View Single Post
Old 02-27-2014, 09:51 PM   #1
gamma001's Avatar
Join Date: Dec 2011
Posts: 4,145
3k post: Overbetting.

The topic of overbetting has been widely overlooked both on the table and in the forums for some time. How many times do you even see an overbet at the tables? Once a session? Maybe twice? Not a lot anyway. Hopefully this post will help you see the advantages of overbetting, help you apply it to your game and understand how to defend against it.

So first letís have a brief overlook at why we bet. There are three reasons:
-for value: to get worse hands to call. Pretty obvious.
-as a bluff: to get better hands to fold. Also obvious.
-for protection/fold out equity/pick up dead money: not as important as the first two and doesnít really apply to large bet sizes since our range is polarised so it wonít be discussed any further.

When we overbet, our range will become polarised to relative nuttish hands and bluffs. On the other hand, Villainís range is mainly going to be bluff catchers i.e. will lose to our nutted hands and will win against our bluffs.

So why would we overbet? There are a number of somewhat interrelated reasons:

-Firstly, the ratio of value bets to bluffs will decrease as our bet size increases, meaning we can bluff more the bigger we bet.

-Furthermore, when villainís range is capped we can overbet to put him in a tough spot with his entire range. For more on capped ranges read TDAs great post on hand reading.

-Additionally, when we have a polarised range we maximise our Ev by getting all in over the three streets with equal sized bets on each street. For anyone who is interested and wants to read further this concept was described and derived in the Mathematics of Poker.

We can work out what percentage of the pot we need to bet to end up all in on the river using this formula **: (final pot-size) = (starting pot size)*(pot growth rate) ^(streets remaining)

So take for example we are on the flop with 7bb in the middle with a 100bb starting stacks and we are HU. This reduces the equation to 200 = 7*R^3 and with a little manipulation we can work out that R = 3.06. This means the pot should grow 3.06 times its size on each street i.e. we should bet 1.03 PSB (1.03 + 1.03 + 1 = 3.06) on each street. That formula is very useful so take note.

So whatís this about our value to bluff ratio? When we bet we offer villain odds to call. The odds offered define our value to bluff ratio from a theoretically optimal standpoint. If the pot is 10bb and we bet half pot villain is calling 5bb to win 15bb which means for him to be indifferent to calling with his bluff catchers he must win 25% of the time i.e. we must bluff 25% of the time and have a nut hands the other 75%. (being indifferent means it does not matter what decision villain makes, both have the same Ev)

However, there is an easier way to think about it:

Whatever you bet is the number of bluffs you should have and whatever the pot is AFTER you bet is how many value hands you should have. So letís say you bet half pot for 5bb; that means you should have 5 bluffs for every 15 value combos, or 3:1 value to bluffs. When you bet pot for 10bb that means 10 bluffs for every 20 combos of value or 2:1. Pretty easy right? Now what about overbettingÖ If we bet 2x pot for 20bb then we have 20 bluffs for every 30 value combos. Hopefully you can see where this is going... as our bet size goes to infinity our value to bluff ratio goes to 1:1.

You might say why does it matter if we can bluff more? The ability to be able to bluff a more in a balanced range increases the Ev of the whole range. We can take hands out of our check fold range and add them to our betting range. Letís further explain:

When we bet a balanced range on the river we automatically win whatever was in the pot. This may seem strange at first by bear with me. If we bet a particular size, take pot for example, and balance that with the correct amount of value bets and bluffs (2:1) villain becomes indifferent to calling or folding with his bluff catchers. Folding, by definition has an Ev of zero therefore calling also has an Ev of zero. Hopefully that made some sense. Now...

At any given point we share the Ev in the pot with the villain. If the pot is 50bb and we have an Ev of 20bb villain has an Ev of 30bb. Thus if villain has zero Ev we automatically win the entire pot. On the river, if the pot is 50bb and we bet a balanced range our Ev is 50bb against an opponent and his Ev is always zero (assuming he has only bluff catchers). I hope that made sense because itís pretty important .

The larger the proportion of our range we can bet the more often we will win the pot. Thus, it is ideal for us to bet as large as possible so we can bet as many bluffs as possible. However, if we bluff more than is theoretically correct Villain can just bluff catch 100% of the time and we lose with our bluffs. If we bluff too little villain can just fold his entire range and we lose value with our value hands. Therefore it is optimal for us to bet a balanced range so we canít be exploited by an optimal opponent.

Thatís probably enough theory for now. Let us not forget that being balanced is pretty bad idea in the majority of situations, especially at the micros and small stakes. It is a much better idea to play exploitatively with the aim of taking advantage of villainís unbalanced lines and poor tendencies. But that does not mean we can take good insight from theoretically optimal play. The most important thing to take from the theory above is: when villain is capped and our value hands are rarely beat, we should be big, very big.

A few examples of good situations to overbet:

We open from CO with 8 8 to 2.5bb and get called from the BB.

The flop is Q T 8 with 5.5bb in the pot.

We bet 4bb and he calls.

The turn is the 2 with 13bb in the middle.

Using the bet size formula above we get 200 = 13*R^2 which means R = 3.92. That implies we should bet 1.46 times the size of the pot.

We bet 19bb and plan to ship blank rivers.

Why is this good spot to overbet? Villains range is capped to one pair hands since he would be raising QT, T8s and sets OTF a very large percentage of the time. The 2 didnít improve any part of his range. We would also want to overbet draws as well as sets and possibly two pair to balance. You might think that we aren't getting value with our sets since villains range is weak. But the benefit of being able to overbet our unmade hands increases the Ev of our entire range (and if villain never calls we can always exploit by decreasing our value to bluff ratio i.e. bluffing more than we should theoretically be able to).

Another example:

We open 2x OTB with A 5 and get called in the BB.

The flop is K 8 7. We bet 3bb into 4.5bb.

The turn is 4 with 10.5bb in the pot.

As before, using the bet size formula we work out we should bet 1.68 PSB.
We overbet 17.5bb into 10.5 bb pot with the plan to jam river.

Again, this is a good spot to overbet because we block the nuts and have plenty of equity. Typically, most villains will be raising most possible flush draws besides 7xdd of which there aren't many combos of. Axdd would also be a check call quite often from villainís perspective but we obviously block that.

Another good scenario to overbet is in 3bet pots as the PFR when we have a big draw. One of the worst situations any poker player can face is being forced to fold a big draw after barrelling the turn in a 3bet pot because we were raised and aren't quite getting the odds to call. Take this hand for example:

We have Q J and 3bet a CO open to 10bb and get flatted 100bb effective.

The flop comes K 8 2 with 21bb in the middle. We bet 13bb and get called resulting in a pot of 47bb.

The turn comes 4 .We bet 26bb and we get shoved on.

We would be calling roughly 50bb to win 150bb which means we need 25% equity to break even, but we only have 20% equity .

We have to fold after putting almost half or stack in the middle. Disaster! There are two ways to deal with this problem. We can check-jam which ensures we realise our equity (but this can be difficult to balance). Or we can make sure we bet all the money in OTT by overbetting. Betting 77bb to win 47 is quite a large overbet so it might be preferential to bet slightly larger OTF so our turn bet isnít quite so big.

The final example is of a hand I played recently (this one is pretty simple ).

Poker Stars $50.00 No Limit Hold'em - 6 players -
BTN: $39.75 - VPIP: 43, PFR: 27, 3B: 4, AF: 1.4, Hands: 84
Hero (SB): $89.45 - VPIP: 21, PFR: 18, 3B: 8, AF: 3.2, Hands: 765669
BB: $69.49 - VPIP: 23, PFR: 5, 3B: 0, AF: 0.9, Hands: 86
UTG: $50.00 - VPIP: 21, PFR: 16, 3B: 7, AF: 3.7, Hands: 616
MP: $9.98 - VPIP: 0, PFR: 0, 3B: 0, AF: 0.0, Hands: 6
CO: $56.80 - VPIP: 18, PFR: 15, 3B: 5, AF: 2.2, Hands: 807

Pre Flop: ($0.75) Hero is SB with T K
2 folds, CO raises to $1.25, 1 fold, Hero calls $1, BB calls $0.75

Flop: ($3.75) 7 8 J (3 players)
Hero checks, BB checks, CO checks

Turn: ($3.75) 9 (3 players)
Hero bets $2.25, BB calls $2.25, CO folds

River: ($8.25) Q (2 players)
Hero bets $85.95, BB calls $65.99 all in

Final Pot: $140.23
Hero shows T K
BB shows T A
Hero wins $117.77
(Rake: $2.50)

The reasons for overbetting here should be quite clear so I wonít dwell too much on it. Villain is quite obviously a fish. Not a drooler but still not a good player in any respect. We have the nuts and villain is NEVER folding the second nuts. So overbet! Doing anything else is leaving money on the table.

Poor situations to overbet are when villain is uncapped. This will occur quite often on dry boards such as K72r or A64r where villain will slow-play his sets and two pairs as a standard to give us a chance to continue bluffing. It also may occur in semi-wet multi-way pots where villain will call hoping to keep a fish or weak reg in the pot with a big hand. Just keep in mind that not all boards provide profitable overbetting opportunities.

Defending against over bets is pretty difficult. This is one of the reasons why it is such a powerful strategy when it is correctly employed. If youíre being harassed by overbets from a particular reg you should defend by ensuring your range is strong on run outs where villain is likely to overbet. This may mean having to slow play sets on wet board. This is something we generally will avoid doing since villainís bluffs have significant equity versus our range. However, when there is a high probability that villain will overbet, calling is worth the risk. Albeit, in some cases a turn or river is just poor for our range and good for villain's which means we just have to acknowledge that fact and call with the best of our bluff catchers. Itís not ideal but we do have to call to prevent villain from being able to bluff recklessly and print money versus us.

Just a few general notes to finish off:
-do not overbet against fish as a bluff unless you have a read that he folds to frequently. Putting large amounts of chips into the pot with a low equity hand against someone that doesn't fold is a no no. Overbet for value vs fish

-Against a good reg who you assume will call your overbet with a somewhat correct frequency you should be a balanced range. Balanced doesn't mean we have EXACTLY the right amount of bluffs and value bets, it just means our range isn't completely out of whack.

-Against reg who will react in an exploitable way to overbets (this will mostly be fold too much) we should take advantage of his leak adjusting our value to bluff ratio (usually by value betting less and bluffing more).

Thank you for reading that wall of text and I hope you took something useful from it

**Referencing Applications of NLHE by Mathew Janda.
gamma001 is offline   Reply With Quote