Hello I am DK. I used to play poker a little bit and now I don't so much but I still like to say words about poker. They may not be right, in fact they may be dead wrong, if they are please tell me ok.

I will try to write some words once a weekish or so until I run out of things to say. Today week 1 I would like to talk about Bankroll

What is a bankroll?

Pretty much everyone knows bankroll management is key to survival in poker. Not everyone actually practices it but they know. But almost nobody knows what a bankroll actually is. Huh? How can you manage a bankroll if you don't even know what a bankroll is?

Most players have some kind of intuitive idea about how much money they should have to play a stake. Savvier ones may use a calculator to figure risk of ruin. They might say "okay with $x k I have a 2% risk of ruin and that is acceptable." The problem is that those calculators assume we are emotionless robots. It is true that a player winning x bb/hr with y stdev has z% risk of ruin... if his winrate never changes. But it does. We are human. We get scared. If we dust off half our roll we feel bad and terrified. We look for safety. We might start avoiding high-variance spots, this hurts our WR. We now might not want to play in a loose, splashy game that plays big, this hurts our WR. We might feel bad after a bad session and not want to play, this hurts our WR. I am not saying this is a bad thing. It may be sensible if not natural. If you are broke you can't make money from poker, so winning less beats winning nothing at all. But when you make that RoR% calculation you are not considering these effects. If losing half your bankroll makes you play like a wuss, your bankroll is actually half of what you said it was.



Bankrolls are like onions. They have layers.



"bankroll is how much money you have to gamble right?" No!! It's more complicated than that. Let's break it down. The outermost layer is how much money you can afford to lose and not give a **** about. In an ideal world, that would be your bankroll. Think of the best poker player you know. He or she absolutely crushes, is an incredibly skilled player, and is always fearless. This is only possible if you are not at all concerned about losing.

But for most of us that is just not realistic.

The next layer is how much money you can afford to lose without it affecting your life. If you are a normal person, losing four figures in a night is going to feel bad. You could have bought something very nice for that much money. Maybe even taken a vacation! But if you are a responsible player, that loss will not change your life. And if you want to play live poker at all you are just going to have to stomach that variance.

Then there is money you can bear to lose, if you change your lifestyle. Maybe you have to cut back on other things. Maybe you can't eat out for a while. Maybe you have to cancel some services. Now that sounds pretty bad but maybe the risk was still worthwhile? That's for you to judge.

The point is that the definition of "bankroll" is going to differ from person to person. A consummate professional's bankroll should be one layer, but might be 2. A responsible, occasional player's bankroll might have 2 or 3 layers. A true degenerate's roll might be like 13 layers, including money he doesn't even have -- whatever he can borrow or steal.

I'm not going to tell you how to live your life. There are so many possible situations in poker. It might even be smart to degen it up, if the reward is great enough. In poker information is key. Knowing what your bankroll really is will allow you to make better decisions.

Hello it is DK again and I have some more words to write.

Today week 2 I would like to talk about Profiling

As you know profiling is when you do not know a player and you haven't seen him play many hands yet so you decide he probably plays a certain way due to being old or asian or woman. It sounds like a really horrible thing in this day and age but many poker players do it because those assumptions are right a little more often than they are wrong so it is a little more useful than assuming nothing at all.

Well what if I told you there is a better way?

Now don't throw profiling in the trash just yet. There are two aspects of it that are still useful.

1. Profiling is still useful for game selection. When you are looking around the room for a table to play at you don't have the luxury of watching a few hands to see how people play so physical appearance is all you get and usually you can just sort of tell if a player is not a grinder at least. There are two kinds of people who will play poker with $1000: people who win at poker and people who have $1000 to blow. Okay and I guess degens are a third group. Point is you should be able to tell them apart just looking at them. The reason this is only useful for game selection is that you don't just automatically win a bad player's money for sitting next to them. You do have to know how to exploit them. Sitting at their table is just the start.

2. Old people really are nits. Like that one is just almost always true. You should absolutely use that stereotype.

Okay so now what do we do instead? Duh. Assume random poker players play like random poker players.

Think about what that really means. What are the characteristics of the average random player?

- Probably calls too loose preflop
- Probably too passive in general
- Probably doesn't value bet nearly as often or as large as they should
- Probably afraid of having big hands cracked

Note that we know nothing about bluff frequency. Believe it or not this is just fine. Every player bluffs. Yes even the 96 year old who is tighter than a cat's bunghole. Maybe he only bluffs once a year but it happens. But if we have a bead on his value bet frequency, then whatever remains is his bluff frequency.

Here is an example. You just sat down at a 1/2 game, $300 effective. UTG+1 raises $12. HJ calls. You call on the button with 99.

Flop ($36) J 8 4
UTG+1 bets $20, HJ folds, you call

Turn ($76) 8
UTG+1 bets $50, you call

River ($176) 2
UTG+1 bets $150

Knowing nothing about UTG+1 I would call. I expect a random player to bet more with Jx or an overpair on a drawy flop to protect his hand. I expect a random player to be afraid of the 8x holding one pair. I don't expect a random player to raise UTG+1 holding 8x. I don't expect a random player to make such a large value bet attempting to get called by worse than one pair and I don't expect a random player to make his stack vulnerable holding just one pair, fearing a river raise. I expect his value range is {JJ} or 8x if he can have it preflop which I find unlikely. If we discount 8x to say 4 combos that makes six total. Meaning for us to call river he needs 3 combos of bluffs. Not very hard to find, how about KQhh, AQhh+? Every additional combo he may have is profit.

We're bluffcatching a random player we know nothing about, making a lot of unfounded assumptions along the way, without having any idea whatsoever about his bluffing frequency, for over 100bb, and I feel pretty good doing it. Because all these assumptions are habits exhibited by 90% of players. It's not at all unreasonable to assume they apply to everyone until proven otherwise. And I think it makes a lot more sense than calling because he's black.

Anyway, if you're scared of lighting your money on fire, try this fun exercise: Go to LLSNL, click on a hand history thread, and skip past the part with the reads. Come up with a decision for the hand and why and then go back and look at the reads and see if that changes your answer. Probably it will for a few hands but probably it won't for a lot of them as well, and that's the point: Making a reasonable assumption until you have something better.