Hi Matthew.

You talk of the two main reasons we bet or raise, being making the pot larger for when we win, and denying our opponent's equity.

I believe there are scenarios in which a "blocker bet" would be profitable but the value doesn't really come from these two reasons.

For example, "setting our own price" with a draw to avoid our opponent pricing us out. If a small bet won't get folds then we aren't denying equity, and if our equity in the pot is <50% then the pot building is costing us more than it's worth. It might still be worth doing however to avoid being forced off our equity.

These scenarios are generally exploitative however. Do you think there is any "blocking" value in betting in a balanced strategy ? Should there be situations in which the value of our bet is derived primarily from avoiding facing a bet from our opponent ?

Thanks.

But you are, and the equity denied doesn't have to be on the current street, your small bet can aid in denying equity on a future street.

Great book Matthew. One question: I’m still not clear on why the first reason to bet is worded like it is. Like if we have any equity, won’t betting always make the pot bigger in case we win? For example if someone has a set, you know his cards, and you have a lower set, betting would still fulfill condition 1 (it clearly wouldn’t fulfill condition 2 but my point is that I don’t really understand how betting would ever not fulfill condition 1). Is condition one met more if we have more equity (so the “in case we win” chance is higher) or if the opponent is more likely to call (so there is greater likelihood of the pot getting bigger)? Could you give an example of when have some equity but condition 1 would not be literally fulfilled?

This is not necessarily the case. In my fictitious example I could assume my opponent never folds even own later streets. Then of course it could be argued my tiny bet is making the pot bigger in case I won, through the value of future betting if my draw hits, but with a small bet I would argue the blocking impact is still the main reason for the bet.

Don't get me wrong, I think the two reasons Matthew gives are an excellent was of summarising the main reasons why we bet. It is difficult to find examples to illustrate my point because they are so comprehensive. However I do think in many cases there is some value to avoiding facing a bet yourself.

Another example:
I am again heads up in a pot with a seriously flawed player. If I make a very small bet on the river with my weak made hand he will call with all better hands and fold all worse hands. If I check he will bet large enough with his entire range and I will be forced to fold. My bet will the highest EV choice because the equity I capture in the original pot more than compensates me for the potential lost bet.

I am not denying equity - my opponent is calling with all better, folding with all worse.
I am not making the pot larger in case I win - if I win, it is because my opponent folds hence I only win the current pot.
The entire purpose of my bet is to avoid being bluffed so I can realise my equity.

hey kids. just finished Part Four: Why We Bet and Raise. i understand the concepts and reasons, however, im having trouble figuring out how we might use this in the construction of a balanced range.

i have tried to construct some ranges for the hand example on page 44. please look over my attempt and help me to adjust it. thanks.

flop 8h 5s 2c

i assume our range is..

88-22
ATo-
K9s-K5s, KJo-K7o
Q9s-Q6s, QJo-Q7o
J9s-J7s, JTo-J7o
T9s-T7s, T9o-T8o
97s-96s, 98o
86s, 87o
75s, 76o

407 Combos (after the flop is accounted for)

Check-Raise:
sets: 55/22 = 6 combos
top pair: A8o, K8 = 21 combos
oesd: 76o = 12 combos
gutters: 97s, 96s = 8
total = 47 combos (11.5%)

Check-Call:
sets: 88 = 3 combos
top pair: all ones we didn't c/r = 57 combos
mid pair: 77/66, A5o, K5s, 75s = 27 combos
low pair: A2o, 44, 33 = 21 combos
gutters: A4o, A3o = 24 combos
total = 132 combos (32.4%)

this has us defending only 44% of the time. is this ok since we should expect the in position player with the range advantage to be able to auto profit some? or do we need to defend wider? maybe some hands like J9s/T7s with bdfd types?

is this c/r too much? is it c/r the wrong choice hands? thanks for any tips. im new to this.

also.. is c/r top pairs.. even just A8/K8 super exploitable? my pro friend says if he sees a player doing this, he will 3bet all his overpairs and several pure bluffs. would that work ?

That 852 board (and the potential check-raise with A8) has come up several times in this thread. I haven't got the book to hand and can't remember which bet-size BTN was using, but FWIW, if BTN c-bets pot after a 2.25x open, Snowie only continues with 40% (assuming 100NL rake).
If the c-bet is only 1/2 pot, then Snowie's BB only folds 33% (or even less if he raises small quite often). Facing a 1/2p c-bet, Snowie would sometimes check-raise with hands like 97s or even QJs. :/

With a solver (and slightly different ranges), you'll no doubt get different sizes, frequencies and strategies, but I think it's generally the case that the BB shouldn't defend at a particularly high frequency when OOP on "dynamic" boards where the turn card will usually greatly benefit the IP player, especially if the initial bet is large. It's gonna be really hard for the BB to get to showdown by calling down with weak hands. Note also that the BB got a better price on seeing the flop than the BTN did, so doesn't have to win quite as often to break even, so the BB is "allowed to" check-fold pretty often.

Hi Matthew,

Enjoying the book.

A question about using multiple flop betsizing depending on the strength of your hand. This idea is completely new to me, it seems tough to manage without getting too unbalanced, but I can certainly see the appeal vs players who won't check/raise enough.

Anyway, lets take the AT4s flop, BTN vs BB, and presume we bet big with the hands you suggest (AQ+ but not AA) and small with weaker Ax, strong Tx, and AA.

What I'm wondering is how to choose which bluffing combos go in which range.

My first thought is that our weaker bluffs should be bet for a larger sizing, since they benefit more from fold equity. So we could bet our offsuit gutshots with no flush card for a big sizing, but bet our gutshots with a one card flush draw for a smaller sizing, since they improve on more turns, and could potentially float a check/raise.

For flush draws, we could bet our weaker flush draws for a bigger size, and our stronger ones smaller.

So basically using weaker draws as more often "one and done" bluffs (obviously not giving up on all of them on the turn, but when we bet big we strengthen his calling range and so should be less inclined to bet the turn).

Am I thinking about this in the right way?

Do you have an example of this that's GTO based? You can justify almost anything from an exploitative perspective, but if you have a GTO example post away.

As explained several times in the book, it's not a binary thing. Pre-flop especially every time you 3-bet you are going to be denying equity and making the pot bigger in case you win. But that doesn't mean some hands aren't much, much better at achieving these objectives than others.

I'll use a medicine example. For administering drugs, the question isn't usually "Are you in pain?" because that's not specific enough. Instead you ask "On a 1-10 scale, with 10 being the worst pain ever and a 1 being just a minor discomfort or irritation, how bad is your pain?" and go from there. Pain is still super subjective, but nevertheless it's a question that's important enough that it can decide how strong of narcotic is given.

If that makes things clearer for you, do the same thing with poker. "On a 1-10 scale, how effective is potting it at denying equity with with QQ on a AK3 flop?" Meh, a 2 out of 10? Likewise, "On a 1-10 scale, how effective is potting it at making the pot bigger in case you win with with QQ on a AK3 flop?" Maybe a 1.5 out of 10 because you pretty much make all the worse hands fold but you can still hit the set or runner runner the straight?

Ok, well if you gave the two reasons for betting a 2/10 and 1.5/10 score, then the bet (with the given sizing) probably sucked.

Compare that to betting the JT on the flop. It's fantastic at denying equity (currently the hand is only jack high), and also if we improve it'll usually be to the flush or straight which makes near the nuts, so our EV will be very very high if we hit. So if we were to ask the same questions as we did earlier and grade them on a 1-10 scale, we'd score it much higher at likely conclude a bet is best.

I don't think there is an easy pattern to follow, or if there is I haven't found it yet. My guess would be no easy pattern really exists though.

So, my short answer to this would be "Most good bluffing hands are reasonable in either range and you'll want a mixed strategy with your gutshots, back door flush draws (BDFD), etc as both small betting and big betting ranges will need the ability to make a straight on a king turn (for example)"

My long answer would be is you look to see what hands you make call and which hands you make fold by betting 40% and what hands you make fold by betting 100% (for now let's just assume two bet sizes). You then look to see if any hands do particularly well against one calling range but not the other, and bluff them in the bet sizing range where they do well against the calls. Likewise, you do the same thing for folds. If one bluff for example has a pretty good chance to improve against the calling range when we bet 40%, but is almost drawing dead when we bet 100%, then bet it 40%.

This is obviously extremely time consuming the result is likely going to remain "a hand that is a good bluff will work well with multiple sizes and be part of a mixed strategy."

As always ArtyMcFly's post is on point, and he even went the time to go the extra mile based on rake. So thanks Arty.

Short answer to the bolded is "no." Medium length answer is "It won't work so long as villain didn't check-raise A8/K8 too much and instead use mixed strategies." Long length answer takes stack depth and more variables into account.

You're not wrong, but that's very exploitable. I don't typically go into detail with situations like this, because once you have very valuable exploitative information, the decisions are usually trivially easy and not something another player can usually learn too much from.

For example, say you're on the river and in position. Villain checks. You know he has air 90% of the time and the nuts 10% of the time and you have a "bluff catcher." If you bet 10% of the pot, villain will go into a DOTA 2 level of nerd rage because he (usually) has to fold to a bet so small and will play awfully for the rest of the night. But checking now has the highest EV for this one isolated hand. What should you do?

Is the above example interesting? In my opinion, no. Is the above example insightful for how to play better poker? No.

The reason I'm typing all this out is to show you're not wrong, but also why it's not something I really talk about. If you want to show a good example of betting where it doesn't really deny equity and doesn't really make the pot bigger in case you win, it needs to be shown from a GTO perspective to really provide interesting insight that we can learn from and use to advance or game and understanding. Otherwise we're just looking at isolated exploitative examples, where it's not hard to create ridiculous situations (not to imply that your example was ridiculous, but we could easily make some even more ridiculous ones up)

MoP proved that block betting (they call it preemptive betting) is part of the optimal strategy in the full street AKQ game with multiple bet sizings. In that scenario the K will sometimes be bet for small sizing (they use 1/4th pot).

Now obviously this is just a simple toy game, but there are river situations in the full game that are very similar to it. Most notably when you are in the BB and the river improves your range. You still have a big chunk of one pair type hands that would rather see a cheap showdown and not be faced with a huge bluff, you also have hands that became very strong and would rather not let villain check behind. In this situation I think it makes sense to develop a small bet strat with bluffs, weaker medium strength hands and the nuts.

Thanks for the reply!

Thanks for the reply Matthew.

I struggle to find an example of an unexploitable bet that derives its value from blocking rather than your two reasons, hence my original question.

However, a situation that I think does demonstrate the value of blocking would be, when we are semi-bluffing with a weak draw (that we would otherwise check-fold) then our bet has more value against an opponent who is likely to bet if we check than against one who will likely check behind, all else being equal. The pot building and equity denying components of our bet are the same against each opponent, however the alternative of checking is more valuable against the passive player (therefore the "blocking" value of a bet is less).

While this example is not from a GTO perspective, I think it does suggest that some of the value of betting in a GTO setting could come from avoiding having our hand turned into a bluff catcher.

Getmeoffcompletely, thanks for your reply. I will try to find my copy of MOP to check this out. Tipton also talks about blocker bets in his books. While these seem to be thin value bets, there does seem to be a "blocking" element to them (hence the name !).

Hey Matt if a short stack in a single raised pot shoves the turn 16bb into 20bb and his weakest bluff has 8% equity don't we have to call 100% of our range since 8% equity makes it 16 bb cheaper? He didn't really risk much money overall to get this bluffing opportunity but he did risk almost 40% of his stack up to this point. Is that enough for use to fold some hands in our range?

Why does 8% equity make it 16bb cheaper?

We may just have different definitions of what a "block" bet is. If you post an example I'll be happy to take a look.

I read in Applications facing a turn bet ip after calling a flop bet that every 1% equity the weakest bluff has when called makes it 2bb cheaper. So a gut shot having 4 outs with 1 street to go has 8% equity. Idk if that can be extrapolated to the situation I mentioned above.

Hi Matthew,

You mentioned somewhere the concept of having a sufficiently high barreling frequency on following streets.

For example, continuing to bet 66% of the time on rivers after a pot sized turn bet.

Does this mean

1. We should be betting 100% of our turn betting range on 66% of rivers?

2. 66% of our turn betting range on 100% of rivers?

3. "Most" of turn betting range on "most" rivers and having "some" rivers where we only bet "a few" hands from oir turn bettign range and "some" rivers where we bet "all" hands from our turn betting range (to the extent that we are still able to be balanced with those river bets)?

In summary when we say "frequency of river bets" are we referring to

1. Number of rivers (out of 46 possible) that we continue to bet on
2. % of our turn betting range that we continue to bet on any/all rivers
3. Combination of both

Hi Matt,

In one of Doug Polk's videos, he says "it's rarely, if ever, correct to bet with 2nd pair on the flop". And he seems to stay true to this thought and he pretty much always checks 2nd pair on the flop.
This seems to contradict some examples in your book, like betting 88 on a K73 (or something) flop.
I have a lot of respect for you both, so I'm curious about it. Basically I'm wondering, how sure are you that betting 88 on the above example is a reasonable option? (I'd ask Doug the same question if I could, fwiw). Thanks

^ At GTO equilibrium, second pair is sometimes bet, often as part of a highly complicated mixed strategy. It's somewhat rare for a solver (or Snowie) to suggest a 100% betting frequency with middle pair, so you're rarely making a mistake if you check it back. In spots where it's apparently a theoretical mistake to check, it's only a very small one.
On a human level (with none of the complicated mixing), it's quite easy to get into the habit of checking back all your mid-strength hands, as per Doug's 'post-flop engine' and general strategic advice, but the full solution to poker does include some 'strange' hands in a balanced/optimal betting range.
e.g. Just about all regs will routinely check back KK and QQ on Axx, but solvers suggest that sometimes they should be bet. I honestly don't think I've ever bet KK on Axx, and I don't think I've missed out on huge chunks of EV by not doing so.

EDIT: I just did a quick BTN vs BB spot with a K73r flop in Snowie. Here's part of the betting range:


As you can see, it recommends betting QQ at very low frequency (4%), with the number increasing as the pocket pair gets weaker (88 at 93%). None of those hands has a pure strat, so on a human level, you might decide to always check QQ and JJ, sometimes bet TT, and usually bet 99/88. You wouldn't lose any EV by checking them all, however, although it might make your range a little easier for a villain to read. One benefit to following Doug's advice (avoiding mixing) is that you develop a much better understanding of what your own range looks like on the turn.

For betting middle and bottom pairs, I believe someone already pointed out that in pio, if these middle and bottom pairs have backdoor straight or flush draws, they are bet at a higher frequency.

It ties in nicely with the

Betting to build a bigger pot in case we win

Logic that is presented in the book.

Also betting these hands ( specially bottom pair) can be handy if you get a runout where a lot of draws get there by the river. In these cases, bluffing with weak middle pairs or bottom pairs makes a lot of sense as they block some two pairs or slow played sets that will always call down.

Sent from my SM-N910W8 using Tapatalk

There is a world of difference between 88 and 75 on a K73.

I don't know Doug personally, but I know he is one of the best NLHE players in the world. Based on what I've seen in PokerSnowie, PioSOLVER, and common sense, it's reasonable to bet second pair on the flop provided the bet sizing is good.

For example, say you never bet 7x on a K73 flop. Is there a counter strategy for this? Yeah, probably. Because now when the turn card comes 7, your betting range won't have any trips, and your checking range will have too many trips. So that likely forces a mixed strategy.

But what fraction of 7x hands should you bet? Do you really need to worry about mixed strategies of 7x against 99%+ of players? Well, the answer to the second question is at least "probably not." So I don't know the context of Doug's original comment, but my guess is he was just trying to convey that is'almost always reasonable to check back middle pair, and it's hard to go too wrong with checking back middle pair (where you can definitely go wrong with betting middle pair, especially if the bet sizing is bad).

Hey Matt any insight on what I said above?