MDF Explained: The Math Behind Calling and Folding

What is Minimum Defense Frequency?

Minimum Defense Frequency (MDF) is the percentage of your range you must continue with against a bet so the bettor cannot show an automatic profit by bluffing with any two cards. If you fold more often than MDF says, every bluff becomes +EV — including napkins like 72o. If you defend at MDF or more, the bluffer's pure-bluff EV drops to zero or negative, and they're forced to balance with real value to make their bets work.

That's the entire point of the metric: it's a defender-side unexploitability floor. It does not tell you which specific hand to call with, and it does not tell you whether calling is +EV in a vacuum. It tells you the frequency at which your range, as a whole, must continue to deny your opponent a free bluff.

MDF lives on the same axis as alpha — they're two ways of saying the same thing. Alpha is the bettor's required bluff-to-value ratio for their bet to be balanced; MDF is the defender's mirror image of that same equilibrium. Where alpha asks "how often does villain need to be bluffing for hero's call to break even," MDF asks "how often does hero need to call for villain's bluff to break even." Same equation, opposite seat.

📖 For the broader framework, see GTO vs exploitative play.

The Formula

The clean form:

MDF = pot / (pot + bet)

Where pot is the size of the pot before villain's bet and bet is the size of villain's bet. Both numbers should be in the same units (chips, big blinds, or pot fractions — pick one and stay consistent).

An equivalent expression you'll sometimes see written out is:

MDF = 1 - (bet / (pot + bet))

That second form is just 1 - villain's risk-reward on a bluff. Villain risks bet to win pot + bet (the pot plus their own bet returning), so bet / (pot + bet) is the fraction villain needs hero to fold to break even on a 0%-equity bluff. MDF is what's left over: the fraction hero must not fold.

A worked example. Pot is 100, villain bets 50. Hero's MDF is 100 / (100 + 50) = 66.7%. Villain risks 50 to win 100; their bluff break-even is 50 / 150 = 33.3%. Hero must continue 66.7% so villain's pure bluff loses money.

MDF by Bet Size — The Reference Table

Two patterns to internalize:

1. Bigger bets demand a smaller defense frequency. A 33% pot stab forces you to defend 75% of your range. A 1.5x overbet only forces you to defend 40%. That's why polarized overbets work — they let villain bluff with hands that have very little equity while only needing hero to fold ~60% of their range.
2. The break-even-fold column is the bluffer's perspective. If villain bets 50% pot and you fold 50% of the time (instead of the required 33%), they print money on 0%-equity bluffs.

MDF vs. Pot Odds — Don't Confuse Them

These two numbers come from the same arithmetic but answer opposite questions.

If villain bets pot, your pot odds are 1 / 3 = 33% — your specific hand needs 33% equity vs villain's continuing range to break even on the call. Your MDF is 50% — half of your range must continue regardless of which specific hand you hold.

A common beginner trap: looking up pot odds, finding their hand has 30% equity vs villain's value range, and folding. That's correct for that hand — but if they fold every hand below 30% equity vs value, they'll often fold 70%+ of their range, far beyond MDF, and donate every bluff villain makes. Pot odds picks individual hands; MDF picks how many hands.

How to Choose Which Hands Defend

MDF tells you the count. It does not tell you which combos. The selection rule, in order:

1. Equity vs villain's value range first. If villain's value range on the river is {TT+, AQ+}, sort your range by raw equity against that range. Defend top-down until you've covered the MDF threshold.
2. Blockers second. Among hands with similar equity, prefer those that block villain's value combos. Holding the A♠ on a four-flush board removes the nut flush from villain's range. Holding a Q on a QJTxx runout reduces sets and KT straights.
3. Avoid blocking villain's bluffs. This is the inverse mistake. If villain's bluffs are missed flush draws, holding a high card of that suit means villain has fewer bluffs in their range — making your call worse. Pick continues that unblock bluffs.

🎯 Stuck on which combos to defend? Paste the hand into DEEPFOLD AI Coach and it will rank your range against villain's likely value distribution.

Worked Example: BTN c-bet 50% on Q72r

Setup: 100bb effective, BTN opens 2.5x, BB calls. Pot is ~5.5bb on the flop. BTN c-bets 50% pot (~2.75bb). Flop is Q♦ 7♣ 2♠ — dry, rainbow, no draws.

MDF = 5.5 / (5.5 + 2.75) = 66.7%. BB must continue with two-thirds of their preflop calling range.

BB's flop range here is roughly: any pair, any A-high with a backdoor, mid-rundowns, suited connectors, broadway combos. Working top-down by equity vs BTN's c-bet range:

Counting combos, the top of range adds up to roughly 65-70% of BB's preflop continuing range — landing right around MDF. Notice that BB doesn't fold every hand without a pair. A bluff-catcher like A5s with a backdoor flush draw still defends, because the alternative — folding it — would push the defense rate well below MDF and let BTN bluff with any two.

Multi-Street MDF: It Compounds

This is where most players leak the most EV. MDF is computed per street, per bet. If villain barrels three streets, your "real" defense versus the entire line is the product of each street's MDF, not the river MDF alone.

Suppose villain bets 66% on flop, 66% on turn, 66% on river. MDF on each street is 60%. Your overall defense rate vs the full triple-barrel:

0.60 × 0.60 × 0.60 = 21.6%

Only 21.6% of your starting range will see a river call. That's correct and expected — barrels narrow ranges fast. The mistake is assuming you need to defend 60% on the river of your turn-calling range, which would over-defend rivers.

A few practical implications:

• Don't float wide on the flop just because flop MDF is high. Hands that float flop with no plan get sliced apart by turn and river barrels.
• Pick hands with playability across streets. Backdoor draws, gutshots with overcards, and pairs that improve to two pair — these defend flop and remain strong if villain barrels turn.
• Polarized players force you to compress. Vs a polarized triple-barreler, your river defense rate within your turn-calling range can be near MDF (because you've already filtered to bluff-catchers and value), but vs a merged barreler the math changes.

Use a solver to verify the compound rates on lines you play often. DEEPFOLD-SOLVER will print per-street defense frequencies for any node, so you can compare your real-table defense against equilibrium.

Exploitative Deviations from MDF

MDF is a floor for being unexploitable. It's almost never the highest-EV strategy. The moment you have a read, you should deviate.

When to fold more than MDF (under-defend)

• Villain underbluffs. Most micro and low-stakes regs do not bluff river enough. Against a player who triple-barrels only with value, your river MDF goes out the window — fold every bluff-catcher.
• Range-faceup spots. If your line screams "weak top pair or worse" and villain's line screams "two pair plus," defending MDF burns money. Pot odds should drive the decision, not MDF.
• Bad price after rake. In live cash with a high rake cap or in micro-stakes online, rake eats into close calls. Tighten up below MDF.

When to defend more than MDF (over-defend)

• Villain overbluffs. Most live recreational players bluff turn but not river — but some bluff every street. If villain's bluff frequency is 40% and value is 60%, defending MDF is wrong; you should call wider because the price is generous.
• Strong blockers in a polarized spot. Holding the A♠ on a fourth-flush river when villain has been barreling means their value (nut flush) is reduced. Even if your raw equity vs value is poor, the blocker makes the call profitable.
• Future-game considerations. Establishing a reputation for not folding can pay back later in the session via villain checking back marginal value.

The cleanest mental model: MDF is "how to not get exploited." Exploitative play is "how to exploit the actual person across from you." The first is a default; the second is the goal.

Common Mistakes

1. Treating MDF as a target instead of a floor. MDF is the minimum. If villain underbluffs, the correct frequency is lower. If they overbluff, it's higher. Defending exactly to MDF against an unbalanced opponent leaves money on the table.
2. Mixing up MDF and pot odds. Pot odds gate individual calls. MDF gates the range. Use both — the pot odds tell you which specific hands to filter in, and MDF tells you how many to keep.
3. Forgetting MDF rescales with bet size. Players who default to "I always call ~60% on the river" are correct vs 66% pot, exploitable vs 33% pot (under-defending), and overcalling vs 1.5x overbets. Look at the size first.
4. Ignoring range composition. If your range arrives at the river with 80% missed draws and 20% strong made hands, no MDF math saves you — your range is structurally weak. The fix is range construction earlier in the hand, not heroic river calls.
5. Compounding MDF wrong across streets. Defending 60% per street is fine on each isolated decision but means only ~22% of your starting range reaches a triple-barrel river. Don't confuse street-by-street MDF with whole-line defense.
6. Defending with hands that block villain's bluffs. Continuing with the K♠ on a missed-spades river removes spade bluffs from villain's range. Better to defend with off-suit Kings or non-spade aces — same showdown value, doesn't block the bluffs you want villain to have.
7. Applying MDF in multiway pots. MDF assumes a heads-up bettor-defender dynamic. Once a third player is involved, the math splinters — defer to range-vs-range solver outputs, not the MDF table.

FAQ

Is MDF the same as alpha?

They're two views of the same equilibrium. Alpha is the bettor's required bluff-to-value ratio for the bet to be balanced. MDF is the defender's required continuation rate to make any 0%-equity bluff break even. The numbers are different (alpha is the bluff fraction of villain's betting range; MDF is hero's fold-resistance), but they describe the same Nash boundary. Most coaches use MDF when teaching defense and alpha when teaching aggression.

How does MDF apply to bluff-catchers?

Bluff-catchers are exactly the hands MDF is designed to keep in your range. A bluff-catcher beats villain's bluffs and loses to villain's value, so its EV depends entirely on villain's bluff frequency. MDF picks the threshold at which you continue with enough bluff-catchers that villain can't profitably add bluffs. If you held no bluff-catchers — pure value vs pure air — you'd just call with all the value and fold all the air. MDF matters in the ambiguous middle.

Should I always defend exactly to MDF?

No. MDF is the minimum to be unexploitable, not the maximum-EV play. If you have any read on villain's bluff frequency, you should deviate. The default is "defend to MDF when I have no information." The whole game above ~$100NL is figuring out which villains underbluff and which overbluff, and adjusting in either direction.

Does MDF apply to preflop?

Loosely. Preflop ranges are wider and the math is muddied by future-street equity, so people don't talk about MDF preflop in the same way. Functionally, the BB's defense vs an open uses similar logic — the BB has to continue often enough that the opener can't auto-profit on a steal — but coaches usually call this "BB defense range" and tabulate it directly rather than deriving it from MDF.

What about MDF when villain is shoving river all-in?

Same formula. If villain shoves 1.5x pot on the river, MDF is 40%. The all-in nature doesn't change the math; it only changes that villain's range is more polarized than usual (no value-cuts left, all chips committed). You still need to defend ~40% of your range — but selecting which 40% matters even more, since you only have one call decision.

How do I learn what fraction of my range a given hand represents?

With a solver or a range tool. Eyeballing it on the felt is hard. DEEPFOLD Basic ($17/mo) includes range visualization for every spot in your hand history, and DEEPFOLD PRO ($25/mo) lets you compare your decision to GTO baseline frequencies. PioSolver, GTO Wizard, and GTO+ all output per-hand action frequencies you can study off-table.

Does MDF still apply if villain bets very small (~10-20% pot)?

Yes, and the table extends: 20% pot → MDF 83%, 10% pot → MDF 91%. Tiny stabs force you to defend almost everything — which is why donk-bets and small probes are so common in modern theory; they let villain bet more thinly because hero can't profitably fold much. The flip side: against a small bet, you almost never raise wide, since villain's range will be very wide too.

Related Reading

• Exploitative vs GTO: when to deviate
• Range advantage explained
• Polarized vs merged ranges
• Bluff-catching principles

💡 MDF is the floor, not the ceiling. Default to it without reads, deviate aggressively with them.

🎯 Audit your defense frequencies hand-by-hand → DEEPFOLD AI Coach