In chapter 21 of The Theory of Poker, David Sklansky focuses on how to think and act heads-up on the final betting round when all cards are out in a one-winner limit game. Once no more cards remain to be dealt, your decisions become purely about betting, calling, raising, or folding for that last bet, and the logic shifts compared with earlier streets.
Core Idea: Endgame Decisions Are About the Last Bet, Not Just “Who’s Ahead”
Earlier in the hand, you often bet because:
-
You want to deny free cards, or
-
You expect to win the entire pot often enough.
On the end, there are no more free cards and only two players left. So:
-
The key question is not only “Do I probably have the best hand?”
-
The key question is:
“If I bet (or raise) and get called, how often do I win that extra bet?”
You can be a big overall favorite to have the best hand, yet betting on the end might still be bad if your opponent only calls when you’re beat.
Bluffing Heads-Up on the End
Bluff Bets
If you have no legitimate hand, your only way to win the pot is to make your opponent fold:
-
Example: Pot is $100, you’re considering a $20 bluff.
-
You’re risking $20 to win $100 → you need your opponent to fold more than 1 in 6 times.
-
-
You decide to bluff only if you believe that fold frequency is realistic.
Bluff Raises
Bluff-raising is much harder to pull off:
-
Pot is $100; opponent bets $20; you call and raise $20 more.
-
You’re investing $40 to win $120 → about 3-to-1 on your bluff.
-
Your opponent is now getting huge odds (like 8-to-1) to call the raise.
-
-
That means they need to be extremely disciplined and capable of folding strong hands for your bluff raise to be profitable.
-
Against typical players, pure bluff raises on the end are almost always bad; folding is usually better when you have nothing.
Last Position: Opponent Checks to You
When you’re second to act and your opponent checks, you can:
-
Check behind (take the showdown),
-
Bet for value, or
-
Bluff.
Betting for Value After a Check
Key principle:
You should only value-bet when you think:
Your hand is a favorite specifically in the subset of hands that call.
So:
-
Even if you were an 80% favorite before the action,
-
If your opponent will fold all worse hands and only call (or raise) with better ones,
your bet actually has negative expectation. -
Classic case: you have an obvious monster that your opponent can see on the board; if they call, you’re almost always in trouble.
So on the end:
-
Don’t ask just “How often am I ahead?”
-
Ask “When they call, do I win more than about half the time (ideally 55%+ to account for possible check-raises)?”
If yes → value-bet.
If no → check behind, even with a hand that was a big favorite overall.
Bluffing After a Check
You bluff if:
-
You think your opponent will fold often enough given the pot size.
-
But in last position, your opponent knows that their check may look weak, so they may call more lightly: bluffs in position often succeed less often than in first position.
Last Position: Opponent Bets Into You
When you’re second and face a bet, you have three options:
-
Fold
-
Call
-
Raise
1. Folding
Fold when:
-
The combined chance that:
-
Your hand is best plus
-
The opponent is bluffing
is less than the pot odds you’re getting.
-
Example: Pot $80, bet $20 → you need to win more than 1 in 5 times to call.
2. Calling
Call when:
-
Your estimate of winning (either because your hand is actually better or villain is bluffing enough) is better than your pot odds.
-
You do not need to be the favorite overall; you just need a positive expectation given the price.
3. Raising
Raise for value when:
-
You expect to be a solid favorite to win that extra bet when called (often ~55%+),
-
And your opponent will actually call the raise often enough.
Against:
-
Average players: they tend to call raises with many hands, so you can raise for value more often but bluff-raise rarely.
-
Tough players: they fold more to raises, so:
-
Bluff-raises sometimes become feasible,
-
But value raises should be more selective, since when they do call, they often have very strong holdings.
-
There’s a special advanced case:
-
Against world-class players, you might raise even with a merely “OK” hand if:
-
The raise will cause them to fold some hands that are currently better than yours.
-
This can increase your overall win probability enough to justify the aggression.
-
But generally, stick to:
Raise only when you believe you’re clearly favored to win the extra bet when called.
First Position: Four Basic Plans
With a legitimate hand and you are first to act on the end, your possible plans are:
-
Bet (for value or as a bluff),
-
Check with intent to check-raise,
-
Check with intent to call, or
-
Check with intent to fold.
What’s correct depends on:
-
Your hand strength,
-
Your opponent’s tendencies (calls more than bets? bets more than calls? timid? aggressive?),
-
The pot size.
First Position: Check-Raising With Strong Hands
Check-raising on the end makes sense when:
-
You are very strong, and
-
There is a good chance your opponent:
-
Will bet if checked to, and
-
Will call your raise.
-
Sklansky frames it as:
-
Compare:
-
Probability your opponent calls a normal bet, vs.
-
Probability they bet when checked to, and then:
-
Call your raise,
-
Or bet/fold to your raise.
-
-
The simplified guideline:
Check with the intention of raising if you believe the chance they bet when checked to is at least about half the chance they would call a normal bet.
You should check-raise less often:
-
Against weak calling stations (they call a lot but don’t always bet themselves), and
-
Against tough players who bet less often and fold more easily to raises.
Sometimes you deviate and simply bet out if you expect:
-
They’ll raise you, and
-
You can then reraise and win three bets total (classic full-house vs. apparent straight/flush spot).
First Position: Fair-to-Good Hands When You’re a Favorite
For decent but not monster hands that are favorites if called, but not strong enough to check-raise, you have two main lines:
-
Bet when:
-
Your opponent will call with more hands than they would bet if checked to.
-
You get more value by simply betting yourself.
-
-
Check and call when:
-
Your opponent will bet with more hands than they would call with.
-
This includes their value hands plus extra bluffs.
-
By checking, you capture those bluff bets instead of forcing them to fold.
-
You never check-fold with a hand that is a favorite if called; folding would sacrifice positive expectation.
First Position: When You’re an Underdog if Called
With a hand that’s behind the calling range, you decide among:
-
Betting,
-
Checking and calling, or
-
Checking and folding.
Key logic patterns:
-
Bet as a smaller underdog instead of calling as a huge underdog
-
Sometimes betting as, say, a 20-to-15 underdog is better than checking and then calling as a 20-to-10 underdog.
-
Betting widens the set of hands they continue with (some worse ones call), making your situation less bad.
-
-
Check and fold when:
-
If you bet and get called, you’re clearly an underdog,
-
If you check, your opponent will almost never bet worse hands.
-
Then paying a bet is simply burning money.
-
-
Check and call when:
-
Your opponent will bet with more hands than they call with, including:
-
Some value hands you beat,
-
Plus bluffs.
-
-
And the pot odds justify the call.
-
Even if you couldn’t profitably call versus their calling range, you can justify calling against their betting range because you now beat some of their bluffs.
-
-
Occasional special case: big pot, timid opponent
-
You might check and call even though:
-
They would call more hands than they would bet,
-
Because they will check behind some better hands you would otherwise pay off, and
-
You also avoid facing a raise in a large pot.
-
-
Practical Examples of First-Position Thinking
Sklansky runs through multiple game examples (draw, hold’em, lowball) that all reinforce the same pattern:
-
Against opponents who call more than they bet:
-
You should lean toward betting your decent hands for value.
-
-
Against opponents who bet more than they call (more bluff-prone):
-
You should lean toward checking and calling, capturing their bluffs and extra betting hands.
-
Summary: A Mental Checklist for Heads-Up on the End
When all cards are out and you’re heads-up:
In Last Position
-
If opponent bets:
-
Fold if your equity < pot odds.
-
Call if your equity > pot odds.
-
Raise for value only if you’re clearly favored when called.
-
Occasionally raise as a bluff vs top players who can fold better hands.
-
-
If opponent checks:
-
Bluff only if they’ll fold often enough given the pot size.
-
Value-bet only if your hand is a favorite against the hands that call, not just overall.
-
In First Position
-
With a very strong hand:
-
Bet or check-raise; choose based on how often they bet and call.
-
Sometimes bet to set up a raise-reraise and win three bets.
-
-
With a solid but non-monster favorite:
-
Bet if they tend to call more than bet.
-
Check and call if they tend to bet more than call.
-
-
As a small underdog if called:
-
Bet when it turns you into a smaller underdog than if you checked and then had to call.
-
Check and call if their betting range includes many weaker hands and bluffs, and pot odds are good.
-
Check and fold if they almost never bet worse hands.
-
Overall, the chapter’s message is:
On the river, your job isn’t just to “have the best hand.” Your job is to choose the line with the highest expectation for that final bet—based on position, pot odds, and how your particular opponent bets versus calls.