I have been doing some research into the RNG in Golden Axe and found that it is based on a timer that resets at the start of each level. If the timer is on the same value, you get the same RNG patterns every time. The timer increases on every frame of the game (60fps) so it is impossible to take advantage of it.

However, I have found that both the 6-Pak and Mega Games 6 versions of Golden Axe have a glitch with this timer. For the first 30 seconds of each stage, the timer is ignored and the RNG always has the same pattern. This is very useful for levels where you can reach the thieves before 30 seconds are up, such as stage 6 and stage 7. The thieves will always use the same pattern, and on stage 7 this is a fast pattern where the thief on the left runs along the top of the screen.

I have tested on all the revisions of Golden Axe I could find and only these two games had the glitch. Both the games are also region locked so 6-Pak can only be played on NTSC-U Genesis (US console), and Mega Games 6 can only be played on PAL Megadrive (EU console). The EU consoles run the game slower at only 50fps so they are not useful for speed running. This means that 6-Pak seems to be the best version of the game to use. I have created a video to try to explain my findings here .

The physical cartridge seems to be relatively common and cheap so I don't think it is any more difficult to obtain than a normal Golden Axe cartridge. The ROM files are just as easy to find as the regular game too. However, anyone playing the game on other platforms will be at a disadvantage, such as Steam/Wii/Xbox/Playstation, as there is no way to play 6-Pak on those platforms.

I think we should update the leader board to give visibility of which submissions are using 6-Pak. But I don't want to make any changes without asking everyone in this community their opinion, so I have created a poll here https://www.strawpoll.me/20361555 .

I will leave the poll open for a couple of weeks and close it on 30th June 2020.

I would also like to propose that we update the submission rules to make cheating more difficult. At the moment the rules are very lax and it would be possible for someone to use various tricks with an emulator to fake a good time. I think the rules on Sonic for Megadrive are a good starting point so I think we should copy those (https://www.speedrun.com/s1). These rules would limit emulator use to 2 known emulators where it is possible to show that no cheats are being used in the video. For other platforms it would be required to show a game reset and/or show the system menu for the console being used.

Again, I don't want to make changes without asking here first. I am mindful of making the rules more strict as new runners could be put off, particularly if they don't read the rules before recording their first run and get their submission rejected.

I have made a second poll here https://www.strawpoll.me/20361596 . I will leave it open for 2 weeks also and close it on 30th June 2020 .

Please comment with your feedback, I don't want to rush into anything or do anything that is not accepted by the community.

I got bored and made a table of health values for all the enemies in arcade mode. Posting it here in case anyone finds it useful.

(With formatting)https://hackmd.io/RXGnpTdSQFOtWgHQr9zNmQ

I agree with Shimegabriel, I would request a new game but keep it in the Golden Axe series.

Sorry for the slow reply. I think the Amiga version is probably best having it's own game on speedrun.com. The Sega CD port is almost identical to the Genesis/Megadrive version which (I think) is the reason it is on the leaderboards here. It was added before I became a moderator. The Amiga version looks like it is closer to the Arcade version, based on the movement/attacks and it is missing the last 2 stages present on the megadrive/genesis.

GG on the run, love the music on the Amiga version.

This sounds like fun! I wasn't able to follow the link though, is it correct? I searched for golden axe tournaments on the site and only found some that had already finished.

Thanks for letting us know, I've updated the submission rules.

For anyone that missed it, well worth a watch

Wow just saw this in time. Good look, I can't wait!

To add onto the previous comments, The MS Dos version is not an identical port of the genesis/megadrive game so I would say it is not fair to play it in a race (it would probably be a disadvantage to play on dos). There are plenty of options for playing the genesis/megadrive version, the Wii version is particularly good. A general point on megadrive, if anyone is planning to play on original hardware the European models run the game at a slower refresh rate making the game slower overall (this is also true for most megadrive games). Finally, as someone who has played the steam version a lot, I ultimately abandoned it as it runs very poorly and can lag on older PCs. I would highly recommend an emulator over the steam game.

A TAS maker ThunderAxe31 has recently been able to make the below damage table for Tyris' attacks. I thought it would be useful to share it. It was created by watching the RAM addresses in Bizhawk emulator, and the address values are below too. Hopefully this will be useful for someone looking for new strategies as you can leave the RAM watch open and see the current health values of each enemy on screen.

From: http://tasvideos.org/forum/viewtopic.php?t=20436

Damage of Tyris Flare's moves 16 huge dropping stab 12 reverse swing 6 jump swing 4 flying kick 4 frontal stab (only against beasts) 4 standing kick (only against beasts or during combos) 4 throw 2 most combo attacks

Useful RAM addresses (BizHawk .wch format) SystemID GEN 00D164 b s 1 68K RAM enemy 1 health 00D1E4 b s 1 68K RAM enemy 2 health 00D264 b s 1 68K RAM enemy 3 health

It's happening December 2018 in Japan http://spiralchariots.com/goldenaxe

In case anyone missed it you can watch the VODs here:

Golden Axe - Streets of Rage 2 -

Probably left this a bit late. But in a matter of hours 2 Golden Axe runners, InTheMug and Anthopants will be running Golden Axe and Streets of Rage 2 at SGDQ.

If anyone wasn't already aware, go check the twitch stream and and cheer them along!

https://gamesdonequick.com/schedule

Best of luck to you both!

Yes you beat the game, and this doesn't affect the end time. This has happened to me a couple of times but only on the steam version. I think it can probably happen on all versions it's just not very common. Certainly made my laugh whenever it does though :)

Almost all enemies will move down when you attempt to "down stab" them. Your options to get around this all require some precise positioning and timing, they are:- -Attack as the enemy enters the screen (some enemies follow the same path every time to enter the screen, and whilst doing so cannot dodge) -Attack when an unmounted animal is on screen, (non-giant) enemies always attempt to mount it and won't dodge whilst doing so -Attack at the bottom of the screen, enemies won't be able to dodge down (but sometimes move up as part of their "surround" AI) -Attack when the enemy is standing up. Enemies all stand up with different speeds so you need to learn each to be able to time this correctly -Attack as the enemy is attacking. You can lure an enemy to charge you if you are far enough away, during this attack they cannot dodge. If timed correctly you can also land a down stab when an enemy is performing a normal swing attack. -Attack from a very specific distance. With a lot of practise, most (but not all) enemies can be down stabbed if you start your jump at the furthest possible distance. This is easiest on skeletons for some reason.

As EightbitSteve suggested, it's worth watching some runs from the leader board to see how these methods are used.

Gave in and had to scratch the itch. I added the ability for runners to submit a time with milliseconds to the leader board. I added milliseconds to the WR run for Tyris. I am willing to re-time anyone's run to add milliseconds on request. I'll leave it up to runners choice when they submit a run whether they provide milliseconds or not, and just time to the accuracy they provide. Unless it's a tie or something. This is reversible though so if it's an unpopular decision I can change it back.

Thanks for the input, I think I'll take your advice and wait until we have a decent reason to time more accurately. The app is nothing really. I write code for a living so whenever I have a repetitive task it usually ends up getting automated :P

Fair enough, I can move the Sega CD runs that were submitted for Gilius and begginer mode into the Sega CD leader board and update the rules again. I can then remove the Sega CD option from the platforms allowed in the other leader boards.

I agree that it should be easy for runners to consistently time themselves, and I hadn't put much thought into it whilst posting. I have stubbornly kept to my own timing method for this precise reason. I can more accurately hit the last split when Death Bringer's axe impales him at the end of his death animation, rather than trying to split as I land the last blow. But I'm not going to argue the case for swapping to that as our current rules are quite established (and it would make a sub 8 min run harder !). I don't think the 3 frames it took for the screen to change in my example is really significant, but I have noticed that some runners don't start their timer as they hit the select button to start the run, and can start shortly afterwards.

I take your point that this isn't really necessary until two runners are tied again. Although I think it would be a shame if someone got a sub 8 min run with 7:59.6 and I had to round it up.

In terms of how hard it is to time a submitted run, I find it easier to frame count with Yua now, where previously I had run my own timer whilst watching the run and hitting the start and end split manually. I just take the start frame and the end frame and plug it into a html app I made http://house-keeper-spots-88787.bitballoon.com/ (30fps for youtube videos, 60fps for my own recordings).

For now I guess we don't need to do anything, I certainly don't think it's worth re-timing all the runs on every leader board.

For now, I've updated the rules for the Sega CD category to be Tyris Flare only. Other Sega CD runs will be added the the leader board for their relevant character category. Wasn't sure if it made more sense to have all Sega CD runs in the same leader board or not.