Explore the caves and shoot The Wumpus with your Crooked Arrow.
Notes:
This plays like the 1981 TI-99/4 game written by Kevin Kenney. The original BASIC text game was created Gregory Yob in 1973.
I made this version as an excuse to implement the maze drawing algorithm, which I've always found intriguing. After that I just wanted to pile on cute things to amuse myself.
I still don't understand why I need information that he's nearby if it doesn't give me anything?
Ending 4: manage to shot onto yourself with your only arrow.
Конор, I think if the smell range was only 1 room it would be too easy.
I love the art and music (the idea of playing music while moving is especially charming). I also really like the core concept, but I’ve been feeling a bit frustrated because it often feels like I have to guess and take risks rather than really reason things out. Maybe reducing the Wumpus scent range to one tile, or giving the player an extra arrow, could help?
Really enjoyed this cart and its a nice variation on the original. The presentation is wonderful and I love the music and art style
Wow, this is very good! I like the style, and it's very fun to play. :D
leafs4n: Yes it is not usually possible to win without taking a risk. I thought about enhancements like multiple arrows, or adding a one-use flare to light up an adjacent tile, or dynamite so you could turn a corridor tile into a chamber tile. Maybe some day I will try to upgrade the game, along with other enemies and hazards.
But I think the game is fine as a simple game with a luck element, meant to be played quickly and repeatedly.
Very fun trying to puzzle out where the Wumpus and the pits are based on the clues! I like that there is some level of chance to the gameplay. It helps add tension in a simple survival horror game without active enemy pursuit where you are as much the hunted as the hunter. Very clever detail that every step you take plays another note of "In the Hall of the Mountain King," which I noticed plays at the start of 1981 version when I looked up some gameplay. This is a great remake, love the character designs :)
Great rendition of this game! As someone who's played several makes and remakes of this, this is one of my favorites.
I'm personally a fan of the puzzle design as is, especially the Wumpus tile being 1/2 spaces away, since part of the puzzle there is figuring out whether a given scent tile is really a dangerous 1 tile or a safe 2 tile. I think if you did want to have adjustments to the difficulty settings, I would let the player adjust how often rooms are generated vs corridors (more rooms being easier, perhaps all the way up to a "tutorial" difficulty where there are no corridors at all).
@cheepicus add this game to another of the Pico-8 carts that I just keep returning to again and again. Regarding your comment above: "I think the game is fine as a simple game with a luck element, meant to be played quickly and repeatedly." You have absolutely nailed it. A simple concept, packaged so beautifully, and captures the concept of "just one more go...". So many wonderful parts to this game particularly the music and movement. Don't change it.
@cheepicus - A minor hiccup in the game. The function to generate Roman Numerals is slightly incorrect. Here's a screenshot:
Try the following correction:
Nice. I do like the addition of win % . I thought I was winning every other game, but apparently I am only on 30%
@phil, I think we forget about the games where an early gamble has to be taken and we lose. Time loss is negligible and the unfairness is easily discarded as we didn't get to be invested in the run. A bit like in minesweeper where your gut feeling success rate is very near from the % of wins for game where you click at least 15 times, but very far from your real %.
50% is probably near from your success rate among those runs you got to invest yourself in. We need more stats and graphs !!!
I'd be curious to know the success rate of a perfect solver.
My most common mistakes are forgetting that the wumpus can be in a hole, and that two different holes can sometimes be reached from the same tile.
.png)
How do you doordash lol
I'm guessing basketball is when a bat drops you into a pit, so maybe it's when a bat drops you onto the wumpus?
I can't seem to increase my success rate any more. (I could always cheat and restart the cart until I get a sure win, but I'm working on the assumption that I must finish every started run)
Looking for advanced tips, what would you choose in that situation and why ?
.png)
@RealShadowCaster - I'd go with my gut and assume the Wumpus is on the bottom row. So, I'd gamble and go to the bottom-left cave and move down to try and give more clues. I'd hope that it heads straight-down. I'd also assume that the caves are both at top-right. But I can't see a strategy here that doesn't involve some guesswork.
@cheepicus - I really can't get enough of this game! I just return to it when I have 2 mins to spare (and spend 15 minutes because of "just one more go...") Here are my current stats (I'm up to 41% now)
@phil, I'm at 36% win, so there's definitely room for improvement on my side.
I guess my "strategy" to walk on bats when I can't make up my mind is not great, even if it is less risky as an action than taking a 50/50 chance, the potential additionnal info must not have been worth the risk in many cases...
What is very surprising to me is the Doordash vs Batsketball ratio.
I'd expect around 2/3 Batsketball and 1/3 Doordash because there's two holes and one wumpus, and that's roughly what I'm getting, but you have a 1/4 3/4 ratio yourself. Given how much you've played, it's got to be an indication of a difference of strategy between us.
I've never been delivered to a wumpus in a hole, so I think these are excluded from the list of potential bat delivery points, and that must skew the probabilities towards Batsketball a bit, but it's likely you get taken by bats more on levels where the wumpus is in a hole. Can't figure out why that should be the case...
Maybe it's the other way around : there's two obstacle tiles instead of three when the wumpus is in a hole, I get more of the map and am more likely to take a chance at shooting, and my 66% Batsketball/Doordash reflect that I'm letting myself get caught by bats on levels I'm stuck early because there is 3 danger tiles, and that would be why I'm getting a 66% when your % reflects the real average...
I think there's an error in the win % formula :
I played 5 games, experimenting with bats :
1 arrow, 2 Batsketball, 1 Doordash and 1 win, and have a win % of 13% instead of 25%
I think that the current formula is
win/(arrow+pit+doordash+wumpus+batsketball+win)
1/(1+2+1+1+2+1)=1/8=12.5% rounded to 13.
In effect, Doordash and Batsketball are counted as double defeat.
The correct formula is simply win/(arrow+pit+wumpus+win)
I think we could have a lot of cool stats if instead of tallying the totals, we saved the sequence of results.
With 6 possible results (or maybe 7 if we add chicken for the cheaters that reset the game mid run), we need 3 bits per run.
With 4 cartdata slots, that's 256 * 4 * 8/3~=2730 history length. A bit less if we add things like date or header info, a little bit more if we pack the slots more. For example, 6^3=216 < 256, so we can pack 3 results in a byte instead of using 9 bits.
The cool thing about cartdata is that the statistics cart can be separate from the game.
@RealShadowCaster Great analysis. I hadn't noticed that my stats for batsketball was so much higher than doordash by ratio. And a bat drop to a wumpus in a hole! Well... I can't say I've noticed where the wumpus was sitting when I lose on a bat-drop.
Occasionally I play "minimal" and try to get to a point in as few reveals as possible to be sure where the wumpus is. This really depends on a lucky layout. More often than not I play to reveal the whole map as much as possible. The more rooms, the easier (generally) and I wonder what my win rate is based on number of rooms...
@RealShadowCaster You're absolutely right about double-counting the doordash and batsketball deaths. Thank you! I'll get a fix in ASAP.
If you get bat-dropped on the wumpus when it's in the pit (canonically it has sucker feet) then the wumpus takes priority. This means doordash should skew slightly above 1/3 of bat deaths. I think it's 10/29 instead of 1/3. So perhaps @phil's records are even more of an outlier!
I actually don't know how to program multicart stuff, so I'm hesitant to hack it into a mostly finished game. If I get comfortable with it I can come back to that.
I did consider tracking other stats but they seemed incidental. (For example, longest arrow kill.) Plus, the gravestone screen is pretty crowded already, and not much text can go on the graves. A second record screen, maybe.
Tracking chickens sounds interesting. I hadn't considered resetscumming at all.
@phil I admit I honestly did not think the "C" roman numeral would be necessary. Incredible!
Kevin Kenney made a great little wumpus version, I'm happy other people appreciate it too.
Fun fact: the TI/99 game could produce isolated unconnected corridor loops:
That's okay because the bat only drops you on chambers. I never saw it produce an unconnected chamber; I think it probably checks for those and retries, which is what I have done. I realized another danger of the TI/99 maze style is an unwinnable game, where the pits block off the wumpus from an attackable position. (Not a problem in the old text game.) So I check for that, but I don't know if the TI/99 version does.
@cheepicus
rns={1000,"m",900,"cm",500,"d",400,"cd",100,"c",90,"xc",50,"l",40,"xl",10,"x",9,"ix",5,"v",4,"iv",1,"i"}
just to be safe : @phil is already at 200 witch is halfway to a display error : 400 is CD, not CCCC :p .
About the stats and multi carts, my idea was to have two independent carts : the game cart that would store the sequence of game results in 4 cartdata slots (currently you use just one and store the sum of each result), and I'd make a statistics cart that would read (not write) the 4 cartdata slots produced by your game and display cool stats. Currently, the available info in the cartdata slot is entirely displayed in game already.
So no multi cart data exchange at all, just two independent carts using the same cartdata slots.
I totally get you don't want to mess with a working game except for bug fixes, and changing the save format might also clear player's current stats. I could fork the game and add stats, if you don't mind, but my pico8 todo list is pretty long already so no rush.
Stats I was thinking of:
most wins in a row
most losses in a row
graph of daily win%
plays per day/week/month/year
...
Updated to 1.2. Record screen improvements:
- Catch reset-scumming
- More roman numeral range
- Note streaks (esp win streaks).
- Hopefully the math is all correct.
I think that's enough record-keeping for now!
Wimpy the wabbit whopped ten wumpuses without worry, WOW !!!
.png)
Looking for advanced tips, what would you choose in that situation and why?
I thought about this longer than any situation in my own attempts, but I would move onto the square I've outlined in blue.
I think it can theoretically contain the wumpus, but it would take some very improbable paths to connect all the near-wumpus squares in the bottom row and not connect to the draft square in the upper right.
The puzzle mechanics here are well-chosen to combine both skill and occasional chance. At first I thought there was too much chance, but then I started really paying attention.
I thought it was neat that this was enough information to solve this puzzle.
This level layout felt a bit unfair, but I made life worse for myself by stepping on bats twice.
At one point I had a 10-win streak. I'm not sure how, since I still guessed quite often After a bit over 100 games I had 60% wins.
Dude this game is fucking impossible how are you supposed to know where anything is?
Because the clues doesn't really matter that much when it's based off of connected rooms and not actual tiles. So the one direction a clue tile can go could just be a stupid hallway and can go where-ever the fuck it likes.
@RealSwordy210 This is why it's really important that the wumpus spoor (the pink stuff) extends TWO rooms out from the wumpus room.
Generally it's safe to explore from two kinds of rooms: empty rooms, and rooms with only wumpus spoor that are connected to a non-spoor room. (And of course, exploring is not safe from a pit-breeze room.)
From this exploration you can sometimes figure out where the wumpus is. Sometimes you can't and you will have to make guesses. With experience you can get better at those guesses.
ok but is it up to 2, or exactly 2?
Because there's a difference between those, and the in-game help menu doesn't do much to clear that tbh :/
And it also sucks that it's just purely random that you can even do anything.
If there was at least an option to generate a beatable map that you can deduce logically with no guess-work
Like, tell me what the hell you would do if the spawning area has like zero safe tile like this shit
Like what I just give up and die just for the game to
clown on how much of a so proclaimed dumb-fuck I am?
And tbh I swear you made it so it generates the map in advance rather than generating the entire
map before spawning you in. Because apparently it feels like a 90% chance I die in what should
clearly be a 50/50 split on a wind tile.
Doesn't really matter though because I think the wind tiles are bugged anyhow as the one here connects to jack-all.
Maybe the dude is actually on the same spot as the hole, and it's just covering it here.
In that case that's just plain-as-day bullshit, and that just makes the
wind tiles even less reliable than they already were.
@RealSwordy210 , each grid has 1 wumpus, 2 holes, and two bats.
Bats are never on holes, and the wumpus can be on a hole, like in your second screenshot.
Wumpus smell is one or two rooms away, regardless of corridor length.
Rooms connected to the wumpus are smelly, and rooms connected to those rooms are also smelly.
In the 1st screenshot, the wumpus smell is connected to a no smell room, so it's a 2 rooms away smell. The wind tells that at least one of the three unexplored path lead to a hole, and yes, you have no other choice than guessing. I'd personally choose up : left is more likely to lead to the same hole the left path below leads to, and right and up have about the same chance to be deadly, but an open tile up has more chance to give more info because it has access to more space. (might be wrong on that one)
I personally don't mind early forced guesses, it's the late game guesses that feel painful to me.
It's the same problem mineweeper has, where a 50% 50% at the end of a huge grid feels very unfair.
Some mineweeper implementations fix the problem by generation grids fully solvable by logic, but they are unfortunately not really fun to play, except for carefully hand crafted small ones where the total number of bombs to find is part of the info you have to use. Even though it IS frustrating, lowering you chance of failure is minesweeper is a very complex challenge that I like, and wumpus has similar optimization problems that I like but that I completely understand someone else might hate.
After my first 100 grids, I had a meager 35% success, and am now plateauing at 55% after around 600 grids.
It would be interesting to write a very good solver that would generate tons of random grids and save the solvable without guess ones for a special game mode of the game that would likely be much more to your taste, but that won't stop the meta aspect of optimizing overall long term %win to be more appealing to me.
Most "no guesses" grids I've encountered are quick wins with very little of the board cleared. I don't think playing a lot of these in a row knowing that's what's coming would be fun. Interesting no guesses boards is no simple task to automatically generate.
Maybe the dude is actually on the same spot as the hole, and it's just covering it here.
I assume the pit is to the south of the room with the wind.
Dude I was so close that time and lost to the 50/50 because why would I have anything nice.
Still haven't won a single game btw, if anyone has any more tips somehow that would be great, I'm tired of getting beaten up by this game.
@RealSwordy210, the room you shot has connection to smell-less rooms going up, down and right, so no wumpus there. Also, you don't have to walk next to the targeted room : the curved arrow can follow the curved tunnel until it reached the next room.
Alternative solve Without elimination :
The room I marked with 2 is connected to the left to a smell-less room, so it's a 2 away smell.
From that room, you can only reach A or B, so one or two of A and B is a 1 room smell.
If it's A, up down and left don't lead to the wupmus so you'd have to shoot right from A to the top right corner.
If it's B, left right and down don't lead to the wumpus, so you'd have to shoot up to the top right corner.
Conclusion, the wumpus in in the top right corner.(and both A and B are 1 away smell rooms)
@RealSwordy210 : I'm sorry to hear you're having such trouble. This version of Hunt the Wumpus can be fun, but you need to be aware of which map states are possible and which are not.
First, does your screen clearly show the pink for "smell" rooms? Your clips show you moving around a lot checking for smell. It should be visible on rooms you have explored.
And yes, sometimes the map generation is unfair. At this point my only available moves were from this "draft" room, with a 50% chance of falling in a pit.
But I got lucky in that one, and explored the rooms below. I now have enough information to win the round. Can you see what to do?
Sometimes I do still explore most of the map. Here there were only a few "draft" rooms, so I was free to explore most of it. But I now have all the information I need.
A general expression of the logic here:
Good luck, I hope this helps you hunt at least one wumpus!
@RealShadowCaster I don't understand why the place I shot couldn't have the dude, because the wind tiles don't say anything about not having a dude there (and vise-versa). Both empty spots had connecting tiles that are both windy and pink
@RealSwordy210
Every tile that is 1 room or 2 rooms away (tunnel length doesn't matter) from the wumpus has smell, and those further away won't have smell. The spot you shot at is connected to a windy but non smelly room (3 in fact, but one is enough) and therefore can't contain the wumpus.
@RealShadowCaster
I feel like that detail should of been way more fleshed out or at least was told in the help menu. ig now it's mainly now just down to if I don't get screwed over by the bats for the 10th time in a row.
I'm actually serious about that, I would post some of the gifs, but lately some of the posts with attachments I've sent here and other places kept disappearing and idk why :/
@cheepicus , just solved a very interesting board and am pretty mad at myself for not taking a screenshot. Most of the board was cleared, yet the positions of one of the holes and of the wumpus where still unknown. I had 6 possible configurations, but 4 of them would lead to an unreachable room, and the two remaining had the wumpus in the same place.
It would be nice and unobtrusive to save the current RNG state before generating the maze, and adding a discreet "copy seed to clipboard" and "play seed from clipboard" (no impact to the stats of course) to the pause menu when on the display map screen. This way, players could start saving a list of memorable maps to be shared with others.
I uploaded 1.3, now with seeds.
Seeds should look like 16.16 Pico-8 hexidecimal strings that would get used with TONUM() and TOSTR(..,true).
SEED=[...]
This shows you what the current seed is. Selecting it does nothing; it's there mostly so you can share seeds with screenshots or whatever. There is no way to enter a seed into the game, but you can type one into a text editor, copy it, paste it into the game, and play it with "PLAY CLIP SEED" below.COPY SEED
This will copy the seed onto the clipboard.PLAY CLIP SEED
This plays whatever seed is on the clipboard. You may have to ctrl-V paste it first for Pico-8 to dare to peek onto the clipboard. Note that this just launches into the game immediately. When playing a seed this way, nothing that happens in the game counts towards the record. (But note that if you start playing a copied seed while you're in the middle of a normal game, that WILL count as a chickenout from the normal game.)[Please log in to post a comment]
