Conversation Analysis

of the Storm” and “Star Wars: Battlefront II”. The word “gambling” similarly appears in some cases, here it is seen in relation to “Star Wars: Battlefront II”. Although not directly negative, media

coverage and studies have generally found the concept to be loaded and undesirable, in connection to microtransactions in video games.

The third category is made up of words, which although sensible in their relation

to“microtransactions”, hold very little significance, when presented at face-value. These include, monetization, customers, currency, assets, market, cash, among others.

The last category is the one of words, that through the utilization of Word2Vec, have been concluded to be a synonym to a context-specific word, as mentioned in our first category.

To further expand on the results from the Word2Vec algorithm, a closer look at the total occurrences for terms, produced by the algorithm is showcased and examined.

The below chart (Figure 16.) shows the frequency of words across the cases, produced by the CBOW and the SKIP-gram algorithm.

Figure 16 - Total amount of words found by SKIP-Gram and CBOW

The chart shows a steep drop-off, and a somewhat high variance in terms of words used. Lootboxes and cosmetics top the chart, and at a first glance, all terms found by the Word2Vec algorithm seem, to some extent, related to the research of this paper.

For the SKIP-gram algorithm, it is, again, seen that lootbox and cosmetics top the charts, and thus gives strong indications, that these terms are by far the most frequent term associated with

microtransactions, across all collected data. This furthermore brings fuel to that fire, that cases in which lootboxes are included, is a heated topic of discussion amongst the users of Reddit. It is also interesting to note, that the term “whale” is identified by both algorithms, which contributes to the thought, that Reddit users might be self-aware of their spending behaviour on microtransactions within a game, and thus are able to have a debate hereof with other users.

The aggregated total of words across all sets, based on both algorithms’ results, is visualized in the below word cloud (Figure. 17). Besides emphasizing the words of relevance for our study, it furthermore illuminates words from the previously discussed four categories. It is seen, that though

not directly connected to our research of microtransactions, that monetization, revenue, and purchases, are topics recognized across all cases, and as such could make the argument for other studies to have a further dive into these areas, but fall out of the scope for this specific analysis.

Figure 17 - Word cloud showcasing important terms

Based on the selected word list, that comprises the topic of microtransactions, it is possible to measure the amount of discourse related to microtransactions for each case (visualized on the left in Figure 18).

Unsurprisingly, there is a high correlation between the amount of textual artefacts (discovered in Figure 12.) and the specific discourse. League of legends tops both charts, although the number of artefacts is far superior in League of Legends, compared to any other case. In Figure 12​​it is shown, that League of Legends more than doubles the amount of textual values produced, compared to the closest case, being Destiny. Based on this, one could have assumed, that League of Legends would thus have a broader audience discussing microtransactions and thus a larger discourse, though our figure, illustrated below, shows differently.

Here it is seen, that League of Legends only marginally edges the case of Fortnite, in their respective amount of discourse towards microtransactions.

Additionally, it is peculiar, that the case of Star Wars Battlefront II, which only had a total of 250.000 comments for the entire SubReddit, consists of more than 25.000 comments surrounding

microtransactions. In other words, 10% of all content on this SubReddit resolves around

microtransactions, which strongly indicates, that this topic was not only heatedly debated, but an integral part of the entire discourse within the community. This revelation from our analysis makes further sense, seen in the light of all the negative backlash the game received, as discussed in our

Figure 18 - Total discourse for each case vs. sentiment for each case

Looking at the bottom of the sentiment analysis, illustrated in Figure. 18, it is noteworthy, that four out of six cases, which had the lowest amount of negative sentiment, all use F2P as their business model. Furthermore, our analysis illuminates the finding, that these cases’ low amount of negative sentiment, comes in spite of the fact, that they are, by our analysis, found to have the highest amount of discourse. As previously mentioned, League of Legends, in terms of artefacts produced, is the most debated case, and thus it is interesting to note, that it is so, while steering away from negative

sentiment towards its use of microtransactions. On the other hand, our analysis of the sentiment for these cases also unriddles, that the four cases with the most negative sentiment, all make use of a business model, that requires users to buy their game up front, while still being confronted with microtransactions after their initial purchase.