator hacks. At that point, you might as well have categories for game genie codes or ram editing in bizhawk. I'd just file this glitch away as a novelty, but save the application of it for a TAS.

I do like the idea of the first frame that your character sprite appears, but that would shave about 1.2 seconds off the run. You'd likely have to start the timer with a -1.25.

It makes sense though because that is when player input starts.

Unfortunately, it is perfectly normal and there's nothing that you can do about it. Most levels don't have enough lag to cause a problem, but others -- like 5-5 -- can be brutal. Trying to kick that last ninja before the final boss door is like a 50/50 chance. Mashing the ninjitsu attack is a solid backup to save your life in the case of a knockback, but of course, it does take a few seconds. Also, you can change your route a little to compensate. For example, on 5-4, I jump over the final three red ninjas instead of kicking them because I don't trust the game to read my attack input.

Just to clarify, it's the Master System version where you can do the early Phantom Castle entrance. To quote from the video:

On the Master System version, if you do this, you'll get a game crash or if you do it one frame later you'll get into the Phantom's Castle early.

There's two, one-frame windows. The first frame causes you to load an invalid level which causes a red screen of death crash. The second frame enters the Phantom Castle early saving 1.6 - 1.7 seconds. All other frames have no effect.

It's risky, but if that 1.6 seconds is the difference between a dead run and a PB, you have to go for it. If I go for it, I mash after the Island cutscene and hope for the best.

I recently discovered a very interesting glitch that completely breaks this game. If you die while getting damage boosted into a life potion -- therefore dying with HP -- the game thinks you completed the current stage and will advance you to the next stage. This works in every stage with a life potion, including all the boss levels and the final boss. This glitch works in both the SMS and GG versions of the game.

I haven't done any runs yet, but this would result in an enormous time save and make the current WR time completely uncompetitive. I'm thinking we'll need any% and glitchess% categories for both the SMS and GG boards if anyone submits any runs with this glitch.

Here are videos showing the glitch in action.

SMS:

GG:

Wonder Boy in Monster Land is one of the few SMS games that don't work with the Genesis controller. This is true whether you're using the Power Base Converter or a flash cart. I play on a Genesis with an Everdrive Pro, so this prevents me from playing the game. I could of course plug in an SMS controller, but given that the D-pad on those things were made by Lucifer himself, I'd rather not.

There is an IPS patch that fixes the game and makes it work on a Genesis controller. Would it be allowed to use this patch and submit runs or am I going to have to bust out BizHawk?

This website explains the fix.

https://krikzz.com/forum/index.php?topic=1203.0

I believe these games don't work due to the Japanese/Export check the games perform. Part of the region check toggles the select line on the controller port. It's supposed to be reset but these games fail to do that.

I patched the games to reset the select line or bypass the part of the region detection code that writes to the select line.

Ultimately I decided to abandon the RNG manip for the bolts. It's mentally demanding and you do lose a second doing the Library before the Clock Tower. What I ended up doing instead is simply playing consistently and making a note of each bolt pattern that I encountered. Instead of manipulating the bolts, I learned to read the pattern and react accordingly. I ended up with these notes. D means the bolt jumps down and U means the bolt jumps up. Based on the first 2 (and sometimes 4) bolts I was able to figure out the pattern I was on.

• DD-DU-UD-UU-DU

• DD-UU-DU-UU-DU

• UU-DU-UD-UD-DU

• UU-DD-DU-UD-UU

• UD-UU-DU-DD-UU

• UD-DD-UUD-UU-DU

The tenth bolt is the one you need to bounce on to reach the wall, so for the last pattern in my list I simply waited for an extra bolt before falling down the ladder to pick up the gem. All the other patterns worked well. Not sure if this helps anyone else since we're all going to play a little differently, but maybe it will.

That said, this RNG discovery was still really helpful. If I hadn't known there was a pattern in the bolts, I never would have considering writing the various patterns down.

I'm a bit late, but I just saw this thread.

Do what you have to do for your health -- that always comes first. Besides, if there truly is any advantage to playing on keyboard, I'm sure it's negligible. It's not like you can move faster or jump further with a keyboard.

I submitted an 18:13 run that "ties" the WR, but I'm actually a few tenths of a second behind the WR. I believe we should enable milliseconds when submitting runs and require milliseconds for top times of runs under 18:20.

For precise timing, I propose the first frame when the Master of Darkness graphic begins to fade out during the title screen (the first frame the colour changes) and end on the first frame when we see Dracula's "leaning back, dead" animation frame.

If there's no item or enemy sprites to capture on the screen, you can go to SMS -> Settings and set Display Objects to false to hide Asterix/Obelix. Assuming you're using BizHawk, of course.

I like the small gap between screens too. It's nice to know where the screen will scroll.

It's an interesting question. It would be nice to keep all the Golvellius versions in one board, but given the substantial changes made to the two ports I'm inclined to agree. Marble Madness takes the same approach by creating a separate board for each port.

This is an old post and the OP hasn't logged in since making it, but just to be clear to anyone reading there's no randomness here. The delay before the fade to black is normal and consistent.

Yeah, this is definitely needed. I just submitted a 15:26, so that's 2 seconds off petaQ's time. A month ago we had a WR tie at 15:34. We're both close and there's not much time left to save in the run. I like your idea of only requiring milliseconds for sub 15:30 times.

Adam_ak and btrim are the full mods, but neither has logged in for over a month.

This page suggests it's an input buffer on pause.

https://www.smspower.org/forums/15144-SMSCastleOfIllusionVersionDifferences

That looks awesome. I love the animated clouds.

Well that would be a game changer. Any footage of the glitch being performed?