From Amateur to Pro Article Three – Statistical Review
Monday, 4 July 2011
When Bluefirepoker coach Alan Jackson took on the challenge of turning me into a winning 6-max online cash game player, he needed to know what he was dealing with.
I sent him all my hand histories and he processed them through his Hold Em Manager (HEM) database. There are different types of coaches in all sports and games. In poker it is fairly common for a student and coaching relationship to exist during sweat sessions alone. Jackson doesn’t do things this way; for him, it’s all about the devil in the detail and he is a staunch fan of the advantages of HEM.
Jackson has the happy advantage of looking at a huge number of winning poker players hand databases. He has a keen eye and he can see statistical patterns that are prevalent in all winning players’ games and those which are missing from the less adept. Armed with this knowledge, he can set a solid foundation of the rules for your game. We call these rules default lines and we only deviate away from them with good reason, perhaps because of game-flow or opponent related considerations.
I had also made a decision to move away from Full Tilt and play on PKR instead. I was losing on Full Tilt and was told that PKR had the softest 6-max cash games in the industry. The only problem was the HEM Heads Up Display (HUD). If you play more than one table, the HUD won’t work. So, during my first 70,000 hands on PKR, I only used HEM to review my game post session. I did this based on the following statistical framework that Jackson believes helps you on the road to financial success.
VPIP (Voluntarily Put In Pot)
This is the money that you have voluntary put into the pot. Folding in the blinds does not count as voluntarily putting money into the pot but limping, calling and raising does.
Jackson set me a range of 17% to 21%. To help me stay within this range I was given starting hand requirement charts, which I used when I played. These charts were explained in the first article we published in this series.
If your VPIP is below 17%, then you are playing too tight. When it is above 21% you have been playing too loose. For more accuracy, it’s advisable to have a hand sample size of 1,000 hands.
If you stay within the starting hand chart parameters, then your VPIP should stay within this range. If you are outside this range, then you need to analyse spots where you have deviated and ask yourself why you have deviated and question whether the reasoning is sensible.
PFR (Pre Flop Raise)
Your PFR is the percentage of times you have entered the pot with a raise.
My PFR target range was 14 to 17%. If this falls below 14%, then I have been playing too tight, and if it exceeds 17%, I have been playing too loose. For more accuracy, it is advisable to have a hand sample size of 1,000 hands.
Deviations are investigated in the same manner as VPIP.
Spread
This is not a statistic you will find in the HEM database. The spread is simply the difference between your VPIP and your PFR. Keeping the spread as tight as you can means you reduce your cold-calling frequency. When I started playing, my target spread was 4% but since my game has improved over time, it has been changed to 5%.
Deviations need a detailed analyse of your cold-calling range. Why are you cold calling? What hands are you using and why?
The spread was covered in article two of this series.
W$WSF (Win when flop seen)
This statistic tells you how often you end up winning the pot when you’ve seen a flop. My target range for this statistic was between 43 to 47%.
If you are averaging below 40% then you are not bluffing enough or are allowing yourself to be bluffed off too many hands.
A good accurate sample size for this statistic is 1,000 hands.
WTSD (Went to Showdown)
This stat tells you how often you get to showdown when the flop has been seen. Cards must be shown at the end of the hand for them to affect this stat.
My range for this stat was 24% to 26%. If you find you are below 24%, then you are getting pushed off too many hands, and if it is over 26%, then you are probably calling too light.
A good accurate sample size for this stat is 5,000 hands.
Vs 3-Bet Call
This stat tells you how often someone calls a 3-bet.
My range for this stat was 20% to 27%. If you find you are below 20%, then you are not defending enough and are playing too tight. If you find you are exceeding 27%, then you are defending too much and playing too loose.
A good accurate sample size for this stat is 2,000 hands.
If you are outside the default stats then you need to analyse your game. When doing this, it is important to divide your analyse in half. The first half of your analysis should focus on the number of 3-bets you call in position and the second half of your analysis should focus on calling out of position.
A good accurate sample size for this stat is 2,000 hands.
Flop C-Bet
This stat tells you how often someone makes a continuation bet after being the initial pre flop raiser.
My range for this stat was 60% to 74%.
Falling below or rising above this range does not necessarily mean you are doing something wrong. If you are c-betting 70% of flops and only 25% of them are successful, then you need to adjust, but it is a great starting point.
Turn C-Bet
This stat tells you how often someone makes a continuation bet on the turn after being the initial pre flop raiser (also called double barreling).
My range for this stat was 40-52%.
Again, this stat needs to be analysed alongside the CB Turn Success % because it’s the success rate of your double-barrels that are important.
Bet River
This stat tells you how often you make a bet on the river.
My range for this stat was 27% to 32%. If you have fallen below 27% then you are not value betting thinly enough and if it is over 32% you are more likely bluffing too much.
A good accurate sample size for this stat is 3,000 hands.
River Call Win%
This stat tells you how often you make a successful call on the river.
My range for this stat was 42-50%. If you have fallen below 42%, then you are calling too much and if it is over 50%, you are folding too much.
A good accurate sample size for this stat is 3,000 hands.
Flop v Raise/Fold%
This stat tells you how often you bet the flop, faced a raise and folded.
My range for this stat was 50% to 60%. If you have fallen below 50%, then you are not folding enough and if it is over 60% you are folding too much.
A good accurate sample size for this stat is 4,000 hands.
Summary
These stats are used for two reasons. The first is to review your own game post-session. You go through each statistic to see where you have deviated from your foundational range. Then you use HEM to analyse why and see if the reasons are resulting in mistakes, before creating strategies for improvement. This is where it’s invaluable to have one-to-one sessions with someone who has seen it, done it and bought the T-shirt.
The second use of these stats it using a HUD in active play. You can study these stats on your opponents and create strategies to exploit their weaknesses or make marginal decisions during the sessions.
In later articles we will go through each statistic in a little more detail.
This series is also a video series that can be found at www.bluefirepoker.com. You can follow my own personal story on my blog at www.leedavy.co.uk. If you are interested in receiving coaching from Alan Jackson then visit www.AJacksonPoker.com.