[Pokerjoe]

But math is everything. Life on Earth is math. Living plants are math. The entire universe is math.

No. Math is a human construct which we apply to the world around us, a construct we use to describe the world around us, but it is not in the world around us. It has a great functional use, but it is an application, a toy, a tool, a fiction, a philosophy, not a reality.

Math geeks often like to think math is everything, because they're good at math and who wouldn't want "what they're good at" to be everything? Artists are the same way, they like to say, "Everything is art!"

But just like artists are typically bad at math (not all; Da Vinci rules), math types are typically bad at art. Notice how they often embarrass themselves in these forums when it comes time to read and write?

Probably if 8th grade algebra students had the arrogance for it, they'd describe calculus as convoluted; math types often describe arguments that way, and for the same reason.

Mistaking calculation for intelligence is a hallmark of the typically not-too-bright math geek. Don't be one. The complete failure of math-types in, for example, economics (which used to be dominated by philosophers) is a shining example of the limits of math, and the general un-intelligence of math types.

Quants don't make money in markets. Quants make money by convincing non-quants that quants can make money in markets. And it's actually the hedge fund marketers that do the real cleaning up. Chartists are full of crap, with no proven ability to win going forward. Yet chartists still abound (and the connection to OP is that his idea is basically trend charting).

As to your point, you're right that this isn't roulette, because it's a market, and inefficiencies are possible. Further, markets have memory. They're self-aware. A roulette wheel, coming black 10x straight, can't think, "Hmm, time for a red!" and casinos aren't going to change the price on black, either.

But markets, observing that MLB faves have been covering more than their due, can react, "Hmmm, time to bet faves!" driving the fave's price up, taking away the profits. And if the market isn't doing that? There's a reason.

Here's the inherent problem (and why z-scores are useless here): the larger the sample, the greater the validity, but the greater the validity, the greater the market's awareness of it, and thus the greater the likelihood the market will act such that the sample isn't valid going forward.

IOW, going forward, no one (and no formula) can say that the past aberration in MLB fave win rates will continue. You are, of course, free to bet it does. But I'd suggest you first understand that the past results can equally well be interpreted to mean, "Wow, baseball dogs are so due!"