PDA

View Full Version : The Fish Factor


rvg72
10-07-2005, 02:35 PM
Everybody knows that table selection is important but it is fairly difficult to quantify. The steps that I’ve outlined try to quantify it and the results are very telling and may give you enough reason to be more patient when determining where to sit down. Scroll to the bottom to see my results broken down by “vs player type”

What do you need? Poker Tracker and Microsoft Access

Please note that all of my DB programming experience is with SQL Server so some of the steps may seem odd to anyone that knows Access really well – it does work but I’m sure there is a slicker way to get the data. Also, there are a lot of steps but each one if very quick.

Steps

1) Using Windows Explorer go to the “C:\Program Files\Poker Tracker V2” folder (or wherever you have Poker Tracker installed) and double click on the file named ptrack.mdb. It should be loaded by Microsoft Access. You might get a security warning, just click on Open.
2) Click on “Queries” under the Objects header
3) Double Click on “Create Query In Design View”
4) Click “Close” on the Show Table popup.
5) Right Click on the grey part of the window and select SQL View
6) Copy the following into this window

SELECT Player_ID FROM players WHERE Screen_Name = 'rvg72’

But replace the rvg72 with your screen name.

7) Click on the Red Exclamation point
8) Make note of the number that is returned – this is your player_ID number in the poker tracker database
9) Right click on the grey area again and select SQL View, Delete the code and replace with the following:

SELECT Screen_Name, (SELECT NZ(Sum(Total_Hands),0) from Session WHere player_ID = PLayers.Player_ID) AS TotalHands, (SELECT NZ(Sum(Vol_Saw_Flop),0) from Session WHere player_ID = PLayers.Player_ID) AS VPIP, (SELECT NZ(Sum(WON_FROM),0) FROM PLAYER_WINNINGS WHERE PLAYER_ID =5 AND OPPONENT_ID = Players.player_ID) AS WonFrom, (SELECT NZ(Sum(WON_FROM),0) FROM PLAYER_WINNINGS WHERE OPPONENT_ID =5 AND PLAYER_ID = Players.player_ID) AS LostFrom, WonFrom-LostFrom AS TotalWinnings
FROM Players;

But Replace the 5 (player_ID = 5) in the two locations with whatever your player_ID is.

10) Click on the X to close this window and save it as Opps
11) Double Click on “Create Query In Design View”
12) Click “Close” on the Show Table popup.
13) Right Click on the grey part of the window and select SQL View
14) Copy the following into this window

SELECT (Select Sum(TotalHands) from Opps where (VPIP/TotalHands) < 0.2 and TOtalHands >= 30) AS TightHands, (Select Sum(TotalWinnings) from Opps where (VPIP/TotalHands) < 0.2 and TOtalHands >= 30) AS TightWinnings, (Select Sum(TotalHands) from Opps where (VPIP/TotalHands) between 0.2 and 0.3 and TOtalHands >= 30) AS NormalHands, (Select Sum(TotalWinnings) from Opps where (VPIP/TotalHands) between 0.2 and 0.3 and TOtalHands >= 30) AS NormalWinnings, (Select Sum(TotalHands) from Opps where (VPIP/TotalHands) between 0.3 and 0.5 and TOtalHands >= 30) AS LooseHands, (Select Sum(TotalWinnings) from Opps where (VPIP/TotalHands) between 0.3 and 0.5 and TOtalHands >= 30) AS LooseWinnings, (Select Sum(TotalHands) from Opps where (VPIP/TotalHands) > 0.5 and TOtalHands >= 30) AS FishHands, (Select Sum(TotalWinnings) from Opps where (VPIP/TotalHands) > 0.5 and TOtalHands >= 30) AS FishWinnings, TightWinnings/TightHands*100*8 AS TightDollarsPerHundred, NormalWinnings/NormalHands*100*8 AS NormalDollarsPerHundred, LooseWinnings/LooseHands*100*8 AS LooseDollarsPerHundred, FishWinnings/FishHands*100*8 AS FishDollarsPerHundred
FROM Players
WHERE player_ID = 0;

15) Click on the X to close this window and save it as Winnings
16) Double Click on “Winnings” in your query tab to return the results

Note: The *100*8 in the query is to 1) give you a per 100 hands number and 2) multiply by 8 making the assumption that on average you are playing with 8 opponents. If you play shorthanded then change the 8 to a more appropriate number. I know this is not a completely accurate way of doing this but it is pretty close – close enough to make the point anyways. Also, I’m defining the player tights completely by VPIP which also isn’t 100% accurate but again, close enough.

Also note that you can change the definitions of Tight / Loose etc by modifying the ranges in the query above.

One last note is that the result is $/100 not BB/100 so adjust accordingly.

So what are my results over about 10K hands? Overall I’m up about 5.7BB/100 playing mainly $2/$4 but now up to $5/$10 (low sample size I know) but I attribute most of this to finding fish and taking their money before someone else does.

Vs Tight Players -$0.91/100
(16434 Hands -$18.75)

Vs Normal +$18.89/100
(17537 Hands $415.00)

Vs Loose +$46.21/100
(9282 Hands $536.16)

Vs Fishies +$103.18/100
(7954 Hands $1025.91)

So try to find the fishes before I do.

rvg

toss
10-07-2005, 02:37 PM
Hmmm very interesting. I'd really like to see this kind of data over something like 300k hands or 500k.

rvg72
10-07-2005, 02:39 PM
That makes two of us - hopefully people with larger databases try this out and share their results.

[ QUOTE ]
Hmmm very interesting. I'd really like to see this kind of data over something like 300k hands or 500k.

[/ QUOTE ]

jskills
10-07-2005, 02:59 PM
Dude - nice work. This kind of ad-hoc analysis is very useful.

I have buddy who wrote a query vs the PT database (Postgres version) that tells you the % of time you lose when you get raised on the turn (I believe the findings indicated youwere beat 70% of the time).

Great stuff.