Using Custom Reports and Statistics

Every report and statistic in PT3 is in a specific section.

Only statistics from the same section can be used in a report - e.g. if the report is in the Holdem Cash Player Statistics section then only stats from the Holdem Cash Player Statistics section can be displayed in that report.

The sections are:

• Holdem Cash Hand – Single Hand Details
• Holdem Cash Player Statistics – Player Statistics
• Holdem Cash Session – Single Session Details
• Holdem Cash Sessions – Summary of Sessions

• Holdem Tournament Hand – Single Tournament Hand Details
• Holdem Tournament Player Statistics – Tournament Player Statistics
• Holdem Tournament Tournament – Single Tournament Details
• Holdem Tournament Tournaments – Summary of Tournaments

_{Note: The only statistics you can display in the HUD are in the Holdem Cash Player Statistics section - or Holdem Tournament Player Statistics for tournament tables.}

_{Note: There are also the same set of sections for Omaha, and these work exactly the same as for Holdem.}

From now on I'm only going to refer to the holdem cash sections, but the same applies to the equivalent omaha and tournament sections.

The previous post explains the process of creating a custom report, and how to use the various options.

This post will walk through the steps needed to create a very simple report in the Holdem Cash Player Statistics section, with screenshots to illustrate.

This is just a basic report to show some stats for all players with more than 100 hands in the database, sorted by VP$IP.

1. Click the Reports tab.

2. Select Holdem Cash Player Statistics from the Section dropdown list.

3. Give your report a Name - my unimaginative name is "Example HCPS Report", but I'm sure you can do better than that..

4. Scroll the Available Stats list up and down to find and double-click the stats you want to see in your report.
I've added Player, Hands, VP$IP, PFR, Amount Won and BB/100.

5. These stats will then be shown in the Report Stats section above.

6. Stats will be listed in the order you add them but you can change the order using the up and down buttons.

You could save and run the report now, but it will currently only show stats for the current selected player and will sort by the player name, so we need to set Filters and Sorting information.

7. Click the blue Filters link to open the Edit Report Filter window.

7a. Untick the Filter on Active Player option so that the report will show stats for all players.

7b. Click the blue Insert link and go to the Statistics tab. Scroll down the list to find "Hands" and select it then click OK. This will enter into the Filter Expression field: #Hands# 7c. After this type " > 100" so that the expression now says: #Hands# > 100

This will filter players based on the value of their "Hands" stat (which is also in the Holdem Cash Player Statistics section).

_{Tip! You can type the whole expression yourself but Insert is particularly useful when you don't know the exact name of the Statistic or Column that you want to use.}

7d. Click the Save button and you should see the filter expression appear in the Filters list in the top section of the window.

_{Tip! At this point you could click New and enter another filter expression.
You could also click the Simple Filters button to restrict which hands are used to generate the stats.}

7e. This report only needs this one filter so click OK to store the filters and return to the main report window.

8. Click the blue Sorting link to open the Edit Report Sorting window.

8a. To sort the report by VP$IP select it in the list on the left and click the right arrow in the centre.
VP$IP will move to the list on the right.
The Sort Method will be 'Ascending' by default - if you prefer to see the highest VP$IP values at the top of the report instead then double click the stat on the right hand side to change to 'Descending'.

_{Tip! If you want to sort by more than one column then add more and use the up and down arrows to change the priority order for sorting.}

8b. Click OK to store the sorting preferences and return to the main report window.

9. The report is now ready so click Save to store it.

10. Click Run Report to populate the report.

The information about a Statistic is split across four tabs.
This section defines each field in the Edit Stat Info section.

Columns are used to retrieve information directly from the database and generate information to be shown in Statistics.
For information on how Statistics and Columns fit together see this section.

Exactly how this works depends on which section the column is in.

For instance the Holdem Cash Hand (HCH) section retrieves information about individual hands so the column will have a value related to a single hand – for instance it might have a boolean (true/false) value to indicate whether the player saw the flop in the hand, or a numeric value to show the size of a bet made on the flop.

Examples

• holdem_hand_player_statistics.flg_f_saw
• holdem_hand_player_detail.amt_f_bet_made

The Holdem Cash Player Statistics (HCPS) section has information from all of a player’s hands. The column effectively iterates through every hand for the player, so rather than recording whether or not the player saw the flop in a specific hand you might have a column to count the hands where the player saw the flop.

Example

• sum(if[holdem_hand_player_statistics.flg_f_saw, 1, 0])

This expression uses the same database field as the HCH stat, but it references it for every hand, and adds up (sum) the result of the if statement for each hand.

The expression works like this:

‘SUM’ adds up the values from all the hands.
‘IF’ checks each hand.

The format is:

sum( if[ expression, then, else ] )

For each hand it evaluates the expression, and if it is true it returns the ‘then’ value, otherwise it returns the ‘else’ value.
Here ‘then’ is 1 and ‘else’ is 0 – this has the effect of counting the hands where the player saw the flop.
In English: if [ the player saw the flop, return 1, else return 0 ].

These 1s and 0s are then added up by the SUM to give the number of times the player saw the flop.

For instance, if the player played 5 hands and saw the flop in the first and third, then you would effectively have: sum( 1, 0, 1, 0, 0 ) = 1 + 0 + 1 + 0 + 0 = 2.

The values returned do not have to be 1s and 0s, but that is the most common usage.
As an example of how else it might be used the following expression will return the total amount bet on the flop as continuation bets, across all of a player’s hands, but will not count bets made when the player was not the preflop aggressor.
 

Example

• sum( if[holdem_hand_player_statistics.flg_f_cbet, holdem_hand_player_detail.val_f_bet_made, 0] )

The Holdem Cash Session section works in the same way as Holdem Cash Hand – it refers to individual sessions.

The Holdem Cash Sessions section works in the same way as Holdem Cash Player Statistics – it iterates through all sessions to calculate total values.

The tournament sections work in the same way.

For a walkthrough of how to create a custom statistic see this tutorial.

Most stats that we build are in the Holdem Cash Player Statistics (HCPS) section as these are the stats that can be used in the HUD and in reports about players' overall stats.

_{All of the following applies to the appropriate Omaha and Tournament sections too, but I will just describe the holdem cash case for simplicity.}

When you build one of these stats you will want to check that it is giving you the correct results and I find the best way to do this is to create a matching stat in the Holdem Cash Hands (HCH) section.
This pair of stats will be based on the same expression, but the way they are constructed is slightly different.

HCPS stats are often based on the number of times a certain action or event occurred, out of the number of times it could have happened, across all of the player's hands. This is normally displayed as a percentage. (You can build other types of stats here too, but these are most common.)
The construction of these 'percentage' stats is described in detail earlier in this tutorial.

HCH stats are based on individual hands and indicate whether the action or event occurred in each specific hand. This is normally displayed as a checkbox to indicate true or false.

As an example I'm going to build a HCH stat to go with the built in HCPS "Steal Success" stat.

The HCPS stat shows the percentage of times the player's steal raises "succeeded", by which I mean the raise won with no further action.
So if the player made 10 steal raises and won immediately 3 times the stat will show "30".

You can add this statistic to the built-in reports on the Positions tab, for instance - either or both the Position Statistics and Position Hands reports.
This will show how often the steal raises succeeded, but to see which 10 specific hands were the steal raises, and which 3 succeeded we need to create a stat we can show in the Position Hands Detail report.

The HCH stat will show a checkbox if the player made a steal raise; the checkbox will be checked if the steal succeeded and unchecked if it didn't.

The HCPS "Steal Success" stat is constructed like this:
( cnt_steal_success / cnt_steal_att ) * 100

The column expressions are:

cnt_steal_success =

sum(if[holdem_hand_player_statistics.flg_steal_att AND NOT(holdem_hand_player_statistics.flg_f_saw) AND NOT(holdem_hand_player_statistics.flg_p_face_raise), 1, 0])

cnt_steal_att =

sum(if[holdem_hand_player_statistics.flg_steal_att, 1, 0])

To build the HCH stat we need to recreate these columns in the Holdem Cash Hands section, but with one crucial difference.
The HCPS stats uses the sum(if[...]) construct to count how often the steal raise occurred and how often it succeeded across all the player's hands.
The HCH section works on individual hands so we don't need the sum(if[...]) part, but the rest is the same, so we create new columns:

flg_steal_success =

holdem_hand_player_statistics.flg_steal_att AND NOT(holdem_hand_player_statistics.flg_f_saw) AND NOT(holdem_hand_player_statistics.flg_p_face_raise)

flg_steal_att =

holdem_hand_player_statistics.flg_steal_att

Note that I change the names slightly too - they are now boolean (true/false) values so by convention we name them flg_ which indicates a 'flag'.

Now to construct the actual HCH statistic we go to the Statistics tab and click New.
Enter your stat name "Steal Success".
The Value Expression can be flg_steal_success - i.e. did the steal succeed or not.

The Format tab is where the real work is done. To make it really clear which hands are the ones where a steal was attempted we're only going to show the checkbox at all if a steal was attempted (i.e. when flg_steal_att is true).

So the Format Expression is:

if ( flg_steal_att, format_bool( flg_steal_success, 'check' ), '' )

This says if flg_steal_att is true then show a checkbox based on the true/false value flg_steal_success otherwise show nothing (''=empty string). See earlier in this tutorial for more information.
Make sure you fill in the Title and Width fields as well, or the stat won't show in your reports.

The Positions tab then looks like this when I add the stats to all 3 reports:

Here you can easily see which hands were steal attempts and which succeeded.
I have included the similar "Steal Attempted" stats for clarity - you will notice that there is one hand where a steal could have been attempted but wasn't, and this isn't included in the steal success stat.

I have also added "x/y" versions of the HCPS "Steal Success" and "Steal Attempted" stats to the top two reports so that you can easily see how many attempts and successes there were.
These are built by duplicating the built in stats and just changing the Format Expression. For example the Format Expression for "Steal Success (x/y)" is:

format( '{1}% ({2}/{3})', format_number(100* cnt_steal_success / cnt_steal_att, 0, false, false), format_number(cnt_steal_success,0,false,false), format_number(cnt_steal_att,0,false,false) )