The mathematical value of a poker tell


By Sean Hansen

A common discussion among many players, especially those who come from the online world, is whether there is any value in live tells. The short answer is yes. The long answer is, yeeeeeeessss! The usual argument is: Poker is all math and tells don’t matter vs. tells matter because I say so.

Yes, poker is all math, 100 percent true. However, tells affect the math. If we’re analyzing our hand strength vs. a range mathematically, and a tell leads us to believe one portion of a range is more likely than another, then we have to weigh their range in that direction to come up with the correct equity calculation. As equities change, so should our betting decisions, ultimately altering our EV.

A simple example: You’re on the river in a $2-$5 game, facing an overbet jam from a player you would never call, given the board texture, and your top-pair hand strength. The pot is $250; he jams $500.

This player is known for overbetting the nuts, so you’re highly disinclined to call. However, you notice after his bet, he scoots back a bit from the table, and “turtles,” putting his head down, kind of shrinking into his chair and looks to be frozen in place, something he’s never done.

Well, I’d say this is a tell of absolute weakness nearly 100 percent of the time (a bluff of say, a missed straight draw), but given this player’s tendency toward overbetting the nuts, even with the tell I have to discount that by, say, 20 percent, meaning I will assume 80 percent of the time from this player, this is a bluff.

Let’s say my tells reading accuracy in general is about 75 percent. That means 75 percent of the time my reading of a tell turns out to be correct. If so, then 75 percent of the time, my call will win 80 percent of the time. So, I have a 60 percent likelihood of winning the pot by calling: (.75*.8) = .6.

Without the tell, our EV would be zero; we’d fold, a zero-EV decision. If we decide to call based on the tell and win 60 percent of the time, the math becomes this: .6(750) + .4(-500); 60 percent of the time we win $750 and 40 percent of the time we lose $500. Solved, that’s plus-$250 in EV, compared to the fold EV of zero.

But wait, how do we know our tells-reading accuracy? How do we know how accurate a given tell is? By practicing and by recording our results. I make notes of when I make a decision based on a tell and I track my results.

I have been doing this for years, I can definitively say what my tells-reading accuracy is and I have learned how reliable certain tells are. There’s no reason you can’t do the same.

— Sean Hansen is founder of Big Slick Poker Academy.