Wednesday, June 11, 2008

Poker variance

I had an interesting theory about poker variance and tournaments.


Lets say you get dealt AKs 10 times, and your opponent gets all in with QQ each time. You win 5 he wins 5, and that's roughly the expectancy of AKs vs QQ. Now realistically each time this happens it's 50/50. Over the long term, it will even out, but in the short term it can go either way. Even then, the most likely scenario you see in poker, is I win one, then he wins one, then I win two, he wins one, I win one, he wins two etc. It generally doesn't venture too far from overall expectancy even in the short term.


Occasionally though, a hot streak happens one way or the other. Suddenly I win 5 in a row. Then at another occasion I lose 5 in a row. Now if I don't ever tilt, and neither does my opponent, we see this as variance and understand for every 10 of these situations that I get, I will win 5. So variance to me is simply, I will get this back later, when luck runs my way for a while.


In tournaments these variance streaks can make or break a tournament. If I win 4 AKs vs QQ all ins in a row, I'm obviously running above expectancy. However, winning these 4 in a row, will mean I'm probably getting myself ready for a final table and really cashing well.


So what does standard poker variance mean to a cash player and tournament player. To a cash player it's simply a way to experience tilt control, to a tournament player it's something you will need to win a big tournament. Interesting way to look at poker variance.

Thursday, June 5, 2008

Stop Loss

A small note about stop losses then. A stop loss is a rule for stopping when you are behind by a certain amount. As a general rule of thumb someone might set 2 to 3 buy ins (BI) as a stop loss. The actual size will vary a lot depending on a lot of factors of course.


Should you have a stop loss? I never believed so. I always thought as long as you are still playing well, and the tables are still in your favour it's OK to keep playing, but I've come to realise it is an important concept. Here are some reasons why a stop loss can be good:

  1. We won't often bounce back fully, meaning we will be down for the night regardless of when we stop now or not
  2. While anywhere between 1 to 3 BI's can be upsetting to lose, beyond that can be much more damaging for the next few days.
  3. If you end up having a big bad running period, you can end up cutting your losses a lot better this way. It will mean that when you get a solid winning streak going you can get ahead a lot quicker than you fall behind on your losing streaks.
  4. It gives a bit more time to analyse your bad plays if that's why you are losing.

There are probably dozens of other reasons, but two sessions spring to mind. These are other players, but they prove the point perfectly. The first one was a solid grinder at CardRunners. He was really crushing the game, and only a very short time away from reaching NL100, already bankrolled for it. One night he lost 3 BI's at 6 max and 7 BI's at heads up. 10 BI's overall, he basically quit poker for a while. He couldn't face playing again in the short term. His next session was 2 weeks later, and he played 50 hands and stopped again. He admits he felt like he was playing OK at the time, so he kept playing, but the real damage was done after seeing how much he'd lost when he finally stopped the session.

The 2nd is a player who had just moved up. He had a rough night, lost 9 BI's and posted all the big pots on a forum. The hands seemed to start off as marginal, but end up as a few ugly calls and shoves. Again even though his mind wouldn't admit it, he was chasing his losses, and trying to hit a big pot to reverse the damage, but in poker, you just can't force the money your way. You have to choose spots to be overly aggressive, and his were bad spots due to the fact he just wanted to start winning some big pots. He was forced to also take a break, and start again at the lower limit, which again has a big impact on your mindset.

Tuesday, June 3, 2008

Poker - Getting Reads

An old archived message I'd heard before pops into my head today. "If you don't know who the fish at the table is, then you are the fish." Only problem is, my fish finding skills are poor at best. Sure I see someone minbet a couple of times, or goto showdown with ace high in a big pot, but I can never sit down at a table knowing who the fish are quickly.

So lets look at what a fish is first. A fish is someone we can more easily make money out of. At lower levels, this is someone making a lot of mistakes, at higher levels this is probably someone with exploitable tendancies. This might mean, we can always bluff them. It might mean they bluff too often. It might just mean they overplay their weak hands or slowplay their monsters too much. Basically we have more than a few spots against a fish where we can expect to make money. If our game is well rounded, we could probably expect to pick up a lot of money off fish when we have position on them, especially if they are the type to bluff too often. The higher the stakes are, the less holes a player needs to be considered a fish, so we might consider someone a fish simply for folding to often to 3 bets at higher stakes.

So how to exploit a fish. The first step is to understand it is +ev to play against fish, but that doesn't mean we win every pot against them. Sometimes it simply means they always tell us when they have a big hand, at which point we could fold easier than normal. You have to use the reads you have, and compare them with the current situation. If he shows strength, you have to be able to fold. If he shows weakness you have to be able to make a move. But don't just expect to win a pot becuase the fish is up against you, still use your poker skills to not pay him off too.

Lets take a look at getting some reads going to help find a fish. The purpose of a read is to know how a villain plays a certain type of hand, so we can get more information about them for later hands. Here are a few things I'm going to be working on to improve my reads:
1. Situational lines. A line is how you played 3 streets with a certain hand, and the parts that make that up. If someone calls the flop, then the turn, then bets the river strong and shows a completed flush on the river, we can see how they like to play draws. If they call a cbet with a flopped set, then check raise the turn, we now know how they play big made hands. Even if you can't watch a pot, becuse you are playing another table, lookup the HH to get a better idea of how big pots played out.
2. Bet sizing. It takes a bit of time to get real information in big pots, but especially smaller pots, it's very easy to get an idea of what someone bets bigger with and what they bet smaller with. A good first step is what do the minbet with (if they ever do)? Then what do they bet full pot size with? This can go either way, so make sure you get these reads.
3. Timing tells. Someone calls pre flop out of position, heads up. The flop comes out A52r, and he checks very quickly. If he folds to a cbet, then his instant check means weakness, if he calls or raises, it means he's trying to get a bet out of us by quickly checking to us. Again if it's a long wait until he checks, then throw out a cbet and find out what that timing tell is. It only takes one or 2 cbets to work out when he does what.
4. Misplayed hands. Once you see a hand showdown, you notice one guy played his hand very poorly. Note down his mistakes, and use that information later, and also to identify him as a potential fish.5. Activity. Obviously the big 3; VPIP, PFR, AF. Very basically VPIP is how many hands a player plays, PFR is how often they raise pre flop, AF is how often they bet as opposed to call post flop. Pokertracker lets you have these stats live, so you can see just how active your current opponent is. If they are not usually that aggressive and are suddenly using big 3 bets on the flop, it's very likely they have the goods, regardless of what other reads are
telling you. A couple of other good stats to have include cbet % and fold to cbet %.
6. How do they react to re-raises? Do they fold too often, or do they call too much? Both of these are vital to know, as we can easily exploit this, or avoid 3 betting them if we know they won't let go of a medium strength hand. If they do though, you should of course raise your really good hands, hoping to get a call too.

So all this goes together to bring our information levels up. If your feeling is he either has a draw or a set, knowing a couple of these reads may make your split decision a lot easier to make the right decision, and that is worth a lot of value.

First session update

OK, so I had a losing session yesterday. PokerTracker has become a very important resource when looking over losing and now winning sessions. Before I just figured I'd had a couple of coolers, maybe a couple of big winners, but sometimes you look at it, and you were in fact a long way behind when the money went in and hit that lucky card. Or in fact that you had no luck when you were miles ahead in a pot, or 50/50. It is absolutely vital to be able to be really critical of your play, because a winning session playing bad, is a losing session in the long run, and likewise the other way.


So on analysis of my play last night, where I lost over 2 buy ins. It felt pretty bad to me, but over 2 BI's were lost due to 50/50's losing or coolers, so overall, it's not as bad as it felt. I see my 5 biggest losing pots were all pretty standard hands where I was roughly 50/50 for all of them, except a set over set situation, where we both flopped a set. I lost the most on that hand. The only thing I regret about last nights play, was trying to force some results in pots I had no right to be in. And almost every time it backfired. I was deliberating trying to push people around, not tilting, but it failed miserably. Not sure if I want to do that again or not now.

I'm back

Hi blog, I'm back, it's real, I'm back for good. I'm finishing up my Cardrunners blog, partially because I may move over to Stox poker for videos soon, and partially because I'd rather not be tied in with a training site for my blogging.

So turbo update time. I'm playing NL50 (25c/50c) on the OnGame network, busy clearing bonuses and getitng a lot of hands/experience, more than worrying about profit at the moment. I'm marginally ahead of breakeven in terms of results, well ahead when bonuses/rakeback kick in. Realistically I'm not far off being fully bankrolled for NL100, but in poker that could be 3 days or 3 months depending on how I play and how things run.


In some ways I'm really finding my feet technically still, but the biggest improvement in my game this year has been mentally. It can be a tough game, but we can easily make it tougher by reacting to situations, and I'll get into that a bit in this blog I think. Overall my most winning style has been tight/semi aggressive, meaning I give up a lot on one pair hands to resistance, but do still bet a bit where I'm not that likely to have the best hand.


So I'll start off with another blog directly after this one, to talk about yesterdays play (losing session) to get the ball rolling.