PLO c-bet frequencies, all 1,755 flops
Most published c-bet advice comes from a handful of example boards. This page aggregates every canonical flop in the 6-max 100bb solve for the most common single-raised pot: button opens, big blind calls, flop checks to the button.
The short answer
After the big blind checks, the button c-bets 52.3% of all flops: 37.0% as a 33%-pot bet and 15.4% as a full pot. The spread across textures is the useful part: ace-high boards get bet 58.3% and the top boards (an ace beside a low or middling pair) over 80%, while trips boards fall to 33.6% and a broadway three-straight like Q-J-T to 33.4%. Before the check even happens, the big blind leads 13.5% of flops - on everything except ace-high boards (3.5%), and on monotone boards most of all (18.3%). Every number is decoded with the same hole-key function the solver tool uses, so both read the same row for the same hand - see the method note below for what that does and does not certify.
The 10 most and least c-bet flops
Reach-weighted over the button's actual arriving range. Size split shown as 33% · pot.
Bet the most
| Flop | C-bet | 33% · pot |
|---|---|---|
| A♠8♠8♥ | 84.4% | 69.9% · 14.5% |
| A♠9♠9♥ | 82.8% | 65.4% · 17.4% |
| A♠T♠T♥ | 81.3% | 71.0% · 10.3% |
| A♠7♠7♥ | 80.9% | 68.3% · 12.6% |
| A♠J♠J♥ | 79.4% | 69.8% · 9.6% |
| 9♥9♦3♠ | 78.1% | 74.0% · 4.1% |
| 9♥9♦4♠ | 77.8% | 74.1% · 3.6% |
| A♠6♠6♥ | 77.4% | 62.6% · 14.8% |
| 8♥8♦3♠ | 75.6% | 70.3% · 5.3% |
| 9♥9♦2♠ | 75.3% | 70.8% · 4.5% |
Bet the least
| Flop | C-bet | 33% · pot |
|---|---|---|
| A♠5♠4♠ | 16.0% | 14.5% · 1.5% |
| 9♠8♠7♠ | 16.9% | 8.4% · 8.5% |
| Q♠J♠T♠ | 18.8% | 7.2% · 11.6% |
| T♠9♠8♠ | 19.4% | 12.2% · 7.2% |
| J♠T♠9♠ | 21.1% | 12.6% · 8.5% |
| 8♠7♠6♠ | 21.2% | 15.0% · 6.3% |
| 7♠6♠5♠ | 21.6% | 14.1% · 7.5% |
| K♠Q♠J♠ | 22.3% | 15.9% · 6.4% |
| J♠9♠8♠ | 22.4% | 16.6% · 5.8% |
| K♠J♠T♠ | 22.4% | 15.7% · 6.8% |
The pattern in the left column is one texture repeated: a paired board - six of the ten an ace beside a low or middling pair, the rest a low pair with a low kicker. The caller's big-blind range folded its bare-ace junk preflop or 3-bet its good aces, so these boards belong to the opener. The right column is one texture too: all ten are monotone, and every one of them also holds three cards to a straight. Flush and straight both live in the caller's range there, and the opener's edge is gone. Trips boards, the lowest class average on the page, sit just outside this list - the fifteen lowest boards are all monotone before T-T-T appears.
By texture
C-bet, donk-lead and check-raise rates per texture class, with each class's share of all flops.
| Texture | Share of flops | BTN c-bet | BB donk | BB raise vs 33% | BB raise vs pot |
|---|---|---|---|---|---|
| 7-high or lower | 9.2% | 50.6% | 15.6% | 13.3% | 7.6% |
| 8-T high | 23.1% | 51.6% | 16.3% | 14.5% | 9.0% |
| K/Q/J-high | 46.0% | 50.2% | 16.4% | 13.2% | 8.1% |
| Ace-high | 21.7% | 58.3% | 3.5% | 11.5% | 6.5% |
| Trips on board | 0.2% | 33.6% | 5.2% | 2.3% | 0.4% |
| Paired | 16.9% | 52.0% | 11.6% | 9.4% | 4.1% |
| Unpaired | 82.8% | 52.4% | 13.9% | 13.9% | 8.7% |
| Rainbow | 39.8% | 55.2% | 14.4% | 14.5% | 9.3% |
| Two-tone | 55.1% | 51.3% | 12.4% | 12.9% | 7.5% |
| Monotone | 5.2% | 40.9% | 18.3% | 5.0% | 1.3% |
| Connected (span ≤ 2) | 3.2% | 39.9% | 14.7% | 8.2% | 4.4% |
| Semi-connected (span 3-4) | 13.6% | 48.7% | 17.2% | 12.4% | 7.2% |
| Disconnected | 66.0% | 53.8% | 13.2% | 14.5% | 9.2% |
Four usable rules fall out. Bet more as the highest card gets higher and less connected to the rest. Bet less as the board pairs into trips or wires into middle straights. Read the donk column as the caller's announcement: it leads on everything except ace-high (3.5%), and on monotone boards most of all (18.3%) - which is also the suit class where the c-bet column is lowest (40.9%). And read the two raise columns as the mirror of the c-bet column: check-raising drops hard as the board pairs (9.4% of paired boards vs a 33% bet, 2.3% once it's trips) and drops again on ace-high - the button's best c-betting texture is the big blind's worst raising texture, because the caller simply has the fewest strong hands exactly where the opener has the most.
Spot checks you can reproduce
The same boards used across our other references, so the datasets can be checked against each other - and against the tool itself.
| Flop | BTN c-bet | BB donk |
|---|---|---|
| A♠K♦4♥ | 64.5% | 2.1% |
| T♠9♠8♥ | 40.1% | 19.5% |
| 7♣6♣5♥ | 43.8% | 19.3% |
| Q♦J♠T♥ | 33.4% | 12.1% |
| A♥Q♥6♠ | 62.8% | 3.2% |
| Q♦8♠8♦ | 45.6% | 14.8% |
The equity side of these boards is on the range advantage page: the boards the caller leads most (T-9-8, 7-6-5) are the ones where its range holds the straights, and the ace-high boards it never leads are the ones where the 3-bettor and opener dominate.
Check-raising the button's bet
Same spot, one street later: the big blind checked, the button c-bet, and now the big blind chooses between fold, call, and raise.
Facing a 33%-pot c-bet, the big blind check-raises 13.1% of flops; facing a full pot bet, 7.9%. Board texture swings this harder than anywhere else on this page: trips gets raised back only 0.4% of the time vs a pot bet, while the top of the check-raise list - disconnected rainbow boards like Q-7-2 and K-8-2 - runs 15.3% or higher. None of the button's ten best c-bet boards (all paired, six of them ace-high) show up anywhere on the big blind's ten best check-raise boards (all unpaired and ace-free): the caller keeps its raising equity exactly where the opener has the least reason to fire big. On A-K-4 rainbow, one of the spot-check boards below, the big blind facing a 33%-pot c-bet folds 32.2%, calls 55.0%, and raises 12.7% - a three-way mix, not a fold-or-call read.
Raised the most (vs pot)
| Flop | vs pot | vs 33% |
|---|---|---|
| Q♦7♥2♠ | 18.2% | 23.6% |
| K♦8♥2♠ | 17.1% | 21.6% |
| J♦5♥2♠ | 16.3% | 21.0% |
| K♦8♥3♠ | 16.1% | 21.8% |
| Q♦5♥2♠ | 15.9% | 19.8% |
| K♦7♥2♠ | 15.8% | 21.3% |
| 8♦5♥4♠ | 15.6% | 18.7% |
| J♦6♥2♠ | 15.5% | 22.1% |
| 6♦3♥2♠ | 15.5% | 18.9% |
| K♦8♥4♠ | 15.3% | 18.6% |
Raised the least (vs pot)
| Flop | vs pot | vs 33% |
|---|---|---|
| A♠A♥9♠ | 0.0% | 0.8% |
| A♥A♦8♠ | 0.0% | 1.2% |
| A♠A♥6♠ | 0.0% | 0.4% |
| A♠A♥7♠ | 0.0% | 0.4% |
| T♠T♥T♦ | 0.1% | 0.9% |
| A♠A♥Q♠ | 0.1% | 2.3% |
| 9♠9♥9♦ | 0.1% | 1.2% |
| A♥A♦7♠ | 0.1% | 0.7% |
| A♠A♥J♠ | 0.1% | 2.1% |
| 8♠8♥8♦ | 0.1% | 1.4% |
0.0% in the "raised the least" column doesn't mean the CSV is broken - it means the solve's own strategy storage quantizes to 0.5-percentage-point steps, so the true rate is somewhere under a quarter-point, or the solve pruned the action outright. Every board on that list is trips or paired aces with a side card (A-A-9, A-A-8, A-A-6): the caller's own range there is capped, so it has nothing to raise for value and nothing worth bluffing with either.
Methodology
Spot: 6-max 100bb, folds to the button, button opens pot, small blind folds, big blind calls; on the flop the big blind checks and the button decides between check, 33% pot, and pot. For every one of the 1,755 canonical flops, each legal four-card holding's solver strategy was read from the solve's own data files and weighted by the solve's own preflop reach range for the button (7,905 frequency mass over 16,432 hand classes). Donk rates are the same computation at the big blind's first-action node. Texture averages weight each canonical flop by its number of raw suit arrangements (summing to 22,100).
Check-raise rates read the two deeper nodes where the big blind, having checked and faced a c-bet, decides fold, call, or raise - one node per bet size. The correct weight for "check-raise vs 33%" is the big blind's own checking frequency for that hand, not the button's frequency of choosing 33% over pot: the button's bet-size split cancels out of a rate that's already conditioned on facing that size, so re-weighting by it would double-count. Internal consistency was checked by summing fold + call + raise back to 100% for all 1,755 flops at both bet sizes (3,510 checks, all within 2 percentage points, the gap being the strategy table's own quantization).
Verification: every hand's row is addressed with the hole-key function the solve's
own manifest names (plo-v2, MonkerSolver
2.3.10) - the same function the solver tool uses, which is
what makes "reproduces the tool board for board" a property of the decode rather than
a hope. That matters because the failure mode here is silent: read a solve with the
previous abstraction's key and you get a real row belonging to a different hand, and
a frequency that still looks like a frequency. Reading this solve the old way moved
the headline averages by under a point, the median board by about one, and the worst
board by nearly twenty. Our flop-aggregate artifacts, a
separate generator with its own node discovery and its own range reader, agree with
this one to float precision on the overall c-bet, donk and both check-raise rates -
a consistency check, not an oracle, since the two also agreed while both were reading
the wrong key. On Q-8-8 two-tone, one of
the spot-check boards, the computed big-blind mix is check 85.2%,
33%-bet 12.9%, pot
1.9%; open that
board in the tool and you should see the same numbers to display rounding. The
check-raise fields share that same decode mechanism one node deeper in the same
tree; the deeper fold/call/raise node was checked against the tool's Study mode on a
live board (entry-node check frequency matched to within 0.05 points) but not fully
walked through to the raise decision in the UI, so treat the check-raise numbers as
verified by construction and by the internal consistency check above, not by a
second independent tool reading. Solve data decoded from MonkerSolver's own files
(data directory 6max_100bb_v3, profile directory plov2_g1_fb100_tb75_rb75); this page inherits that
solve's tree and assumptions. A re-solve published behind the same name retires every
number on this page at once, and so does a re-pack into a new abstraction - the
profile KEY stays put through both, which is why the directory and the codec are the
things printed here. Generated
2026-09-11; regenerate with node scripts/study-guides/gen-cbet-frequencies.mjs --solve 6max_100bb_v3.
SolvePoker, "PLO C-Bet Frequencies", solvepoker.app/plo-cbet-frequencies, updated September 2026. Questions
- How often should you c-bet the flop in PLO?
- In the most common 6-max single-raised pot (button opens, big blind calls, flop checks through to the button), the 100bb solve continuation-bets 52.3% of all flops after the check: 37.0% as a 33%-pot bet and 15.4% as a full pot bet. Any fixed personal frequency far from that, in either direction, is a leak against strong players.
- Which flops get c-bet the most in PLO?
- Ace-high boards, and ace-high boards with a paired side card most of all. The top of the table is A-8-8 two-tone at 84.4%, with A-9-9, A-T-T and A-7-7 close behind - all above 80%. Ace-high as a class runs 58.3% vs 50.2% for K/Q/J-high boards.
- Which flops should you not c-bet in PLO?
- Monotone boards above all - 40.9% as a class, and the entire bottom ten of the table is monotone, with A-5-4 checked back 84.0% of the time. Then trips boards (33.6% as a class, the lowest average on the page) and broadway three-straights: Q-J-T gets c-bet just 33.4% because the caller's range holds the straights.
- How often does the big blind donk-bet the flop in PLO?
- 13.5% of flops overall in this spot, concentrated on the boards that favor the caller: T-9-8 two-tone gets led 19.5% of the time and monotone boards 18.3%, while ace-high boards are led 3.5%.
- What c-bet size does the PLO solver use?
- This solve offers 33% pot and full pot on the flop. After the check it picks the small size about two-thirds of the time (37.0% of all flops vs 15.4% potted). Pot-heavy boards are the dynamic ones where equities swing; the small size carries the dry, static textures.
- How often does the big blind check-raise in PLO?
- 13.1% of the time it's facing a 33%-pot c-bet, 7.9% facing a full pot bet. Board texture decides almost all of it: trips gets raised back only 0.4% of the time vs pot, while a disconnected rainbow board like Q72 gets raised 18.2%.
- Are the boards you check-raise the same ones you c-bet?
- No overlap at all. The button's ten best c-bet boards are all paired, six of them ace-high (A-8-8 two-tone leads at 84.4%). The big blind's ten best check-raise boards are all unpaired, rainbow and ace-free - Q-7-2 rainbow tops that list at 18.2% vs a pot bet. The caller keeps its raising equity exactly where the opener has the least reason to fire a big sizing.