Reverse Implied Odds and Trouble Hands
Implied odds have an evil twin: the money you lose on later streets when you make your hand and it's second best. This lesson identifies the classic trouble hands and gives you the one question that sorts them.
Assumptions: All examples use a 6-max online cash game at 100bb effective stacks ($0.50/$1 where dollar amounts appear) with no rake unless a different setup is stated.
Everything in this module so far has been about money flowing toward you on later streets. Now flip the sign. Reverse implied odds are the future money you lose when you hit your hand — top pair, a flush, a straight — and someone else has hit a slightly better version of the same thing. The pot odds looked fine, the hand "connected," and you still shipped sixty big blinds the wrong way.
This isn't bad luck. It's a property of certain starting hands, as predictable as the set-mining math, and it has the same asymmetric structure as implied odds — just pointed at your stack instead of your opponent's. Implied-odds hands win big when they hit and lose small when they miss. Reverse-implied-odds hands win small when they hit and lose big when they hit slightly worse than someone else. The miss isn't the danger. The almost-perfect hit is.
The mechanics of getting out-kicked for a stack
Why do these hands lose big specifically? Because the strength of your hand controls how much you pay, and second-best hands feel strong. When K♣J♦ flops top pair, nothing on the board warns you. You have exactly the hand you were hoping for, so you call the flop, call the turn, and talk yourself into the river call — while an opponent holding A♠K♠ is enjoying implied odds you are funding. Domination converts your hit into his payday: same pair, permanently better kicker, three streets of value collected from a player who can't see the problem.
Run the numbers on the classic disaster.
The hijack opens K♣J♦, calls a cutoff 3-bet to 9bb, and flops top pair on K♥8♦3♣. Against the part of the 3-bettor's range that continues big on this board — AK and KQ — a simulation puts K♣J♦ at about 13% equity. Thirteen. With top pair on a dry board. There are 16 live combos of AK and KQ in that range once your cards and the board are removed, and against every one of them you are drawing to three jacks.
Now count the cash flows, because reverse implied odds are a counting exercise exactly like the deficit was:
- When you're behind (AK/KQ): preflop call 9bb + flop 6.5bb + turn 14bb + river 30bb = 59.5bb lost, and no street offered a natural exit — the bets were all "reasonable" sizes into a hand that felt strong.
- When you're ahead (his AQ-type hands that missed): he continuation-bets 6.5bb once, gives up, and you win roughly 16bb — the preflop pot plus one bet. Ace-high doesn't pay top pair three streets.
Lose ~60 when you're wrong, win ~16 when you're right. For the call-down to break even you'd have to be ahead almost four times as often as behind — against a 3-betting range built around AK and KQ. The domination started preflop: K♣J♦ against A♣K♦ alone is about 25%, and unlike a flush draw's 35%, that 25% comes with no disguise and no payday when it wins. That's the signature of reverse implied odds: bad equity and bad money flow.
The trouble-hand census
Three families produce most reverse-implied-odds disasters at 100bb. Notice they all look like value:
Dominated broadways: KJo, KTo, QJo, QTo, AJo from early seats. Their problem is structural: the hands that play big pots against them — opening ranges, 3-betting ranges — are stuffed with AK, AQ, KQ, and AJ. When KJo makes top pair, the better top pairs are disproportionately present. This is why the site's charts open KJo from the hijack but fold it to early-position aggression and why "it's a pretty hand" has cost more stacks than any draw ever did.
Weak suited aces: A4s, A6s and friends — for their pair, not their flush. The flush these hands make is the nut flush, and that half of the hand carries genuine implied odds. The trap is the other half: flopping top pair of aces with a four kicker against ranges full of AT+, then treating it like the strong hand it visually resembles. Weak suited aces are fine speculative hands when played for the flush and the wheel and money furnaces when their owner falls in love with a dominated ace.
Small suited cards and small flush draws: 84s, 63s, J5s-type hands. These make flushes that aren't the nuts and never will be — and a non-nut flush is the most expensive second-best hand in poker, because flushes don't fold.
That third family deserves its own crime-scene reconstruction.
Anatomy of a baby flush
You defend 8♥4♥ in the big blind and flop a flush on Q♥9♥2♥. Feels like christmas. Now do the arithmetic the hand block summarizes. The hearts you don't hold — A, K, J, T, 7, 6, 5, 3 — combine into 28 possible flush combos for your opponent. Any combo containing the A♥, K♥, J♥ or T♥ makes a better flush than your Q-9-8-4-2: that's 22 of 28 combos, or 79%. When the money goes in flush-over-flush on this board, you are crushed four times out of five — and flush-over-flush is precisely the matchup that stacks people, because neither player folds.
What about when he doesn't have a flush? The best realistic aggressive hand, top set of queens, still has about 34% equity against you — ten outs twice to a full house — so even your winning scenarios are sweaty. Map the button's 3-betting-the-flop range honestly: better flushes (many), sets (drawing live against you), and the occasional A♥-high blocker bluff. Your 8-high flush beats the bluffs and the sets-for-now. It loses a stack to the category the action specifically announces. Calling the flop raise and folding to continued aggression — or just calling down small — loses less than most players manage here, but the real lesson sits one street earlier: 8♥4♥'s "best case" flop was never as good as it looked, and that fact belongs in the preflop decision. (At a discounted big-blind price, defending 8♥4♥-type hands can still be fine — the site's BB-vs-BTN chart includes 85s but draws the line well above trash like 84s. The point isn't never to play small suited hands; it's to know what their flush is worth before the pot gets big.)
Reverse implied odds scale with the action
The same hand's reverse-implied danger isn't fixed — it grows with the strength of the range attacking you and the size of the pot you're building. KJo limped around to you in the big blind is a fine hand: the limpers' ranges are weak, the pot is small, and your top pair genuinely is the best hand most of the time it appears. KJo facing a 3-bet is the disaster you just watched, because 3-betting ranges concentrate exactly the AK/KQ/AQ hands that dominate it, and the pot is three times deeper before the flop arrives.
This gives you a practical dial rather than a blacklist. Ask: how strong is the range I'd be hitting against, and how big will the pot be when we collide? Trouble hands are playable against weak ranges in small pots and toxic against strong ranges in big pots. That's why the same KJo appears in the site's hijack opening chart (attacking mostly-folding ranges in a small pot) and vanishes from every chart that responds to serious aggression. Position helps too, for the same reason it helped implied odds in reverse: in position you can keep the pot small with a dominated hand and escape cheaply when the action tells you you're second best; out of position the better kicker decides the pot size for you.
Pricing the reverse component: discount your outs
Reverse implied odds also live inside draws, and you already own the tool for handling them: the deficit calculation from lesson one, with one adjustment — discount outs that can make you second best.
Take a nine-out flush draw facing a 5bb bet into a 10bb pot. With clean outs the math is friendly: 9/47 ≈ 19% per card, deficit = 5 ÷ 0.1915 − 20 ≈ 6bb of future money. But if your draw is to an 8-high flush against a range that holds plenty of bigger hearts, two of those nine "outs" are really invitations to lose a stack. Re-run the numbers at seven clean outs: 7/47 ≈ 15%, deficit = 5 ÷ 0.1489 − 20 ≈ 14bb — more than double. And the honest accounting is worse than the deficit alone shows, because the discarded outs aren't neutral: the times the flush comes in second best, you don't merely fail to collect, you pay. A baby flush draw should be played for its price, not its implied odds; the implied half belongs to whoever holds the ace of the suit.
The general rule: count an out as clean only if, when it arrives, you'd be happy to play a big pot. Outs to dominated pairs (the jack in KJo's three-jack "redraw"), outs to baby flushes against flush-heavy ranges, and outs to the dumb end of a straight all get discounted or thrown out — and the deficit they leave behind tells you what the call is really worth.
The diagnostic question
Every hand in this module — printer or furnace — can be sorted with one two-part question asked before you call:
"When I hit, who pays me? And when I hit, who beats me?"
Run it against the hands you now know:
- 2♠2♣: When I hit (a set), top pairs and overpairs pay me, and almost nothing beats me. Both halves point the right way: positive implied odds.
- 8♠7♠: When I hit (straights, mid flushes), overpairs pay me; bigger flushes occasionally beat me, but my straights are usually the nuts on their boards. Net strongly positive.
- K♣J♦ vs a 3-bet: When I hit (top pair), worse hands pay one small bet; AK and KQ beat me and charge three streets. Negative — fold preflop and feel nothing.
- A♥4♥: Split verdict — flush half positive (nut flush, gets paid by smaller flushes and sets), pair half negative (dominated ace). Play it for the half that's positive.
- 8♥4♥: When I hit my best hand, only bluffs and worse-flush rarities pay; 79% of opposing flushes beat me. Negative, and no discount price fully repairs it in raised pots.
The question works because it forces you to name actual hands on both sides of the ledger instead of grading your cards in a vacuum. "Who pays me" measures the implied half; "who beats me" measures the reverse half; the difference is what your speculative call is really worth. A hand can flop monsters frequently and still be a long-term loser if its monsters are the second-best kind — and a hand can hit rarely yet print because every hit is the nuts and every customer is locked in.
Make the question a reflex and the trouble hands lose their disguise. KJo stops being "two paint cards" and becomes "a hand that wins 16 and loses 60 against the range that's actually playing back at me." The baby flush draw stops being "suited!" and becomes "a 79% chance of second place in the biggest pots I'll play with it." Implied odds and reverse implied odds are the same calculation with the sign flipped — and from now on, you compute both before the chips go in.