4 ms·

The number of mentions needs to be normalized by the number of job posts. Example: - 2021-01: posts=842, python=194, ratio = 194 / 842 = 0.23 (mentions per po

The number of mentions needs to be normalized by the number of job posts. Example: - 2021-01: posts=842, python=194, ratio = 194 / 842 = 0.23 (mentions per post) - 2025-01: posts=487, python=87, ratio = 87 / 487 = 0.18 And then if you want to see a trend, do a moving 6 months average. [2021-01] https://news.ycombinator.com/item?id=25632982 https://news.ycombinator.com/item?id=25632982 [2025-01] https://news.ycombinator.com/item?id=42575537 https://news.ycombinator.com/item?id=42575537

math_dandy 2y ago
Number of job posts itself seems like a better measure of the state of the job market than mentions or mention-density. What aspects might mentions capture that number of postings wouldn’t?
- internet101010 2y ago
  It would capture the shift in need for each language.
- snek_case 2y ago
  There's also the popularity of HN itself and the popularity of posting jobs on HN that's an important factor. You could both normalize the figures and include a separate graph for job posting density per month or something like that. Total posts on HN monthly would also be interesting to visualize. Is it trending up or down?
  - bb88 2y ago
    There's also the popularity for HN based upon the particular language. It's possible there's skew where rust becomes more popular say, and gets the bulk of the new posts, but python or java aren't super interesting anymore. So those people stop showing up to HN. I'd like to see the popularity of posts by language. I bet we'd see a lot fewer Java and Python posts compared to Rust.
mountaineer 2y ago
Yeah, I found the percentage of job posts with the keyword in https://www.hntrends.com https://www.hntrends.com when I was publishing it, was effective.
a-dub 2y ago
probably more like number of job posts that include at least one programming language for conditioning. (if you want to look at relative trends). distribution over job class as well as an accurate count of actual jobs per post here also probably makes sense for measures of overall activity and longitudinal analysis. could maybe skip the deep counting if you can show that it's static across the dataset.
wslh 2y ago
Optionally, the novelty factor plays here: C++ created 40 years ago, Python a little bit later, and Rust less than two decades ago. The selection of programming languages should be augmented/benchamrked with the TIOBE index. [1] https://www.tiobe.com/tiobe-index/ https://www.tiobe.com/tiobe-index/

math_dandy 2y ago

Number of job posts itself seems like a better measure of the state of the job market than mentions or mention-density. What aspects might mentions capture that number of postings wouldn’t?

internet101010 2y ago
It would capture the shift in need for each language.
snek_case 2y ago
There's also the popularity of HN itself and the popularity of posting jobs on HN that's an important factor. You could both normalize the figures and include a separate graph for job posting density per month or something like that. Total posts on HN monthly would also be interesting to visualize. Is it trending up or down?
- bb88 2y ago
  There's also the popularity for HN based upon the particular language. It's possible there's skew where rust becomes more popular say, and gets the bulk of the new posts, but python or java aren't super interesting anymore. So those people stop showing up to HN. I'd like to see the popularity of posts by language. I bet we'd see a lot fewer Java and Python posts compared to Rust.

mountaineer 2y ago

Yeah, I found the percentage of job posts with the keyword in https://www.hntrends.com https://www.hntrends.com when I was publishing it, was effective.

a-dub 2y ago

probably more like number of job posts that include at least one programming language for conditioning. (if you want to look at relative trends). distribution over job class as well as an accurate count of actual jobs per post here also probably makes sense for measures of overall activity and longitudinal analysis. could maybe skip the deep counting if you can show that it's static across the dataset.

wslh 2y ago

Optionally, the novelty factor plays here: C++ created 40 years ago, Python a little bit later, and Rust less than two decades ago. The selection of programming languages should be augmented/benchamrked with the TIOBE index. [1] https://www.tiobe.com/tiobe-index/ https://www.tiobe.com/tiobe-index/