Episode 204: Angela Bassa

Angela Bassa is the director of data science at iRobot and an occasional poker player with more than a passing interest in Libratus and its implications for artificial intelligence. In this interview, we talk about poker’s intersections with data science and artificial intelligence, as well as what’s going through a data scientist’s mind when she sits down for beer and poker with friends.

Strategy this week comes to us from Nate’s recent play in a Foxwoods $1K!


Coaching Carlos

Tournament Poker Edge

NYC Live Event with Just Hands Poker


0:30 – Hello & Welcome
11:38 – Strategy From Nate at Foxwoods
41:41 – Interview: Angela Bassa


Hand 1

Blinds 150/300/25. Effective stacks 30K.

Hero limps QQ UTG, UTG1 raises to 1200, UTG2 calls BB raises it 3800, Hero calls, UTG1 folds, UTG2 calls.

Flop J97tt. BB bets 7K, Hero?

Hand 2

Blinds 600/1200/200. Effective stack 72K.

Villain opens HJ to 2800, Hero calls JTo on CO, Button calls, SB calls.

Flop (11K) 9h 5c 2h. Check, check, Hero bets 5300, folds back to HJ who calls.

Turn (22.5K) 8s. HJ checks, Hero?


8 thoughts on “Episode 204: Angela Bassa

  1. Hi guys,

    Was hoping Nate could write a little about what his plan for the river was if villain had called turn in hand 2.

    What he would consider good river barrelling cards if the OESD missed and if/how much he would have value bet had one of his overs paired. I’m interested in the factors contributing to these decisions.

    Keep up the good work,

    Long time listener Mark from Australia

    • In practice a lot of this would be read-based. I have a lot of value hands on the turn, enough to balance out plenty of bluffs on the river. That said, 16 combinations of JT / some combinations of QT and QJ / etc. will add up to plenty of bluffs, so I probably shouldn’t bet everything on the river. I really doubt he’d check-call twice with a set or check-call the flop with a gutshot, and (for those reasons among others) it’s not so hard to bet 98 for value even if a 6 comes, so I think the 6 is a good bluffing card. Board-pairing cards ought on average to be pretty good for me, too.

      That analysis aside, I was pretty comfortable in my physical read of the situation and likely would have made my decision (how big to value-bet if I hit my hand / whether and how to bluff if not) based not only on board texture but also on that read.

      Sorry not to have a better answer here!

  2. Thanks for taking the time to reply mr meyvis!

    It was the thought process I was interested in and you answered it perfectly!

  3. Fun interview. Reflexively object to the buzz around Watson “beating” humans at Jeopardy. A huge proportion of the skill edge in Jeopardy comes from buzzer timing, not just figuring out the punny questions to reach the correct answers. Am i supposed to be impressed that Watson, a computer, can send electrons through a wire faster than Ken Jennings and Brad Rutter can mash a button with their human hands?

  4. Another good interview! I’m always interested in what people elsewhere in the tech sector think about stuff.

    I’m personally not bullish on self-driving cars. While everyone probably means something a little different when they talk about “us having self-driving cars”, I’d specify it as this: I don’t think we’ll see a majority of new car sales in the United States, or cars on the road in the U.S., being self-driving cars within my lifetime (say, 40 years). I think the theoretical and engineering challenges are larger than most people believe, and there are also significant economic, social and moral challenges as well. I do think it’s entirely possible there may be some profitable high-end self-driving cars in production by then, but most people won’t be able to afford them.

    Of course, there are a lot of people a lot smarter than I am working on these problems, so what do I know?

Comments are closed.