An Analysis of Ethical Human Choice: Game Theory through the Lens of the Trolley Problem
Advisor: Ms. Aprajita Singh
Onsite Mentor: Sarah Jin
The trolley problem is a classic moral dilemma that has plagued people around the world from different cultures for centuries. My project designs a moral dilemma in which a subject is prompted to choose between two paths for a train, each containing a certain chance of having some people tied to the tracks, leading to their inevitable death. An example track could have either a 50% chance of 2 people or a 10% chance of 10 people on either track. To put people in this position is to ask them to balance risks of death for seemingly anonymous people. This project will analyze the differences which drive different decision-making patterns, specifically through lenses of gender, age, and socioeconomic status. The purpose of this project is to study the sociology of human ethics using the mechanism of decision making.
Project Blog
Week 0: The Ethics of Murder?
Hey everyone! I’m Heeya, a senior at BASIS Independent Fremont, and I’m so excited to be sharing my project with you all!
My project is more of a personal query, an intriguing idea I had– analyzing the ethics of murder (hypothetical murder). I’m sure you’ve all heard of the Trolley Problem— a moral dilemma in which the subject is prompted to choose between two paths for a train, each containing a certain number of people waiting for their inevitable death. My project tweaks the problem slightly so that each path contains a certain chance of having people tied. For example, an example track could have either a 50% chance of 2 people or a 10% chance of 10 people on either track.
To put people in this position is to ask them to balance risks of death. I’m excited to analyze the differences which drive the different decision-making patterns, specifically through the lenses of gender, age, and socioeconomic status.
I could bore you with all the reasons I’m qualified to pursue this project, but at the end of the day all that matters is that this project is asking a genuine question I’d like answered, and that’s a curiosity that will drive me to the end.
So until next time, which track would you lead the train onto?
Week 1: Heeya Tries Her Hand At Drawing
I’m backkkkkkk! Welcome to Week 1 of my Senior Project! This week I worked on the design layout of my game - which will depict the modified Trolley Problem for people to easily interact with the situations and give accurate results. Since I’m currently in India (visiting relatives), I decided to draw out my characters and design before transferring them to the Game-Dev platform. It was so fun; I got to draw multiple sketches and really practice my artistic side. You can imagine me and my cousins all huddled around an ipad drawing different versions of the trolley and debating whose is the best (my brother won, annoyingly). Here are a few snippets of the layouts I decided on, although I’m sure they’ll be subject to change. I’m excited to see where it goes!
I’ve also been thinking about our next steps. I dipped my feet into the pool of game development by taking some tutorials in Construct3- that’s the game dev platform I’ll be using- and found it’s quite user-friendly! Next week, I’ll be taking a more thorough look at the platform and beginning the true process of constructing this interface for users to interact with and play my game!
Until then,
Ta-ta from India!
Week 2: Deconstructing Construct3
Welcome back to my journey in game development! This week I began the journey to create my online game using the platform Construct3.
To start, well, there isn’t much to start. Although I’ve dabbled in Python, HTML, css and Java, my experience in digital programming is limited. I most certainly haven’t dealt with any game development in the past, so this week was an adventure to say the least!
I started with some basics, a few youtube videos and online tutorials. Quickly I got sucked into the game development world via youtubers like GDC (Game Developers Conference) and Blackthornprod, and although it’s a brand new world to me, it seems like I’ll have more than enough resources to be able to complete the project successfully and learn a new skill along the way.
So, having established solid ground, I embarked on the platform itself, starting out with the characters. However, as I tried to import my lovely drawings from last week, it disallowed me from copy pasting the Canva images. Unfortunate, I know. But we persevere!
I redrew each of my characters, and focused on their functionalities. I played around with the toggle for the lever, switching it from left to right, and how to change the way the railroad responds to stimulation. I’m still working around some emotions for the person, but that’s a minor issue. My trolley is successfully moving on its railroad, although I’m working on the change from alive human beings to unalive ones.
As I more-or-less finish up the design bits of my game, next comes the technical aspects. I’ll be working on functionalities of the game that simulate the behavioral differences which I’ll be testing for (ex. continuous or intermittent play), and overall stats for each play. But more on that next week!
Week 3: So Many Errors!
Hey friends! This week was a rollercoaster to say the least.
I started it off on a plane back home! A whole 23 hours of living within airports and airplanes gave me a throbbing headache and a horrible case of jetlag. Most of the work done this week was accomplished at ungodly hours- 11PM to 3AM usually. But we managed!
Armed with boundless enthusiasm and a few YouTube tutorials under my belt, I progressed on my journey to create the next gaming masterpiece.
Building on from pieces last week, I needed to continue finalizing the basic functionalities of most of the characters. However, I was about to face my first enemy: the sprites. Having dove headfirst into the world of sprite implementation, the creation of characters within the platform, last week, I was ready to complete their construction. But perfecting the functionalities on these was easier said than done.
First, there was the case of the disappearing trolley. One moment, my friend on wheels was riding slowly but steadily. The next, poof! Gone without a trace. Turns out, I forgot to set the collision boundaries properly, and my trolley decided to take an impromptu reroute to the void. Cue the existential crisis as players grappled with the weight of their decisions, only to watch helplessly as the trolley careened off into nothingness.
As I worked to reconfigure the trolley, it came upon a separate problem- not sticking to the route! Turns out, my attempts to implement a dynamic track-switching mechanism had backfired spectacularly, resulting in a trolley with a mind of its own. Lesson learned: never trust a trolley with a thirst for chaos.
Well then, that wraps up my adventures for this week. I’m excited to update you all again next week!
Week 4: User Interaction?
Oh my god, the week I’ve had! To be fair, it started last week.
To my faithful followers, I am so sorry to say I was just able to get up my last week’s blog. After three whole weeks of faithfulness from my laptop, always remembering the password to the blog website, my dear love Chrome decided to leave me hanging high and dry last week with no way of retrieving my password. Although there was much panic at the time, the crisis has been averted now, the password is safely written down in a special place.
To catch you up on my coding affairs, I was playing around with the main setting for the game when I realized, players had no way of understanding how to navigate the system! Thus this week was a whole bunch of work on separate pages to make the system more user-friendly.
I started with a simple “Landing Page.” The page just welcomes users to the system, and allows them to choose between an “Instructions Page” and starting the game.
The Instructions Page details how the game works, starting with the basics. In case you’ve forgotten, let’s go over it one more time! The page starts with displaying the main characters: the person, the lever, the track, and the trolley. As the page explains the role of each character, helpful arrows direct the user to exactly what is being explained.
For the person, we start with their duties and role as part of the game. the person stands very close to the lever, which is first set to the default mode.
Then, it is shown that the trolley is able to move along the railroad that has been drawn. One important thing I made sure to include is the fact that the trolley will move along the predetermined path without input from the user. This is important because it’ll later become relevant when exploring the roles of those who choose to be passive players versus active players. It’s important for users to know their role so they, even subconsciously, make that choice in their head.
In the end, after the players have clicked through the pseudo-tutorial for the game, they are prompted back to the landing page, and can then choose to begin the game!
And that’s about it! OH BUT WAIT. I almost forgot the most exciting news of all! In my endeavors to work out this instruction page, I figured out how to import the earlier Canva images that I drew before!!! I was so happy to not waste all that hard work, and now my game looks just a little bit more like what I envisioned it earlier.
Next week, I’ll finish up the game for sure, completing any nuances and especially implementing intermittent play between games. By the way, any ideas on what we could use to entertain users in that buffer time? I was thinking of One Direction Music videos, but I’m open to suggestions!
See you next week !
Week 5: PHASE ONE- COMPLETE
Friends! I have wonderful news! Guess who’s done with their game!!!!!! It has been completely developed and is ready for release!
But before we get to that, let me tell you about my adventures with coding it. Construct3, I’ve realized, is pretty easy to navigate. Quite simple actually, except for the fact that I CAN’T COPY PASTE OBJECTS. For every single thing you will see on the screen here, I had to create a separate object, and give it separate properties and build it from group up. Over, and over again.
But, we made it through. All objects were created and were moving in the right places, and users were able to use the game! Splendid! HOWEVER, we now were working more on the intricacies of the game, and my biggest competitor was the timer.
For every situation through which the user is prompted to make a choice, they are given exactly 8 seconds to quickly toggle the lever, or choose to not participate. Now, I wanted to have a countdown so that users knew how much time they had left, and thus were able to make decisions wisely. However, I also had to code it in a way that after ten seconds, given that the user had not chosen to toggle the lever, the trolley would continue on its own path. To my computer science friends, sounds like a simple if-else statement right? WRONG.
Construct3 runs the entire event page(code section) every single second, as in, it repeats itself every second. There is also no built in if-else statement. So, I first tried to create a variable that would track whether or not the button has been clicked. It was a global variable: trolleyRan, set to 0 initially. Once the lever had been clicked, trolleyRan was set to 1. If trolleyRan was still equal to 0 later, the trolley would then start its original path. In normal code, where the computer runs through each line consecutively and waits for the first to complete before progressing, this should’ve worked fine. However, since the variable was set to 0 to begin with, for every second in which Construct3 runs every line of code, the line that was checking for trolleyRan == 0 would always be true, and thus the trolley would race down the original path without waiting for the countdown to end. What a dilemma! If you're confused, trust me, I was too. Feel free to leave any questions(or suggestions) down for me below, and I’ll try to answer them to the best of my ability. Although fair warning: I am not even close to a computer science major.
And that fact emerged quite quickly in my next phase. As I tried to extract my data directly from the game itself, I ran into a plug-in called gameAnalytics. Trust me, Google, Chatgpt, and Safari all told me it would be easy to navigate. Spoiler alert: it was not. I was stuck for hours, figuring what in the world an SDK is(again, not a computer science major), and trying to implement the events into my own code. However, my friends, I hate to admit it, but I am not a perfect person. Even after a long night of the limited youtube tutorials, I was forced to admit defeat. I surrendered to a simpler method: just making a userID which would indicate each choice the users made.
And that wraps it up. I’m so proud that I was able to completely develop this, without any prior knowledge and a limited skillset. It’s been a journey and now i'm officially a gamer 🎮.
My last ask is that you too play my game! It can be found here: https://heeyadas.itch.io/trolley-problem and is open to the public! Remember to submit your userID afterwards!
See you all next week, when I hope to have a little more insight into human behavior.
Week 6: To The Masses....
Hey friends! Welcome BACK from Spring Break! How was your break? Mine went pretty well, I must admit. A whole lot of laying around doing nothing.
Anyway, back to the project. I am sad to report I don’t have much to update you all on. This week was a bit of busy work really. First, I had to distribute my survey out to the folks of the world. To start, I decided to use the platform Prolific! And I must say, it was a whole lot more user-friendly than Construct3.
Prolific allowed me to choose a couple select demographics that I wanted my survey to go out to, and target a certain amount of people. I chose 50 to start with, just to test the waters. We’ll see next week if the number will be expanded.
Further, with the help of my internal advisor, I distributed the survey to people within BASIS Independent Fremont itself! Next on my list, I want to ask some BASIS teachers if they’d be willing to have some of their students take the survey. I think it’d really help with getting people in the age 10-18 demographic.
I’m afraid that’s about it! For my next steps, I’m really hoping to get cracking on that research paper I’m writing to present my findings. While I wait for my participants to take the survey, I’ll be tackling the Introduction, and a bit of the Method section.
Once again, I would love it if you would also take a hit at my trolley game! I promise it’s a lot of fun and really makes you think! Here’s the link: https://heeyadas.itch.io/trolley-problem and remember to submit your userID when finished. Thank you again.
We’re nearing the end folks, the last stretch is upcoming!
Week 7: Parallel Parking and Survey Making
This week’s title may confuse you- but don’t worry, I’m also permanently confused. Over the last week or so, I’ve truly embarked on my journey to master the skill of driving. Easy enough, except I cannot for the life of me figure out how to park properly. I’m great at just about anything but that. And well, as any good driver knows, if you can't park properly the first time, it turns into a whole bunch of little corrections, back and forth, until you finally nail the landing.
To be quite honest, this project is feeling a little bit like my subpar parking. Starting from the beginning, it's been a series of little corrections to my trajectory. But not to worry, we persevere!
This week’s challenge was Prolific surveyors. Now, when I was setting up the survey, I noticed some nonsense about “collecting Prolific IDs.” I didn’t need any such thing, so I simply declined the option to collect. However, I did not realize that Prolific surveyors would have access to their IDs regardless! Thus, at the end of my own survey, when I asked surveyors for this userID, Prolific users gave me their Prolific ID, rather than the one that I specifically prescribed to them myself!
Now this was quite an issue. The userID I created was extremely specific and important for my study. I first thought of sending out a message to all surveyors. After doing that, I realized not all surveyors would be capable of finding the code again.
Thus I began my next quest, making a simple good survey to send out- a simple version of the game I made myself. This was not a difficult task, but rather completed quickly and efficiently. After depositing this information into Prolific again, I was able to collect more data on surveyors, this time without any issues.
Next week, I’ll be finally compiling all the data into a readable version of my results. As we edge upon the finish line, it’s really time for me to get cracking on that paper!
I’ll see you all next week! Stay curious!
Week 8: Light At the End of the Tunnel
Guys, I’m not going to lie, this week has been largely unproductive. It’s been such horrible weather and all I want to do is sleep. Not to mention I got sick. Again.
Some highlights: We had a workshop on presentation skills and how to sculpt our final talk. What a scary thought. We talked about the different merits of previous speeches, and how to prepare ourselves for our own talks! It’s crazy to think this will all be over in three weeks.
As we wrap up, I want to give my most sincere thank yous to everyone who has helped my project. Thank you to every person reading this blog, and especially to those who took the time to fill out my survey a couple weeks ago. More thanks to each of my advisors and of course special thank yous to Ms. Nagpal for organizing this all and making it so accessible and understandable to us.
Although I was supposed to spend this week analyzing data, it looks like that will be more of a next week activity. I did get through a little bit more of my poster, but the bulk of that will be wrapping up next week hopefully.
I’m excited to see through the end of this! It’s been a journey (although it is at a bit of a standstill now.) I’ll see you all next week, with hopefully more to add!
Thank you everyone!
Week 9: Progress At Last!
Hey guys! What a week! I’m proud to say, although chaotic, this week did yield some productive work, in the end. Amidst Senior Prom, Commit Day, and other exciting events, also came some exciting analysis of our data!
Let’s take a look at what we found. In fact, why don’t you take a guess? We can make it a little game. I’ll say a potential outcome, and you try and guess whether it’s true or false.
Elder people were more likely to take risks throughout the game.
Go ahead! Guess!
Surprise, surprise! It’s false! Those who marked their age over 35 were the largest group to consistently choose the option that was the least risky. Now, if I were to make a guess as to why, it isn’t super difficult to come to a conclusion. As we age, it is a common phenomenon for people to choose more stability in their life, and stray away from risks that they may not be able to take. In addition, our sense of morality is more developed, allowing us to make complex decisions about the lives of these characters. Together, these factors play into the overall decision-making of this group, making one of the most strong trends that was displayed in the analysis of my data.
2. Men were more likely to continue risk-taking behavior throughout the game.
Time to guess!
Correct! This one is true! Although women were more likely to initially choose the option that has a higher chance of people on the track, it was observed that men kept choosing that option past the first scenario. Now this one’s a little more complex in its psychological reasoning. What do you think? I need, say, another week to figure it out!
And, thats a smoooooth transition into my next (and last!!) week plans. I’ll be finishing up the psychological reasonings for each trend that I observed, and writing up my report for the entire project! It’s a sentimental end, but it must end soon!
See you soon!
Week 10: The Last of Its Kind
Wow, here we are guys! Finally at the end. It’s been a journey! Let’s cover what I got done this week.
I finished up my research report, finally adding in all the psychological reasonings for my data. Unfortunately, I was not able to get enough variation in my data to correctly analyze the effects of socioeconomic status on risk-taking behavior. But! We were able to notice some distinct trends through the lenses of gender and age, so quite a lot of interesting information came out of that!
I’d tell you more, but I think you’ll have to come to my presentation to find out. Shameless promotion: It’s next Saturday at 10:15 AM! Don’t worry, I’m also not usually awake then, but I hope my talk will liven you up! At the symposium, I’ll also have my game, live for you to play at any given moment! You’ll get the chance to experience the emotional turmoil of murdering innocent people- the product of about five weeks worth of effort. That was a doozy.
At last, there will also be my research poster, which will display a summary of all my findings and my work over the weeks. I’m still working on the poster currently actually! The deadline is this Monday, but I’m hoping to have it finished by the end of today.
I’m at a loss for what to write. I’ve had so much fun sharing this journey with you all and am so thrilled you guys came along with me. Although I’m fully aware I could just be talking to no one, the idea of having an audience out there patiently waiting for my next blog has kept me going at times.
As I wrap up on high school, and start my next journey at college, I feel this was the perfect way to end my high school career. I hope, for the students doing this next year, you choose a project that truly excites you, the way mine did for me. It was a privilege to explore my interests with you all and I could not thank you more.
To all my readers,
Thank you, and Farewell.