Articles Poker Forums

you gotta narrow that shit yo!

also poin, this statement:


"Wait, I'm confused. Take the first hand. How can you decide if CTS is more likely to be value betting vs. bluffing if you don't assign ranges? I thought the way you figured out which was more likely was by enumerating the ranges, like "he'd value c/r with hands {1, 2, 3, 4, 5} and he'd bluff c/r with hands {6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20} so he's 4x as likely to be bluffing."

is not accurate. The reason being that hes 100% of the time vbetting hand 1-5, whereas his bluff frequency is variable. So, he could be bluffing 0% of the time with hands 6-20, or he could be bluffingi 100% of the time with them, in which case he's 4x as likely to be bluffing. My point is that there's an easier, more reliable way of going about this.

BalugaWhale-

I think this from my first post:

"you want us to think something like "well, villain has a decent idea of where *we* are, so if |B|/|V| was really that big, he'd be a spew monkey" and then re-adjust our idea of how exactly {R} gets split up into {B} and {V}."

was me sort of understanding what you meant in your first response. But now I think you're saying it's even easier: there are some situations where, if we really think about it, we can figure out that his range is almost entirely made up of vbetting hands, so we can just figure out our equity vs. them. And that can be true even if his vbetting range is quite narrow. Yes?

Also, your second post is obviously right. I left out the weighting in my post because I was trying to simplify things. Couldn't you tell by how simple my post was?

This article sucks imo.

"Well, the truth is that he has two different ranges. One is for value betting and one is for bluffing. If you can figure out which one is more likely, you can test your hand against that range. It’s only if you think that each is equally likely that you have to evaluate your hand against his entire composite range. Guess what? It’s pretty damn rare when it’s equally likely someone is valuebetting you or bluffing you."

1. He has one betting-range on the river. Of course, that range has two parts/segments: the hands that he bets for value and the hands that he bets as a bluff.
2. Do you mean with "which one is more likely" which part of the range is bigger / contains more combos? It's really weird to say "how likely is a range?". Personally I dont know what it means at all! Cf. saying "how likely is it that he has a set" or "how likely is it that he is betting a better hand"? I conclude that your use of the word "range" is very unusual in this article.
3. So let's say it is "more likely" that he is betting for value (that's almost always the case for a good player). Now you conclude that you beat only 25% of that betting-for-value-range. If I understand you correctly, you write that it's not necessary to think about his total betting-range, because it's more likely ("not equally likely") he is betting for value or something. So you fold, because he pots the river and you have to win 33% so 25% is not enough... But we both know that this is bullshit: if he bluffs only sometimes it's an easy call.

If your opponent bets the river, you are by definition playing against his complete betting-range.

I think what you want to say is something like this: at every decision point, the range of your opponent changes. He opens 15% utg (range changes from random to 15%), cbets 70% of that 15%, and reraises 30% of that 70% of 15%. Dont make the mistake to think that if the range op your opponent is loose at some point, his range is necessarily loose at another point. In some spots, players have no bluffs in their ranges etc.

I think poin was getting at the same points with his posts. Usually I like your articles Baluga, but this one is not good at all imo. Your use of the terms is non-standard and confusing (without good reasons afai can see) and it's not very clear what the message is.

You aren't saying that we should choose whether he is more likely to be betting for value or as a bluff, then eliminating the bluff (or value bet) range, right? If we decide that he is almost never bluffing, we figure how we fare versus his Vbetting range, then factor in the discounted chance that he is bluffing, right? For instance, he is most likely Vbetting, I am 25% versus that range, but there is a 25% chance he is bluffing, of which I am 100% against. Therefore I am .75(.25) + .25(1.00) = 44% to win when I call. It would be a mistake to not factor in some chance of him bluffing if indeed there is a chance, right?

hey teamdamsko,

sorry you didnt like the article. Let me see if i cant clarify my point a little bit--

The actual math of handreading is more or less impossible to do. Let's say im facing a bet that gives me 1.4:1 on the turn. If he has hand X, I have 10% equity. If he has hand Y, I have 40% equity. If he has hand Z, I have 70% equity. So, if he has hand X 50% of the time, hand Y 20% of the time, and hand Z 30% of the time, am i getting odds to call? I can't do that off the top of my head. Maybe you can, and maybe some people can, but I'm almost positive that most every successful poker player does this more by "feel" than by actual calculation. Now let's move to the river.
Obviously on the river he has one large composite hand range. In that range are two separate ranges-- one for value betting and one for bluffing. We agree on that. However, my point is that in the midst of this, he actually only has one hand. And its our job to figure out, to the best of our ability, what actual hand he has. When we look at it in theory, he should show up with a bluff a certain percentage and a strong hand a certain percentage. However, we also need to realize that he only shows up with ONE hand each hand. So how do we determine more accurately which hand he has? By starting with the "is he v-betting or bluffing" question and moving on from there.
Rarely is it the case that someone's entire range is ONLY value betting or ONLY bluffing. That said, I see people screw up weighting those ranges so significantly, that I think it's far easier to start from the question of intent and proceed to the question of weighted ranges, as opposed to skipping the first question and proceeding to the second.
Hopefully that makes more sense.

PS: I wrote everything but the last paragraph of this response a little tipsy, so if its awful I can try again later to make it make more sense.

colonel mustard-

of course. The point of the article is to say that if you try to weight the range before you identify likely intent, you often mess up the range weighting significantly.

Thank you. I have been applying this idea lately and it has been quite helpful.

Just joined today. One of first articles i read. And right from begining - on river beting only bluffing or trying to get value. Its not tru at all. You can make bet to block your opoent possible big bluff. You can bet for your image purposes. In hu sng you can make loose bet on river even if you almost know for sure you will be caled simply to set up your opoent for later when blinds will be biger and desicions more important for bouth of you. So... dont make so radical statements - its stpid.

piranija-
a few things
first, I'm not dealing with SNGs here in which it might be correct to play badly early on in order to win later when it counts. I'm talking about a cash game where its generally correct to play well all the time.
As for betting so as not to get bluffed, I don't believe this is a very legitimate reason. If you think youre opponent is likely to bluff a lot when you check, then you should check-call. If you think your opponent is only likely to value bet if you check, then you should check fold. If you have a hand that has no showdown value (say, 87 on a K9642 board)you're not betting to avoid being bluffed, youre betting to make a better hand fold.
There is a third reason for betting, called capitalization of dead money, which applies any time we bet any street ( to varying degrees), but it is usually a secondary reason and not a primary one like valuebetting or bluffing.

ok, thanx for answer. Its wery stright forward aproach but if its your way to see river bets - what ever works

My brain is trying to process this, but Im not sure I fully understand.I get the point that we cant assume he would be bluffing with 100% of his range that he got to the river with,but still how do we decide how likely he is to be bluffing? You mean we need to try and figure out a bluffing % and then assign a vbetting range which he will bet 100% of the time, and weigh them according to that?

Just joined today, first article I've read in this forum. Thank you, it was very insightful.

Disregarding the nitpicking about "complete ranges" and "split ranges", I think it's just useful to have another way to think about handreading. This concept of first asking yourself the question of "intent" before evaluating your opponent's complete range can only help the process.

ok, thanx for answer. Its wery stright forward aproach but if its your way to see river bets - what ever works

thats the whole point...its straightforward, which means its easy, which means $$$

Interesting article Baluga.

I wonder how does the idea of merging your ranges on the river fits in your view of the subject. Sometimes people value bet extremely thinly on the river in order to merge their ranges between pure value bets and bluffs. That should allow us in theory to be able to bluff more frequently..Let me know what you think..