"We need a fade tracker" — okay, but does fading actually pay?

The vig sets the bar

Standard −110 odds mean you risk $110 to win $100. Break-even win rate:

So fading someone only profits if they lose more than 52.38% of the time — not just "more than half."

Losing ≠ fade-worthy

A picker who wins 47.6%–52.4% loses money themselves AND loses you money when faded. Inside that band, the sportsbook beats both of you. That's most public pickers.

What a real edge looks like

To earn 5% ROI fading someone, they must genuinely win only ~45% against the spread over a large sample. Sustained 45% is as rare as sustained 55% — being reliably wrong requires the same skill as being reliably right.

Sample size fools everyone

A 12–20 cold streak looks fade-worthy, but a true 50% picker hits stretches like that routinely. Run the simulator a few times at 50% and watch "obvious" hot and cold seasons appear from pure noise.

1 · Demand the sample

Fewer than ~300 tracked picks tells you almost nothing at these margins. The standard error on 100 picks is about ±5 points — wider than the entire edge you're hunting.

2 · Selection bias is the product

You heard about this picker BECAUSE of a memorable cold streak. Trackers built on famous streaks start at the peak of the noise.

3 · The book always takes 4.5%

Every fade pays the vig. Whatever the picker does, roughly 4.5 cents per dollar wagered flows to the house. That's the only guaranteed edge on the field.