In the library with the
 

candlestick holder, but WHO did it ?

*

A friend (GF's friend to be precise) just gave me her records for 2005; they were quite impressive though incomplete. Here's what I do have . . .


3- 6 limit holdem: 145,000 hands, W/R 2.07 big bets/100 hands

5-10 limit holdem: 47,000 hands, W/R 1.68 big bets/100 hands

8-16 limit holdem: 29,000 hands, W/R 2.49 big bets/100 hands

3- 6 limit stud : 78,000 hands, W/R 2.61 big bets/100 hands

*

She showed a profit at all other limts/games (2-4 and 3-6 Omaha/8, 1-2 and 2-4 [blinds] NL holdem, 1-2, 2-4 and 3-6 [blinds] Hi-only PL Omaha, and some 2-4 and 3-6 hi-draw) but the total number of hands at all these combined was less than 50,000, and if she broke these down by game she didn't tell me of it. (Note: The profit from these "other" games was nominal).

I referred to these records as incomplete since no effort was made to calculate S-D., no tallies were kept regarding swings - either up or down - and, there were no records for individual sessions.

Here is my/her question.

Having zero knowledge of the types of games but precise knowledge of the win rate (for what most would agree is a fairly large sample), how accurately can we estimate her S-D and/or confidence level ?

Using the 3-6 holdem as an example (since it has the largest sample size) . . .

What is the consensus estimate of the following -

1. Likelyhood that this person is at least a 1 big bet/100 hand winning player ?

- I am under the assumption that this is a virtual certainty.

2. Likelyhood that she is at least a 1.25 big bet/100 hand winning player ?

- I have this as upwards of 95% but I am very open to the idea that I am dead wrong about this one.

3. 1.5 b-b/100 ?

- Far more likely than not (75% ?) but this is the one about which I'll not be at all suprised to find I was mistaken in my estimate.

4. If we can formulate a reasonable guess as to her S-D that would be great as well.

- If it helps, she is nothing if not solid though by her own admission somewhat risk averse; her win rate is (and likely will continue to be) held down by this unwillingness to push small edges.

. . . Then again, if she can win a few hundred over the next week she'll clear 50k for the year after first learning the game in 2004 - I wish I'd made that in my 2nd year - or 3rd, or 4th. (Or 12th [img]/images/graemlins/cool.gif[/img]).

*

There is consistency here; the drop when moving from 3-6 to 5-10 is as would be expected, and the surge at 8-16 can be explained away as extremely dilligent game selection. (In this case, "selection" is a slight misnomer since it refers to whether or not to play, not to which game to sit in; there is rarely more than one 8-16; often there are none.)

Thanks to all who take the time to tackle this one; I'd like to help her if I could.

She's my fiance's best friend - MAJOR suckup points are at stake here. [img]/images/graemlins/grin.gif[/img]

Re: In the library with the
 

[ QUOTE ]
Here is my/her question.

Having zero knowledge of the types of games but precise knowledge of the win rate (for what most would agree is a fairly large sample), how accurately can we estimate her S-D and/or confidence level ?

[/ QUOTE ]

I think that a typical SD for limit holdem is around 15BB/100 hands. It won't be precise. Hers could be higher or lower, but given your description of her play (not pushing marginal edges, etc.) I would expect her SD to be lower if anything. In other words, I think we can safely assume that 15BB/100 is a worst-case scenario SD.

[ QUOTE ]
Using the 3-6 holdem as an example (since it has the largest sample size) . . .

3- 6 limit holdem: 145,000 hands, W/R 2.07 big bets/100 hands


What is the consensus estimate of the following -

1. Likelyhood that this person is at least a 1 big bet/100 hand winning player ?

- I am under the assumption that this is a virtual certainty.

[/ QUOTE ]

Using the 15BB/100 estimate for SD, I get 99.7%

[ QUOTE ]
2. Likelyhood that she is at least a 1.25 big bet/100 hand winning player ?

- I have this as upwards of 95% but I am very open to the idea that I am dead wrong about this one.

[/ QUOTE ]

I get 98.1%

[ QUOTE ]
3. 1.5 b-b/100 ?

- Far more likely than not (75% ?) but this is the one about which I'll not be at all suprised to find I was mistaken in my estimate.

[/ QUOTE ]

I get 92.6%

[ QUOTE ]
4. If we can formulate a reasonable guess as to her S-D that would be great as well.

- If it helps, she is nothing if not solid though by her own admission somewhat risk averse; her win rate is (and likely will continue to be) held down by this unwillingness to push small edges.

[/ QUOTE ]

See above. Also, here is a SD poll in the archives that shows that the median SD for the respondants was about 15 or 16 BB/100.

I hope this gives you some idea of what her true winrate might be. It would be good if someone could verify my numbers just to be sure.

< Thumbs up >
 

The 90%+ confidence level for "at least 1.5/100" is not what I was expecting; I was anticipating 70-75 and was leaning toward thinking it might be as high as 80 but 92.6 is a very pleasant suprise.

Thanks a bunch. [img]/images/graemlins/smile.gif[/img]

BTW, re. win rates for 3-6 . . .

What figures have been tossed around here in terms of the highest possible ?

I would think while there are a few "freaks" who might attain some monster numbers plaing one table, 3 per 100 would be about as high as a multi-tabler could realistically expect to attain; am I correct ? . . . Or am I confusing MY potential with that of others ?

Both she and I play mostly at STARS so assume no rake rebate(s), which leads then to the question -

How much rake does the typical winning 3-6 player pay per 100 hands ?

- I have always assumed something in the neighborhood of $10/100 for a full (9 or 10 handed) game; does this sound about right ? If so, a 25% rebate would be nice and would add substantially to the bottom line but would not turn too many losing players into winners, nor would it be enough to make a very small winner want to contemplate "quitting his (or her) day job".


Happy holidays,

- Chris

*

Note: I emboldened "winning" since in all but the rarest of instances winners pay less rake then non-winners. Also, her wont to pass on marginally profitable opportunities will lead to her paying even less rake. When I asked how much rake I was thinking of a 1.5-2.0 bb/100 winner whose style would best be described as middle of the road as regards agression.