The PokerTracker 3.0 beta test is open to the public now. I downloaded it from their site last night and ran my last month’s data through it. All the key numbers matched the numbers in my old PokerTracker DB, which is good news. It comes with a built-in idiot-proof install of the PostgresSQL database, which is a good thing. It does some weird clustering when it finishes loading in the hands, which is a bit slow, but otherwise the new program looks like a upgrade. As a nerd, I appreciate some of the technical improvements. The original app is clearly built with the PowerBuilder application developmet platform, which is rather dated. The new one looks much more slick. All of the pieces are not yet in place — you cannot run the custom reports and the integrated HUD isn’t in place yet and I had some trouble with the filters. It looks like the MTT section needs some work too.
The advantages are numerous. The windows are dynamically resizable and everything is sortable and customizable. There are cool new graphs and the filters have a ton more options than they used to. I think the new filters are going to help a lot of players to find leaks in their game. The hand detail window can group by starting hand combo (the old way) or it can group by what you ended up with. The session tab is a ton better, because you can see sessions defined the old way or all multiple tables played at the same time combined together. Not only will this help you to look at your results the way you really think of sessions, but it will help me a lot when it comes time to settle up with the tax man. There is a nifty graph of your net results over the course of the session so you can relive it all in graphical form. In general, it is fast and responsive and seems very well thought out. I need to see everything working, but my first impression is a good one.
I would warn people that this is beta-test software in a sense that people who don’t do this for a living don’t often see. You may be used to seeing nearly complete programs when you get a “beta” version. This isn’t like that. This is honest-to-God test code. Pieces are missing (with comments like “this will have …” to explain what you are missing) and some stuff doesn’t work quite right. I still think PokerTracker is going to be the gold standard by which competitors are judged.
My poker results were quite ordinary. I had some very bad hands in the middle and was down $1,100 at one point. I rallied nicely at the end to book a tiny $90 win over 380 hands and another 500 VPPs in a two hour, two table session. I’m now at 11,000 VPPs, well ahead of pace.
I had a few interesting hands in this session. The one that sticks in my mind is one where I three-bet with JJ from the cut-off against a solid TAG from MP2. He’s a good, thinking player whose range is fairly wide there. The flop was Q75r and he check-called. He also check-called a duece on the turn. The river brought an Ace and he checked to me. My instinctive reaction was “He probably just caught his Ace and is trying to check-raise me.” Later on, I realized this is unlikely. He knows that the Ace is a scare card and that I might well check behind, so if he was chasing AK/AJ that passively, he would donk the river to prevent me from taking the free showdown. He had to hold some medium or small pair or flopped a pair with a suited connector type hand. I checked behind on the river, which I think was a pretty poor play. I was certain to be good and although he wouldn’t call with a lot of hands if I three-barreled there, he would call some of them.
I was unhappy with my play there and in a few other spots where I paid off even though I simply knew I was beat. Limit poker can fool you into mistakes because it is often correct to call when you are almost certainly behind, because if you are ahead just one time in ten, you will profit. Some players take that too far and I’m guilty of that sometimes. When you are simply certain that you are beat, you don’t have to pay that last big bet to see the proof.