Sklansky bucks calculator (php script)

[runout_mick]

[ QUOTE ]
I missed your girlfriend's graph, it didn't load earlier.

[/ QUOTE ]

Nope, I snuck it in there on an edit.

Thanks for all the work you've put into this, it's much appreciated. Now I've got a new toy to try and figure out.

[meleader2]

so not only am i running bad, but i'm super aggro in the middle and decide my hands aren't good enough at the end?

this is month of may

[Vilica]

I think this shows that it is a good thing I play limit and not nl.

[DWarrior]

OK, this is my db from January. According to the statistics, over 45% of my winrate is luckboxed.



Red line is when I got coaching, and if you look closely, the green line has been outperforming the cyan line pretty steadily.

The pink lines are just to show the discrepancy increase, because the up/down shifting may look confusing. Then I just moved them closer together above the green line to make comparing them easier.

Is it possible that there is a tiny error in the calculations that is causing the discrepancy?

[Phil153]

Hey DWarrior,

It's great that you're asking these questions. I've spent a lot of time investigating accuracy, but it's always possible I missed something.

Steadily increasing lines are my biggest worry that something is slightly off, and I've spent a lot of time going through them by hand. I didn't release earlier for this very reason. But the variance you have is very possible - for the first 30K you ran exactly as expected, which means that the variance possible is at least as great as what you experienced in the second half, regardless of accuracy.

[hoppscot22]

i think that simply means that you are winning a lot of money without going to showdown

[Phil153]

I also have a couple of questions:

- Did you change sites when you started coaching? It's always possible I missed site specific data in the database.

- What does your "all in before river" line look like?

Thanks for your help.

edit: And one more question. Do you play a site/stakes with a 5% rake structure? If you don't, the program could improperly assigning rake to later streets such as the river, where you often have a better hand.

I haven't had time to implement rake tables for the sites or for limit, and my testing showed that it had a negligible impact on results. Perhaps there is something in the way you play that's affecting it. Or not. I would like to investigate further though.

[DWarrior]

lol, I'm gonna stop editing now.

I'm gonna read your reply then possibly post the example as a separate post [img]/images/graemlins/laugh.gif[/img]

[DWarrior]

OK, this is the scatterplot in question. The reason I removed it from the previous post is because I don't exactly understand what I'm looking at so I may be totally off. I read it as money won/lost plotted against sklansky $ from river bets only.



This is Scatter, BB vs SB, River

Shouldn't money bet on the river exactly match sklansky dollars? If you graph that and hover over some of the points, you'll see that a lot of the time when you win money it exceeds the Sklansky $ (for example won $17, sklansky $16).

Assuming my understanding is correct, how do you explain these:

The strangest points are the ones in the center (winning when having 0 or -1 sklansky), like the right-most center point is $44 won, -$1 sklansky.

[Phil153]

The river SB measures how much you put in vs your expected share of all the money that went in on the river. Against one opponent, they should be equal as you see in the two straight lines. Against more, they'll be higher.

The few that are slightly less than $0 are due to split pots + the effect of rake, which is removed from the pot before SB is calculated. From your $ I'd say you'd play small enough stakes that the rake is still being taken out on the river.

Example (5% rake), on river.

Pot: $20 - $1 rake. You bet $20, he calls $20, you split. Winnable pot is now half of $40 - $2 in rake. = $38, so you lose $1 in SB.

Unless I'm misunderstanding you, I don't see a discrepancy here.