Critism of "Sukupuolirepresentaatiot Magic: The Gathering".

I was for fun searching for Magic: The Gathering AI bot papers on Google Scholar when this gem popped up. Sukupuolirepresentaatiot Magic: The Gathering translates as Gender representations in Magic: The Gathering. It was rather surprising since I wasn't searching for SJW papers and it was in Finnish. And It was some bachelor's thesis from 2019. I had to know how this ended up front of me and was this any good.

The first strike for me was in the abstract of course. The researcher only looked at the War of the Spark extension and in that extension, only Planeswarkers were looked at making total to 39 cards. War of the Spark extension has 265 cards. That is roughly 15 percent of the cards of War of the Sparks were look at. This calculation, of course, includes cards like Bleeding Edge which is rather neutral for gender representations so 15 percent is the theoretical minimum. The thing is that the thesis makes conclusions for the whole Magic: The Gathering. Both 2019 and 2020 starter sets were 300 cards each. Even if 39 cards were a representation of War of The Spark it would still be a long way from a representation of whole Magic: The Gathering. No, I'm not writing this because the conclusion was there is sexism. After all, the thesis concludes there isn't any. I'm saying this because logic doesn't really flow for me were the conclusion sexism found or not. But hey I haven't read the methodology yet. Maybe it could turn me around. Thesis ask four questions:

1. Can planeswalkers be divaded by gender?
2. Is there a quantifiable difference between genders?
3. Are weapons gender divided?
4. Do cards function differently based upon gender?

The first question is too vague and includes every other question. Second and third I like since they are precise and they actually could be useful measures when thinking about what kind of cards Magic doesn't have. Sexism logic is still a bit of a mystery. The fourth one I have questions about categorization. As a trend, every card should have different effects. So is the question for every woman there is a man with same but more powerful effect? I don't really know? Well, I looked ahead to results and it is kinda what I said. No more specific method is given but an example of women using healing magic is given. Since I'm now at the results of the fourth question author notes that card colour is more influential as that relates to game mechanics.

Gender in the thesis is the two-category system (man and woman). Identification is done by characters clothing, body shape and facial hair. On unclear scenario MTG-wiki, pronoun usage, and game makers blogs and interviews. This to me is an alright method defining gender. The author doesn't and is apologizing for the fact since the two-category system can be prolematized. She says later that she did this because it simpler analysis considering the scope of the thesis. Scope in a sense is a problem because bachelor's thesis but I still going to nip this. A simple two-category model is fine as most of the time we can't determent anything about the behaviour of the character in the cards. We then only can determent gender by body shape really or outside of the card image sources like the lore. But beyond this, we are using planeswalkers simple because other cards maybe are classes of entities so Gender becomes even more ambiguous. At this point I don't know has author even read her own sources. Do they really say that if a card says goblin but the picture is clearly a male we don't have to think that woman can't be goblins? I can give her that planeswalkers do have more lore behind them so they can be more important but does group name really matter more in this analysis considering the picture?

The reason why I hit this was answered when the thesis talks about generally about Magic: the Gathering. Apparently, someone has proven that Magic may be the most complicated game in the world because of game's complex countability in a Turing machine. I don't want to be rude as this clearly isn't the author area of expertise. She shouldn't have made this citation, however. No. Source cited is paper proofing that two-player game of magic is Turing complete. This has some implications that are somewhat beyond me as well. According to the paper's conclusion, this implies that ... identifying the outcome of a game of Magic in which all moves are forced for the rest of the game is undecidable. which would indicate the most complex game (or least one of). I just don't think Turing completeness can be said to be: the game is complex in a Turing machine. It indicates that the game of magic is an automata that can emulate a Turing machine. Turing completeness implies between automata that what is hard on Turing machine is hard on compared automata (if Church–Turing thesis). Why did you cite this anyway? You don't use it for anything.

I don't really have anything to say about the conclusion. It is there is no statistical power to say there is any meaning full difference. Proofs of the implications are left to citations. In conclusion, I probably should challenge my self. How I would do this study properly under the assumption that citations are correct. Meaning that this kind of analysis could have indicated sexism. The biggest problem of the thesis is that there isn't anything to compare result gotten. Analysis should have made over time. So study should look at most if not all cards of Magic. The statistic should be collected for every set and sets should be ordered by time. Also, the statistic shouldn't be a card but the characters in the card. This way if there is a change it is detectable. For categories, I would use male, female, non-humanoid, unknown/undefined/not enough information because without outside information known more than male/female is hard. The Non-humanoids category is there because sexual dimorphism for fantasy animal species can be hard. If the amount of work is too much use a computer to make the classification or just look at the subset of the cards and make your claims about that subset rather than faulty generalization! This would still be rather hard for a bachelor project.