I have recently started playing seriously again after a long time away from the game. I used to be a good, profitable, TAG. Now my game isn't as profitable as it was and I'm having to change some things. One thing I've noticed about the new era of poker is that it is "wrong" to bet into the preflop raiser on the flop if you flatted OOP preflop.

I know a lot more players cbet with a very high% of their hands now so I sort of get the concept; "If i'm strong why not x/r or even x/c, after all he's "suposed to bet". And if i'm weak then betting the flop might look weak inviting a re-raise from preflop raiser so I'm better off x/folding weak, x/calling marginal hands, or x/raising as a bluff."

I'm not convinced, I'm not sure, but not convinced.

If my donk bet will be read as weak then why not do it with big hands?
If my donk bet looks to strong then why not do it as a bluff?

I feel like it should "level" itself out and leave you in a position where you should be able to make donk bets profitably. I'm all for usually checking to the raiser I always have been. What about these spots;

I have AdJd and flat a MP open.
flop comes JsTs2c. Do I really want to let him peel the turn for free? What if he has AsKs. Do I want to let 99 off the hook when a spade hits freesing my action when I maybe get a call on the flop?

OK, and what if I flat preflop from the CO with 9d9c vs an open from MP, and the flop comes; 2s3s4h. It's super easy for him to have two over cards, nine spades, and the other 3 5's. He likely has between 6 and 18 outs (but I don't know which) Do i want to let him draw for free and fail to define opponents hand?

Maybe I am just pointing out the obvious and most players would agree that there are times to donk bet, like when you have equity but the board is very wet. But I'd also like to address the concept of donk betting fairly often and what the downside is. Whatever the bet appears to be it should be possible to level with it, no??
Or is it something totally different that I don't understand?

Why don't I want to donk bet my middle pairs (As8s) on Td8c2s?, for instance. Is it because my opponent would likely have cbet letting me just call and get the same money in vs a weaker range? If so then why not donk bet bluff with 33 on that same Td8c2s?

Or am i just way off base and misunderstanding a more fundamental reason for not donk betting?


THoughts???

bet folding A8/33 on T82 is pretty standard depending on opps.

I would say donking or x/c x/r depends on stack sizes a lot. like lots of people donk sets and combo draws going for a 3b all in, but they don't balance this with medium strength hands and maybe bluffs

There's definitely nothing wrong with donking just try to balance somewhat. Be warned if you do it with medium strength hands vs aggressive players then you will often be put to the test so make sure you are prepared

Yeah donk betting isn't recommended cause ur repping either a marginal one pair hand or a monster like a set most of the time. Against aggro opps, they can raise u with air and u won't know what to do unless u actually have the monster. Donk betting can work well in certain situations but note that it is rare. For example if u have a monster against an aggro opp who likes to raise donk bets leading into them could be best. Or if ur up against an Opponent heads up who plays pot control with overpairs and doesn't bluff and calls only with tptk+ or a big draw, donk betting gives u value on ur monsters and allows u to steal the pot when u have nothing

Against a recreational player who will not exploit your weak checking range and will not c bet a lot then donking is fine. The better players will recognize that you're slightly weaker when you check call or check raise and exploit you by value betting thinner when you check.

fwiw last night on 'poker after dark" Dwan donk/called a shove by Galfond with something like K4 on a 24T board. Betting to induce a bluff raise imo.

I think one of the problems with donking the flop is that it's hard to balance. (Tom Dwan can manage it, though, haha!). You'd want to do it with some monsters, some bluffs, and some draws. If you're mostly donking middle pairs like your example, expect to be pushed off your hand. You'll flop middle pair a LOT more than you'll flop a set or better, so it's hard to get the balance right.


If you are thinking about donkbetting strategy, you might be playing oop too much.

I think most good players will agree that donk-betting is sometimes a good play - in spots where you have a monster and expect villain to misinterpret your bet and raise, say, or spots where you expect villain will often check behind and you can't afford to give a free card. If you search the forum, you'll find other threads about this.

I don't like it in general but there are rare spots where it is the best play.

When we check to the preflop raiser we have 3 options: fold, call and raise. We have to do our best to balance all these possible lines so as to not be easily exploitable. It is a fairly difficult task to do this, but I think its manageable for anybody doing their homework away from the tables.

When we mix in a donking range we open up a whole other subset of hands that require balancing. Once we donk, we will either get a call, fold or raise. When we face a raise we have another 3 options: fold, call and raise.

Adding in all these different lines to our repertoire complicates things a great deal, and could be a very bad thing against equal or better skilled opponents with position on us.

Instead of our range having 3 parts when facing a bet after checking (fold, call, raise) we have now broken it into two groups (checking, donking) and those 2 groups are further defined into smaller groups by folding, calling or raising against villain's bets or raises.

By doing this, we are narrowing our own range and defining our hand on behalf of the villains we face.

That is the gist of why I don't like it.

So when should we use a donk bet?

IMHO we should sometimes donk bet in spots where we are not concerned with balance, and it appears to be the line that will be most +EV.

Most of these spots will arise where we are HU in a pot with a fish, or there is a big fish in a multiway pot, or against semi competent players that you have very little history with but reason to believe that donking is best.

Donking into players that are good relative to your own skill that you play with regularly is probably bad.

For online players using a hud, it is also worth mentioning that you will likely never build up a meaningful sample on how a player reacts to a donk bet because 1) so few people actually do it, and 2) most donk bets (at least that I see) are ridiculously small - like 1bb! - so any data you have may be skewed since 1bb donk bets can basically be treated as a check most of the time. Having little information about how they react to a donk bet compared to the often vast amount of intel you can pick up on a regs cbetting tendencies, ( board texture, sizing, frequencies) it seems silly to want to venture into unknown territory rather than take lines where you have the most information.

If leading wasn't referred to as "donking," it wouldn't have such a bad reputation. Yes, some players do it poorly: fish donk a lot using weird sizing and a lot of decent players donk with an unbalanced marginal range. But depending on opponents, stack sizes, board texture, it can be the best play.

How could this be bad against a skilled opponent with position on us?


How are we narrowing our range? I don't really get what you're saying here.

At the same time, if we never donks then we are telegraphing our play to our opponent. Theoretically speaking being able to donk, whether you do it or not, could obviously never lower the value of the game for you.


Still don't get how it could be bad against good players.

I believe what he's saying is that it makes the game more complicated than it already is. Not that it is strictly bad, just that it probably shouldn't be added until you are playing very well with a more standard approach to the game. If your opp is decent/good he will be cbetting a high% and you'll have to balance your call/fold/raise vs cbet range to avoid an exploitable range. If you add donk betting to your list of standard plays then you have to balance your range of donk/call, donk/fold, donk/re-raise, x/call,x/fold,x/raise. If you rarely donk you only have to balance for x/call, x/fold, x/raise. So you don't want to make it so much a part of your game that you have to either balance it or risk being exploited unless you are much better than your opponent (able to actually balance twice the options). But, if you almost never donk bet then you can use the donk bet when its just CLEARLY the highest EV play and not have to worry about balance because it rarely comes up.

I think that's a very good argument. It says to me, personally, that its OK for me to donk bet fish because they won't exploit my donk bet range and it will be OK to do on very rare occassion for a specific reason against other players but I won't want to do it very often unless I get so good that I've pretty well mastered the fundamentals and I can add donk betting to gain extra edge. So, in my case, prolly leave it alone for a while. (accept vs. Fish).

I stopped doing it all together when I noticed my game not being as profitable as it used to be and I think my game suffered because a) there are times to do it and I wasn't even doing it then and b) I'm playing small stakes mostly and medium stakes live vs fish so I don't have to worry as much about balancing my donk bet range. I usually do it as a steal, with a big hand on a wet flop, or with a monster vs super stations. I don't think I'd consider it vs an aggressive opponent who's going to mostly cbet anyway. I'm not good enough, plain and simple. No Dwan.

well raising makes the game more complicated then only betting or calling, but no one would suggest only calling, betting or folding.

Fwiw, since initiative has no inherent value in poker, if it's correct for a Hero to a c-bet when out of position, then it's also correct for the Hero to donk bet in the same situation in that the hand distributions are the same in both situations.

In general I don't feel that I'm skilled enough to donk into anyone I consider to be a quality player. In live low stakes (1/2, 2/5) play however, there are plenty of reasons to donk into fish (covered previously).

One place that I'm never sure what to do is when I flop a hand where I feel that my non-donk options risk turning my hand face up. The most common example is when I'm in a 3-way pot with a fish in early position and I'm in the middle, having called a pfr by a late position player, holding Ax suited or suited connectors. I flop the flush draw, and the fish checks.

I know that if I check the raiser will bet, and I am reasonably sure that the fish will call with almost any pair, or any draw. I have a hard time in these spots, as I feel like my options are to a) donk into 2 people, potentially building a big pot on just the draw oop, b) check/call, pretty much killing any implied odds if I hit the turn (the face-up scenario), or c) c/r the inevitable flop c-bet, again, building a big pot on a draw, out of position, and also turning my hand face-up against a thinking player.

I don't mean to derail the thread, but this is the type of situation that I'm most often tempted to donk into...well that, and flopping the immortal stones in a slowish game where I'm just trying to get a bit more value.

I think you could have some bluffcatchers in your check call range for this scenario. With the bad player in I think the c bettor will have to play pretty honestly on the turn and river. This has the benefit of balancing your range with some semistrong hands and probably getting you some free cards with your draws.

When we add several more lines to our decision tree, there will be less combos for each line. So our rangewill be more narrow for each line, respectively. This is more problematic vs good players cause they will be able to deduce your range and develop a very good idea of what you have when you donk.

When you donk, you give him more information before he acts, so hesitate could lower the value of your game. Checking to the of raiser almost always doesn't let him deduce your range, so his decision should be a little tougher and he'll have to be more balanced. It slightly negates his positional advantage rather than enhancing it.

I agree there will be few times a donk is the best vacuum play. And sometimes he'll check behind and a bad card will come. But the free card will help us a lot of the time and having his 4bb c bet in the pot before we reveal anything is very valuable.

Overall, the non sinking strategy will let us call wider preflop, maintain balance, make better turn/river decisions and get to SD more often.

We could go through all of your points one-by-one, but I think it might be easier to address the issue from a different tact. I think you would agree that if "initiative" doesn't have inherent value in determining optimal poker strategy, then there is no reason to make a distinction between donking and c-betting out of position (which should really just be called betting out of position). And it's my contention that optimal poker only cares about ranges, equity, position, stack sizes. And so if we should never donk bet, then it also true that we should never bet when out of position.

I'm sure that there's a strategy that includes donking that could crush any game, probably for more than a non donking strategy could. So you're right in that sense. However, as someone with average(brag) intelligence I find it easier to estimate and adjust ranges based on game flow. Because of this, I disagree with the last sentence. If we were to give up the lead out of position then we disrupt the game flow and make hand reading more difficult on ourselves, and probably make hand reading easier on our opponents.

This isn't to say that I never donk, but if I were playing with pros I'd probably avoid it for the sake of information hiding.

So you want to use a different term? How about "spackle" or "spit" or "snatch" or "Munch"?

We're talking about two different things. The OP asked if there was anything fundamentally wrong with donk betting. At a game theoretical level of poker, donk betting vs betting out of position is not distinguished from each other.


I've heard this argument made before, but I don't really understand it. To me it just sounds like you want to do something because you're comfortable with it which doesn't sound like a good reason to base a decision. If you're not sure how your opponent will react to it, then start doing it and start to see how he reacts so you can exploit it. And just like every other spot in poker if you don't want to be exploitable, then you're going to have to balance your play. And to be honest, I think we make our play much more predictable to the Villain if we check to him 100% of the time. I don't understand how that cannot be the case.


No. I'm just saying that you're drawing a distinction where it's not relevant.

Let's say you devise a balanced donking strategy with a range made up of air, medium strength hands and monsters. This range may be balanced, but when you check you will be weaker because you will have to have some check folding hands in your range. I think this is giving up more information than checking 100% where you get to fold some air or call and raise with balanced ranges. This leads me to think that the first step in determining a balanced donking strategy should be to identify our junk hands and divide them equally between check or bet, so neither range is weaker. Then we can fill in the range with the appropriate value hands to make our bluffing optimal. I'm not sure but I think this would lead to an unbalanced checking range since we're just playing with the leftovers.

Ok I thought you were suggesting we use a different term instead of abolishing the term definition and all. You make good points, carry on.

The way you balance a bet when out of position is the exact same way you balance any other bet or raise for value.

I'm not sure how our checking range becomes weaker, it just contains less hands which doesn't necessarily make it weaker. If we are significantly deep, then we're going to need to have nutted hands in both our donking range and our check-call range. But this is also the case even if we check 100% of the time and then have either a check-raising range or a check-calling range, so it's not as if checking eliminates this need to have to have the nuts in all parts of our ranges.

It seems like we need to figure out our continuation range, and then consider our options(call or raise) and take each action with all hands at a certain frequency. Doesn't this lead us to a bluffing frequency that isn't optimal? Or does this just leave us with some junk hands that we include in our range at the correct frequency, throwing out the rest(what's to say that the rest is the correct folding frequency?)? I think this is about as deep as I can go into this. So I will bow out and continue reading. Thanks for playing. Good luck approaching gto.

RainbowBright, I agree with everything you have said in this topic (eg. initiative doesn't have inherent value, +1).

Only on this point, I like to add that there is a little difference: your range.
When you are the raiser preflop on most flop you have a stronger range than the caller, and you have a lot of nut/semi-nut hand (top set, overpair, TPTK, nutFD).
The preflop caller usually have a capped and weak range, so donkbetting on most situation works against him (you need less nut hand in a x/? range than in a b/? range, because you are risking more money, so who has the strongest range should be the one that bet).


I agree that never donk-betting can't be optimal, but this is a decent approximation that can be made by non-optimal player that try to reduce the complexity space of poker, using their time improving other part of their game.
For example I almost never donk-bet, and I haven't found yet how to integrate donk-betting in my game. But sooner or later I have to do some work on it.
On the other hand, if few has studied how and when donkbetting, can be expected that most villain react in non-optimal way to donk-betting, so there can be a valuable thing to study.

yeah, I agree with all this.

I disagree with this. At least the way I play, I only flat raises with hands that are ahead of villain's range and play well postflop (but not strong enough to 3-bet for value), hands like 99. I suppose it's the same for everyone who 3-bets a polarized range.

This is mostly true for reg vs reg hands though, and it only applies to HU pots , when there are already callers, you can profitably setmine with the smallest of pairs.

One thing that i've forgot to tell is that when defining who has the strongest range almost only the best part of his range count (the hand that one can plays for the whole stack for value).
You can be flatting a lot of middle to strong hand, but I suspect that you rarely flat QQ+/AK, so on most flop you don't have enought strong hand for being the one that start betting.

eg. raised pot on K72 rainbow:
raiser can have KK-77-22-AA-AK 100% of times on his range, there are 9 combo of set, 6 combo of OP and 12 combo of TPTK, say that he openraise 20% (265 combo), mean that 10% of his range is TPTK or better.
flatter can have 77-22 100% of times on his range, there are 6 combo of set, say that he coldcall 8% of hand (100 combo), mean that 5.6% of his range is TPTK or better. Flatter don't have any other hand with decent equity (the next one are TPMK+backdoor FD), and having only 5% of his range that is strong don't help i'm a lot.

These are simplification, it's not so easy, obvious set here is a lot better than TPTK, but you should count that raiser have also top set in his hand, overall I say that here raiser have a stronger range, or a range that can be profitably played aggressively.

Disclaimer: it's a complex topic, and I'm far from having well understood it, just trying to explain what I think, that can obviously be wrong

But it does matter that villain has initiative cause hes far more likely to bet. And donking specifically refers to situations when villain has initiative, so....

Fwiw, I think it's an OK strategy to almost never bet at an unraised pot oop either. This is pretty much the way I play FL.

Its not nearly as simple as you seem tothink to balance a donking range and keep your other ranges in tact. Especially at game speed and especially on certain textures. And this is without considering all the other benefits of not donking.