In Matt Janda’s book “Applications” he calculates pot odds differently than I’ve seen before, say in Harrington. For example, pot is $50, opponent bets $30. $80-$30 odds. I get 37% of the time we need to win to be profitable: 30/80, but they do the math 30/80+30 = 27.3%. I don’t get it. Why?

When you win, you'll win 80
When you lose, you'll lose 30
Let's call your win percentage W. So you lose 1-W.

EV = W*80 - (1-W)*30
agreed? You'll win 80 with probability W, you'll lose 30 with probability 1-W

To find the breakeven point, find W=0
0 = W*80 - (1-W)*30
80W - 30 + 30W = 0
110W - 30 = 0
110W = 30
W = 30/110

It might also help to think about the fact that you 80:30 odds is in "odds format" which has a specific conversion to a percentage. With X:Y odds, it's the same as Y/(X+Y). Why? Think about a drink recipe, like 2 parts OJ, 1 part vodka. The ratio of OJ:vodka is 2:1

There are 3 "parts" total, right? So 2 of the 3 parts are OJ, 1 is vodka, so 2/3 OJ

I’m Russian so your vodka analogy helped. I.e. 33% vodka 1/3. 30/80+30= 27.3% of the time I’d have to win to make the call. So if the pot is $50 on the flop, donkey bets $30, and I have an OESD, it’s a call because I make my hand 33%ish times and my pot odds are 27.3%?

No, not exactly, because the money you are calling only gets you to the next street, and your odds to make the straight on the *next card* are worse than 27%. This is a classic problem in poker where you are trying to compare immediate odds to showdown winning chances.

There are 2 ways to approach the problem. One is to compare immediate pot odds to your chance of hitting your hand in the next card. The other is to compare showdown winning chances to your *effective* pot odds. If pot odds is "how much do I risk, vs how big is the pot" then effective pot odds is "what is my total future risk, vs the total future size of the pot." The effective pot odds method is better, but harder to estimate. The immediate chance of winning odds are easy but they will lead to some folds because they leave out important factors such as "if I hit, I can extract extra bets" and "if I call the flop, he may check the turn or bet small" and so forth.

Also, I encourage people to think in odds format instead of percentages. It's easy to figure out your pot odds because they're in front of your face... pot is X, call is Y, pot odds are X:Y. To make this work, just learn the common equities in odds format instead of percentages.

Like if you are going to win 25% of the time, then it is 3:1 against you winning. If the pot odds are better than 3:1, you call. This way you do all your calculations/memorization away from the table.

Another approach is to look at implied odds. I used the following:

Pot = 50: V Bet = 30. Pr(Hit straight on turn) = 17%. Pr(Win|Hit) = 85%. Pr(Villain calls turn bet |Hit) = 70%

Applying an advanced implied odds model using these values, you will find that if you bet at least 160 on the turn if you hit that will be a +EV action assuming the effective stack is that big. The required implied odds are about 8 to 1. If you don't hit on the turn, you fold and lose your flop call amount.

Note that by having a win given hit probability of less than 100%, the model includes reverse implied odds. The result is exact if the future bet is all-in or if the action starts on the turn. It is approximate otherwise but still better than just using immediate odds.