davidus 2019-05-01 11:51
Nice little game!
Foon → Ludum Dare Explorer → LD44 → Click Heartbeat
By hangacs
| Category | Rank | Score | Count | |
|---|---|---|---|---|
| Overall | 1142 | 2.40 | 23 | |
| Fun | 1133 | 2.23 | 23 | |
| Innovation | 369 | 3.32 | 22 | |
| Theme | 1086 | 2.32 | 22 | |
| Graphics | 1041 | 1.95 | 23 | |
| Audio | 667 | 2.27 | 22 | |
| Humor | 948 | 1.70 | 19 | |
| Mood | 1071 | 2.15 | 22 |
Nice little game!
i think this is a neat concept! i like the idea of risk and reward using only numbers and math. i do think it is a little easy, though, since you can just keep sending the same number and wait several beats to consider the next card. perhaps, every beat, you could shuffle the cards and add new ones in?
Nice work, but it seems to be lacking in gameplay. This could be a fine educational game, though.
Very interesting concept! I love math puzzles and like dealing with the number theory and this seemed especially related to the Collatz series which I like a lot.
One criticism I have is that there's no backspace button, so some big number turned out to be divisible by 3 and then you're destroyed and can't do anything about it. I strategized with pen and paper around the number 4 as the input and managed to beat the game that way.
@snowdozer I'm consider adding more path or something. But random the cards will make the game totally random and not playable, though it's really random now because player can only consider few cards.
@gentel I think 1/3 numbers can be divided by 3, but too many node totally become chaos in my brain. I just add some random nodes and test random inputs, and sometimes I get very large number.
I just randomly placed the nodes and let the machine go but it was fun to later trace back how the numbers were working. It took me a little amount of time to figure out I was supposed to click the numbers on my keyboard. I was originally trying to click the cards.
@dk5000p Thanks for your feedback. I find that I should write a more detailed instruction because limited by my skill and time the game really lacks interaction.
I love finding weird & unique games during LD... this was something different, which is rare. I was terrible at it, but I really enjoyed being terrible at it!
Evaluation...
1 - Mechanics and Learning: Learning the main mechanics has been challenging. I know I can interact with card numbers just after each heartbeat. After I get to the maximum number of nodes added, as I understand it, it is not possible to replace previous nodes. I tried the game twice, the first time I started trying to control the numbers I was filling; on the second try I first tried to fill the nodes; both approaches failed miserably. It really lacks a little instruction (maybe Math is taught relatively differently where I studied and that prevents me from seeing something).
2 - Flow: I was distressed all the time because I could not understand what I was doing. There was no visual feedback to know if I was advancing or not, I think. I suggest you focus on a little more about it; or separate in two stages: an extremely simple, clear objective that already teaches to use the cards, another with the degree of difficulty presented (perhaps).
3 - Narrative and Polishment: The interface is very simple (either in the sound part or in the graphic part). I believe that, in time for jam, it is good enough to show the concept. The main polishment suggested would be to amplify the feedback (either sonically or graphically) to facilitate understanding and progress in the game.
4 - Culture: The game has the potential to attract school-age players if it is a set of levels based on mathematical challenges.
5 - Monetization: I believe that the game could be delivered for free via smartphones, having extended version paid (with more daily departures, metrics etc).
@gustavo-christino Wow! Thanks for your suggestion.
1. At present this game is designed for computer rather than normal man. For me it's either easy or hard to win. As node number increasing, input is more likely change exponentially.
2. Maybe change the goal from a sum to largest number is better. Or add other mode such as limited time or limited clicks. It's hard to use computer to solve this problem because many nodes and many cards will combine so many possibilities, so maybe I should add a global record system. However it's obvious that this does not attact majority of player.
3. I'm new to unity and especially GUI. I designed some visual feedback effects and some statistics but it takes time to implement. I'll consider to deliver a mobilephone game but it will take a long time.
The game is nice and has potential! Good job! :)
It's a great concept, maybe some polishment and some actual instructions on how to play and this game would be a great mobile to puzzle around
Great work
cheating.png
Hey, very nice little programming task you provided here. After trying twice and getting to know the game I just wrote a little something < 100 lines of code that will randomly look for better and better ways to generate the highest possible number, starting from either 0, 1 or 2 (we do not want to click too often, right?). Nice little challenge and even without a programming background this makes for an interesting puzzle game :-)
One thing to improve though: put the control scheme at the start of the game, I don't want to have to read your description or any other external manual. The game should let you know how to play it.
Another thing: my program thinks that rule 4 is really unnecessary, all the good combinations have no rule 4 in them.
cheating_more.png
Here is the best my program found, start with 5 heart beats:
cheating_even_more_with_5.png
@gonutz Wow! That's really amazing. I have 2 questions——is this the best solution for "2"? I'm not familiar with algorithm but can this be called "Greedy Algorithm"? I guess if you want the best solution you have to test all possibilities.
Actually my algorithm is stupid, I just randomize all cards every try. As you can see in the last screenshot, it starts out with two 11s which can obviously be improved :-)
Considering the number of combinations, there are 4 cards for 21 spots, that is 4^21 = 2^2^21 = 2^42 which is 1024 times the numbers representable by a 32 bit value. I have recently done a unit test over a full 32 bit input, it took about 5 seconds. This means we can probably test any card combination for one specific input in about an hour and a half. Since you have to click the hearts, considering we don't write a bot to click them a zillion times, let's say a user can click them at most 30 times in one heartbeat. That would be 45 hours, so in two days we would know the best answer for sure.
That said, it would also be fun to write a little simulated annealing where you randomize the card combinations less and less over time and then try changing only one or two cards at a time instead of all at once. This would get us very good results more quickly. Maybe I will do this tonight, it sounds like a fun exercise :-)
I am confused :| I feel stupid. Will try again with a clear mind XD This is a bit abrupt, but I like the idea and realization, heartbeat sound is kinda relaxing but this is kinda nerdy and mysterious and i don't feel I would be able to beat it without randomness. Interesting approach :)
@knarf Relax! No one can do a good job without using computer.
@gonutz I've tested up to 11 nodes for number 2 as input. It seems that N[number of node] = a*Log(M[max output]), which show no different with expectation. If I find the best node step by step, number will stop to increase at one point, which suggests that I should add more randomcity to my card.
I am not quite sure what you mean by N[number of node] = a*Log(M[max output])...
@gonutz i decided to build on your idea of using a computer to do the work for us!
here's the result, with a base of 26. (my finger hurts) heartbeat2.png i haven't proved for sure that my strategy creates the highest number, but it's pretty darn high.
one thing to notice is that you can go into an endless loop of applying the third card. if the remainder of the number divided by 3 is 2, then every time you apply the card it'll result in another number with a remainder of 2. or, to put it mathematically, ``` if n % 3 = 2, then (n * 2 + 1) % 3 = 2. ``` this means that the number will never be divisible by 3!
of course, this only ever increases the number by a factor of 2. the second card can increase the number by so much more, but it has the potential of throwing us off of our remainder of 2 journey. that's why i wrote a quick program that checks if the second card throws us off track, and applies the third card otherwise. and so we end up with this number!
i did add a part to my program checking if the fourth card was useful, but it never seemed to be. the first card wasn't that useful either, since it would often lead to dead ends where the only way forward is to drop down using the second or third cards. (there are very few cases where decrementing the counter is beneficial to us, but note that they do exist, i think). i do think there's a possibility these cards are useful, but it's kinda low and it would take some more work to figure out.
edit: noticing quite a few possibilities with fourth card -> second card. probably the earliest one is by starting off with 32, fourth card (128) then second card (821), and continuing using my above strategy, resulting in a number that is 11 digits long! there are also a million other later ones that use fourth to second, like 392 -> 8651 or 4712 -> 84881; tons of room to go higher than my original result.
Nice analysis, your game has a lot more potential than you initially thought I think :-)
With these results you could now go and create different cards, maybe add some levels with higher scores, drop in fixed cards at some positions which the user cannot change... There are a lot of ways to build on this game. Maybe for the next LD? :-)