Guess the salary | #8 ETown Hackers

Guess the salary for jobs in various locations. How close can you get?

Comment

Our original idea was to do a guessing game with how much different people paid in taxes, as we know how big of an issue wealth inequity is. The poverty rate in the USA is 11.4% (as of the 2020 Census), so wealth inequality is especially important right now. However, we found that we couldn't access data on how much people pay in taxes, so we decided to use wages of different professions in different areas. We spent a while refining and formatting the data. Getting the livable wages took a while since we had to manually get data for lots of counties in order to get the average livable wage for the area.

We also spent a while figuring out the thresholds for the scoring. In the beginning we had decided on having a lives and point system, where you get points based on how close you are to the answer. If your answer is close but not correct, you lose one life but still can get some points depending on how close it is. If your answer is incorrect, you lose two lives and get no points.

However, as we started playing the game, we found that it was too hard. We spent a while tweaking the values and changing up our system until we found something that worked out well. Our main changes were:

1) Increase number of lives to 7 (initially was 5) 2) Use a percentage of the wage as our threshold. This makes it easier to guess the higher paying jobs since the threshold is higher. We added a minimum value to prevent the threshold from getting too low for low paying jobs. 3) Start with higher thresholds and shrink them as the the rounds progress. This makes it easier to get started.

Our final thresholds were:

Rounds 1-5:

  • Close: 20% of wage with minimum of $7,000
  • Correct: 10% of wage with minimum of $5,000

Rounds 6-10:

  • Close: 15% of wage with minimum of $5,500
  • Correct: 6.5% of wage with minimum of $4,500

Rounds 11+:

  • Close: 10% of wage with minimum of $4,500
  • Correct: 5% of wage with minus of $3,500

Note that the thresholds are for the difference. So for example, if the first round had an answer of $40,000, the thresholds would be:

  • Close: $32,000 to $48,000 (20% of $40,000 = $8,000)
  • Correct: $35,000 to $45,000 (10% of $40,000 = $4,000 but increased to the minimum of $5,000)

Our frontend is built with HTML, CSS and JS, without any sort of framework. The backend is built using Typescript and NodeJS, and is connected to the frontend using web-sockets. The app is hosted on Google App Engine, and the domain name (wages.tech) was purchased from domains.com. The source code is publicly available on GitHub.

Albert did a lot of the backend work and Eli did a lot of the data collection and processing. We both worked on the frontend. In addition, we both worked together to solve problems like the thresholds mentioned above as well as giving each other feedback on our individual work. Note that most of the work was done over VSCode live scare, so many of the commits show under Albert even though we both contributed.

Share this project:

Updates