Monday, October 19, 2015

A Big Advantage of the Poker Pro - Evaluating Hands and Losses

Every hand of poker you ever play will be unique. There will never be the same set of players with the same cards, stack sizes, and mental states. But situations that are similar to past situations come up constantly. One huge advantage I have over my competition is that I know how to look back at hands and sort out in detail if I made mistakes and how big they were. Anytime I can I find spots where I thought what I did was correct was not or the reverse I learn something new and increase my advantage.

I should spend more time evaluating wins, but it's human nature to look back when things have been going south.

I've put in 68 sessions at $2/$3/$5 this year and through session #62 there was only one instance where I lost two times in a row. #63 was a loser #64 was a loser, then #65 was the mega amazing $5,700 crush everyone for 7 hours win from my last post, which I followed up by promptly losing 3 times in a row.

In my last session I dropped $1,600+ in a good game and I've been thinking a lot about why.

Losing 5 out of 6 times isn't all that improbable. In fact if I win two thirds of the time there is still a 1.5% chance that I'll lose 5 of the next 6 sessions I play. So every 200 sessions I should expect 3 runs like this. Maybe I'm just hitting that bad run of cards that is inevitable. That's one explanation.

The other is that something has changed about my play or the play of my opponents. In fact I think I've narrowed down one contributing factor. I've been far too willing to make high variance plays, or put another way I've been far too willing to gamble.

Here is a perfect example from my last session. I called $10 vs a straddle with 66 and after a few more calls a wild player moved all in for about $200. Everyone else either folded or looked like they were going to fold and I was faced with a decision to risk $190 to win about $240. I called and lost to AJ.

The amateur level analysis is: I got it in when I was a small favorite. I didn't win but I made the correct play.

My in the moment analysis was: If I was up against overcards I'd be somewhere in the range of 48%-55% to win depending on the exact overs and getting better than even money that would be profitable. If my opponent was equally likely to have 22-55 as 77-TT that would be a wash as sometimes I'd be way ahead and others way behind (I discounted JJ+ as those hands would probably not shove). I should be a tiny favorite against his range and I'm getting better than even money.

My post session analysis is: Looking back I'd actually be much more likely to see 77-TT than 22-55 and that's the problem. If I'm up against a range of AT+, KQ, 55, 77-TT then I'm only 42.6% to win when I need better than 44.2%. Couple this with the non zero chance that I'm going to get called by someone behind me and this is a pretty big error.

I had a similar situation where someone with about $250 in his stack made it $150 to go vs a raise to $25 and a call. I called and on an A 8 5 flop I check called $100 and lost to 88.

The amateur level analysis is: I thought I was up against a small to medium pocket pair and I was. I was right.

My in the moment analysis was: This guy was also wild and I'd just seen him do something similar with AJ. I had AQ and decided to call and shove the last $100 on any flop fully expecting to get called, but thinking there was some small chance that I might unload a medium pair on a flop with a couple of overs or lose AK if we both missed. Essentially I knew I was risking $250 to win $300 in a spot where I was very likely to be about 45% to win, but there was some tiny amount of equity in him folding in some very specific spots.

My post sessions analysis is: Turns out if I was against AJ+, and 55+ I'd be 46.5% needing 45.5% for this to be profitable. If he's capable of making this play with worse aces or KQ my equity jumps up a bit, but if someone else calls behind me I'm usually cooked.

These are spots where if you've evaluated the situation accurately and perfectly then you have a small edge. The problem is if you've screwed something up - like maybe either of these guys are frustration shoving AA or KK because they're sick of losing or one of the guys who looks like he's ready to fold actually has a huge hand and comes along - then you've made a horrible miscue.

It's really easy to look at those two hands and think "Shit I lost two flips. I got it in close to 50/50 with some dead money in the pot both times and it's unlucky that I didn't win one of them!" That's what I was thinking on the drive home. But a deeper dive reveals it's either correct to fold or the margins or so razor thin that all you're doing is adding to your swings. If I'd just folded both I would have been out $10.

Two other big hands I had were even more gambly.

On the first I had 66 in the big blind and 6 of us saw a flop for $20. The flop came down 8 7 5 with two clubs giving me an open ended straight draw. It got checked to the raiser who bet $60 into the $120 pot. I called and then a very aggressive player with a big stack made it $300 to go. The raiser folded and it was back to me. I had about $475 left.

What I came up with in the moment is "That's a pretty big bet, this guy loves to push draws, that is probably a flush draw which means I'm actually ahead. If it's not then I should have 8 good straight outs and maybe a 6 is good also." So I shoved.

I ended up risking $475 to win $715. I missed the straight and lost to 87. But does that mean it was a bad play?

The rank amatuer analysis here is: Boo! I lost the pot! I'm stupid and/or the universe hates me.

The solid amatuer analysis here is: I was 34.2% to win and 1.6% to tie. Call it 35% to win which would mean I'd need $830 or more in the pot to make it profitable. This was a clear mistake.

The pro level analysis here is: My opponent's hand range probably any two clubs, 55-99, 87, 86, 76, A8, A6, T9, 96, 64. If I plug that into a simulator and I get that I have 41.65% equity against that range meaning I only need the pot to be $665 to make this a profitable move. But what if I've misread this guy a little bit. If he's only making that play with two pair or better I'm only 26% which is dreadful. The breakeven point on this is I need 39.9% equity so the "What if I fucked up the analysis just a little bit" factor of safety is not there. This is not a big mistake, but I'm really adding to my swings by doing this.

I just learned a lot by doing that analysis!

In my other gambly hand there was a straddle and a very loose, sort of aggressive player who is not good made it $50 to go. I had JT of clubs and we took the flop 6 way. The flop came down Q 8 5 with 1 club giving me a gutshot with a backdoor straight draw. The raiser bet out $100 into the $300 pot, which seemed weak to me. Normally I don't expect people to continuation bet into 5 opponents with air, but I thought this guy might. He had about $400 left behind and I thought about moving all in, but I was sitting on a $1,500 stack and one player left to act had me covered. I decided to make it $300 to go thinking that would be just as effective. The others folded and the raiser went all in. I called another $200 and lost to AQ.

The rank amatuer analysis here is: Boo! I lost the pot! I'm stupid and/or the universe hates me.

The solid amatuer analysis here is: This is good and bad. I made a strong move which is good, but ultimately I got it in there risking $500 to win $800 when I was only 23% to win which is atrocious. At least when I called the last $200 to win $1,100 I was getting the right price.

The pro level analysis here is: In order to sort this out we need to decide what his preflop range is, which hands he's betting on the flop and which hands he's calling our raise with and how often we win when called. I'd put his preflop range on 66+, any two cards T and higher and any Ax suited (this is very loose). There are 197 combinations of two cards that make all those hands and only 60 of them make hands that can call a raise (AA, KK, QQ, 88 and all of the one pair of Q hands). That means 69% of the time I'm going to win the $400 in the pot uncontested, 24% of the time I'll lose $500 and 7% of the time I'll get it all in and win $800.  If you sort all that out my equity in the hand is +$212, meaning if I made this play some huge number of times I'd be better off $212 per instance than if I'd folded. This illustrates the power of bluffing with a draw even a shitty one like a gut shot and back door flush draw.

But there are some major assumptions here. The biggest one of which is that my opponent is going to bet 100% of the time on the flop. Let's say he only bets AA, KK, QQ, 88 and all of the one pair of Q hands. Then I've put myself in a horrible spot of losing $500 77% of the time and winning $800 23% of the time which has an equity of -$201! ACK! 

So where does that leave us? It turns out the key here is more psychological than mathematical. Is this the kind of guy who would bet $100 into $300 with a hand that he's willing to stack off with? I thought he would have bet more if he had a good hand. It also leaves us with the notion that if a player is betting 100% (or close to it) of the flops, putting them to the test is a great idea.

No comments: