TShock User Survey
In Terraria 4.2.10, I snuck in an update that added a link to a Typeform survey. The survey is now closed, and after 325 responses, it’s finally time to discuss the results.
This survey is, of course, not entirely a complete sampling of server owners or plugin developers. Many of the 1,200+ online servers are running older versions of TShock, and many server owners do not update immediately to new versions. With that being said, however, the data is still interesting to look at.
How would you rate TShock?
The first two questions in the survey were directed at the overall quality of TShock, and its user submitted plugin catalog.
The average rating for TShock was 4.65/5. However, 94% of responses were either 4 or 5 on the scale. 5% of responses were in the lower half, rating TShock 2 or 3, and 0% of responses (only one person) rated TShock as a single star.
The plugin catalog rating tells a slightly different story, at a rating of 4.20/5. Only 81% of responses were 4 or 5, meaning that 19%, or roughly one in five users considered the plugin catalog average or below in quality. 15% of the low ratings were 3, meaning that only 4% of users rated it entirely negatively on the scale.
While it’s clear that the plugin catalog could use some work in this department, only 32% of users said that they would pay a one-time fee of $15 for a hypothetical plugin that lowers the barrier for plugin development. This percentage is somewhat higher than the amount of individuals who rated the plugin catalog poorly, and indicates that it would indeed be a beneficial move to at least place effort in lowering the barrier to entry for new plugin developers. The question asked based on overall quality, not quantity. In fact, numerous plugins have under 50 downloads in the new resource manager, when compared to the most downloaded plugin, Essentials+, at nearly a thousand downloads.
How easy is TShock to setup? To get new players?
Two questions were directed specifically at server owners, namely referring to TShock’s setup and player acquisition.
Asking about the setup process for TShock, 21% of respondents rated that the process was difficult. This response is somewhat unsurprising, as the wiki, existing temporary authentication system, and other areas of TShock are less than friendly for newcomers. On Linux, this problem is even higher, as a recent issue and many new Slack users report problems with the newest versions of Mono running TShock correctly. Undoubtedly, this area could use some major work, and, unlike the somewhat pie in the sky question regarding plugin quality, the setup process can be directly modified to be easier by contributors.
Unlike TShock’s setup process, which we have a direct influence on, player acquisition is still a rough patch that we are unable to directly solve. 72% of people reported that player acquisition was on the easier end of the scale, with 28% of TShock users reporting that it was “so/so” to “difficult.” This is obviously a difficult problem to find the source of. TShock historically holds approximately 10-15% of online Terraria players, however, not all Terraria players play massive multiplayer servers, and many play single player alone. This could be due to player base exhaustion (not enough players to join servers), or due to lack of advertising and availability. Unfortunately, only time will tell what precisely is leading to the player acquisition problem being so high.
How easy is it to write TShock plugins?
As previously mentioned, 32% of TShock users reported that they would pay for a hypothetical plugin that made development easier. Yet, with this statistic in mind, 34% of users indicated that plugin development was the hardest option on the scale. 36% of users indicated that plugin development was on par with other mod APIs, including Bukkit, TConfig, and GMod. To put it in perspective, 43% of people considered developing for TShock more difficult than other mod APIs that operate in much the same way as TShock’s.
While this appears to be a significant hurdle to overtake, only a small subset of individuals – 75 respondents, to be precise, reported to be plugin developers, in comparison to the 203 responses to the difficulty question. More than double the amount of plugin developers answered the question regarding plugin difficulty. This could be explained by several factors. Namely, if someone attempted to develop in the past, but gave up and stopped, or if people have seen the source and documentation and then give up. Without further research into how the question was asked, it appears somewhat dubious that more than double the number of respondents decided that the process was difficult, when only a small subset were reported as plugin developers.
Is TShock a good community?
Respondents were asked to rate TShock’s leadership and support experiences. Unfortunately, these questions were asked prior to the removal of a key member of the TShock development staff, meaning that the results may be skewed due to decisions made around this point. Take them with a grain of salt.
The average rating for TShock’s leadership was 4.18/5, with only 77% of respondents rating it 4 or 5. When asked to rate the support experience, respondents answered 4 or 5 only 75% of the time, indicating that the leadership is somewhat better than the support experience, but only by a slim margin. The overall support rating was 4.13/5.
Finally, TShock users were asked whether or not they felt comfortable asking for help in the community. 80% of respondents said they did, and 20% said they did not.
While some results have been skewed due to events transpiring in the middle of the survey, and others have dubious answer quality, it’s possible to create goals from these results.
TShock needs to improve its plugin API. TShock needs to improve the process plugin developers go through to develop plugins, as well as improve its documentation to make it easier and quicker to get questions answered.
TShock needs to improve its support experience. Some initiatives have already started, but this area will only improve with time.
TShock needs to improve its community, by taking steps to ensure a positive environment with openness. A 20% rate of questions going un-asked is unacceptable.