Are you referring to the emotional language or the lack of supporting argumentation? These are different points. Let us take an example that usually does not cause any controversy: “the QoL is good in GD.” How do we know that this is objective, however? You wrote “As soon as you describe something that cannot be measured, you’re subjective.” but there is no metric for QoL that can be measured. One could give many examples of specific QoL features. How many examples are required to make it convincing, though? One could compare those features with the features in other games. How many other games would be sufficient? One could refer to many posts by the players praising QoL improvements or to posts by the developers when the features were introduced. How many posts would be sufficient and is it important to include all of the links explicitly? The readability is also important, after all.
If I do not provide any supporting argumentation, then it is clearly subjective. This is why I am especially curious to learn which of the statements are unsupported in your opinion. The only aspect where it is difficult for me to find strong objective arguments is voiceovers. Sure, there were several posts in GD forums celebrating the replacement of Ulgrim’s voice in the expansion and lamenting that the original actor who voiced Kasparov passed away. (These posts were probably not migrated to the new site.) Sure, there are four comments on https://www.reddit.com/r/Grimdawn/comments/7h7uge/unplayable/ stating that the voice acting is mediocre or worse and no comment contradicting that statement. Sure, I can find many posts and youtube videos about excellent voiceovers in other games (hundreds of posts about Diablo 1/2 but also many posts about indie games such as PoE or Enderal). Here is one example: https://www.reddit.com/r/pathofexile/comments/soigz8/favorite_voicelines_in_poe/; check how many comments agree with that statement. I am not aware of a single post saying that the quality of voiceovers is good across the board in GD. Google search shows a single youtube video where the voiceovers in GD would be in the title and that video has zero comments. Would it be better if I mentioned all of the above in my comparison as supporting argumentation?
I provided additional supporting argumentation in response to Medea’s statement that the pace of combat is not bad in TQ. Would it be useful if I included this argumentation in the review itself?
Or is the issue that I should better separate between the objective and subjective? In about four places, I write “my personal opinion is …” stating that I am not trying to make it objective, in contrast to the rest of the review. However, these statements appear in the same paragraphs as the rest. Perhaps I need to separate them in a better way?
The question of emotional language is different. Suppose we agree that I am allowed to say “the QoL is good in GD”. Am I allowed to say “excellent” or “superb”? There is no metric or commonly accepted scale for QoL, after all. Or perhaps I can only state that “GD has that particular QoL feature while TQ does not have it”, without any attempt at a generalization? I daresay it would make for an informative but boring review.
At the end of the day, it is impossible to make the review 100 percent objective, short of making it empty. Look at the amount of emotional language scientists are using in popular articles and talks. The best I can do is to address concrete criticism thanks to your specific feedback, should you choose to provide it.
What I meant by the average is the average number of runs between two drops of the same unique or two items with the same affix. Alternatively, the average number of drops of the same unique over X consecutive runs. My claim is that the deviations are not as large in GD compared to TQ. How do I substantiate this?
Ideally, one needs to run a large number Y of trials, each trial consisting of a large number X of consecutive runs. (Alternatively, one could use the sliding window technique but these are technical details.) The scale must be large indeed but one does not actually need to know the ground truth or look up anything in GrimTools in order to assess the average and deviations. Furthermore, it is much easier to notice that something is wrong than to prove that the drop generator is perfect. In TQ, there were many posts by players farming a particular item, say stonebinder cuffs, observing complete lack of drops over many runs and multiple drops over a small number of runs. Eventually, someone meticulously recorded drops over a reasonably large number of runs (an order of 1000, IIRC) for each of the popular farming spots and presented the findings. The findings supported the observation in a consistent way: if 0 is the most popular value of drops for a run, then consistently having more runs with two drops than runs with one drop would be highly unlikely for a good drop generator. Most likely, it means that the generation is too affected by the seed. Someone has even posted a fix for the generator, which was later included in TQDefiler. There were a lot of discussions about the fix, with the consensus that it made it better but did not fix the problem completely. Later, a fix (or perhaps the same fix) was included in AE. The players continue to complain but to the best of my knowledge, nobody has created a large data sample for AE.
In early GD days, there was a post by Medierra about the improvements to the drop generator but as far as I remember, it did not include technical details. I am not aware of the same volume of complaints and published test results in GD as we had in TQ. Typically, this is an indicator that the players do not perceive it as a big problem. Admittedly, it is not the same as to mathematically prove that the drop generator is better in GD.
