Bill Connelly's S&P Stats for Houston Week 3 Ranked 34th

Up from last weeks 39th

I don’t understand the math here. It looks like we are projected to win all our remaining games (with the USF game being essentially 50/50), but then there’s only a 13.6% chance we’ll finish 11-1 or better?

Because they’re projections. He’s not saying he’s absolutely correct. Although we’re projected to win there could be an upset here or there.

Interesting thing to note: Were not projected to play anyone ranked higher than us. This of course could change as the season progresses. Memphis and USF are the greatest chance of that happening.

I get that, but I don’t get how you calculate potential upsets into the equation. And hence the reason I went to law school instead of taking the advanced statistics route.

Maybe he’s calculating in his percent error? (Somehow…yea I wasn’t a stat major although I did enjoy the two classes I needed to take in order to receive my degree)

It’s like going to Vegas and doing a 10 team parlay. You can pick all favorites and still have a very small chance of winning. Chance is that one or more of your favorites in your parlay lose. All it takes is one loss to wreck it. Same thing here, but with two losses. We have a 50/50 better shot at each individual game, but that’s not the same as going 10-0 the rest of the way.

If you had a 75% chance to win every game you would have a 3% chance at going undefeated. That’s because you have a 75% chance for one game but a 75% of 75% to go 2-0 and a 75% chance of 75 percent of 75 percent to go 3-0.
Make sense?


Not really, but probably because I’m a lost cause on this stuff. I’ll take his (and your) word for it.