Colt Ramsey

Background

Colt holds an advanced degree in Mathematical Statistics from a prestigious Texas university. His career has centered on building models from messy, real-world data and validating them against observable outcomes. He is the kind of analyst who sees a public dataset and immediately wants to know what everyone else got wrong about it.

He built Lottery Truths from the ground up: the data pipeline that scrapes and validates 6,000+ historical draws, the Poisson splitting model that adjusts for jackpot-sharing risk, the polynomial regression that predicts ticket sales, and the automated system that keeps everything updated. Every step of the process is under his direct control because doing it right required exactly that.

Why This Site Exists

I built Lottery Truths because I was tired of bad analysis being treated as authoritative.

The sports gambling era has turned every loud voice with a platform into a self-proclaimed expert. Confidence and volume is rewarded and statistical rigor is optional. People make real financial decisions based on analysis that would fail an undergraduate statistics course.

The lottery space is somehow worse. Most "Powerball vs. Mega Millions" comparisons are just reformatted odds tables. They ignore that ticket prices are different, that the answer changes every draw, and that a billion-dollar jackpot is worth far less per ticket than the headline number suggests because of splitting risk. They present a static answer to a question that is inherently dynamic. It is lazy, and people deserve better.

So I built something better. A model that accounts for every rule change since 1997, uses real sales data to adjust for jackpot splitting, normalizes everything per dollar, and updates automatically. It shows its work. If I got something wrong, you can find the error, since the methodology is fully documented.

I have no financial incentive to make either game look better than it is. There are no affiliate links on this site, no "play now" buttons, and no sponsors. The analysis says what the math says.

Approach

Every claim on Lottery Truths is backed by a calculation you can trace. The methodology page documents the full model: the Poisson splitting formula, the polynomial sales prediction, the era-by-era rule handling, and the validation pipeline that checks every data point before it enters the system.

I tested models that did not make the cut and documented why. Google Trends as a ticket-sales predictor. Random Forest and Gradient Boosting ensembles that overfit beautifully in-sample but collapsed on out-of-sample data. Weighted regressions that tried to pool data across pricing eras. Each one failed the same test: does this actually predict better than the simpler model? If not, it does not ship.

Transparency is the entire point. If the model is wrong, I want to be the first person to know.

Personal

Based in Texas. Private by nature. He is a frequent visitor to Las Vegas, drawn equally by the math of the casino floor and the quiet of the surrounding Mojave Desert.

While he has never cheated at a casino game, his strict adherence to advantage play and statistical discipline has earned him bans from numerous tables and a handful of casino properties. To him, a casino ban isn't a mark of dishonor—it is just confirmation that the math works.

He would rather ride through the desert than post on social media. No Twitter, no Instagram, no TikTok. The work speaks for itself.

Questions about the methodology or data? Reach out at colt@lotterytruths.com.