Names of places in Croatia

[FlatAssembler]

My friend, can you explain to me, how could I be wrong? I have been thinking about that for more than a year, and I can think of two ways I can be wrong, none of which seem likely.
1. Using collision entropy, although it seems like a very natural choice, is inappropriate here. Markovnikov Chains, or some more appropriate model of the language, would perhaps provide a very different result from using collision entropy with birthday calculations. Which seems unlikely as two PhD informaticians (Anđelko Lišnjić and Franjo Jović) have reviewed my paper and have not noticed that.
2. Some phonosemantic hypothesis is true. In that case, the entire historical linguistics is, well, not based on good principles. If some phonosemantic hypothesis is true, looking for regular sound correspondences to establish that two languages share a common ancestor is also unreliable, and not just the method of analyzing toponyms I came up with.

In either case, how is the mainstream methodology less wrong?

B mine...and, exactly. There's nothing to explain to you which hasn't already been explained, from specifics to principals to generalities, and yet you remain wholly and completely oblivious.

I don't understand, if my paper about applying informatics to the names of places is so obviously wrong that even you can see that, why did it get published in two peer-reviewed journals? Why did a PhD linguist ask me to send her a copy? Why didn't two PhD informaticians whom I sent a manuscript of my paper find an error in my reasoning? Why did quite a few people who know something about informatics, including a PhD statistician teaching computer science, also say my arguments seem good to them? Why did my communicology professor tell me my arguments look good to him? It's clearly not Flat-Earth level wrongness, that anybody who is remotely related to the field can explain you why it's wrong.