Here's a follow-up to my first post...
This is all $2/$4 Party Poker April-May 2005...
After data mining 37,000 hands.
Looking for winning players...
Is not as easy at it might seem...
Because a lot of bad players can have a good month.
So I've been using this filter...
To filter out players that "play correctly":
>300 hands, > 9% Pre Flop Raise, VPIP < 22%
You get 36 players as follows:
Break even after 219 hours play as a group...
Not impressive.
Now let's look at a subset of these 36...
Those with VP$IP under 15%:
9 of the original 36...
This is a small and losing group.
After 57 hours: -2.28 bet/hour.
This group is too tight for $2/$4...
And is not getting enough action to be profitable.
If you eliminate the Rocks...
You get 25 players at VPIP 15 to 22 with PFR > 9%...
With a BB/100 of +0.83
This is interesting...
Because you could code a Bot that does this.
Of course...
Most of the money is made with good Post Flop Play...
So taking it to +2.0 and higher BB/100...
Is an art form that takes time and ** talent ** to develop.
And looser than 22% ????
Here it is:
Only 3 players, but you can eliminate the 51% guy...
So only 2 players out of 38 take VP$IP into the mid-20s.
But it can work if you play well.
All this tells you...
16-25 VP$IP and >9% PFR is a starting point...
The easy part.
On this foundation you develop solid Post Flop Play.
I'm now exporting 100% of the data from PT...
And may have some interesting analysis...
On the next few weeks.
For example...
I now have 39,000 "known starting hands"...
(Only about 13% of starting hands are known...
Usually the ones that make it to the showdown)...
And can start confirming the advice in SSH.
Hold 'em has amazing complexity as you dig deeper...
For example...
Based on my limited sample...
More money is made with suited connectors like 98s...
Than with Gapped Suited Broadway hands like QTs...
Because both good and bad players then to overplay the latter.
rm+
[img]/images/graemlins/cool.gif[/img] [img]/images/graemlins/cool.gif[/img] [img]/images/graemlins/cool.gif[/img]