PDA

View Full Version : what is formula of "bet vs missed flop cb out of position" stat



Rattanat
10-21-2010, 10:25 PM
So how I halculata "bet vs missed flop cb out of position" stat?

Patvs
10-23-2010, 01:07 AM
http://forums.holdemmanager.com/custom-reports/3595-formulas-every-hm-stat-updated-nov-12-2009-a.html

Rattanat
10-23-2010, 07:05 AM
http://forums.holdemmanager.com/custom-reports/3595-formulas-every-hm-stat-updated-nov-12-2009-a.html

I have read it several times, but dont find it? :confused:

Also this are missing:
Bet vs missed flop CB in position
Bet vs missed turn CB in position
Bet vs missed turn CB out of position

TheZepper
10-25-2010, 04:09 PM
These are HUD stats only - not Report stats - and therefore their formulas are NOT posted in the Formulas for every HM Stat thread (or anywhere else that I'm aware of).

Furthermore, the Bet v Missed CB HUD stats are all incorrect, so I'd advise you not to use them anyway

Rattanat
10-26-2010, 09:58 AM
Furthermore, the Bet v Missed CB HUD stats are all incorrect, so I'd advise you not to use them anyway

Interesting, I use lot of them.

What are the problems of these stats?

TheZepper
10-30-2010, 03:41 AM
The formulas used to calculate these 5 HUD stats are incorrect.
For example, the BetvMissed Flop CB OOP HUD stat (ie - Bet on Turn) presumably equates the statement "Facing Flop CB = False" with MISSING a CBET ...... this conclusion is incorrect in that this statement is true in limped pots that are checked thru as well as hands in which the PFR fails to bet on the flop.
There are other difficulties relating to these stats - most notably, identifying the relative positions of the PFR and PFC after the flop is very problematic (especially in multiway pots - the PFR has to act after the PFC on the flop, but doesn't necesarily have to act last).

I couldn't find it last night, but I think that this is the current formula for the BetvMissed FlopCB OOP HUD stat:

Stat GroupName="TN HUD" ColumnName="TN BETvMssFLCB Made HUD%" ValueExpressions="sum(case when ph.flopfacingcbet = False and Flop.LasttoActonStreet = False and ph.Turnplayeractiontype_id between 1 and 44 then 1 else 0 end) as TNBETvMsdFLCBPossxHUD; sum(case when ph.flopfacingcbet = False and Flop.LasttoActonStreet = False and ph.Turnplayeractiontype_id between 23 and 44 then 1 else 0 end) as TNBETvMsdFLCBMadexHUD;" Evaluate="TNBETvMsdFLCBMadexHUD/TNBETvMsdFLCBPossxHUD" ColumnHeader="Turn BET\nvMsFLCB\n HUD%" ColumnFormat="0.0%" ColumnWidth="61"

Send me PM if u'd like to discuss these stats in more detail.....

jj12
12-01-2010, 11:19 AM
Can the support reply about what is the formula for this stat? I need it for HU play, and find the values incorrect. I have created a custom filter (in Reports) for the situation where the flop checks through and villain is OOP; I am looking at the "bet turn stat" for villain, in these situations. The values and the sample sizes are very incorrect for the HUD stat, when compared with the filter stats in Reports.

Thanks

morny
12-04-2010, 03:31 PM
Ok i can understand this is a concern, ive limited availability today but ill spend some time writing up a bug report for this and try and get this fixed.

TheZepper
12-05-2010, 03:41 AM
jj12: You're thinking on the right track, but your filter is incorrect because it doesn't limit the stat to situations where a player is OOP Postflop as the PF Caller. Fortunately, in HU play, the Turn Bet v Missd Flop CB OOP stat is fairly straightforward because the relative position of the PFR and the PFC are fixed. Although I can't do anything about correcting the stat in the HUD, I believe the correct stat FOR HU PLAY only is:



Stat
GroupName="Turn"
ColumnName="TN BETvMssFLCB HU OOP%"
ValueExpressions="sum(case when PKH.positiontype_id_firstraiser >= 0 and ph.preflopplayeractiontype_id in (48,50,52,55,58,60,62,66,70,72,74,77,80,82,84) and ph.Flopplayeractiontype_id = 22 and Flop.LasttoActonStreet = False and ph.Turnplayeractiontype_id between 23 and 44 then 1 else 0 end) as TNBETvMsdFLCBMadexHU;
sum(case when PKH.positiontype_id_firstraiser >= 0 and ph.preflopplayeractiontype_id in (48,50,52,55,58,60,62,66,70,72,74,77,80,82,84) and ph.Flopplayeractiontype_id = 22 and Flop.LasttoActonStreet = False and ph.Turnplayeractiontype_id between 1 and 44 then 1 else 0 end) as TNBETvMsdFLCBPossxHU;"

Evaluate="TNBETvMsdFLCBMadexHU/TNBETvMsdFLCBPossxHU"
ColumnHeader="Turn BET\nvMssdCB\n OOP"
ColumnFormat="0.0%"
ColumnWidth="61"

You need to add "<" before the stat and "/>" after the stat (without parentheses) to run it as a custom stat.

Hope this helps.

jj12
12-05-2010, 07:56 AM
How can I use the definition you gave me? Do I paste it somewhere?

Well, I play almost exclusively HU. The filter is correct. I take a 50 hand session look through all the hands manually and compare to the value of the filter. I sent Morny a hand history (of only 11 hands) where their stat shows 66% bet vs missed flop cb with sample of 3, whereas reading the hand histories shows the correct value is 100% bet vs missed flop cb with sample of 1. It is easy to find a case where HM breaks.

Please fix this as its very important stat, that I am interested when deciding whether to cbet the flop or not.

jj12
12-20-2010, 02:38 PM
Hi,

Is there any update on this stat? I sent you a trivial 10 hand example that results in a wrong stat. I really need this stat for playing HU.

Thanks