Fundamentally, why do we bet multiples of the pot size pre-flop, and fractions of the pot size post flop?

We also bet fractions of the pot pre-flop.

SMH, okay, why do we typically raise around 200% pot pre flop and only bet 50-100% pot post flop as standard?

If we min raise preflop, we're betting 40% of the pot. Math is the same as check-raising postflop.
To overbet preflop we'd have to open to more than 3,5x.

Im not sure if you don't understand the question, or you are just straight messing with me, but I posted in Poker Theory because Im looking for an explanation of why our post flop bets are so much different in sizing than pre flop.
I understand that there are exceptions to the rule, Im just asking why the standard is 3 x pre flop followed by 2/3-3/4 post flop.
Preflop pot is 1.5bb

Standard open is 3bb (200% pot)
Standard flop bet is 60-100% pot

A 3bb open pre-flop is still less than a pot-sized raise (except in SBvBB, where it's exactly a PSR, since it gives the BB 2:1 odds). A pot-sized raise from any position except for the SB would be 3.5bb. If UTG opens for 3x, that's an 80% of pot raise.
FWIW, many strong players frequently open for 2.25bb (half pot) or even minraise. Post-flop, common sizes can be anything from quarter pot to 5x pot.

The optimal size depends on what the ranges look like and what you're trying to achieve. e.g. If you have 10bb in a tournament, you'll often open jam 3.6x pot with a wide range, because most of your EV comes from fold equity. (You can profitably play a wider range by jamming 10bb, than than by minraising from that stack depth, since a minraise doesn't steal the blinds very often). If you have 200bb, a smaller size (i.e. not open shoving!) will make more sense.

EDIT: More generally, I think 3bb became a "standard" opening size, because it was the 'happy medium' between going "too big" and not getting action, and "too small" and getting called by multiple players. In short, it thins the field in a way that is beneficial to the raiser. Once you see a flop, you're usually heads up, so you don't need to bet big to disincentivise multiple players from calling, because they already folded pre.
In soft/passive games, picking a larger opening size might be appropriate. In tight/aggro online games, smaller sizes work better, since players know they should still fold pre with junk hands even when facing small raises.

It's about the way strong hands reduce the profitability of the opponent's draws by betting or raising.

AK vs 87s preflop? not sure I think it's like 60%/40%. It takes a big bet to reduce profitability of 87s effectively.

Strong top pair vs flushdraw on the flop? probably like 65/35%. It takes a slightly smaller bet to reduce the profitability of the flushdraw effectively.

Strong two pair vs pair+flushdraw on the turn? probably like 75/25%. It takes a slightly smaller bet still to reduce the profitability of the strong draw effectively.

I think because of how close equities run preflop as opposed to postflop.
See Bob's post previous to mine.


Making a re raise opens the action again and allows the original bettor the opportunity to push you off your equity when they wake up with something. So whenever you make a re-raise (pre or postflop), the fold equity / value you get should be sufficient to counteract the fact that sometimes you will get raised off your hand, and that villain has good odds with some of his hands.

This is a darn insightful question.

Intuitively I would say you won't fold out many weak players with a minraise pre - which opens the door to getting outdrawn with ludicrous hands.
If you minraise pre you might as well limp because many weak players are then playing bingo and not poker. They're not thinking long term but simply "I'm getting to play/see a flop for just 2bb instead of sitting on my thumbs for 2 minutes - that's cheap fun! Maybe I'll even get lucky!"...what exactly would that accomplish? Even with aces you (or at least I) don't want to play against 7 people post-flop.

With really novice players (like the guys I started a home game with - where none of us ever had played before) you need to step it up to 4/5bb (or even higher) to fold out anything at all.

With a minraise you are, as an average player, just inviting expert players in who will outplay you post (and know it). I guess they'd not even look at their cards if they're in position. A 3bb raise limits that edge somewhat.

So I think the 2.25-3bb range seems like a good compromise between making all player types fold weak hands (so you can start putting people on hands...which you can't do if you just invite everyone in), average players continue with weak-ish hands, and also setting a foundation for the exponential rampup if you want to build a big pot.

Another factor is that you get to see 3 cards with the investment...so I think it's only natural that - proportionally - a player would be prepared to invest more for the rendered information.

It also occurred to me last night that preflop raising is charging the opponent to see three cards, while flop and turn betting is charging the opponent to see just one card at a time.

Interesting. How does the pre flop range equity change if the game were to flop just one card, two cards, three cards, four cards, or all five cards?
I know that people tend to study pre flop range equity in relation to showdown value, but i wonder if there is a linear relationship between the equites as each individual card is shown.

(Apologies if my terminology is incorrect, but I assume you get my gist)

Thank you all for your responses. Im humbled to hear from Arty, Bob, and Robert - 3 of my favorite posters here!
The reason for my initial question is that I am trying to figure out some un-exploitable system which would allow me to frequently change my RFI bet sizes in a way that would be difficult for my opponents to figure out why. It doesn't need to be a +EV strategy, so long as doesn't define my range in any way.
I have experimented with adjusting my RFI sizing based on my position at the table, but this is quite easy to pick up on. I have also picked different RFI sizes at random using highest hole card suit, or the minute number at the time I bet. I found that incorporating different seemly random RFI sizes into my play for part of my session really seems to throw a lot of people off their game, and they make higher frequency post flop mistakes as a result, especially people who are multi-tabling as they have to do a bit more investigation than usual.
i assume because they think that there is something unique about my range given my unorthodox action prefop. Ill play a session like that for a while and then go back to more conventional sizing. It seems to work out well for me, so if anyone has any good ideas or links to other discussions about this that would be really helpful.

hi,bro,What software should I use to build preflop range with straddle?thank you?

No clue. Sorry.

If you open raise with a larger sizing when UTG, when your range is more clearly defined, how can your opponents exploit this? You are charging them extra to see a flop in exchange for this information. I would vary the size larger for looser tables, start at about 5x if this gets two callers or less. You could mix up the sizing a little to appear random, but really your range of hands is what matters, and you need your bet size to fold out a lot of equity, or force villains to pay to play. You could do this for the first two seats, then go 4x the next two seats, 3.5 the rest. This sounds very balanced to me, now all you need are hands for each range.

The preflop equities are based on making the best 5 card hand out of 7 cards. If you change the game so that only three board cards come out? Then the value of high cards goes up and the value of drawing hands goes down; it's harder to make a straight or flush with 5 cards compared to 7.

If we change the game to Cincinnati, which is holdem with six rounds of betting and one card coming out at a time, then the correct preflop raise size should be smaller in theory. With one card coming out at a time, it's much harder and more expensive to draw out vs a stronger hand. In Texas holdem, you can flop a straight, a flush, or a full house. In Cincinnati, you cannot.

Look at the other extreme: two hole cards to start and all 5 community cards come out at once in a megaflop, so to speak. I would think that this game features large preflop raises in the effort to protect vs drawing hands.

Ive never heard of Cincinnati, but will read through some posts/articles later tonight. Sounds like and interesting structure.
Is there a poker game which falls in the middle where each player gets 2 hole cards and there is a preflop round followed by a 3 card flop with no turn or river?
I think id be interested to see how the hand equities differ pre flop between flop only starting hands vs starting hands with 7 community cards. Im sure someone has done this before. Any idea where to start looking?

Not sure if it's played somewhere under a different name, but the creators of Libratus use this game as an example in their paper. They refer to it as "No-Limit Flop Hold'em".

https://www.cs.cmu.edu/~sandholm/saf...ed.aaa17WS.pdf

Very interesting read, thank you nolispeifaflaatoi.

Just want to circle back a little bit. Firstly, I really think you are doing great by changing things up in your game in specific ways just to see what happens, assuming you are playing small enough stakes of course. This is a good way to incorporate game theory into your game. After all, you can only think and talk poker long enough before you have to test and prove it on the table.

However, be careful with drawing conclusions from small samples. Poker can be quite mesmerizing over small samples. Also, sifting through a large database of hands to find patterns can be equally misleading. This is where game theory helps, to keep the ship upright in the storms of chance and variance.

So, you want to be random in RFI sizing and throw off your exploitive opponents. This makes them think more often, possibly giving up or making mistakes if the opponent is multi-tabling. Great. However, everything has plusses and minuses in poker. What if opponents don’t care or notice the different sizings?

Lets suppose that part of the random sizes is to open for less than normal size. You may in fact not be doing this, but lets just suppose. This *should* lead to more calls or more 3-bets from opponents. Now we have not altered our range at all in preparation, but are facing more multi-way pots and being tested by 3-bets more often. I know you said slight -EV is ok, and I actually like your thinking but this is not the way to do it.

Lets suppose that part of the random sizes is to bet larger than we normally would. I suspect this is more likely what you meant anyway. So what do we give up in exchange for more folds from opponents and more likely winning more blinds uncontested? Well, when opponents call or 3-bet we are not prepared for this and our range is not as strong as we need it to be. Especially when oop, this will definitely be -EV depending on how you continue in the hand if oop.

If you want to randomize your sizing in exchange for a tiny amount of EV, then just literally randomize, but on the higher side not lower, and only a small enough amount to be noticeable to villains who care, but not detrimental to the dynamic of the hand versus villains who don’t care. You could flip a coin and add 1 or 2 BB to any RFI and be totally fine (and give up some EV).

But what I really want to encourage you to think about is this:

If you are willing to give up a little EV preflop and recover it later in the hand what is the best most cost effective way to do this? Humans are crafty exploiters, even the ones that are in a player pool that is selected specifically for being “beatable”. They will just make a note that you are “random RFI guy” OR WILL NOT CARE ANYWAY. Either way your plan could be wasted.

But what about your hand selection in each seat? What if you randomly infrequently insert a speculative hand into that otherwise narrow range which you are trying to avoid anyway? If you pick the right hands, which should be chosen as “easy to play” post flop, this could be very costly to that 12-table reg that has stats that assume what you can and can not have in a particular seat. Also, since you are the raiser, play these special random hands more liberally post flop especially when card removal gives you a hint to lead and c-bet the flop with an aggressive size bet. Now you are capped way less often, and have prepared your range to connect on more flops. Now you have greatly increased your chances of getting a fold from villain, or playing a larger pot with a hand that can surprise and stack off a reg.

Of course, the logic goes full circle. If these hands work well this way, why not add them permanently? In the long run they are break even at best and probably lose some money by themselves. They would not pass muster when you sort your database by EV per starting hand. That is not why they show up randomly, to just lose a little money. They shore up your entire range and force villains to play you differently. They *allow* you to make those larger and more frequent bets BECAUSE YOU ARE NOT CAPPED.

As usual, only try this at small stakes you are willing to ride some swings with, even go down a stake to have fun with it.


Anyway, maybe this wall of text is TL;DR.

What about your original question itt? Why do we go from full pot bets preflop to making 1/3 to 1/2 size bets postflop? Do you have some ideas or are you genuinely just asking the question?

Not sure if earlier posts clarified this for you, but your calculation is wrong here. By this logic, if someone bets $4 into a $6 pot, then raising to $6 total would be a pot-sized raise, when in fact that's not even a legal raise size. The best way to think of it is to think of a raise as calling their bet and placing another bet on top of it into the new pot size without the dealer dealing another card (or as they used to say in olden times, "seeing" your bet then raising). So you "see" the $4 bet, making it a $14 pot, then you raise another $14 for a $18 raise total.

Long way of confirming/explaining ArtyMcFly's response to you that a $3 raise is less than a pot-sized raise (the pot is 2.5bbs once you "see" the bet, and then you're adding 2bbs on top).

In any case, in modern-day online poker (where games are more theoretically oriented than live games), the biggest sizing differences are between flop sizes and river sizes. There are also big differences between RFI sizes and 3b sizes. Without giving a complete explanation of all the justifications for every size on every street, I'll illustrate some of the main factors at play here:

1) Range polarity. This is the most fundamental concept for understanding what type of ranges prefer to bet large versus what types of ranges prefer to bet small or face small bets. A range largely made up of draws and marginal showdown hands wants to see more cards and/or get to showdown as cheaply as possible. A range that includes both strong value hands and air wants to bet large with both poles: the value hands want to extract maximum value and the air benefits from large bet sizes because it gives the opponent worse pot odds, making for a higher bluffing threshold.

Note that some of these concepts are meaningless on the early streets (there's no such thing as "air" preflop, for example), and other concepts are meaningless on later streets (there's no such thing as a "draw" when there's no cards left to come.)

2) Range definition. A player's range before raising first in is any two cards, and he is opening that range into various players whose range is any two cards. When a player 3bs, for one, they are making that choice when the alternative was to flat, which tends to polarize their range for going with the more aggressive option. Also, they are making this decision into a player who has defined their range by raising, which could be anywhere from the top half of hands (the button in a 3+ handed game) to the top 10% of hands (early position of a full-ring game).

You may notice this means if you have an open limping strategy, opening larger becomes a more viable strategy because your RFI range consists of less of the hands that perform poorly facing a large 3b (eg: in middle position of a full-ring 100bb+ game, this might include offsuit broadway hands, loosely-connected low-rank hands, and low pocket pairs).

Similarly, OTF, while there's usually a player with a generalized range advantage and a generalized polarity advantage, both ranges still vary widely and often include anything from the nuts to air and a variety of hands in-between. By the time you get to the river, oftentimes one player has been the aggressor throughout the hand while the other player has been on the defensive, which (depending on the runout, bet sizes, etc) tends to make one player's range strongly polar while the other player's range is decidedly medium-strength.

At the risk of venturing a bit away from theory, the above is the reason that it's a reasonably consistent heuristic that the first postflop bet made by any player in any spot where ranges are reasonably varied (eg: SRP involving players from mid-to-late position) is generally going to be on the smaller side--ie: rarely exceeding half pot, and generally being smaller than that. There tends to be a larger array of "motivations" (again, sorry for the non-theoretical language) for the bet that might include protection, thin value, or even just "maintaining initiative" (taking a line that keeps hero's range uncapped and making the decision tree larger for villain so that hero's range is more polar on later streets and villain's range is more middling) in order to keep bluffing options open for later streets, etc. Whereas on the river (particularly OTR with the option to end the hand with a check), betting is generally for pure value or for a pure bluff and simply checking with any hands in-between has no drawbacks.

3) Equity realization. Preflop is by-and-large a slightly ahead/slightly behind scenario, and allowing a player to call gives them an opportunity to realize much of their equity because they get to see three cards before you can bet again. Consider how this matters in combination with range definition: when you 3b a HJ opening range from the blinds, much of their range is ahead of AKo in terms of hot-and-cold equity, and even against AA, those hands get to see 3 cards, any one of which can suck them out on your preflop nuts; it takes a much more preclusive price to mitigate this preflop than it does with a flop bet.

On the river, while it is an absolutely ahead/absolutely behind scenario, a player realizes their TOTAL equity with a call. On the flop, there are still two betting rounds left before showdown and a call only allows you to see one more card.

Similarly, OTF, while there's usually a player with a generalized range advantage and a generalized polarity advantage, both ranges still vary widely and often include anything from the nuts to air and a variety of hands in-between. By the time you get to the river, oftentimes one player has been the aggressor throughout the hand while the other player has been on the defensive, which (depending on the runout, bet sizes, etc) tends to make one player's range strongly polar while the other player's range is decidedly medium-strength.


Good post!!

One question-I don't understand the term Generalized Polarity advantage as used in the above paragraph? thanks for any response

It really is all about the possible branches in the game tree. Through trial and error it seems that 2-3BBs seem to work preflop. I think an important factor is the pricing out of the BB.

When the BB is facing a 2x open, he can call quite a lot of hands. When facing a 3x open, he can call a moderate amount of hands. When facing a 4x open, he can choose to be patient and wait for good hands to play without making much of a mistake. With a limp, he can play everything.

When people make abnormally large preflop bets like 6x or 8x, I feel they are very easy to play against. You can choose to be patient and wait for a good hand to play back at them and you won't be making a mistake. When the sizing is 2x-3x it sometimes forces you to play.

Let us pretend that there is no SB and BB and everything is just ante'd. In this case the correct play may be to just limp. Thoughts on this?

I once played a heads up poker variant where the best play was to check everything while OOP. The IP player would then be correct to bet 3xish to 4xish with everything. The OOP player would then have to fold, call, or raise depending on his hand. The positional aspect in this game was such an advantage that it was profitable to raise all hands. 3x-4x was that sweet spot of indifference without risking too much. The game tree just worked that way.

Nice post. If no blinds and just antes then the correct open size is still a function of the odds required to call, and thus still a function of the size of the pot. Most ante only games are limit, such as stud and the position changes based on up cards.