SplitSuit's value-betting guide draws the distinction between obtaining worse action and making better hands fold. This article makes the first task auditable: one explicitly weighted river range, three stipulated size responses, and an equal-input comparison with checking. The numbers are original teaching assumptions, not a solver output or a claim about how players actually call.
Absolute Strength Is Not the Calling Range
Top pair, a set, and a flush are hand descriptions. Their betting value depends on the board, the range that reached the decision, and the response to your chosen size. A flush on a paired board may face full houses. A modest pair may get called by weaker pairs against one opponent but not another. Before betting, describe the hands you are targeting rather than relying on the name of your own hand.

Distinguish the range that arrives, the range that calls, and the range that raises. A checked range can contain weak hands that never pay another bet, traps that raise, and bluff-catchers whose decisions depend on size. Counting every worse hand in the arriving range as a source of value overstates the case. Counting only the hands revealed at showdown can distort it in a different way.
The same idea applies to a read such as “this player calls too much.” Too much against which size, on which street, with what hands, and in what sample? It can be a useful hypothesis, but it is not a complete river calling range. An honest worksheet preserves that uncertainty and tests more than one plausible response rather than choosing the response that makes the bet look good.
A Fully Specified River Model
Here is an original No-Limit Hold'em river worksheet. You are BTN, acting last, and hold K♦Q♥ on K♣9♦5♠2♥7♣. Your opponent checks. The pot is 100bb, both players have 200bb remaining, and there is no rake, jackpot deduction, or tournament payout effect. If you check back, the hand goes directly to showdown. You hold one pair, kings, with a queen kicker.
We do not claim to derive the arriving range from a real preflop/flop/turn history. Instead, the next table defines a small hypothetical river range so that the arithmetic can be checked. “Raw combos” counts all exact legal holdings in the named class after removing your cards and the board. “Weight per combo” stipulates how much of each legal combo reaches this decision. The resulting weighted mass is not a count of literal hands in a live session.
| Opponent hand class | Raw legal combos | Weight per combo | Weighted mass | Result versus your K♦Q♥ |
|---|---|---|---|---|
| KJ, suited and offsuit | 8 | 1 | 8 | Worse |
| KT, suited and offsuit | 8 | 1 | 8 | Worse |
| A9, suited and offsuit | 12 | 1/2 | 6 | Worse |
| QJ, suited and offsuit | 12 | 1/2 | 6 | Worse |
| AK, suited and offsuit | 8 | 1/2 | 4 | Better |
| K9, suited and offsuit | 6 | 1/2 | 3 | Better: two pair |
| 99 | 3 | 1 | 3 | Better: set |
| 55 | 3 | 2/3 | 2 | Better: set |
| Total | 60 | — | 40 | 28 worse / 12 better |
The combo arithmetic is visible. K♣ is on the board and K♦ is in your hand, leaving two kings. Each can combine with four jacks for eight KJ holdings or four tens for eight KT holdings. Three nines remain, giving twelve A9 holdings and six K9 holdings. Q♥ is unavailable, leaving three queens and twelve QJ holdings. Three remaining nines give 3 choose 2 = 3 pocket-nine combos; the pocket fives work the same way.
The weighting step is different from card removal. A blocked combo is impossible. An unblocked combo given half weight is possible but stipulated to appear half as much as a full-weight combo. Treating those as the same operation would hide an assumption. The poker ranges guide explains that distinction more generally.
What Checking Is Worth Under These Assumptions
There are no ties in the defined range. You win against 28 of the 40 weighted units and lose against 12. Checking back therefore wins the existing 100bb pot with probability 28 ÷ 40 = 70%. Its expected value from this river decision, with no further money invested, is 0.70 × 100 = 70bb.
This convention treats winning the existing pot as a positive result and losing at showdown as zero additional return. Earlier investments are already in the pot and are the same for every river option. We are not claiming that checking earns a 70bb profit measured from the beginning of the hand. If you use a different accounting baseline, apply it consistently to every option; the difference between the options stays the same.
A bet must be compared with this 70bb check EV, not merely with folding or with an imagined zero-value check. Because BTN acts last, checking here closes the action. That removes the additional complication of an opponent betting after your check. An out-of-position example would need that response branch and could give checking a different value.
Bet Size Changes the Calling Range
Now stipulate three responses. In this first model, the opponent never raises, all better hands call every size, and only worse hands fold. Against 25bb, KJ and KT always call, A9 calls half the time, and QJ always folds. Against 60bb, KJ always calls, KT calls half the time, and A9/QJ fold. Against 100bb, KJ calls half the time and all other worse hands fold. These are assumptions established before the calculation.

| Your river size | Worse weighted mass that calls | Better weighted mass that calls | Weighted mass that folds | Your winning share when called |
|---|---|---|---|---|
| 25bb | KJ 8 + KT 8 + A9 3 = 19 | 12 | 9 | 19/31 = 61.29% |
| 60bb | KJ 8 + KT 4 = 12 | 12 | 16 | 12/24 = 50% |
| 100bb | KJ 4 = 4 | 12 | 24 | 4/16 = 25% |
The 25bb size does not get “nineteen combos” in a literal deck-count sense. Its nineteen is weighted calling mass: eight full-weight KJ units, eight KT units, and half of the six-unit A9 group. This is why the original weights and the size-dependent calling assumptions both remain visible. Neither the arriving range nor its response can be inferred from your final hand alone.
To compute bet EV, use three branches from the same decision point. When the opponent folds, you win the original pot: +100bb. When a worse hand calls a bet of size B, you win the original pot plus the opponent's new call: +(100 + B)bb. When a better hand calls, you lose your additional bet: −Bbb. Your own bet is returned as part of the pot you win, so do not count it again as profit.
For 25bb, the calculation is (9/40 × 100) + (19/40 × 125) − (12/40 × 25) = 22.5 + 59.375 − 7.5 = 74.375bb. Checking is worth 70bb under the same assumptions, so the 25bb bet adds 4.375bb relative to checking. That is the modeled difference, not an empirical win rate or a guarantee for a real opponent.
| Option | EV under the stipulated response | Difference from check |
|---|---|---|
| Check back | 70bb | — |
| Bet 25bb | 74.375bb | +4.375bb |
| Bet 60bb | 70bb | 0bb |
| Bet 100bb | 50bb | −20bb |
For 60bb, 16/40 × 100 + 12/40 × 160 − 12/40 × 60 = 70bb. For 100bb, 24/40 × 100 + 4/40 × 200 − 12/40 × 100 = 50bb. The largest bet collects the most money when one of its few worse callers pays, but that does not compensate for the better hands continuing and the lost calls from weaker groups.
The 25bb bet is best among these four modeled options, not proven best among every legal bet size and line. We have not defined the response to 10bb, 40bb, 150bb, or an all-in. A solver or a broader study would need those additional options and a complete model. The worksheet teaches the comparison; it does not pretend that three chosen sizes solve the river.
When “More Than 50% When Called” Applies
The first model has a helpful shortcut because it is unusually restrictive: the river is complete, checking closes the action, there are no ties or raises, every better hand calls, and every fold comes from a hand you already beat. Let W be the worse weighted mass that calls, L the better mass that calls, T the total arriving mass, and B the bet. The difference from checking is B × (W − L) ÷ T.
Under those assumptions, you add value over checking when more than half the calling mass is worse: W > L. Each worse caller gives you an extra B; each better caller costs you B; worse hands that fold give you the same original pot that checking would win. At W = L, betting and checking tie in this simplified model. That is why the 60bb size has the same EV as checking.
It is not a universal rule that “over 50% equity when called means bet.” Remove the no-raise assumption, let checking induce bets, allow better hands to fold, move the decision to the turn, or change the future strategy, and the comparison needs additional branches. A threshold derived from one model cannot establish the value of an action in another.
A Raise Branch Can Reverse the Result
Keep the same board, holding, 100bb pot, 200bb remaining stacks, and forty-unit arriving range. Consider the 25bb bet again, but change the response. The nineteen worse calling units stay the same. Of the twelve better units, AK and K9 contribute seven calls, while the five units of 99/55 raise to 100bb. In addition, half the QJ group's mass—three units—bluff-raises to 100bb; its other three units fold. The three A9 units that did not call still fold.
For this variant, you fold to every raise. That rule is part of the model, not a recommendation. There are now six folds, nineteen worse calls, seven better calls, and eight raises. When you bet and fold to a raise, you lose the new 25bb whether the raiser had value or a bluff. The calculation becomes (6/40 × 100) + (19/40 × 125) − (7/40 × 25) − (8/40 × 25) = 65bb.
| Response to 25bb | Weighted mass | Your outcome from the decision |
|---|---|---|
| Folds: A9 3 and QJ 3 | 6 | +100bb |
| Worse calls: KJ/KT/A9 | 19 | +125bb |
| Better calls: AK/K9 | 7 | −25bb |
| Value raises: 99/55 | 5 | −25bb because you fold |
| Bluff raises: QJ | 3 | −25bb because you fold |
| Total | 40 | EV 65bb |
Your winning share when called has risen to 19 ÷ 26 ≈ 73.08%, because some better hands now raise instead of call. Yet the complete bet-fold line is worth 65bb, below the unchanged 70bb check EV. The bluff raises turn some hands you would beat at showdown into lost pots. That is a concrete counterexample to treating “ahead more than half the time when called” as sufficient for a bet.
If you call some raises instead, calculate that different strategy with its extra payment and the raising range. Do not keep the 65bb result while quietly changing your response. Nor should a player conclude that an opponent actually bluff-raises QJ this often. The three-unit bluff-raise weight is a sensitivity assumption that shows which missing information matters.
Thin Value Is About the Margin, Not a Fixed Hand Class
“Thin value” describes a close value-betting case, not a permanent list of hands such as second pair or top pair with a weak kicker. The 25bb bet in our no-raise model has more worse than better calls, but its advantage over checking depends on only seven units of net additional worse action. Alter the response weights and that margin can shrink, disappear, or reverse.
For a simple sensitivity test, leave the 25bb size and twelve better calling units unchanged but reduce the worse calling mass from nineteen to eleven; all other worse hands fold. Under the original no-raise assumptions, the difference from checking becomes 25 × (11 − 12) ÷ 40 = −0.625bb. The same hand and same size now lose slightly relative to checking, because the calling response changed.
This is a more useful study question than “Can I value-bet top pair?” Ask which calls support the margin. Is it weak Kx paying? Is it some A9 bluff-catching? Do you have evidence for those calls at this size, or are you adding them because you want the bet to work? The closer the modeled comparison, the more honestly it should display its dependence on uncertain inputs.
Value Before the River Has More Moving Parts
On the flop or turn, hands that are worse now can improve. Betting can obtain a call and also deny live equity to folds. Your hand can improve or become weaker relative to a changing board, and both players may bet again. GTO Wizard's reasons-for-value-betting article describes bets as drawing value from both growing the pot and denying value, rather than forcing every holding into a pure value-or-bluff category.
Jonathan Little contrasts betting and checking on a three-diamond turn in a 50 BB example, in English with English captions. It illustrates why worse calls and later action matter; the video's particular solution is not a range chart or an independent validation of our river model.
For example, a current overpair may welcome a call from a draw, but the future runout and later sizes affect what that call is worth. A turn call is not the same event as a river call with no cards left. Conversely, causing a worse hand with live equity to fold can have value even though you were already ahead. Include that effect in the line comparison without using “protection” as a reason to bet any amount.
You also need the alternative check strategy. In position, a check may take a free card or close the river. Out of position, a check can face a bet that you might call, raise, or fold to. A model that assigns zero to every check loses these distinctions and can make almost any positive-return bet look attractive. Compare complete options from the same decision point.
Blockers Affect Value Targets Too
Your cards remove both better and worse opposing combinations. Holding a king in this worksheet reduces the available KJ, KT, AK, and K9 hands at once. That is not automatically good or bad for a value bet. Removing worse calls can reduce value; removing better continuation can improve the comparison. The effect depends on which exact groups remain and how they respond.
Do not import a blocker slogan from bluffing without checking its purpose. A card that removes some strong calls may help a bluff, yet the same card can also remove the weaker hands a thin value bet needs. Enumerate the relevant combinations and retain their action weights. “I block the nuts” is not a substitute for asking what calls your bet.
A River Value-Bet Worksheet
| Before you bet | What to write down |
|---|---|
| Decision state | Exact board, holding, position, pot, remaining stacks, and deductions |
| Arriving range | Legal combos, inclusion weights, and how the prior action supports them |
| Value targets | Specific worse hands that call each proposed size |
| Better continuation | Better calls, raises, and any credible better folds |
| Your raise response | Fold, call, or re-raise, with the extra cost included |
| Check alternative | Showdown, induced bets, future cards, and your response |
| Sensitivity | Which uncertain weights would reverse the comparison |
Use the hand-analysis guide to preserve the action record. The GTO poker guide explains why a model's strategy depends on its inputs. Neither a successful call nor an opponent's fold establishes that your chosen size maximized value. Record what happened, then review whether your assumptions were justified before seeing the result.
The practical habit is simple: name the worse calls first, include the better continuation, and compare with checking. If the answer depends on a speculative caller or an ignored raise branch, make that dependency visible. A value bet becomes easier to evaluate when the source of its value is a range you can inspect rather than a label attached to your own cards.
Sources and Model Boundaries
- SplitSuit — “How to Value Bet in Poker (With Examples)”, reviewed October 28, 2026. Definition, Value Betting Example, and Value Bet Considerations support the worse-continuation question and the importance of size. Its player labels and example responses are not empirical evidence for this worksheet.
- GTO Wizard — “Reasons for Value Betting in Poker”, December 6, 2022; reviewed October 28, 2026. Sources of Value supports the distinction between growing a pot and denying value before the river. No solution frequency, chart, or numerical result is reproduced here.
The exact range, combo weights, size responses, raise branch, EV convention, and calculations are original educational models. They do not represent measured opponents, solved play, or Poker-Eye performance. The no-raise model and the raise variant are deliberately separate; real decisions require the actual action history, ranges, legal options, and future strategy.

