clock menu more-arrow no yes mobile

Filed under:

Prediction Revision

I noted on Friday that I would be revising my methodology for the prediction parts of each game preview. With Buckeye fans feeling that I'm getting a bit too smarmy with my Longhorns, this seems like a good time to make the change.

The problem with the current model is that I'm forecasting an outcome, and a precise one, at that, by taking a shot at the score. It's a fool's errand, on several levels.

For one thing, I'm not going to be predicting any losses at this point. Why would I? We haven't lost in ages and we haven't seen anyone play yet. I believe we're as likely - probably more likely - to lose a game, or even two, as we are to run the table. But with the possibility of going undefeated well within the bounds of reason, I'm not likely to predict a specific loss.

Second, we don't have nearly enough data yet to being doing anything as precise as picking scores. It's just unrealistic, and uninformative.

Retired. For now.

With that in mind, we'll revise the predictions from here on out to outline the different ways the game could play out. "Texas will win if: blah blah.  Texas will lose if: blah blah." By covering the range of likely scenarios, good and bad, we can offer something of substantive value.