Following my last post on the December 2018 analysis of the ProtonDB data, I went through the free text comments (i.e. what people actually wrote about the performance of the game they tried with Proton), and based on 3541 comments, comprising more than 47 000 different words (after removing stopwords like “the”, “at” and the like), I created this little wordcloud of the most popular terms. The bigger the word, the more frequently it was used in the dataset.
What do we see? Well, that the most common terms (note that I removed “game” and “steam” intentionally from that list), are positive and relate to games “working”, “running”, sometimes “perfectly”. Of course, there’s clearly “issues” and “crashes” visible as well but from a bird’s eye view it looks like more positive than negative.
I will probably include that from now on in every monthly dump – if something pops up we may see some visible changes in the word cloud.
You might want to check out the following articles too!
BoilingSteam lets you access our content for free, but writing articles is a constant investment. We don't use ads or sponsporship, help us make our activities sustainable by donating via Patreon or LiberaPay if you prefer it anonymous. You can follow what we do via our newsletter, our RSS feed, our Mastodon profile or our Twitter feed. We also have Peertube, Youtube and LBRY channels. If you'd like to chat, you can also find us on #boilingsteam:matrix.org. (what is Matrix?)
Similar to what Van said; I have left several comments such as “no issues” or “no issues with …” for cases where I didn’t experience issues that others might have had previously, so it might be worth removing that word as well.
good feedback about “no issues”. I will make an exception for that one. If you see something else let me know.
Ok after checking again, the frequency of “no issues” is certainly there, but is completely dwarfed by the number of real “issues”. If I plot “no issues” independently of “issues”, “no issues” is barely visible on the word cloud. In other words, it does not change the big picture.
I think it could be misleading as for example “runs” could mean things like “runs pretty badly” 😉
indeed. but most of the comments associated with runs are “runs fine”
Is there software for doing this or did you have to make this manually? its impressive looking as an overview. is the source something you are able to share?
I think its fair to say that the games that get commented on mostly will be the ones that people are most excited about want to share good news 🙂
this was interesting, thanks
I use R and the wordcloud2, tokenizers and stopwords packages. This and having the data in the right format is sufficient. the rest is minor tweaks.