1-Up!
Game Page: http://speeddemosarchive.com/AlexKiddMiracleWorld.html
SS deathless run
Verifier Responses
Verifier 2 has more comments below
Decision: Reject
Reason: I had a tough time with this one, but after reading the verifier comments and watching it a few times myself, I just felt like it was a bit too sloppy considering the length of the run.
This run will be available for a month. After that this link will no longer work.
http://queue.speeddemosarchive.com/queue/verificationfiles/35/
SS deathless run
Verifier Responses
Quote:
Audio is great other than a buzzing sound in the background that changes along with the video signal.
Video is surprisingly good considering the system only ever had coaxial output.
Run seems free of cheats.
This was my first game for my first very own console, and so I have both great love for and familiarity with this game, and the runner seems to share my history with it. Therefore I kind of wince every time I see any room for improvement, which I do in this run. The runner's notes do mention how odd this game's momentum can be, but I know there are slight compensations you can do to make it work for you (I have also seen this level of control in other speedruns). It might be that I judge this run a little too harshly because of my experience with the game, but I see a lot of room for improvement from more efficient movement. Each chance for improvement equals less than a second of time gained, but time does add up. Other possible improvable parts include when the runner punches the paralyzing block in the water section of level 1, something I'm not sure is necessary to get past the frogman quickly, crashing the motorcycle early in level 15, and falling down twice in the room in Janken's Castle (at 12:52) in the video.
All this does add up, and rougly estimated I do think this speedrun is improvable by at least 30 seconds. It seems to me that this run is only marginally faster than the PAL version speedrun by Konnan would be if it ran at NTSC speed, with the only major improvement being the final boss skip. For a game as short as this I think there are missed opportunities where the runner could have trained to do screens much quicker. One example is not waiting at the room with the 3 flames on the floor in level 11 (07:49 in the video). There is a non-TAS run by "talbot1939" on Youtube that does this room just as fast as Keirden's TAS does it.
Despite my nitpicking I must confess this is a very good run in a game that is very easy to fail at. The improvability I suspect of this run crashes with how much I desire this game to get some well-deserved love on the site. This leaves me rather tentative. I wish the runner's notes included something about how long the runner worked on this, and how safe he played it.
As for my timing:
Player presses start: 0:12
Player assumes control: 0:16
Game completed: 15:37
Run time from player presses start: 15:25
Run time from player assumes control: 15:21
But enough floundering. I have to give a verdict.
My desicion is to ACCEPT this run... but I would like to encourage the runner to try for a sub-15 minute time in the future.
Video is surprisingly good considering the system only ever had coaxial output.
Run seems free of cheats.
This was my first game for my first very own console, and so I have both great love for and familiarity with this game, and the runner seems to share my history with it. Therefore I kind of wince every time I see any room for improvement, which I do in this run. The runner's notes do mention how odd this game's momentum can be, but I know there are slight compensations you can do to make it work for you (I have also seen this level of control in other speedruns). It might be that I judge this run a little too harshly because of my experience with the game, but I see a lot of room for improvement from more efficient movement. Each chance for improvement equals less than a second of time gained, but time does add up. Other possible improvable parts include when the runner punches the paralyzing block in the water section of level 1, something I'm not sure is necessary to get past the frogman quickly, crashing the motorcycle early in level 15, and falling down twice in the room in Janken's Castle (at 12:52) in the video.
All this does add up, and rougly estimated I do think this speedrun is improvable by at least 30 seconds. It seems to me that this run is only marginally faster than the PAL version speedrun by Konnan would be if it ran at NTSC speed, with the only major improvement being the final boss skip. For a game as short as this I think there are missed opportunities where the runner could have trained to do screens much quicker. One example is not waiting at the room with the 3 flames on the floor in level 11 (07:49 in the video). There is a non-TAS run by "talbot1939" on Youtube that does this room just as fast as Keirden's TAS does it.
Despite my nitpicking I must confess this is a very good run in a game that is very easy to fail at. The improvability I suspect of this run crashes with how much I desire this game to get some well-deserved love on the site. This leaves me rather tentative. I wish the runner's notes included something about how long the runner worked on this, and how safe he played it.
As for my timing:
Player presses start: 0:12
Player assumes control: 0:16
Game completed: 15:37
Run time from player presses start: 15:25
Run time from player assumes control: 15:21
But enough floundering. I have to give a verdict.
My desicion is to ACCEPT this run... but I would like to encourage the runner to try for a sub-15 minute time in the future.
Quote:
Audio and video are good. I don't seem to hear the buzzing sound Verifier 1 mentioned.
I had worked on a speed run through this game for a short while some time ago, before dropping the project in favour of other games. Alex Kidd is far from an easy game to speed run. Inertia can make it tricky to land on the desired spot, and many mistakes result in instant deaths and restarts. I commend the runner for taking up this game. For the purpose of verification, I spent a lot of my day testing various things in an emulator (because I only have the 50Hz SMS copy at hand).
Here's a level-by-level commentary on the run (split into 4 sections so that it's easy to open and close).
Overall, I see around 20-25 seconds worth of improvement, which is a non-negligible amount of time. My initial thought after watching the run was "This is somewhat sloppy in more than a few spots.", and after dedicating an entire day to scrutinise it, I am still unsure what the verdict should be. Perhaps I am unnecessarily harsh on the run, because it's a rather difficult game. I will spend a bit more time thinking about it and seeing what the other verifiers think.
I had worked on a speed run through this game for a short while some time ago, before dropping the project in favour of other games. Alex Kidd is far from an easy game to speed run. Inertia can make it tricky to land on the desired spot, and many mistakes result in instant deaths and restarts. I commend the runner for taking up this game. For the purpose of verification, I spent a lot of my day testing various things in an emulator (because I only have the 50Hz SMS copy at hand).
Here's a level-by-level commentary on the run (split into 4 sections so that it's easy to open and close).
Level 1 - Mt. Eternal
The mistake at 0:30 should've resulted in a reset. There is no need to break the two blocks, as one can simply jump over them after picking up the rightmost bag of money. Here's a video of it: [CLICK ME]
Rather than picking up two bags in the surface section, you can also quite quickly pick up two bags of money near the exit to the level. The time difference for this is marginal, and breaking blocks underwater can be tricky, so I don't blame the runner for not doing so. Not punching the funny-block while underwater is a lot safer, without being slower than punching another block, due to that damned frog, so the runner picks the right route here.
Level 2 - Mt Eternal 2
Good stage. It's trivial to get through it quickly, as there are hardly any obstacles for the bike, and the runner made no mistakes. Curse those walls of text before the Janken fights.
Level 3 - Lake Fanthom
Good stage. Unfortunate bump into the octopus. The swimming controls are awkward, though, since Alex keeps floating up all the time.
Level 4 - Island of St. Nurari
Very unfortunate mistake here. The runner recovers from the mistakes well, though, and not too much time is lost (a bit over one second). This level is a piece of cake, and the missed jump wasn't hard. Damn the walls of text in this game.
The mistake at 0:30 should've resulted in a reset. There is no need to break the two blocks, as one can simply jump over them after picking up the rightmost bag of money. Here's a video of it: [CLICK ME]
Rather than picking up two bags in the surface section, you can also quite quickly pick up two bags of money near the exit to the level. The time difference for this is marginal, and breaking blocks underwater can be tricky, so I don't blame the runner for not doing so. Not punching the funny-block while underwater is a lot safer, without being slower than punching another block, due to that damned frog, so the runner picks the right route here.
Level 2 - Mt Eternal 2
Good stage. It's trivial to get through it quickly, as there are hardly any obstacles for the bike, and the runner made no mistakes. Curse those walls of text before the Janken fights.
Level 3 - Lake Fanthom
Good stage. Unfortunate bump into the octopus. The swimming controls are awkward, though, since Alex keeps floating up all the time.
Level 4 - Island of St. Nurari
Very unfortunate mistake here. The runner recovers from the mistakes well, though, and not too much time is lost (a bit over one second). This level is a piece of cake, and the missed jump wasn't hard. Damn the walls of text in this game.
Level 5 - St. Nurari 2
Missing the money bags is an easy mistake to make, because the helicopter's altitude is not that easy to control. I experimented with an alternative route here for a while (going down instead of grabbing those two bags of money between the two red blobs), but it ended up being hardly quicker than the route taken by the runner while being a lot more risky. Apart from the 1.5-2 seconds lost near the first money bags, the level was fine. The section above the water has some really interesting design, by the way. It simply forces you to do the same short section 4 times before suddenly making a rice-ball appear. Swimming is much faster, fortunately.
Level 6 - Village of Namui
Not grabbing the helicopter is a good choice. That thing is a pain to control in this stage, and is not quicker than simply walking. I'm not sure why the current SDA run does make use of it, to be honest. You should've grabbed at least a few of the bags of money from that huge floating pile.
Level 7 - Mount Kave
This level has a lot of tiny, sloppy mistakes. It started off well, but turned into a bumping galore, wasting precious time.
Level 8 - Blackwoods [typo in the runner commentary]
You shouldn't slow down near the very start. You can hold right throughout the entire stage here, until you reach the bossfight. At the bossfight, you can simply drive straight through him and start punching as soon as your bike breaks. Here's a video to show you how to do this stage: [CLICK ME]
Missing the money bags is an easy mistake to make, because the helicopter's altitude is not that easy to control. I experimented with an alternative route here for a while (going down instead of grabbing those two bags of money between the two red blobs), but it ended up being hardly quicker than the route taken by the runner while being a lot more risky. Apart from the 1.5-2 seconds lost near the first money bags, the level was fine. The section above the water has some really interesting design, by the way. It simply forces you to do the same short section 4 times before suddenly making a rice-ball appear. Swimming is much faster, fortunately.
Level 6 - Village of Namui
Not grabbing the helicopter is a good choice. That thing is a pain to control in this stage, and is not quicker than simply walking. I'm not sure why the current SDA run does make use of it, to be honest. You should've grabbed at least a few of the bags of money from that huge floating pile.
Level 7 - Mount Kave
This level has a lot of tiny, sloppy mistakes. It started off well, but turned into a bumping galore, wasting precious time.
Level 8 - Blackwoods [typo in the runner commentary]
You shouldn't slow down near the very start. You can hold right throughout the entire stage here, until you reach the bossfight. At the bossfight, you can simply drive straight through him and start punching as soon as your bike breaks. Here's a video to show you how to do this stage: [CLICK ME]
Level 9 - Blackwoods underwater
Good stage. It's pretty much hold-right-to-win, but that last frog can be a real pain if you bump into him.
Level 10 - Bingoo Lowland
Good stage. Around 6:15 you flew up, instead of simply shooting the block and going straight through the wall. The money bag you grab this way wasn't really necessary and could've, for instance, been obtained in level 6 instead. Nothing big, though. I've always hated those clouds, and you got through them perfectly. I agree with the runner that crashing the copter on the cloud would be an unnecessary risk. It's not terribly hard to perform, but given the time it saves, I would not recommend doing it.
Level 11 - Radaxian Castle
Shame you didn't get the correct two rocks in the first jump punch. At 7:35 you could've broken another rock to move a bit further to the left to avoid the bat crouch. Here's an (imperfect) video for this: [CLICK ME]
In the room with the three flames, you only have to wait for the first one to pass by. You can use the ladder to go down as soon as the left-most flame has bounced off. This isn't trivial to do, but not terribly difficult either, and given the amount of time it saves I think it does belong in a run for this game. Here's a video of the trick: [CLICK ME]
I'm a bit puzzled as to why you jumped at 8:01. Getting nervous from getting so far into the game?
The remainder of the level is good. The use of invincibility is quite necessary to get past that last, darned flame. Good work. Gotta love the appropriate strategy of beating Stone Head with two rounds of 'paper'
Level 12 - City of Radaxian
That first jump was quite a bit too late. Other than that, perfect execution. Well done. Beating Scissors Head with two 'stones' is as amusing as ever. That last punch was a very close call!
Good stage. It's pretty much hold-right-to-win, but that last frog can be a real pain if you bump into him.
Level 10 - Bingoo Lowland
Good stage. Around 6:15 you flew up, instead of simply shooting the block and going straight through the wall. The money bag you grab this way wasn't really necessary and could've, for instance, been obtained in level 6 instead. Nothing big, though. I've always hated those clouds, and you got through them perfectly. I agree with the runner that crashing the copter on the cloud would be an unnecessary risk. It's not terribly hard to perform, but given the time it saves, I would not recommend doing it.
Level 11 - Radaxian Castle
Shame you didn't get the correct two rocks in the first jump punch. At 7:35 you could've broken another rock to move a bit further to the left to avoid the bat crouch. Here's an (imperfect) video for this: [CLICK ME]
In the room with the three flames, you only have to wait for the first one to pass by. You can use the ladder to go down as soon as the left-most flame has bounced off. This isn't trivial to do, but not terribly difficult either, and given the amount of time it saves I think it does belong in a run for this game. Here's a video of the trick: [CLICK ME]
I'm a bit puzzled as to why you jumped at 8:01. Getting nervous from getting so far into the game?
The remainder of the level is good. The use of invincibility is quite necessary to get past that last, darned flame. Good work. Gotta love the appropriate strategy of beating Stone Head with two rounds of 'paper'
Level 12 - City of Radaxian
That first jump was quite a bit too late. Other than that, perfect execution. Well done. Beating Scissors Head with two 'stones' is as amusing as ever. That last punch was a very close call!
Level 13 - City of Radaxian 2
Such a silly stage. Being quick enough to avoid the last dragon is not trivial, and probably not worth the risk at this point in the run.
Level 14 - Kingdom of Nibana
Hardest stage in the game. Amazing work, TAS-like. 10/10. Disregard all my other commentary, this alone makes the run worthy of being accepted.
Level 15 - Road to Janken's Castle
Losing the bike too early is very unfortunate, especially since you only barely missed the jump. There is no good recovery from this mistake, and the time loss is quite substantial.
Level 16 - Janken's Castle
This level is hard as balls, thanks to Alex' inertia. The runner misses several ring shots throughout the level. Why on earth did the runner jump at 12:40 instead of just breaking the bottom block? Not landing on top of the left platform in 13:00 is a bit of a nasty mistake. The jump at 13:16 is a huge pain. Wouldn't be surprised if that ruined several of the runner's recordings. 13:58 is a bit sloppy, again. Skipping the final boss is a huge time saver, and a quite hilarious 'ending' to the run.
Level 17 (if we can even call it a proper level)
Shame you missed the first frog punch. Cost you about 3 seconds.
Such a silly stage. Being quick enough to avoid the last dragon is not trivial, and probably not worth the risk at this point in the run.
Level 14 - Kingdom of Nibana
Hardest stage in the game. Amazing work, TAS-like. 10/10. Disregard all my other commentary, this alone makes the run worthy of being accepted.
Level 15 - Road to Janken's Castle
Losing the bike too early is very unfortunate, especially since you only barely missed the jump. There is no good recovery from this mistake, and the time loss is quite substantial.
Level 16 - Janken's Castle
This level is hard as balls, thanks to Alex' inertia. The runner misses several ring shots throughout the level. Why on earth did the runner jump at 12:40 instead of just breaking the bottom block? Not landing on top of the left platform in 13:00 is a bit of a nasty mistake. The jump at 13:16 is a huge pain. Wouldn't be surprised if that ruined several of the runner's recordings. 13:58 is a bit sloppy, again. Skipping the final boss is a huge time saver, and a quite hilarious 'ending' to the run.
Level 17 (if we can even call it a proper level)
Shame you missed the first frog punch. Cost you about 3 seconds.
Overall, I see around 20-25 seconds worth of improvement, which is a non-negligible amount of time. My initial thought after watching the run was "This is somewhat sloppy in more than a few spots.", and after dedicating an entire day to scrutinise it, I am still unsure what the verdict should be. Perhaps I am unnecessarily harsh on the run, because it's a rather difficult game. I will spend a bit more time thinking about it and seeing what the other verifiers think.
Verifier 2 has more comments below
Quote:
Excellent Quality.
All stages go smoothly, some minor exceptions
Money that get collected is exactly enough for what he needs, or is just unavoidable.
Overall, fun to watch, game play is well executed and is fast.
Decision Accept
All stages go smoothly, some minor exceptions
Money that get collected is exactly enough for what he needs, or is just unavoidable.
Overall, fun to watch, game play is well executed and is fast.
Decision Accept
Quote:
A/V is fine. No cheat detected.
Like previous verifers already said all stages gone pretty smoothly, but they still could be some little improvement. No big mistakes (previous verifiers already listed all the improvement that can be done). Nothing much to say about it. This is a good run to watch (we need more SMS run !!) and i think it's fit the SDA standard. Decision : accepted.
My timing :
Starts : 0:12
Gain control: 0:16
Finish : 15:37
Like previous verifers already said all stages gone pretty smoothly, but they still could be some little improvement. No big mistakes (previous verifiers already listed all the improvement that can be done). Nothing much to say about it. This is a good run to watch (we need more SMS run !!) and i think it's fit the SDA standard. Decision : accepted.
My timing :
Starts : 0:12
Gain control: 0:16
Finish : 15:37
Quote from Verifier 2:
Addendum to original commentary.
Verifier 1 encouraged me to watch the TAS for this game, as there was a very interesting trick present in it, which can be found here.. This trick allows you to skip part of the castle. It may look difficult to perform, but it took me roughly 10 minutes of testing to figure out exactly how it works, i.e., it's not hard.
One simply has to pause/unpause the game while the screen is scrolling and it will suddenly keep on scrolling. You then have to jump when the screen is focusing on the room /below/ the one you are in, which will make the screen scroll in the opposite direction again. Then you simply have to use the pause buffering again to make Alex reappear on screen. This trick works at every time when the screen is scrolling. There is zero risk of the game resetting, and if your X-position is correct (which is not hard at all) you will have perfect consistency in this trick. I can make a video of it, should the other verifiers require it.
Whether this trick is quicker than just running through depends on how often you'd have to pause-buffer (the less, the better), and how we should deal with timing the pause buffering, but it's quite certainly faster than running through the subsequent rooms when done right.
An additional comment is that the runner takes an inefficient route at 13:36. You don't have to go all the way to the left, but can simply jump straight to the top-right platform (with the frog on it) from the platform left of the first frog. This is not a hard jump and is a bit quicker than the route the runner takes. I suppose I can forgive the runner for not doing it, as there's still a small risk of death if you somehow miss-time the necessary punch.
Verifier 1 encouraged me to watch the TAS for this game, as there was a very interesting trick present in it, which can be found here.. This trick allows you to skip part of the castle. It may look difficult to perform, but it took me roughly 10 minutes of testing to figure out exactly how it works, i.e., it's not hard.
One simply has to pause/unpause the game while the screen is scrolling and it will suddenly keep on scrolling. You then have to jump when the screen is focusing on the room /below/ the one you are in, which will make the screen scroll in the opposite direction again. Then you simply have to use the pause buffering again to make Alex reappear on screen. This trick works at every time when the screen is scrolling. There is zero risk of the game resetting, and if your X-position is correct (which is not hard at all) you will have perfect consistency in this trick. I can make a video of it, should the other verifiers require it.
Whether this trick is quicker than just running through depends on how often you'd have to pause-buffer (the less, the better), and how we should deal with timing the pause buffering, but it's quite certainly faster than running through the subsequent rooms when done right.
An additional comment is that the runner takes an inefficient route at 13:36. You don't have to go all the way to the left, but can simply jump straight to the top-right platform (with the frog on it) from the platform left of the first frog. This is not a hard jump and is a bit quicker than the route the runner takes. I suppose I can forgive the runner for not doing it, as there's still a small risk of death if you somehow miss-time the necessary punch.
Quote from Verifier 1:
I went back over the run, and I fear I was looking at it through rose coloured glasses, because I see more and more room for improvement. Especially with the skip Verifier 2 has found to be easier to pull off than I thought, as well as the hesitation in the 3 flame room of the blue castle (07:19 in the video) and crashing early on the last motorcycle stage, together with all the spots that lost 1-2 seconds or less adding up.. I now estimate a minimum of 40 seconds of improvability, over 46 if I'm strict.
I would have really loved to know how long the runner spent getting this time, how many tries it took him, but in the light of this...
I regrettably have to REJECT this run.
I truly hope the runner can put the time in and get a new run with some of these spots straightened out. A sub 15 minute time is definitely within reach.
I would have really loved to know how long the runner spent getting this time, how many tries it took him, but in the light of this...
I regrettably have to REJECT this run.
I truly hope the runner can put the time in and get a new run with some of these spots straightened out. A sub 15 minute time is definitely within reach.
Quote from Verifier 2 again:
I agree with the above conclusion. As much as I'd love to see a new run for this game go up (-the SMS really deserves a lot more love-), I feel that the run has too many flaws to go up in its current state. The game is hard to run, so I could certainly forgive some mistakes, but there are simply too many of them.
I sincerely hope that the runner will revisit this run and at least obtain a sub 15 time, whilst avoiding the bigger mistakes.
My final verdict is thus, unfortunately: REJECT.
I sincerely hope that the runner will revisit this run and at least obtain a sub 15 time, whilst avoiding the bigger mistakes.
My final verdict is thus, unfortunately: REJECT.
Decision: Reject
Reason: I had a tough time with this one, but after reading the verifier comments and watching it a few times myself, I just felt like it was a bit too sloppy considering the length of the run.
This run will be available for a month. After that this link will no longer work.
http://queue.speeddemosarchive.com/queue/verificationfiles/35/
Thread title: