How Donk Betting Can Up Your Winrate

Many players navigate huge chunks of their poker career without ever employing donk bets. After all, they are called “donk” bets for a reason. Historically, it is only bad players that employ such bets as part of their poker strategy.

Is donk-betting as bad as it’s made out to be? We might be surprised to learn that donk-betting plays a vital role in theoretically correct play. In other words, if we don’t employ donk-bets in certain scenarios, we are not maximising our winrate.

Here, we will consider the main situations where we should consider making a donk-bet.

Let’s start by defining our term:

Donk Bet – A postflop bet made out-of-position by a player who is not the aggressor on the previous street.

So, in order to donk-bet the flop, the definition implies we need to be the preflop caller. We might call an open-raise, a 3bet, or a 4bet – the key idea is that we did not take the last aggressive action preflop. In the case of donking the turn, the definition implies we closed the flop action with a call OOP (although we don’t necessarily need to be the preflop caller).

Traditionally, players have always been taught to check every street to the previous aggressor, so many are completely closed off to the idea of donk-betting.

We’ll discuss two general areas where donk-betting should be considered:

- Exploitative donk-betting
- Theoretically correct donk-betting

Table of Contents

Exploitative Donk-Betting

In scenarios where we expect to encounter a high frequency of continuation bets (on any postflop street), donking has limited application. After all, given that our opponent will nearly always c-bet, we could get a check-raise, which is often more profitable.  

Donk-betting starts to become a valuable exploitative tool when we don’t expect our opponent to be continuation betting with a high frequency. We want to prevent him from being able to check back often and realise additional equity. We also want to avoid missing value with our stronger holdings, so leading makes sense.

Our donk-betting strategy is hence partly reflective of what we think regarding our opponent’s c-betting aggression. If he c-bets too much, we never donk. If he c-bets too little, we’ll consider opening up an aggressive donk-betting strategy.

The second scenario where donk-betting becomes a valuable exploitative tool is when our opponent is playing poorly against donk-bets.

There are some obvious fundamental errors we can look out for:

- Folding too much to donk-bets
- Calling too much against donk-bets
- Raising too aggressively against donk-bets

Folding/Calling too much – Fold to donk-bet frequency should be roughly 40% on each street - assuming we are talking single-raised HU pots. If we see our opponent is folding 70% of the time to turn donk-bets, then we clearly need to be opening up a bluff-heavy donk-betting range on the turn.

In fact, it’s possible that we should never be check-folding the turn, since any hand we are planning to check/fold can be donk-bet profitably. Many players don’t bother putting fold-to-donk stats on their main HUD panel, or even checking their popups for such info. Good players are on the lookout for every available exploit, however.

Raising too aggressively – There is a poker school out there that gives the following advice - “Always raise flop donk-bets and fire every turn”. Now while arguably highly unsophisticated advice, it does have its merits against clueless opposition.

Theoretically Correct Donk-Betting

Assuming we were looking to program some perfect GTO poker robot, it would be incorrect for us to ignore donk-betting possibilities completely. If we forced the robot to always check to the PFR, it would never be able to play perfect GTO poker, since GTO poker involves donk-betting. 

The easiest way to conceptualise this is to understand that GTO poker doesn’t care who has the initiative - i.e. it doesn’t care who was the aggressor on the previous street and who was the caller. It only cares about the current setup of each players’ range (along with other variables such as stack-depth etc.) 

As a very general rule, when one player’s range is stronger, he will tend towards doing more betting. It’s important to take this statement for what it is. Most players who talk about range advantage, don’t understand it on a decent level (or any level at all). Having a significant equity advantage does not guarantee that one player will do most of the betting or even any betting at all.

Several different components allow us to describe one range as “stronger” than another, and pot-equity is simply one small part of that. Currently, a large amount of misinformation on the topic exists - circulating amidst training videos and even amongst high-stakes pros. When we use the term “stronger”, we’ll assume the relevant variables have been taken into consideration, but we won’t break them down into specifics in this article.

If the OOP player A has a stronger range than IP player B, player A does not want to be checking his entire range. Why? Because player B can exploit this by checking back very frequently and peeling free equity. It’s necessary that player A funnel some of his hands into a donk-betting range so that his checking range isn’t strongly exploitable. Now when player A checks, his checking range we’ll be weak enough where player B is still incentivised to attack with a reasonable frequency.

Conversely, if player A’s range is already weak, donking doesn’t make too much sense. Player B will already be incentivised to bet aggressively IP, and player A weakening his checking range even further is not going to help with his plight. He’ll instead pool every available defending resource into his checking range and make his best stand possible. Despite this, he’ll still find himself forced to fold at an above average frequency, even when following a GTO strategy.

Note that this does not mean that player A made any sort of mistake. It simply means that some flops or turns and rivers, favour one range over another, and good GTO players need to adapt accordingly. The classic example that is usually given is when a relatively high second card on the board pairs. 

Imagine BB calls against a BTN open and the board texture is as follows:

Board Texture:Q♦ J♠ 6♥

BTN fires a continuation bet and BB calls.

Turn Card: J♣

BB donk bets.

The basic reasoning here is that BTN might not cbet his Jx holdings on the flop. They are not especially vulnerable and hence play reasonably as checks. BB, on the other hand, will be calling all his Jx holdings against a cbet. The turned Jack is hence more favourable for BB than for BTN. He becomes like player A in the theoretical discussion above.

Summary of When to Donk Bet