Puzzle & Action: 2do Arukoto ha Sand-R, known in US arcades as Puzzle & Action: Treasure Hunt, is a 1995 arcade game co-developed by Sega AM1 and CSK Research Institute (CRI). This is the third game of the Puzzle & Action series of games where players progress by completing mini games to reach the end. A year later this game was ported by CRI and is pretty much a perfect port of the arcade game. This TAS uses the Saturn version of the game since it's both arcade perfect and MAME's emulation of Sega Titan Video games are imperfect.
The main plot is simple: the main characters are treasure hunters who found a huge treasure chest. However, a band of thieves steal it from behind them. Now the treasure hunters travel the world in order to get their loot back. They gotta get past each boss's pawns through completing weird mini games and then challenge each boss in their own separate challenge.
At first I thought this game would be pretty easy to make a TAS for, albeit a bit long. Unfortunately I seem to have the unfortunate problem of just running into interesting RNG manipulation challenges. I originally tried to make a "1 player, all games" run that I mostly completed but at a certain point it became pretty clear that if I'm doing annoying RNG manipulation to get the last few games I might as well instead just make a proper 2 player run.
- Emulator used: BizHawk 2.8
- Saturnus core
- BIOS used: JP version 1.0
- May contain puzzles and/or actions
Important technical note
This TAS was made with the JP version 1.0 of the Saturn BIOS. At the very minimum, this affects loading times and may cause desyncs with the incorrect BIOS.
General mechanics and techniques
Game difficulty rank
This game has a difficulty rank system that makes the games progressively harder the farther in the players are. The game also ranks down anytime they fail a mini game round. The difficulty select at the start also affects the game's starting rank. What influences rank is simple. Completing mini games successfully increases rank up until it maxes out, and failing mini games decreases rank. difficulty rank influences in a one or multiple ways:
- Increases target quota needed for completion
- Reduces time given to complete a mini game
- Adjusts the factor in the mini game to increase difficulty
These changes will be explained for each mini game when relevant.
RNG factors and manipulation
RNG is usually dependent on the timing of inputs. Delaying inputs in places is used to influence RNG when it can be helpful. For what mini games show up, the only part that I know influences the selection is at the start of a stage where the frame the player shows up to the selection determines availability. The problem is that later stages have a LOT of mini games to get through so I mostly influenced the selection to the point that the first few games are on the quicker side.
2 player changes and impact
Some mini games get adjusted to the presence of 2 players and requires both players to cooperate to get times similar to playing solo. This leads to a wash for some mini games in terms of completion time. The main benefit of using 2 players is that the impact of RNG for multiple games. This reduces the amount of RNG manipulation I do end up doing. Also worth noting that in the even that both players reach the goal on the same frame, priority goes to player 1.
Mini game comments by stages
The mini games will be divided by order of appearance in the TAS, then unplayed mini games. I don't know the proper names for each one so they're named by whatever description I could think of that's most helpful. The exception are the boss stages, taken from the English version of the arcade game.
|Game number||Mini game||Information|
|1||Coin shooter||The goal of coin shooter is to shoot a coin to propel it to a given target height. Higher difficulty rank raises the goal. The fastest method to achieve this is to shoot the coin with 2 bullets in rapid succession whenever possible. Using 3 frames to fire twice (A, blank frame, A) is very subtly slower than using two frame (A, C) but the second time this mini game shows up reveals that it still doesn't matter except at extremely high heights.|
|2||Castle shapes||The goal of castle shapes is to identify the floor plan of 3 castles as the camera rotates the model on screen. This worst case RNG scenario for a 1 player game is when the correct choice is far to the right, although it's not a big frame loss. With 2 players this is mitigated to only the center choice being the slowest as player 2 starts on the right side. This is the fastest mini game by far, especially since there's no "Ready, Go!" startup before each castle. Difficulty rank only seems to change layout complexity.|
|3||Towel heater||The goal of towel heater is to reach maximum heat by rubbing a towel. This is done by alternating A and C every 3 frames. Difficulty rank only affects starting heat level which isn't the biggest issue even in real time. There isn't much in the way of messing around with the unsuccessful player's inputs but I do what I can.|
|4||Find crooks||The goal of find crooks is to select the windows while the criminals pop up. Difficulty rank affects the quota to reach for completion. RNG does affect when and where the criminals pop up but they pop up often enough that it's a relatively fast game even at maximum rank. This is the 3rd fastest mini game.|
|5||Train pusher||The goal of train pusher is to shove the passengers into the doors by alternating A and C. Difficulty rank and RNG affects how closed the doors start at before player input for each round. That amount of potential time lost is less than a second in the worst case so just perfect play is used. The fastest door order is the starting door, then the right most, then the left ones in order they are reached. I tried my best to make the unsuccessful player's inputs interesting.|
|6||Powerful Hercules||The goal of the first boss stage is to mash A and C to squish Hercules into the other wall. Pretty simple straightforward but this game is scaled to require more inputs from 2 players.|
|Game number||Mini game||Information|
|7||Triangles||The goal of triangles is to identify which of the 6 tetrahedrons is different based on their faces as they rotate. Difficulty rank I'm pretty sure affects the given time and similarity of the different shape. The player cursors persist between rounds so the next triangle can be selected as soon as possible via maneuvering in the precious round. For 1 player the worst case RNG scenario is having the correct triangle appear in opposite far ends every round. For 2 players, each one starts on one side and the cursor persistence can make selections ever so slightly faster. This is the 2nd fastest mini game.|
|8||Sky counter||The goal of sky counter is to count the number of flying objects and select the correct answer. The number of volleys and amount are primarily affected by game rank, although the precise number is affected by RNG. The cursors persist between rounds so the other player can be set up closer to the correct answer during the previous round if needed.|
|9||Train pusher||It's train pusher again.|
|10||UFO search||The goal of UFO search is to observe where the UFOs stop and select their location to blow them up. Difficulty rank primarily affects the number of UFOs and their speed and time moving around. RNG determines where they can stop and some manipulation can be used to slightly reduce the number of moves, though it's by frames. Unusually I kept getting a lot of patterns where both players could reach the correct selections at the same time. That's not always the case and would lead to either player having a slight speed advantage by frames.|
|11||Hungry lions||The goal of hungry lions is the split the big piece of meat by shooting the middle with bullets until a path from the top to bottom is made. This is scaled in 2 players by making the bullet holes smaller compared to 1 player mode. The main RNG is where the initial pull of the meat goes, which is a problem for a 1 player game. For 2 players this always means one gets to shoot sooner than the other, mitigating the effects of RNG. This doesn't mean that the first shooter is the one to get the winning shot, though.|
|12||Coin shooter||This repeat shows that 3 frame double shots are on par with 2 frame double shots, where the winner was doing the 3 frame double shot pattern. Otherwise it's the same game|
|13||Find crooks||More crooks get busted.|
|14||Shaved ice||The goal of shaved ice is to spin the directional inputs clockwise or counterclockwise to shave the ice as fast as possible. I'm pretty sure the only thing difficulty rank affects is the time limit as it's a very simple game. Direction doesn't seem to influence completion time.|
|15||Tricky Fredi||The goal of this boss stage is to rapidly alternate left and right to break out of the binding ropes. Only one player needs to complete their escape to win. Also very simple in execution.|
|Game number||Mini game||Information|
|16||Hammer fire||The goal of hammer fire is to save the people on the building by hammering the burning part of the building. Difficulty rank primarily affects the floor count and how long each part is on fire. There's more ability to affect RNG by intentionally having one player whiff or simply not hammer, letting the other get a better fire floor spawn on their next swing. This isn't as much of an option for 1 player games where their RNG is determined on the frame the game's selected.|
|17||Triangles||Even more of the fastest triangles out there.|
|18||Fish cubes||The goal of fish cubes is to get the fish cube on the starting square right side up on the exit square. Difficulty rank affects the complexity of the puzzles and RNG slightly affects possible puzzles in those pools. For the most part the best solutions at high rank are similar in moves required that I didn't bother with puzzle manipulation too much.|
|19||Castle shapes||More castles means more layouts to select.|
|20||Towel heater||Once again, the two are back at it with the spicy towels.|
|21||Sky counter||The sky count-em-up is here again.|
|22||Train pusher||push push push|
|23||Moai match||The goal of Moai match is to figure out and select the matching Moai with the floating one, identified by their shape. Difficulty rank increases the similarity of the possible selections. RNG affects the correct location although both players start in the center. Having to select the far end ones is slower but even then the top Moai has to complete a rotation regardless of where the correct answer is.|
|24||Ball puncher||The goal of ball puncher is to punch the metal ball as it swings away until it gets high enough to hit the target in the middle. Difficulty rank affects the weight of the ball and the initial swing, culminating in a 10 ton ball with no swing at max rank. Even with perfect inputs the time required is still very tight.|
|25||Hungry lions||Separating meat with a gun, just like real life.|
|26||UFO search||Those UFOs can run but they can't hide.|
|27||Attractive Cat woman||The goal of this boss stage is to simply alternate A and C to run forward, adjusting vertical location with up and down. There are hearts that will knock the player back upon collision. Only one player needs to get to the goal to win so the other player can get a heart, as a treat.|
|Game number||Mini game||Information|
|28||Castle shapes||Spin them castles one last time.|
|29||Moai match||Oh my, it's more Moai!|
|30||UFO search||pew pew pew|
|31||Find crooks||That's a lot of criminals that got in there.|
|32||Hammer fire||Even with RNG manipulation this is still rough to get a fast time in.|
|33||Ball puncher||Trying my best to make the unsuccessful player inputs more interesting.|
|34||Coin shooter||Shooting a coin all the way up to 100 meters is pretty impressive.|
|35||Fish cubes||Rotate cubes for more fish. Includes showing off a slightly slower solution in one of the rounds.|
|36||Train pusher||The absolute limit of pushing here.|
|38||Hungry lions||A few last big meats for the road.|
|39||Towel heater||rub rub|
|40||Barrel rescue||The goal of barrel rescue is to shoot the oncoming barrels in their target spot. This can be done on the first frame but for the most part only the last barrel matters in when it's shot down. It's fairly time consuming in how long it takes for the barrels to be vulnerable. Even then the timing and hitboxes are pretty wack and make minimal sense. Definitely one to avoid in real time play.|
|41||Closest fish||The goal of closest fish is to identify which fish is |
|42||The boss of evil, The King Shark||gotem with the round start DP|
The goal of the final boss is to uppercut the fist when it's low and to avoid the fist when it starts slamming. Each dragon punch is done at the first opportunity for a double hit when it's related to getting damage done. After the last ground pound of the rampage it's possible to walk into the fist and get positioned for an immediate double hit shoryuken. The rage periods are mostly determined by a mix of the damage done between chances and boss HP thresholds.
|Jar opener||The goal of this jar opener is to, uh, open the jar. This is done by timing an input when the glowing logo fills up the outline. Difficulty rank I think affects the timing window but otherwise this is a quick and simple game. Shame I never seemed to roll this one in the TAS.|
|Road trip||The goal of road trip is to pick up the girl past all the traffic lights in the given amount of time. Difficulty rank affects the amount of lights and the time remaining. Unfortunately this means that at high enough rank and with bad RNG, this can be made impossible on the first try even with perfect inputs. This should be avoided at high rank in real time play due to that unfortunate factor.|
|Hot balloons||The goal of hot balloons is to remember what's on each balloon and identify the correct one given by the game. Difficulty rank probably affects balloon movement but the kind of slow cursor movement and having the unfortunately likely RNG roll of selecting the farthest balloon every time makes this one very slow.|
|Hair grow||The goal of hair grow is to identify the bottle which has the hair growing formula. This takes a while since the animations take a while. Difficulty rank increases the complexity but it's still a pretty easy game. It's just not a fast one, especially if RNG gives more test drops and/or puts the correct bottle far away from the cursor.|
|Monkey climb||The goal of monkey climb is to guide a monkey to the banana bunch. This involves walking around the tower to working ladders and alternating A and C to climb them. Difficulty rank increases the tower complexity. Regardless of difficulty this mini game takes a while due to the animation of the towers and general height of the towers.|
|Hat spin||The goal of hat spin is to keep track of the coin among the spinning hats and identify its location. The spinning takes a lot of time and the cursor speed is pretty slow. Difficulty rank increases spin speed mostly, as well as changing directions. This is by far the slowest mini game of the bunch, especially if RNG decides to make the coin end up as far as possible from the player.|
Rough order of mini games by speed
These were determined from the abandoned 1 player TAS and don't account for the full extent of rank and RNG, biggest offender being closest fish. Even then I try to give the possible variance from educated guessing. Variance can be positive or negative. These also assume that the full mini game explanation is in effect.
|Mini game||Frame count||Frame count variance|
There's a very obvious way to improve this. Simply get better RNG in mini game order to get the fastest games possible. However, this is much easier said than done since it's hard to say a given set order is going to be faster through a big chunk of mini games that all have to be completed. If someone could figure out the order without having to play through all the games first would make routing a better mini game order much easier.
A note to encoders
Once again since this an arcade port, I have an alternate input file that includes putting in initials at the end.
Samsara: Claiming for judging.
Samsara: Something about the art style in this game is just so... Sega Saturn, even though this is a port and not the original, and it's the third game in a trilogy, and the designs had already been around since Bonanza Bros... But man, it just feels so Saturn to me.
The run looks good. I appreciate the usage of 2 players to both fully optimize each minigame for time and add variation to each one, even making repeats interesting to watch. Accepting!
Also, I have absolutely no idea how the title should be displayed on the site. Everything I look up gives a slightly different romanization. Whoever publishes this, remember both the alternate file and to have fun deciding which title is correct!