Hey. You’ll probably never see this since your article’s now blown up, but I think that I have a unique perspective given my background and I have a different critique than what others are offering.

I appreciate the effort you put in at this first stab. As an engineer who’s also a history Ph.D. dropout, I can appreciate when technical folks engage with the humanities. That said, there are a few issues that I haven’t seen addressed so far in the comments:

1. Lots of folks are talking about the quality of your data set, but no one is talking about the underlying historical sources that feed any data set on battles. Even if Wikipedia were a perfect representation of all the historical data that we have available to us, the fundamental problem is that so much historical data is forever lost to us. Entire swaths of history are basically total patchwork/guesswork/inference, battle stats included. Especially when you get into ancient stuff. This isn’t an edge case, this is a core, defining attribute of history that historians have been trained to deal with for hundreds of years. History isn’t exclusively written by the victors, but the dried pressed wood pulp and calf skins that it’s written on is almost exclusively and selectively preserved by the victors. A statistical model can’t handle this.
2. Even among the data available to us, huge amounts are contradictory or cut from whole cloth. One side will say that there were 10k troops and the other side will say there were 100k troops. Etc. This problem occurs up to our present day, even with all our modern technology, because often the root cause is political. Each side has its own interest in skewing numbers. Just look at https://www.iraqbodycount.org as a testament to the ongoing pertinence of this problem. This is again a fundamental problem not solvable by a purely statistical model.
3. What counts as a “win” is highly subjective and up to whatever the historian decides. Was the Tet Offensive a “win”? For whom? To make matters worse, Wikipedia just picks one opinion — either a win or a loss.

One final point that’s more about your tone.

I know you mean well with this project, but you should be aware of the political/social context of your work. By your own admission, you’re an amateur, and that’s no crime. However, there are so many examples of well-meaning technical people blundering into humanities fields that scholars have been working at for centuries. Generally speaking, these technical people will make huge, epic pronouncements about their work (e.g., “Napoleon was the Best General Ever, and the Math Proves It”) while scholars, who are much more aware of the nuances of the work, will avoid such pronouncements. It would be like if, upon trying your hand at a biology experiment for the first time, you published an article entitled “Bread Cures Cancer, and the Data Proves It.” Or if you did a little bit of math and came up with a non-general solution to the three body problem and wrote “This Equation is a General Solution to the Three Body Problem, and the Math Proves It.” Except that it’s the humanities, so nobody cares and it’s anyone’s game, right? Because it’s “just” history, right?

Generally speaking, technical people who wander into humanistic fields as dilettantes make far, far more money (e.g., data scientist in metro area), are more adept at navigating media outlets (e.g., Medium) than scholars and their voices tend to have a lot more weight (e.g., 8.8k claps and counting). This all comes at a time when graduate students are making $17k/year working full-time in the humanities, graduating to become part-time adjuncts at community colleges with no health care, barely scraping by after spending 9 years on their Ph.D.’s. They are struggling just to put food on the table, let alone becoming Medium famous, while still maintaining the same level of rigor and integrity in their work that you’d expect from any of their colleagues in the sciences.

My point is, and I say this with love as a fellow engineer, don’t be a bull in a china shop. Please consider treading more lightly in the future, doing a bit of literature review and connecting with scholars who have been studying these issues for hundreds of years. Many humanists are deeply resentful of the general ongoing project to dismantle their careers and only keep STEM in the universities. The tone of your work, and the tone of others like it, doesn’t help.

But also please keep going. We need more technical folks who are willing to engage with the humanities in creative ways. You may want to google “digital humanities.” There’s a whole field people doing stuff like this.