PokerTracker

Hi!

I have been looking around the forum and could not find an answer to my question, so I decided to ask it myself.

I am trying to build an MTT statistic that will be calculated for every hand (so I suppose it would be a T(ourney)/hand statistic) so that I can plot it on a graph as an overlay stat.

The formula I am trying to write down should calculate the amount of chips won scaled by some coefficient that takes into account some sort of money value and should look something like this:

custom_stat = amt_chips_won * amt_buyin/ starting_stack

What I need:

The information I cannot find is the 'starting_stack' amt. I found the code for how to get the number for summaries (for example here) but not as a hand statistic. Can somebody help me how to write a starting stack stat that can be accessed in the above calculation?

Thank you!!

sam_hawkins wrote:I am trying to build an MTT statistic that will be calculated for every hand (so I suppose it would be a T(ourney)/hand statistic) so that I can plot it on a graph as an overlay stat.

If you want a statistic calculated for every hand and display results in a graph it will need to be a player statistic and not a hand statistic. A hand statistic is for use in a hand report (listing individual hands) and only player statistics are available for graphs.

sam_hawkins wrote:The information I cannot find is the 'starting_stack' amt. I found the code for how to get the number for summaries (for example here) but not as a hand statistic. Can somebody help me how to write a starting stack stat that can be accessed in the above calculation?

A players starting stack isn't something that is stored directly in the database schema but there is SQL in the thread you linked however creating/using that type of SQL is beyond my own knowledge I'm afraid.

Thank you for your reply! I expected that some sort of modification of the SQL code I linked will be the answer to my question but unfortunately I only know python myself. Do you think anyone will be able to help from the PT4 team?

Or, alternatively, where can I found the SQL codes for other stats so I can try to figure it out myself? For example, where can I find the code for 'amt_chips_won' ?

You can see the code for all existing columns in 'Configure -> Statistics -> Columns'.

Yeah I found that, thanks! I think I managed to create the proper column for amt_starting_stack but when I create the custom stat, I cannot find it as an overlay stat in the graph (I think this is related to where the information comes from in the database) so I'm not sure I can do this on my own.

Please export your custom statistic, compress it and attach it here.

Here is the statistic, dunno if it is correct. The column starting stack is

Code: Select all

first_value(tourney_hand_player_statistics.amt_before) over
(partition by tourney_hand_player_statistics.id_tourney order by tourney_hand_player_statistics.date_played)

I'm don't think that that will work as a Player stat. Your expression wants to have a single value, but across all hands for a player there would be multiple values so the result would be undefined.
For example in a Player stat "amt_won" is the total amount of chips won or lost across all of the hands, but your "amt_starting_stack" is [trying to be] the starting stack in a single hand, and that won't work.

You may be able to get it to work by doing the whole calculation in a column, but you're mixing different levels of data (hand data and overall tournament data) and I think that still won't work.

WhiteRider wrote:I'm don't think that that will work as a Player stat. Your expression wants to have a single value, but across all hands for a player there would be multiple values so the result would be undefined.
For example in a Player stat "amt_won" is the total amount of chips won or lost across all of the hands, but your "amt_starting_stack" is [trying to be] the starting stack in a single hand, and that won't work.

You may be able to get it to work by doing the whole calculation in a column, but you're mixing different levels of data (hand data and overall tournament data) and I think that still won't work.

Thank you, that's the kind of insight into how the stats work I was looking for. Could you please explain why exactly is it a problem that the stat takes different values across different hands? I think I am not exactly understanding what is different from 'amt_won' -- there, my understanding is that it calculates the amt of chips won in each single hand and then when graphing you just want to graph a cumulative sum. Still, if I check amt_won for several individual hands, it will return the amt of chips individually for each hand. Am I mistaken somehow? Because that is exactly how I would like the amt_starting_stack column to work, a value assigned to each hand separately (but with the added property that this value will be the same for all hands for a given tourney), except it would not make sense to plot the cumulative sum.

I do understand the mixing of 'summary' data with 'hand' data, but I would be surprised if that was not possible, just define the summary data as hand data that is stationary over some subset of hands, right?