Incorrect Pot Splitting

Report bugs found in Poker Mavens

Incorrect Pot Splitting

Postby Jimulacrum » Sat Feb 20, 2021 8:49 pm

The two winners in this hand should have each gotten 39, but instead it went 40/38. Seems like the software splits each half of the pot independently and gives the out-of-position player the extra chip both times, where it should alternate to make it a truly even split for the exact same hand both ways.

Hand #16656-62 - 2021-02-21 01:43:24
Game: Limit Omaha-5 Hi-Lo (100 - 500) - Stakes 3/6
Site: Roscoe's Wiggle Room
Table: Saturday Circus - Table 1
Seat 1: PaulG (138)
Seat 2: Jimulacrum (253)
Seat 3: podo (121)
Seat 4: kirchhausen (124)
Seat 6: ClazyRichAZN (205)
Seat 7: RomandaGreat (508)
Seat 8: JoseRijo (351)
RomandaGreat has the dealer button
JoseRijo posts small blind 1
PaulG posts big blind 3
** Hole Cards ** [7 players]
Dealt to Jimulacrum [Td 3c 2s Ts 8c]
Jimulacrum calls 3
podo calls 3
kirchhausen folds
ClazyRichAZN folds
RomandaGreat raises to 6
JoseRijo calls 5
PaulG calls 3
Jimulacrum calls 3
podo calls 3
** Flop ** [3h 5d 4h]
JoseRijo checks
PaulG checks
Jimulacrum checks
podo checks
RomandaGreat bets 3
JoseRijo folds
PaulG calls 3
Jimulacrum folds
podo raises to 6
RomandaGreat calls 3
PaulG calls 3
** Turn ** [4c]
PaulG checks
podo bets 6
RomandaGreat calls 6
PaulG calls 6
** River ** [Jc]
PaulG checks
podo bets 6
RomandaGreat calls 6
PaulG folds
** Pot Show Down ** [3h 5d 4h 4c Jc]
podo shows [Ah 9h 2d 2h Js] (Hi: a Straight, Ace to Five Lo: 5432A)
RomandaGreat shows [2c Ad 3s Qs 5c] (Hi: a Straight, Ace to Five Lo: 5432A)
podo splits Hi Pot (20) with a Straight
RomandaGreat splits Hi Pot (19) with a Straight
podo splits Lo Pot (20) with 5432A
RomandaGreat splits Lo Pot (19) with 5432A
Rake (0) Pot (78) Players (PaulG: 18, Jimulacrum: 6, podo: 24, kirchhausen: 0, ClazyRichAZN: 0, RomandaGreat: 24, JoseRijo: 6)
Author of SOHE, Dramaha, and Live Straddle!
Jimulacrum
 
Posts: 7
Joined: Thu Jan 14, 2021 9:12 pm

Re: Incorrect Pot Splitting

Postby Kent Briggs » Sat Feb 20, 2021 9:24 pm

Jimulacrum wrote:The two winners in this hand should have each gotten 39, but instead it went 40/38. Seems like the software splits each half of the pot independently and gives the out-of-position player the extra chip both times, where it should alternate to make it a truly even split for the exact same hand both ways.


There are two pots in a hi-lo game and they each had 39 in them at showdown. Because you apparently have the "Smallest chip" setting at 1, it had to split those odd numbers at an imbalance, with the early position getting the extra chip. If you had gone with a 0.01 or 0.05 or 0.25 smallest chip, it would have given 19.50 to each player in each pot for an equal split all together.
Kent Briggs - [email protected]
Briggs Softworks - http://www.briggsoft.com
Kent Briggs
Site Admin
 
Posts: 5278
Joined: Wed Mar 19, 2008 8:47 pm

Re: Incorrect Pot Splitting

Postby Kent Briggs » Sat Feb 20, 2021 9:33 pm

And just to emphasize the point, imagine Player A and Player B split the high pot and Player C and Player D split the low pot. Given the self-imposed constraint that fractional chips are not allowed, it would be impossible to split those 78 chips 4 ways equally.
Kent Briggs - [email protected]
Briggs Softworks - http://www.briggsoft.com
Kent Briggs
Site Admin
 
Posts: 5278
Joined: Wed Mar 19, 2008 8:47 pm

Re: Incorrect Pot Splitting

Postby Jimulacrum » Sat Feb 20, 2021 10:40 pm

I understand that fractional chips would allow us to escape the error in this instance, but it is still an error. This is not how the pot would be split in this spot in a real poker game. When the whole pot is divisible, and both players have the exact same high and low hands, they each get half the pot. In this case, the splitting mechanic is getting in the way of the desired outcome instead of producing it.
Author of SOHE, Dramaha, and Live Straddle!
Jimulacrum
 
Posts: 7
Joined: Thu Jan 14, 2021 9:12 pm

Re: Incorrect Pot Splitting

Postby Kent Briggs » Sat Feb 20, 2021 11:15 pm

It would only be an error if you wanted to treat the special case of the same two players splitting both sides of the pot differently than you would all other cases. It would be easier for a human dealer to just split the total pot in half but that doesn't necessarily make it the correct way.
Kent Briggs - [email protected]
Briggs Softworks - http://www.briggsoft.com
Kent Briggs
Site Admin
 
Posts: 5278
Joined: Wed Mar 19, 2008 8:47 pm

Re: Incorrect Pot Splitting

Postby Kent Briggs » Mon Feb 22, 2021 12:51 am

Out of curiosity, I went looking through Robert's Rules of Poker (included in the help file) to see if this situation was mentioned and sure enough under General Poker Rules -> Ties - 5c it says:

"In high-low split games, the high hand receives the odd chip in a split between the high and the low hands. The odd chip between tied high hands is awarded as in a high game of that poker form, and the odd chip between tied low hands is awarded as in a low game of that poker form. If two players have identical hands, the pot will be split as evenly as possible."

That last sentence bolsters the OP's claim but complying with it perfectly is a problem. My pot distribution routine announces the hi-pot winner(s) and moves half the chips to them. Then it announces the lo-pot winner(s) and moves the remaining chips. If each half has an odd chip then the only way you could make that perfect split (in this one special case) would be to violate the Smallest Chip setting (temporarily) or reverse the normal odd chip tiebreaker in the lo pot (but only for this special case). I don't like either of those options.
Kent Briggs - [email protected]
Briggs Softworks - http://www.briggsoft.com
Kent Briggs
Site Admin
 
Posts: 5278
Joined: Wed Mar 19, 2008 8:47 pm

Re: Incorrect Pot Splitting

Postby Jimulacrum » Tue Feb 23, 2021 2:16 pm

I can appreciate that it might be tricky to code, but as you can see, "my" way (a proper equal split) is the proper way to handle it in a real poker game, and poker software should do the same.

I guess the upside is that it's only mildly problematic and only happens in a small number of cases, so probably a low-priority fix. But it ought to be fixed at some point.
Author of SOHE, Dramaha, and Live Straddle!
Jimulacrum
 
Posts: 7
Joined: Thu Jan 14, 2021 9:12 pm

Re: Incorrect Pot Splitting

Postby Kent Briggs » Tue Feb 23, 2021 3:58 pm

Jimulacrum wrote:I can appreciate that it might be tricky to code, but as you can see, "my" way (a proper equal split) is the proper way to handle it in a real poker game, and poker software should do the same.


In a face to face game you're not documenting the high and low pots in a hand history file, which needs a consistent format. That is the real tricky part. Reversing the tie break order appears like a solution for two players (i.e., hi pot splits 20/19 and lo pot splits 19/20) but that scheme falls apart when there are more than 2 players in the split.
Kent Briggs - [email protected]
Briggs Softworks - http://www.briggsoft.com
Kent Briggs
Site Admin
 
Posts: 5278
Joined: Wed Mar 19, 2008 8:47 pm


Return to Bugs

Who is online

Users browsing this forum: No registered users and 3 guests