Why undoverfit

Every model that learns from data must choose how seriously to take what it sees.

Overfit too hard and you mistake every small wrinkle of the past for a law of nature. Underfit and you flatten everything into a line that explains nothing and predicts even less. Between those two failures sits the small, honest place I want this blog to live in — close enough to the data to learn something real, far enough to leave room for what comes next.

I called it Undoverfit because I keep meeting that word in my own head. In statistics, in finance, in the way I read news, in the way I evaluate people and ideas, in the way I think about Noroom — the project I am building to give people their data back. The same trap, in different costumes.

So this is the deal. I will write about technology, data, statistics, finance, entrepreneurship, side projects, and whatever quiet thing catches me on a given week. No daily dispatch. No optimization for the algorithm. Just notes, written as honestly as I can, in a place I own.

If you find one useful, that already makes it worth it.