Magician Lord is a side-scrolling platform game developed by ADK (Alpha Denshi Corporation) that was released in 1990. It was a launch title for both the SNK Neo Geo MVS (arcade) and AES (home) systems. It was also one of the pack-in games for the Neo Geo AES, after SNK expanded the Neo Geo beyond the rental market--for those who could afford the $600. I was not one of those people, but I did play this game quite a bit at the arcade. The hero, Elta, equipped only with a pointy wizard's hat and fancy book-learnin', must battle his way through 8 Stages and defeat Az Atorse, the God of Destruction. Magician Lord is (in)famous for its difficulty, colorful and detailed graphics, grotesque and imaginative enemies, and excellent music. But it is perhaps best known in Western markets for its poorly translated cutscenes, accompanied by a weird, echoing voice-over as the evil Gal Agiese taunts Elta. I always felt this only added to the overall eerie feeling of the game, but the consensus is that it is simply hilarious.

Game Objectives

  • Emulator used: FBA-RR v0.0.7
  • Plays at Difficulty Level 3
  • Aims for fastest game completion
  • Aims for maximum score without sacrificing Frames
  • Takes damage to save time
  • Abuses programming errors
  • Manipulates luck
  • Contains 1 speed/entertainment tradeoff
  • What imprudence!

Excerpts from Elta's Book of Magic

"Momentum Maintainum":
The most important technique used in this TAS, "Momentum Maintainum" comprises the fundament of all of the other most useful techniques used. Very simply put, after Elta has gained momentum, pressing Left or Right for 1 Frame every other Frame allows him to maintain his momentum. This is but one example; Elta's momentum can be maintained with numerous combinations of input. This technique is used almost continuously throughout this TAS.
"Levitatum":
This technique allows Elta to "levitate" to the left or right, or to hover at his current location, without a surface to stand on. This is essentially what "Momentum Maintainum" becomes after Elta has walked off of a ledge. It functions in the same way, with fewer options than with "Momentum Maintainum". Pressing Left or Right for 1 Frame every other Frame causes Elta to "levitate" to the left or right at his current momentum. Pressing Down for 1 Frame every other Frame while using "Levitatum" cancels horizontal momentum and allows Elta to hover in place. If Elta jumps or fires, then "Levitatum" is cancelled.
"Elevatum Levitatum":
This technique allows Elta to "levitate" upward to the left or right. When Elta takes damage, he is pushed 17 pixels vertically and 17 pixels horizontally backward. After taking damage, Elta will remain in the "damage taken" posture until he lands on a surface, or until the invulnerability timer expires. While Elta is in the "damage taken" posture, not input can be entered. However, if Elta is standing slightly below the level of a platform when he takes damage, he will be pushed upward and land on the platform on the first Frame on which he bisects it. When Elta bisects the platform above him, the "damage taken" posture is cancelled, and input can again be entered. More importantly, Elta's momentum from taking damage can be maintained--both horizontally and vertically--by using the "Levitatum" technique. This allows Elta to elevate to the left or right at an impressive 4 pixels per Frame: "Elevatum Levitatum"!
"Pass Through Stone":
This is not merely one technique, but an art of study unto itself. Using these various techniques, Elta is able to pass through a solid object in a number of ways.
"Walk Through Walls":
This technique allows Elta to pass through a vertical surface. This is essentially what "Momentum Maintainum" becomes when Elta has reached a wall or other vertical obstacle. It functions in the same way, with fewer options than with "Momentum Maintainum". Pressing Left or Right for 1 Frame every other Frame allows Elta to maintain his momentum as he passes through a vertical surface. This is the most basic example of "Pass Through Stone".
"Jump Through Walls":
This technique allows Elta to pass through a vertical surface at a particular entry point and trajectory. If Elta is jumping or falling toward a vertical surface while facing the opposite direction, he will pass into the vertical surface. If the B button is pressed to jump left or right, and then on the next Frame the joystick is pressed in the opposite direction that Elta is facing for 1 Frame, the jump will continue to move forward in its original direction, but Elta will be facing backward. This allows Elta to easily jump backward into a vertical surface. To jump through a vertical surface in close quarters with maximum horizontal momentum, stand at least 1 pixel distant from the surface, press the A button for 1 Frame, delay for 1 Frame, press the joystick toward the surface for 1 Frame, and finally press the joystick away from the surface for 1 Frame. This will build momentum and turn Elta around before he collides with the surface, allowing him to move within it. Using the "Jump Through Walls" technique will invariably result in a "Surface Zip" on the first Frame that Elta has entered a sufficient distance within a surface.
"Surface Zip":
This technique allows Elta to "teleport" through a smaller solid surface instantaneously, or to hurtle through any larger size of solid surface at tremendous speed. When Elta has entered a sufficient distance within a surface, and then either cancels his momentum or changes his Y location, the game will attempt to push Elta out of the surface in the opposite direction that he is facing. If the surface is fully within the screen, Elta will be "teleported" instantaneously to the opposite end of the surface. If the surface extends beyond the screen boundary, the game will push Elta to the screen boundary (either X -21 to the left or X -108 to the right) and then scroll the screen at the breakneck speed of 11 pixels per Frame until he exits the surface. This is theoretically true of both vertical and horizontal surfaces, but great care must be taken when using a "Surface Zip" through a vertical surface, as it will invariably push Elta down to the bottom screen boundary if it is not fully within the screen--killing Elta. During an extended "Surface Zip", the direction that Elta is moving can be changed by pressing the joystick in the opposite direction that he is facing. As long as the surface is not fully within the screen, the direction of the "Surface Zip" can be changed as many times as desired until Elta exits the surface.
"Out & In":
This technique allows Elta to bypass the surface he is standing on. If the surface is a solid block, rather than a thin floor, Elta can also Surface Zip out to the edge of the surface, in addition to bypassing the surface he is standing on. It requires a corner where a vertical surface and a horizontal surface meet. Jump toward the vertical surface and turn backward so that Elta will pass through the vertical surface. The jump should be only far enough to just clear the corner of the horizontal surface. Allow Elta to fall into the wall, but change the direction he is facing just before he is Zipped out of the surface. Changing the direction he is facing at the last moment will cause Elta to be Zipped back out of the vertical surface the same way he entered it, but below the level of the horizontal surface from which he jumped, thereby bypassing the "floor".
"Slingshot":
This process is roughly the same as for the "Out & In" technique, except here it is used to Surface Zip horizontally along the floor, without falling completely through it.
Jump Through/Under/Past Vertical Surface:
If Elta uses Walk Through Walls to enter a vertical surface, and he then jumps, he will be instantly pushed out to the bottom of the vertical surface. However, if Elta has horizontal momentum, and the B button is still being held and charging a jump when Elta clears the right or left edge of the vertical surface, Elta will immediately begin to rise as per the strength of the jump without losing his horizontal momentum.
"Preemptive Strike":
When Elta opens a door, the transition to the Stage's Guardian Area begins. During this transition, there is 1 Frame during which input can be entered. The Frame on which to enter input for a "Preemptive Strike" is 1 Frame prior to Elta's Y position being set. This occurs 4 Frames prior to the screen beginning to fade to black. Entering any input except Up on this Frame will prevent the Forced Crouch when the next area begins. Generally, the best option is to press Right or Left or the B button on the "Preemptive Strike" Frame, and then hold Right or Left for 2 Frames after Elta spawns in the Guardian Area, but this is dependent on circumstances.
The Frame on which Elta spawns in a new Stage can also be used for a "Preemptive Strike". Elta always spawns at a substantially higher Y location than the level of the surface below him, such that he is "dropped" into the Stage. However, on the "Preemptive Strike" Frame, the game considers Elta to be standing on a surface. It is generally slower to enter input on this Frame, due to momentum issues, but it is used to great effect on Stage 4. It is also used before certain Boss battles in order to achieve a higher jump than would normally be possible, for entertainment purposes.
Another opportunity to use the "Preemptive Strike" technique is in Stage 8. Between each of the Guardian battles, the screen fades to black, similar to the transition to a Guardian Area in other Stages. Using the "Preemptive Strike" technique in this situation is superior for 2 reasons compared to using it during a Guardian Area transition. Because a new area is not loaded between Guardian battles, a projectile fired on the "Preemptive Strike" Frame will remain on the screen after the next Guardian battle begins. Also, opening a door causes Elta to be unable to take any action until the "Preemptive Strike" Frame. Because Elta is free to move around the area before the transition to the next Guardian battle begins, and because a projectile fired on the "Preemptive Strike" will remain after the transition ends, Elta can place a projectile virtually anywhere on the screen, and so can hit the next Guardian on the first possible Frame. This is useful not only for the damage, but also for manipulating the AI of certain Guardians.
The following input combinations can be entered on the "Preemptive Strike" Frame:
Right or Left
B button
B button and Right or Left to preemptively start a jump to the right or left
A button to fire a preemptive projectile
A button and Right or Left to fire a preemptive projectile at a particular angle
"Jumpus Interruptus":
Many horizontal surfaces in the game can be passed through from underneath, such that Elta can jump through and land on them. Normally when Elta jumps up through such a surface, he will continue to move upward as per the strength of the jump and will not land on the surface until he has begun to fall back down toward it. Elta's Dragon Warrior Form is peculiar in that firing his weapon during a jump sets memory address 00104065 to a "falling" status. By firing his weapon after holding the jump button for the desired number of Frames, Elta will immediately land on any horizontal surface through which he passes, rather than needing to wait until after the jump has passed its peak. The weapon need only be fired for 1 Frame to set the "falling" status. Because a stronger jump has a greater rate of elevation, it is best therefore to use the strongest jump possible to jump up through a surface which still allows time to fire the weapon, regardless of the actual elevation needed to successfully make the jump. This technique conserves horizontal momentum.
Instant Light Pillar:
The vertical location at which the screen locks for Stage 4 Guardian battle can be manipulated by Surface Zipping into the area. This allows Elta to land a bit lower on the screen than would normally be allowed, and also causes a solid surface to remain at the top of the screen that would normally be scrolled off-screen. This causes a glitch with the placement of the light pillar that descends after a Guardian has been destroyed. Normally after a Guardian is destroyed, the screen would center on Elta, the light pillar would descend from the top of the screen, and Elta would begin to ascend only after the pillar has reached his location at the bottom of the screen. In this case, the game considers Elta to be located within the light pillar on the first Frame that it spawns, allowing him to begin ascending instantaneously, without even needing to wait for the screen to center on him. This saves 47 Frames. This also causes the Minions to be Surface Zipped through the "ceiling" at the top of the screen as they are summoned by Warlock. Not that Warlock has time to summon anything in this TAS...
Prevent Forced Crouch:
If Elta falls for 10 Frames or more, he will be forced into a crouching posture when he lands. This forced crouch can be cancelled by pressing any button or direction on the joystick, but doing so requires at least 1 Frame. Using the Dragon Warrior Form, a forced crouch can be prevented by firing his weapon 4 Frames prior to landing. This technique cancels all horizontal momentum, but in a situation where Elta lands in front of a door and there is no need for momentum, it is still useful for entering the door on an earlier Frame.

Comments

To complete the game, Elta must complete 8 Stages and then defeat Az Atorse. Stage 1 through Stage 7 each consist of an Outer Area, followed by a Guardian Area which must be accessed via a particular door in the Outer Area. At the end of each Guardian area, Elta must defeat a Guardian (a sort of "Mini-Boss"). After defeating a Guardian, a Boss battle ensues. Upon destroying a Boss, Elta retrieves one of the 8 books, and the Stage is completed. At the end of Stage 7, Gal Agiese reincarnates each of the 7 Guardians. Stage 8 is simply the Guardians on Parade (a sort of "Boss Rush"), wherein each Guardian from the previous Stages must again be defeated in succession. When all of the Guardians have again been defeated, Elta is allowed to take a Form of his choosing (except his normal Form), and the battle against Az Atorse himself begins.
In the game (except during the Final Boss battle), the Hit Points colored blue represent Elta's Hit Points. If Elta loses all of these blue Hit Points, he loses a Life. When Elta collects an Orb, he assumes a new Form, and gains additional Hit Points equal to Elta's previous maximum Hit Points. The additional Hit Points provided by the new Form are colored pink. If Elta loses all of these pink Hit Points, he will be reverted to his normal Form. Prior to the Final Boss battle, Elta is allowed to take a Form of his choosing (except his normal Form). For the Final Boss battle, Elta is provided only with Hit Points equal to his normal Form, regardless of the Form that is chosen. In this case, the Hit Points colored blue represent Elta's overall Hit Points. If Elta loses all of these blue Hit Points, he loses a Life, and he loses his chosen Form. In this TAS, it is imperative that the Dragon Warrior Form be retained throughout, and as such, the "Elevatum Levitatum" technique can only be used 3 times.
Difficulty Level 3 is chosen for this TAS, as it provides Elta with an additional Hit Point. This is the maximum number of Hit Points that it is possible for Elta to have. This additional Hit Point allows the "Elevatum Levitatum" technique to be used one extra time, thereby saving substantial Frames. It also looks kinda cool.
After a Guardian or Boss has been destroyed, the game will not advance until Elta is standing on a surface.
After a Boss has been destroyed, a book will slowly descend from the top of the screen down to Elta's location, ending the Stage. As such, it is faster to have Elta standing at a higher elevation when a Boss is destroyed in order to minimize the number of Frames required for the book to move to Elta's location. Only Stage 3 and Stage 7 have an elevated platform in the Boss room.
A jump started immediately or soon after landing from a previous jump may have a slightly different arc than a jump made after walking for several Frames.
Entering a door may take slightly longer, depending on the manner in which Elta comes to first be within range of the door. Pressing Up following a Frame of delay or after landing from a jump seems to produce the best results.
If the B button is pressed to jump left or right, and then on the next Frame the joystick is pressed in the opposite direction that Elta is facing for 1 Frame, the jump will continue to move forward in its original direction, but Elta will be facing backward.
Elta cannot fire on the Frame at the peak of a jump.
Collapsing floor falling terminal velocity is 7 pixels per Frame. This is .75 pixels per Frame faster than Elta can normally fall.
When Elta takes damage, he is pushed 17 pixels vertically and 17 pixels horizontally backward.

The Forms

Aside from his normal Form, Elta can assume any of 6 other Forms, depending on the combination of Orbs in his possession. Elta begins each new life with 1 Red Orb. Red, Green, and Blue Orbs can be found in treasure chests throughout the game. Unfortunately, most of my notes regarding the various Forms were lost, but suffice it to say, there is only 1 Form worth mentioning in detail.
Dragon Warrior:
If Elta possesses 2 Red Orbs, he assumes the Form of Dragon Warrior. Dragon Warrior has the 2nd highest jump of all the Forms. He also has the smallest collision box by far. Dragon Warrior does 3 Hit Points of damage per projectile. The range of the weapon is rather limited, but this is certainly Dragon Warrior's only weakness, and a minor one at that. The weapon can be fired at any of 16 different angles for pinpoint accuracy. By holding down the A button, 1 projectile will be fired per each 4 Frames, and the direction of fire can be changed without releasing the A button, creating a sort of "flamethrower" effect at a rate of .75 Hit Points of damage per Frame. However, by pressing the A button for 1 Frame every other Frame, the normal rate of fire can be doubled. This changes an already powerful Form into a real monster! But still Dragon Warrior is not satisfied. While holding the A button, if any direction is pressed on the joystick other than that which is already being held (or that which Elta is already facing), an additional projectile will be immediately fired. By changing the direction being held every Frame, 1 projectile can be fired each Frame, thereby quadrupling normal damage output to 3 Hit Points of damage per Frame. He is the ultimate power in the universe.
Move while Firing:
Hold Right or Left for at least 2 Frames such that Dragon Warrior begins moving. On the next Frame, hold the A button and press the joystick in any other direction except Down or the initial direction. On the next Frame, continue to hold the A button and press the joystick in the initial direction again. While continuing to hold the A button, alternate between pressing the joystick in the initial direction, and any other direction but Down. Thus, Dragon Warrior can shimmy to the right or left while firing his weapon at optimum rate. Alternately pressing the initial direction and then Up and the initial direction provides more rapid right or left movement than other input combinations. Using the proper combination of directional input, Dragon Warrior can move left or right while firing either in the direction he is facing, or in the opposite direction. Following a Surface Zip, this technique can briefly be used without Frame loss.
Maintain Secondary-Intermediate Firing Angle With Minimal Projectiles:
The first projectile that Elta fires is fired at a primary or intermediate angle. This is one of 7 angles while he is standing, or one of 8 angles if he is jumping. After this first projectile has been fired, the joystick can be pressed in a different direction while continuing to hold the A button, and another projectile will be fired on the next Frame. This second projectile will be fired at a secondary-intermediate angle. Continuing to hold the A button and pressing the same pattern of directions on alternating frames will fire 1 projectile per Frame while maintaining a specific angle of fire. This is quite useful, but in many situations, not every projectile fired will hit an enemy, or it is otherwise problematic to fire a projectile on every Frame. Considering how easily Lag Frames may be generated by firing so many projectiles so rapidly, it is generally best to minimize the number of projectiles fired to only those which are critical (shots in a TAS shouldn't miss anyway, right?). This creates a problem, in that if the A button is released even for 1 Frame, the angle of fire is reset to one of the original 8 directions, which can cause Elta to miss subsequent shots which could have hit if the A button had not been released, and a secondary-intermediate angle of fire had been maintained throughout. To maintain just the right angle for just the right Frame while firing minimal projectiles, continue to hold the A button as normal, but instead of alternating the joystick direction on every Frame, do so on every 4th Frame. If the A button is held for 5 Frames without a direction change, another projectile will fire at the same angle as the previous one.
Aim First Projectile After Jump:
A projectile cannot be fired until at least 1 Frame after the end of jump input. However, if the A button is held down after the B button has been held down for at least 1 Frame, a projectile will fire on the earliest possible Frame after jump input has ended. The A button must remain held until the projectile has fired. This projectile will fire straight ahead of Elta. However, on the Frame before the projectile is fired, if any direction is pressed other than that which Elta is facing, the projectile will fire to the right at an intermediate angle between right and up-right (67.5 degrees). The advantage here is that this angle of shot cannot normally be achieved without first firing a projectile at a primary or intermediate angle. This would require an additional Frame, plus it would introduce an unnecessary object on the screen, which could cause Lag. This technique can also be applied in any situation wherein the A button can be held without firing a projectile, such as when the projectile limit has been reached, or on a "Preemptive Strike" Frame.
If too many objects are present on the screen, the game may generate Lag Frames as needed to compensate. The emulator does not register these Lag Frames, they are all managed by the game itself; this makes them eminently more difficult to identify and remove.
It is not necessary to fire a projectile in order to alter the angle of a projectile fired on the subsequent Frame. Simply holding the A button on the Frame prior to the angle change is sufficient to alter the angle of the next projectile fired. As such, even if it is not possible to fire a projectile at a particular angle on a particular Frame (e.g. because the angle of fire has not yet been changed), the A button can still be held at that angle, and the angle can be changed on the next Frame. This will fire just 1 projectile, at the desired angle.
Dragon Warrior jumps 6 pixels higher if he is standing still than if he jumps while he is moving. This can also be achieved by releasing Left or Right on the first Frame that the B button is held down. This allows for the higher jump with minimal loss of forward momentum. Though this higher jump does travel the same maximum distance as a maximum-strength normal jump, it arrives at that location 4 Frames later. Nevertheless, the higher jump is faster under certain circumstances. For jumps of less than maximum strength, the higher jump is always faster, which is useful for an area with a low "ceiling" or other situations requiring a shorter jump. The higher jump is also faster for a jump that lands at a higher elevation than the elevation from which the jump started.
When executing a jump with Dragon Warrior on a "Preemptive Strike" Frame, the joystick must be pressed right or left in addition to pressing the B button to prevent executing a higher jump.
Dragon Warrior is not to be confused with Erdrick from the Dragon Warrior games by Enix. In Magician Lord, Dragon Warrior is actually a dragon. Who is a warrior.
X walking speed is .625 pixels per Frame (1-0-1-1-0-1-1-0 pattern). This is 10% slower than Normal Form X walking speed.
Screen X walking speed is 1.25 pixels per Frame (1-1-1-2 pattern). This is 10% slower than Normal Form X walking speed.
X jumping speed is .9375 pixels per Frame (1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-0 pattern). This is 50% faster than X walking speed.
Screen X jumping speed is 1.875 pixels per Frame (2-2-2-2-2-2-2-1 pattern). This is 50% faster than X walking speed.
"Surface Zip" speed is 11 pixels per Frame.
Maximum upward vertical jump speed is 3 pixels per Frame.
Maximum "Elevatum Levitatum" speed is 4 pixels per Frame.
Falling terminal velocity is 6.25 pixels per Frame (6-6-6-7 pattern).
Ladder climbing speed is 1.875 pixels per Frame (2-2-2-2-2-2-2-1 pattern).
Maximum moving jump height is 38 or 39 pixels.
Maximum standing jump height is 44 or 45 pixels.

The Enemies

Enemies have more or fewer Hit Points, based upon the game's Difficulty setting. Values indicated here are based upon Difficulty 3.
Mosquito has 5 Hit Points and is worth 10 points.
Teleporter Skeleton has 3 Hit Points and is worth 20 points.
Long Arm has 20 Hit Points and is worth 100 points.
Electro Bird has 15 Hit Points and is worth 40 points.
Wanderer Skeleton has 3 Hit Points and is worth 20 points.
Trilobite has 9 Hit Points and is worth 30 points.
Frog has 5 Hit Points and is worth 10 points. Elta cannot be damaged by nor collide with a Frog.
Frogman has 35 Hit Points and is worth 200 points. Frogman begins transformation after completing 1 jump. The transformation takes 47 Frames. During the first 33 Frames of transformation, projectiles cannot collide with Frogman, so he cannot be damaged. After these first 33 Frames, Frogman can again be damaged, but he cannot be destroyed until the transformation has completed, even if he has sustained sufficient damage to do so. Even if sufficient damage has been done prior to the completion of the transformation, Frogman must be damaged at least once more after the transformation has completed in order to be destroyed.
Skull has 9 Hit Points and is worth 50 points.
Floating Eye has 9 Hit Points and is worth 300 points.
Armored Dog has 15 Hit Points and is worth 100 points.
Cat Girl has 15 Hit Points and is worth 100 points.
Chariot Knight has 15 Hit Points and is worth 100 points. Chariot Knight's lance is impervious to damage and will deflect Elta's projectiles if they are fired at a low trajectory. Chariot Knight must be hit in the head or from behind in order to be damaged. A low jump trajectory while moving toward Chariot Knight at maximum momentum provides an optimum angle of fire. As such, a 2 Frame jump toward Chariot Knight is ideal.
Destructible Floor has 33 Hit Points and is worth 0 points.
Treasure Chest has 1 Hit Point and is worth 0 points.
Sword Master has 63 Hit Points and is worth 200 points.

Stage By Stage Comments

Stage 1: Dale of Evil Gods

Outer Area:
Momentum is typically maintained by a 1-Frame jump once per each 4 Frames, alternating between facing right and left with each jump. Rather than jumping up onto the slightly raised stone floor, Elta instead uses "Levitatum" to continue to move to the right, just below the level of the stone floor's surface (Elta's waggle becomes a wiggle). After levitating to the right a sufficient distance, Elta fires at the Trilobite, cancelling "Levitatum" and causing him to fall to the right. Elta lands on the outstretched hand of the Shiva statue. Why this Shiva statue, and only this Shiva statue, in an inaccessible area, has hands with a solid surface is anyone's guess. Perhaps it is a remnant of an incomplete, unimplemented section of the area. Regardless, the statue is the only thing which makes it possible to bypass the upper portion of the area, and so I say thank you, O Auspicious One.
Guardian Area:
"Levitatum" is used to safely cross the pit of fire, thus bypassing the entire upper portion of the Guardian Area.
Guardian Sword Skeleton has 63 Hit Points and is worth 1600 points. Sword Skeleton can only be damaged while he is in attack posture or while he is jumping.
Boss Face of Evil has 96 Hit Points and is worth 10000 points. Stalactite can be damaged, but it is futile, as it has infinite Hit Points. Face of Evil is not immediately vulnerable, so the battle begins with some fireworks and then "Levitatum" perilously close to the Boss whilst avoiding the falling Stalactites. Elta slowly charges up his fire breath, and then unleashes it upon the Boss.

Stage 2: To the Evil Mine

Using "Elevatum Levitatum" by taking damage from Ball and Chain would save 234 Frames. Using this technique in the Outer Area would also cause 2 Mosquitoes to spawn in the Guardian Area rather than just 1. This would require 13 more Frames to set up "Elevatum Levitatum" in the Guardian Area, further reducing the Frames saved thereby. Compared with other opportunities on the game, this is not sufficient to justify the loss of a Hit Point.
Guardian Moon Demon has 81 Hit Points and is worth 1600 points.
Boss Pulsating Eye has 127 Hit Points and is worth 10000 points.

Stage 3: Highway Leading to a Foreign Space

Guardian Alien has 127 Hit Points and is worth 1600 points.
Boss Alien Dragon has Hit Points and is worth 10000 points.

Stage 4: Castle of Devils

Elta spawns 69 pixels above the floor.
Guardian Warlock has 127 Hit Points and is worth 1600 points.
Boss Hellhound has 255 Hit Points and is worth 10000 points. Bad dog! Stay!

Stage 5: Anderground Passage of Terror

In the Guardian area, use "Slingshot" to Surface Zip along the floor. Remain within the floor and continue moving to the right. Just past Electro Bird 2, take damage from Frogman and use "Elevatum Levitatum" to move up and to the right. Continue moving up until Screen Y Major 1 is reached. This unlocks horizontal scrolling to the right. At this point, jump to begin the Surface Zip through the wall to the right. Elta must either jump or continue climbing for a bit to allow time for the floor to scroll up so that he does not simply fall to his death after the Surface Zip. This method saves 557 Frames compared to climbing up and down the various ladders to the Guardian room.
In the Guardian area, 2 groups of Frogs spawn as Elta advances. 2 Frogs in group 2 transform into Frogmen. The 2nd Frog from the right of group 1 is destroyed to manipulate one of the Frogmen in group 2 to land as far to the right as possible before he transforms. If any Frog in group 2 is damaged before the critical Frogman has jumped, he will jump to the left instead of to the right, which is far less desirable.
Guardian Axe Warrior has 127 Hit Points and is worth 1600 points. After Axe Warrior spawns, he will jump after 32 Frames have passed, unless he is damaged before these 32 Frames have elapsed. If Axe Warrior is damaged before he starts his first jump, he will assume a defensive posture and become impervious to damage for 64 Frames. As such, it is faster to wait for Axe Warrior to jump before beginning to attack. The time spent waiting for Axe Warrior to jump is used to move closer to him.
Boss Tower of Terror has 96 Hit Points and is worth 10000 points. Tower of Terror projectiles have 1 Hit Point and are worth 0 points. Tower of Terror can only be damaged by hitting the center eye. The center eye has a protective shell which opens and closes periodically so that they eye can fire a projectile; the center eye can only be hit while this shell is open. Even while it is open, the shape and location of the shell near the center eye, combined with the large size of Dragon Warrior's projectiles, make it difficult to hit the center eye consistently while Elta is standing on the ground. As such, a jump is used instead, which also allows Elta to move closer to the center eye while he is firing, while maintaining maximum momentum. Tower of Terror's projectiles can be destroyed, but they can also block Elta's projectiles, preventing them from hitting the center eye. The Boss's first projectile is destroyed with a shot while Elta is falling from the spawn point. A 5 Frame jump is used after falling from the spawn point, which keeps Elta's angle of fire just below the path of the Boss's second projectile, allowing uninterrupted fire at optimum rate. A 6 Frame jump or greater means that some of Elta's projectiles would need to hit the Boss's second projectile. A 4 Frame jump or less would not allow Elta to remain airborne long enough to fire sufficient projectiles to destroy the Boss, which would mean more Frames of delay waiting to land and fire again. Nevertheless, a Boss with only 96 Hit Points which can be immediately engaged is a rare treat from this game--just 86 Frames from Elta's spawn to final input.

Stage 6: Corridor Leading to Hell

Guardian Hell Mistress has 63 Hit Points and is worth 1600 points.
Boss Hell's Creation is comprised of 3 destructible components. Hell's Creation Eye has 81 Hit Points and is worth 10000 points. Hell's Creation Mouth has 81 Hit Points and is worth 1000 points. Hell's Creation Tongue has 81 Hit Points and is worth 1000 points. Destroying Hell's Creation Eye first destroys Hell's Creation entirely, though no points are then given for the destruction of the other 2 components. From the ground, Hell's Creation Eye can only be hit with maximum efficiency from Player Position X 80.5, and only by using an Up-Right firing pattern. Hell's Creation Spawn has 3 Hit Points and is worth 0 points. The movement of Hell's Creation Spawn is dependent upon the amount of damage sustained by Hell's Creation Eye and Hell's Creation Mouth.

Stage 7: Gal Agiese

In the Outer Area, the screen will not scroll vertically until a particular X location has been reached. As such, Elta cannot simply fall through the castle wall on the earliest possible Frame or he will simply fall to his death after the Surface Zip.
Guardian Grand Knight has 127 Hit Points and is worth 1600 points. Grand Knight is quite lethal, and would have been quite a bit of work to TAS, save for 1 fatal flaw. Grand Knight adjusts his posture to a low guard when Elta crouches. By crouching for at least 1 Frame at least once per each 4 Frames, Grand Knight is held alternating between the high guard posture and the low guard posture. In this state, Grand Knight is unable to move or defend himself--easy pickings. :)
Boss Incarnation of Az Atorse has 191 Hit Points and is worth 10000 points.

Stage 8: God of Destruction Reincarnates

Unlike the first battle against the Stage 6 Guardian, Hell Mistress, Elta's projectiles can penetrate her light pillar. This allows Elta to destroy Hell Mistress well before she descends to the ground. This is clearly unintentional, as the light pillar can still deflect many projectiles, but by firing at a particular angle on a particular Frame, Elta's projectiles can penetrate the light pillar with impunity. Note that her light pillar will still cause damage to Elta if he collides with it.
Momentum is maintained between Guardian battles.
The Form selection screen can be manipulated to appear 1 Frame earlier if Elta is moving when the screen transition begins, or if any input is entered just prior to the start of the transition.
Az Atorse hovers at the center of the screen when he first appears. From the left, the nearest safe location from which to fire at Az Atorse is X 39.5. From the right, the nearest safe location is X 80.
There are 2 laser spikes on the left and 2 on the right of Az Atorse. The spikes are destroyed sequentially, regardless of what part of Az Atorse Sub-Body is hit. Az Atorse Main Body can be damaged, but it is futile, as it has infinite Hit Points. After Az Atorse Sub-Body has been destroyed, Az Atorse Main Body will release 8 Orbs which rotate around it. If Az Atorse Sub-Body is destroyed as quickly as possible, Az Atorse Main Body will still float to the right edge of the screen before returning to the center and releasing Orbs. Because there is no great rush to destroy Az Atorse Sub-Body, the coup de grâce is delivered in dramatic "Ninja Gaiden" fashion. Az Atorse Main Body will only be destroyed if all 8 Orbs are destroyed, or if its self-destruct timer expires.
Az Atorse Sub-Body has 255 Hit Points.
Az Atorse Spawn has 1 hit point and is worth 0 points.
Az Atorse Main Body has infinite Hit Points.
Rotating Orb has 33 Hit Points and is worth 100 points.
Final Frame of input is Frame 26263.

Other Comments

By popular demand, here is the alternate ending to this TAS. In this version, Az Atorse is destroyed directly, rather than waiting for him to self-destruct. This required 23 additional Frames (and 1720 re-records), but it saves the viewer from watching nearly 3 minutes of inactivity before seeing the ending sequence.
If you enjoyed this TAS, please encourage your TASVideos.org legislators to reconsider allowing submissions created with MAME-RR! This would allow access to the vast library of arcade games that MAME-RR supports, and ensure continuing arcade mayhem from myself and other Players who are familiar with the platform(s).

Noxxa: Judging.
Noxxa: The situation here is best summed up by this post by adelikat. We have, in this situation, two submissions that are considered identical timewise. The other submission was published earlier this month and aimed for fastest input; this submission aims for fastest game completion. Both of these are perfectly and equally valid ways to end a run under this site's rules. For comparison and judgment purposes, both of these runs are identical in time.
Back to adelikat's post: as this submission does not improve on time compared to the current publication, we need to look at this in a different way: like a playaround obsoleting another playaround. That is an uncommon situation, and in such a case, we only obsolete a published movie if there is a significant consensus that the new run is a large improvement on entertainment compared to the currently published run.
Now, in the case of his submission, that happens to be the case: this submission is near-universally agreed upon as being much superior in entertainment value compared to the published run. That this run takes less than half a second of extra input time in order to eliminate a 3-minute waiting period after input ends as in the current run, definitely forms a significant difference in entertainment value for the final boss battle, or even for the overall run.
Therefore, accepting to Moons to obsolete the currently published run.
feos: Publishing...


Joined: 11/6/2011
Posts: 7
feos wrote:
If we suddenly start considering this run an improvement, then I have a question: what exactly prevented doing so in the initial submission? If nothing, then there's no point in "obsoleting" it. Otherwise I can make 100 submissions of a single game with slightly edited endings, each of which would be "more entertaining". And I'm not even mentioning what Masterjun can do to abuse such an approach.
That said, I don't think any rule should be hard and fast, especially on a site whose goal is more or less entertainment. In US law for instance, Judges can throw out a case that fits the letter but not the spirit of a law, like the incident where Buzz Aldrin hauls off and decks a conspiracy theorist for harassing him. To the letter of the law, Aldrin commits assault, but in the spirit of the law, the level of harassment and clear intent to sue or prosecute Aldrin showed in the video (he was recorded asking the cameraperson "did you get that?") went against the spirit of the law; To protect citizens from a violent threat. In the same vein, I think it'd probably be best to defer to input time with exception for situations like Magician Lord, where a few frames saved adds a significant amount of time to the realtime.
Skilled player (1743)
Joined: 9/17/2009
Posts: 4986
Location: ̶C̶a̶n̶a̶d̶a̶ "Kanatah"
c-square wrote:
I have to vote on the side on Entertainment. I like this one better. One thing to consider in this whole argument: What would the argument be if the submissions were done in reverse? What if the shorter movie with the longer input was submitted and accepted first, and then the slightly shorter input version that took three extra minutes to complete was submitted next? Are we comfortable saying that you can obsolete any TAS on this site by ending input a bit earlier, even if it adds hours (or theoretically, days) to the TAS itself?
I believe this was addressed before as not a real improvement.
Player (146)
Joined: 7/16/2009
Posts: 686
If this was done as an edit to the previous submission, there'd be no debate. The people who are against this new movie obsoleting the old one only appear to object to the method (obsoletion instead of movie file replacement), nobody actually seems to say this movie isn't better. I think that makes it pretty clear this movie should obsolete the previous one.
Former player
Joined: 6/30/2010
Posts: 1107
Location: Zurich, Switzerland
I personally don't care too much what version ends up being published. I love the cheap cutscenes of this game and thought that ending the input that early was pretty clever. Publishing both would also work for me, but in that case, this movie could only go to moons. And it would be pretty arbitrary for a separate category.
Current project: Gex 3 any% Paused: Gex 64 any% There are no N64 emulators. Just SM64 emulators with hacky support for all the other games.
Skilled player (1417)
Joined: 10/27/2004
Posts: 1978
Location: Making an escape
As much as I'd prefer to have this be the main movie on this site, I think it's just not worth the hassle of everybody fighting over it to really argue for obsoletion. At this point, I think the best solution might be to have a little blurb in the existing movie description that mentions and provides a link to what the final boss would look like without waiting.
A hundred years from now, they will gaze upon my work and marvel at my skills but never know my name. And that will be good enough for me.
adelikat
He/Him
Emulator Coder, Site Developer, Site Owner, Expert player (3576)
Joined: 11/3/2004
Posts: 4754
Location: Tennessee
The situation here is that we have a published movie that aims for time and a submission that does not improve it in terms of time. Imagine a Super Mario Bros warps movie that is the exact number of frames as the published movie but does more entertaining antics and tricks along the way. The movie needs to be judged similarly to a "playaround" obsoleting another playaround. It can be done, but it is hard. The same thinking needs to be here. Is this movie a clear entertainment improvement over the published movie?
It's hard to look this good. My TAS projects
Warepire
He/Him
Editor
Joined: 3/2/2010
Posts: 2178
Location: A little to the left of nowhere (Sweden)
adelikat wrote:
Is this movie a clear entertainment improvement over the published movie?
I definitely think so, while it was clever to end input early it murdered the entertainment.
Player (26)
Joined: 8/29/2011
Posts: 1206
Location: Amsterdam
adelikat wrote:
The same thinking needs to be here. Is this movie a clear entertainment improvement over the published movie?
Yes, absolutely. The published movie is seven minutes of awesome followed by three minutes of boredom where nothing happens. The submission here is seven minutes of awesome followed by an epic endgame sequence. That's a clear and obvious improvement. For a very long time, TASvideos has allowed Speed/Entertainment Tradeoffs. This is a clear example. Now if this were a Vault tier movie, then one might make the case that speed trumps everything else; but as it's in Moon Tier, the entertainment should definitely be the deciding factor.
Noxxa
They/Them
Moderator, Expert player (4128)
Joined: 8/14/2009
Posts: 4090
Location: The Netherlands
I don't think this should be seen as an speed-entertainment tradeoff. Both this run and the other run aim to complete the game as fast as possible - they just use different definitions of 'fastest completion'. Reaching the ending as fast as possible is just as valid a way of completion as aiming to end input as fast as possible (though some may prefer one method over the other). That doesn't have anything to do with tiers or speed-entertainment tradeoffs. The question here is whether this, generally considered more entertaining, method of ending the game is significant enough to warrant replacing the current publication with it.
http://www.youtube.com/Noxxa <dwangoAC> This is a TAS (...). Not suitable for all audiences. May cause undesirable side-effects. May contain emulator abuse. Emulator may be abusive. This product contains glitches known to the state of California to cause egg defects. <Masterjun> I'm just a guy arranging bits in a sequence which could potentially amuse other people looking at these bits <adelikat> In Oregon Trail, I sacrificed my own family to save time. In Star trek, I killed helpless comrades in escape pods to save time. Here, I kill my allies to save time. I think I need help.
Site Admin, Skilled player (1255)
Joined: 4/17/2010
Posts: 11495
Location: Lake Char­gogg­a­gogg­man­chaugg­a­gogg­chau­bun­a­gung­a­maugg
JSmith wrote:
feos wrote:
If we suddenly start considering this run an improvement, then I have a question: what exactly prevented doing so in the initial submission? If nothing, then there's no point in "obsoleting" it. Otherwise I can make 100 submissions of a single game with slightly edited endings, each of which would be "more entertaining". And I'm not even mentioning what Masterjun can do to abuse such an approach.
This isn't just a slight improvement. It's about 25% more entertaining (per minute) than the published movie. And it's moon tier, which means that sort of speed-entertainment tradeoff is a good thing.
I know, and I agree that it looks way better, but still, why not do it in the first place?
Kibibit wrote:
That said, I don't think any rule should be hard and fast, especially on a site whose goal is more or less entertainment. In US law for instance, Judges can throw out a case that fits the letter but not the spirit of a law, like the incident where Buzz Aldrin hauls off and decks a conspiracy theorist for harassing him. To the letter of the law, Aldrin commits assault, but in the spirit of the law, the level of harassment and clear intent to sue or prosecute Aldrin showed in the video (he was recorded asking the cameraperson "did you get that?") went against the spirit of the law; To protect citizens from a violent threat. In the same vein, I think it'd probably be best to defer to input time with exception for situations like Magician Lord, where a few frames saved adds a significant amount of time to the realtime.
We are trying to have rules that match the spirit of TASing and all, but we also need rules that won't be abused if someone decides to take the work. My example above wasn't resolved.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Samsara
She/They
Senior Judge, Site Admin, Expert player (2241)
Joined: 11/13/2006
Posts: 2823
Location: Northern California
feos wrote:
I know, and I agree that it looks way better, but still, why not do it in the first place?
£e Nécroyeur wrote:
It is as Patashu says, I end input on the earliest Frame that allows Elta to survive long enough for Az Atorse to self-destruct. This is done with respect to TASVideos.org rules regarding movie length. Depending on viewer reaction, I may submit a "Fastest Game Completion" version of this TAS, have it debated in the Forum for a month, and see it ultimately rejected and placed in Gruefood Delight. :)
You can't really blame him for ending input early after that whole debacle.
TASvideos Admin and acting Senior Judge 💙 Currently unable to dedicate a lot of time to the site, taking care of family. Now infrequently posting on Bluesky
warmCabin wrote:
You shouldn't need a degree in computer science to get into this hobby.
Former player
Joined: 8/1/2004
Posts: 2687
Location: Seattle, WA
adelikat wrote:
The same thinking needs to be here. Is this movie a clear entertainment improvement over the published movie?
I think we need to have a clear definition of what we are judging here. The input files that are submitted are essentially what the authors are offering, and I have almost always solely judged submissions based on what the authors decided to deem part of the run. Nécroyeur obviously had the intention of ending the input earlier and letting the boss time itself out, and I feel like that is the point at which we stop judging a submission (save that it still completes the game, which this one does). Otherwise, the NES Monopoly run should theoretically be more entertaining if someone comes in and speeds through the rounds that would currently end the match, rather than letting the RNG run its course. There are plenty of other runs where input ends earlier in order to reduce input length, yet this has never been an issue until now?
Radiant wrote:
For a very long time, TASvideos has allowed Speed/Entertainment Tradeoffs.
Yeah, but these tradeoffs are to be decided by the author, not the Workbench. It isn't our place to tell them to trade time for entertainment if that was not their initial goal, and it has already been stated that the initial plan was to timeout the final boss and reduce the input length. e: Just to poll the thread: If you know that the input ends early and lets the final boss timeout after a certain amount of time, do you still sit through the entire duration of the fight with a frown on your face or do you simply skip through the fight, knowing that the input has ended already? If you do the former, do you also sit through the credits for every single submission that you judge? Would you feel cheated if you found out that the author could skip the credits at the end of a game?
hi nitrodon streamline: cyn-chine
Joined: 7/2/2007
Posts: 3960
Zurreco wrote:
e: Just to poll the thread: If you know that the input ends early and lets the final boss timeout after a certain amount of time, do you still sit through the entire duration of the fight with a frown on your face or do you simply skip through the fight, knowing that the input has ended already? If you do the former, do you also sit through the credits for every single submission that you judge? Would you feel cheated if you found out that the author could skip the credits at the end of a game?
It depends. How interesting are the boss's antics? How much effort goes into setting up the early-input-end? How much input time is saved by timing out the boss rather than beating them directly? How long does the boss take to time out? In practice, for the vast majority of runs, you cannot save gratuitous amounts of time via early-input-ends, because the developers took pains to ensure that the boss does not have large safe zones for the player to hide in. Thus, ending input early is generally just a matter of fighting the boss "normally" up to 99% done, then launching your last attack from as far away as possible. This can be pretty entertaining (see the any% Gimmick! run, for example, which uses a carefully-aimed shot combined with AI manipulation to hit the boss multiple times). Perhaps more importantly for this discussion, I can't think of any situations where I found myself actually bored by the lack of input, except for this game. Well, River City Ransom came close, but at least the boss was actually doing something while we waited. Here, there's a large safe zone, as a consequence of the boss having a very simple attack pattern that makes no particular effort to actually aim attacks at the player. The boss doesn't do anything interesting while waiting for timeout; the player doesn't have to do anything particularly interesting to set up the safe zone; it takes a huge amount of time for the timeout to occur. And worst of all, the input still ends when the boss is less than half a second from death if the fight had been persecuted "normally", which makes the amount of time saved by ending early grotesquely tiny compared to the amount of time spent staring at the boss and waiting for it to die on its own. All of these factors combine to make this instance of ending input early pretty dull.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Site Admin, Skilled player (1255)
Joined: 4/17/2010
Posts: 11495
Location: Lake Char­gogg­a­gogg­man­chaugg­a­gogg­chau­bun­a­gung­a­maugg
Guys, both ending types are valid and up to the author. We had a huge discussion about it, and people halved. Neither is better, both are good. It's just, "we want to avoid meaningless publications". If this particular run is proven to be a "significant improvement", no problem, obsoletion can happen. But we must figure out how to handle such cases in future, to not let the rules have holes. Maybe the statement for these should be "I consider it a significant improvement, even though time-wise the very run is exactly the same, and want it published". Because we need to make sure that you realize that it in no way is a time improvement, if someone adds a few presses to run the ending earlier. If this is understood, and the new run still feels way better, then we should obsolete.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Joined: 7/2/2007
Posts: 3960
feos wrote:
Maybe the statement for these should be "I consider it a significant improvement, even though time-wise the very run is exactly the same, and want it published".
That was what I said with my first post in this argument -- this run is an entertainment improvement over the predecessor. And we're targeting Moons here anyway, right? So entertainment is what matters.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Samsara
She/They
Senior Judge, Site Admin, Expert player (2241)
Joined: 11/13/2006
Posts: 2823
Location: Northern California
feos wrote:
It's just, "we want to avoid meaningless publications". If this particular run is proven to be a "significant improvement", no problem, obsoletion can happen. But we must figure out how to handle such cases in future, to not let the rules have holes.
Well, the acceptance and publication of the previous movie happened very quickly, even though there was debate over the ending. Delaying judgement in favor of looking at an alternate file would most likely handle every case such as this. Otherwise, it just seems like it would be a game-by-game basis, backed up by the opinions of the people watching. In this case, I'd agree that this is a significant entertainment improvement over the previous movie even if the entire rest of the movie is exactly the same.
TASvideos Admin and acting Senior Judge 💙 Currently unable to dedicate a lot of time to the site, taking care of family. Now infrequently posting on Bluesky
warmCabin wrote:
You shouldn't need a degree in computer science to get into this hobby.
Warepire
He/Him
Editor
Joined: 3/2/2010
Posts: 2178
Location: A little to the left of nowhere (Sweden)
Samsara wrote:
feos wrote:
It's just, "we want to avoid meaningless publications". If this particular run is proven to be a "significant improvement", no problem, obsoletion can happen. But we must figure out how to handle such cases in future, to not let the rules have holes.
Well, the acceptance and publication of the previous movie happened very quickly, even though there was debate over the ending. Delaying judgement in favor of looking at an alternate file would most likely handle every case such as this. Otherwise, it just seems like it would be a game-by-game basis, backed up by the opinions of the people watching. In this case, I'd agree that this is a significant entertainment improvement over the previous movie even if the entire rest of the movie is exactly the same.
I 100% agree with this.
Personman
Other
Joined: 4/20/2008
Posts: 465
We should be trying to minimize input length. If this leads to a boring movie for some people, we can have a separate encode, or those people can learn how to use their video player's "fast forward" (or "stop") feature. Watching a boss flail around and kill itself for three minutes is quite entertaining to me. If it were an hour, I wouldn't watch the whole thing, but the fact that it existed would be so hilarious that I would be even more entertained, and I'd probably watch ten minutes of it by skimming through. This isn't an argument for accepting the "early" input end based on entertainment — it's a counterargument to people saying we shouldn't because it's not entertaining. I don't think a subjective issue like this should be decided on entertainment grounds; I think we should find out what the shortest input is, and then make room for alternate versions/encodes if there's a major entertainment issue (like we do with camhacked sonic runs, for instance).
A warb degombs the brangy. Your gitch zanks and leils the warb.
Skilled player (1673)
Joined: 7/1/2013
Posts: 453
Alternate versions, right! It's important, of course, to avoid meaningless publications. But this isn't *entirely* meaningless (and certainly not to me). And frankly, if an alternate version is not allowed on the front page, it seems likely that it will still end up being published in Gruefood Delight or Obsoletely Fabulous. Published. I am conflicted in making a TAS just as others are conflicted in judging it. We are all bound by the rules and guidelines of TASVideos.org, and I do my best to adhere to them, but there is some duality even to the rules. I feel like "complete the game as quickly as possible" and "make the game as entertaining as possible" are both Rule #1. A quandary arises when it is impossible for a TAS to meet both conditions. Hyperbole aside, I'm not making 100 Submissions of the same game with slightly tweaked endings. I made 2 Submissions of this game, both of which were received favorably, and the 2nd of which was authored purely based on viewer feedback. I don't feel that this is any kind of abuse of TASVideos.org or its users.
Editor, Expert player (2480)
Joined: 4/8/2005
Posts: 1573
Location: Gone for a year, just for varietyyyyyyyyy!!
How about publishing both runs and allowing future publications of similar kind, ONLY IF the frame difference between the endings is equal to that of Magician Lord. This way we would only get a few extreme/notable cases published. It would also be an easy rule to follow: The extended input must shorten the AVI by at least 3 minutes, otherwise it is rejected.
Editor, Skilled player (1824)
Joined: 8/25/2013
Posts: 1200
Aqfaq wrote:
How about publishing both runs and allowing future publications of similar kind, ONLY IF the frame difference between the endings is equal to that of Magician Lord. This way we would only get a few extreme/notable cases published. It would also be an easy rule to follow: The extended input must shorten the AVI by at least 3 minutes, otherwise it is rejected.
That sounds good to me.
effort on the first draft means less effort on any draft thereafter - some loser
Post subject: Movie published
TASVideoAgent
They/Them
Moderator
Joined: 8/3/2004
Posts: 15629
Location: 127.0.0.1
This movie has been published. The posts before this message apply to the submission, and posts after this message apply to the published movie. ---- [2641] Arcade Magician Lord by £e_Nécroyeur in 07:17.73
Joined: 1/23/2006
Posts: 31
Late to the party, but happy to have seen this "better" ending in my opinion. I do agree with basically everything said as an overall and not picking single statements. It's hard to even answer the poll that way. I guess I can say I like the idea of "Run ends on last input needed to reach the ending in the fastest way possible." If your too weak in an Adventure/RPG to beat a boss but can time it out, do it - if someone can optimize that, get to the boss and beat it to start a credit roll faster by leveling up then that's the accepted, right? If you drop onto the axe from SMB without input, but the timing is exact same if someone taps a frame more right or just stops the pixel needed for the hit-box, it's the same. Enough people watch and analyze the playing style of the submitted runs that I'm pretty sure that will be caught right away if it's a problem that input ending sooner or later makes no difference in frame count. Putting this back to Magician Lord, I did watch the 3 minutes of the boss flailing around. I kept thinking "He's really going to avoid dying before the timeout? Is RNG involved in any of this or it succeeds 100% of the time?" It still murdered the entertainment for me though - the objective of any game to is to win, and if you can chop even over 30 seconds of the "movie" you see on youtube off, then it's worth it. Addendums, footnotes - explaining that it can be done or tagging a seperate video with a "trick" to end input early are all ways to give as much information to the viewer as possible, and those of us who not only watch but read the comments/submission text love finding out about the tricks. On just a recommended youtube link without other information, it's going to leave people confused. Save time for the viewer - we'd never watch a hollywood movie where they just watch the bad guy drive in circles for *minutes* before they somehow off themselves... and the protagonist is basically laughing directly at them the entire time. For that reason, maybe shoot for the lowest common denominator, that the average person watching these get it straight from youtube and not the site? Just my 2c. Hope it shows a long time lurkers/avid watchers viewpoint. :) Thanks for the movie and improvement, again.
Skilled player (1673)
Joined: 7/1/2013
Posts: 453
Thank you all for commenting, voting and rating! I'm happy to see the more popular edit of this TAS given prominence on the site. Let's just hope I don't find another game with an option to end input early... Now I can get back to work on [Undisclosed TAS]!
Skilled player (1673)
Joined: 7/1/2013
Posts: 453
For those watching the YouTube Encode, please note that the caption at the 16 mark states Played by ?e Necroyeur This is not correct. My nickname is £e Nécroyeur I have sent a message to feos regarding this, and hopefully this Encode will soon be replaced with a corrected version.