When an NBA First Half Goes Way Under, Bet the Under Again

My favorite tools at The Action Network are probably the FantasyLabs Trends tool and our Bet Labs tool. Any DFS or betting situation or theory I'm curious about, I can easily look up whether my hypothesis is right.

Here's an example: I guessed that in NBA games in which the first-half over/under missed drastically — by 10 points or more — that it would be profitable to hit the second-half over and take advantage of regression.

Is that actually true? Or has the market properly priced that in?

It's actually the opposite that's true: The market overadjusts, and it's actually been profitable in this situation to take the under in the second half after it went way under in the first.

The numbers below are on games with a full-game over/under of 210 or higher.

But what about in games where the closing over/under is already high? Are those games more likely to regress and lead to over value in the second half?

Here's how the second-half under has performed in games with a full-game total of 219 or higher and the first-half total missed by 10-plus points:

Oh, baby. Unders in this situation have returned a robust 10.2% ROI since 2005.

It's been especially good the last two seasons:

• 2017-18: 36-19 (65.5%), 2.23-point margin, 24.8% ROI
• 2018-19: 41-8-3 (59.4%), 2.68-point margin, 12.7% ROI

The most recent match was Friday in DeMarcus Cousins' debut with the Golden State Warriors against the Clippers. The first-half total went under by a whopping 22.5 points. You'd think that would regress, but the second-half under also hit. The market is seemingly overadjusting to regression.

This is a good system to build and track throughout the night, especially if you're following along in our app and see a game go way under the first-half total.

And, of course, go test your own hypotheses. You might just stumble upon some great betting or DFS value.