[isetcap]

I think these tools were designed by this guy...

http://www.homestarrunner.com/vcr_sm.html

[jjgold]

I think they would look cool all on big page, maybe each category color coded.

Great Stuff here for the gambler

[Ganchrow]

The Poisson calculator now provides the option for both one and two variable calculations.

Let me know if anyone finds any errors.

[Tchocky]

Excuse me if I sound like an idiot but the only ones I understand are the parlay calculator and U.S./Decimal odds converter. Can someone please explain all the other ones?

Freeplay Value calculator description

[pags11]

excellent work as always ganchrow...

[RickySteve]

Very nice to have such useful tools in one convenient place. Thanks a lot, Ganchrow.

[BetByInternet]

nice.

[rezster]

a valuable resource, thanks for posting

[jjgold]

I use these tools all the time, we should have a link on the homepage also.

[Aces]

Nice

[TLD]

I’ve been using the “Free Play Calculator.” My unofficial rule is to hold out for at least 80% value on a free play moneyline scalp.

[bookie]

You know what would be cool to see add to the scalpe value calculator, or as a thing on its own...

A scalp bet dividing calculator. If line one is X and line two is Y to assure equal profit (loss) on each outcome bet amount A on X and B on Y.

[Ganchrow]

Originally Posted by bookie

You know what would be cool to see add to the scalpe value calculator, or as a thing on its own...

A scalp bet dividing calculator. If line one is X and line two is Y to assure equal profit (loss) on each outcome bet amount A on X and B on Y.

Unless I'm misunderstanding you, what you're describing is already what the "Scalp Value Calculator" does.

[bookie]

You're right. I hadn't played with it enough to realize that.

[Tchocky]

Originally Posted by Ganchrow

I'll start off with the Freeplay Value Calculator, which, as its name implies, figures the value of a freeplay bonus based upon betting and hedging at specifioed prices.

For example let's say that a book offers you a $200 freeplay which you bet at +200 and hedge -215, You'd enter $200 into the "Freeplay Size" box, +200 into the "Freeplay Line" box, and -215 into the "Hedge Line" box.

When you click "Calculate" you see that the proper amount for the hedge bet would be $273.02 and your locked-in profit would be $126.98. This means that you would have captured 63.49% of the notional ($200) value of the freeplay.

Now compare that to the same $200 freeplay, bet at -110 (entered into the "Freeplay Line" box) and hedged at +120 (entered into the "Hedge Line" box). After clicking "Calculate" you see that the proper hedge bet under these circumstances would be $82.64 and the locked-in profit would then be $99.17. This means that you would have captured 49.59% of the notional value of freeplay.

The conclusion which could then be drawn from this is that betting a freeplay at +200, and hedging at -215 would be a better usage of the freeplay than betting at -110 and hedging at +120.

Thanks for the explantion, Ganchrow.

[Ganchrow]

Originally Posted by Tchocky

Thanks for the explantion, Ganchrow.

No problem.

Just let me know if you want an explanation of anything else.

[isetcap]

ganchrow, you've got some great tools...

[Korchnoi]

An interesting result that is clear only really when you hedge out your freeplay risk is that you should prefer a "bad" longshot bet to even a great even money bet.

Even if you can find a bet +120 and hedge it out +100 (thus finding a bet with 4.76% edge) you only capture something like 54% of the freeplay value. Compare this to betting a longshot at +800 and hedging it at -900 (losing 1.76%) -- your freeplay value is something like 76%.

The only difference between a free play and regular bet is that in a freeplay you don't get your initial stake back. Therefore, you maximize your freeplay value by reducing the extent to which the fact that you don't get your original stake back doesn't much matter.

On many sites there are restrictions on how long your odds can be on a freeplay. On BetCris the max payoff of a freeplay is $5000 and the max freeplay bonus is $500. I put mine all on the Patriots to win the Super Bowl at +1000 (which happens to be both a longshot and a bet with edge). I'll probably let it ride for a while.

Thanks for the tools Ganchrow!

[Ganchrow]

Originally Posted by Korchnoi

An interesting result that is clear only really when you hedge out your freeplay risk is that you should prefer a "bad" longshot bet to even a great even money bet.

Even if you can find a bet +120 and hedge it out +100 (thus finding a bet with 4.76% edge) you only capture something like 54% of the freeplay value. Compare this to betting a longshot at +800 and hedging it at -900 (losing 1.76%) -- your freeplay value is something like 76%.

Very, very true.

To present an even more dramatic example, if you could find a +100 bet for your freeplay that you could hedge out at +1000 (representing a gigantic scalp of 69.23%), your profit from using the freeplay would only be 90.91% of the freeplay value -- the same profit you could expect betting the freeplay at +2000 and hedging at -2100.

Just as you said -- stick with longshots when betting your free plays.

[RickySteve]

The return can even be driven above 100% by finding huge moneylines which can be scalped. This is a rare scenario, but lovely when it occurs. (Hint: basketball playoffs)

[LLXC]

These are some nice tools! Thanks

[jjgold]

I use these tools all the time, no site has all these different options of calculators.

[cPaxus]

Ganchrow,

Thank you. Saved them on my Laptop. Useful tools.

[Ganchrow]

Originally Posted by cPaxus

Ganchrow,

Thank you. Saved them on my Laptop. Useful tools.

Glad you liked.

People need to let me know if there are any other tools they'd like me to put together for them.

[betso]

Ganchrow or anyone else, could you please explain to me the usage of the Poisson, Reverse Bet and Round Robin Calculators?

I am familiar with the Poisson distribution for random numbers, but not really how it is being used in a practical betting context. Thanks.

Poisson calculator description

[betso]

Thanks Ganchrow.

Originally Posted by Ganchrow

One variable Poisson calculator:
...

Solution:

1. Select "One Variable" radio button
2. ...
3. We see that the probability of hitting the under ("Less than 15½") is 55.7847%, corresponding to a fair money line of -126.17

Very good example. Just to make sure I am with you, I got 54.76% for the probability here.

Originally Posted by Ganchrow

Two variable Poisson calculator:
...

Solution:

1. Select "Two Variable" radio button
2. ...
3. We see that the probability of hitting the Nets -1½ ("A +1½ < B") is 52.0527%, corresponding to a fair money line of -108.56
4. ...

Cool. I fiddled a bit in Excel with these numbers and for a moment thought that this should be equivalent of using the one variable with expected average (15.2+1.5) 16.7 and proposition 17, but does not come out the same. The reason why this approach could be wrong is that both the proposition and expected average in this case are Poisson random variables. Is this correct?

Btw, is there anywhere I can find historical data for analysis of the numbers mentioned in the example (3pt attempts etc)?

[Ganchrow]

Originally Posted by betso

Very good example. Just to make sure I am with you, I got 54.76% for the probability here.

54.76% is correct. I must have miscopied the data from the calculator (original post fixed).

Originally Posted by betso

Cool. I fiddled a bit in Excel with these numbers and for a moment thought that this should be equivalent of using the one variable with expected average (15.2+1.5) 16.7 and proposition 17, but does not come out the same. The reason why this approach could be wrong is that both the proposition and expected average in this case are Poisson random variables. Is this correct?

That is correct. What you'd want to do would be create a matrix where the value in the i^th row, j^th column = the probability of seeing exactly i of event A and j of event B, which (because the two events are independent) = P(exactly i occurrences of event A) * P(exactly j occurrences of event B). (In theory there should be an infinite number of rows and columns. In practice, the numbers converge rather quickly.)

Then to get the probability that A + 1½ < B, you'd just sum over all values in the matrix such that i - j < 1½. This is exactly how I've implemented it in the Java Script.

Originally Posted by betso

Btw, is there anywhere I can find historical data for analysis of the numbers mentioned in the example (3pt attempts etc)?

DatabaseBasketall has game logs going back several years. I'm not sure where you can find historical prop bet lines, although that shouldn't really be necessary.

Chapter 9 of Stanford Wong's Sharp Sports Betting gives several examples of how to use the Poisson distribution to model relevant prop bets.

[RickySteve]

Originally Posted by Ganchrow

Let's say a book is offering up a prop bet on a combination of events, each one of which you to believe to be Poisson

... and uncorrelated.

[Ganchrow]

Originally Posted by RickySteve

... and uncorrelated.

... and independent.

[betso]

Originally Posted by Ganchrow

...
That is correct. What you'd want to do would be create a matrix where the value in the i^th row, j^th column = the probability of seeing exactly i of event A and j of event B, which (because the two events are independent) = P(exactly i occurrences of event A) * P(exactly j occurrences of event B). (In theory there should be an infinite number of rows and columns. In practice, the numbers converge rather quickly.)

Then to get the probability that A + 1½ < B, you'd just sum over all values in the matrix such that i - j < 1½. This is exactly how I've implemented it in the Java Script.
...

Cool little algorithm, this helps my understanding.

Originally Posted by Ganchrow

DatabaseBasketall has game logs going back several years. I'm not sure where you can find historical prop bet lines, although that shouldn't really be necessary.

Thanks, the data looks good, it will come in handy for analysis.

Originally Posted by Ganchrow

Chapter 9 of Stanford Wong's Sharp Sports Betting gives several examples of how to use the Poisson distribution to model relevant prop bets.

I gotto get that book.

[betso]

Thanks for the tools Ganchrow.

Attached is an implementation in Excel for some of the calculators above (the discussions on this thread and going through their code helped me understand their use).

[Ganchrow]

Originally Posted by betso

Thanks for the tools Ganchrow.

Attached is an implementation in Excel for some of the calculators above (the discussions on this thread and going through their code helped me understand their use).

Very nice, but hasn't anyone ever told you to always use Option Explicit in VB.

Next up you may want to try implementing a multiple simultaneous bet Kelly Calculator.


Good job all around.

[betso]

Originally Posted by Ganchrow

Very nice, but hasn't anyone ever told you to always use Option Explicit in VB.

Thanks, this is basically the first VBA code I write, so Option Explicit is new to me. Seems like a sensible thing to do though. Thankful for any other comments (and possible bugs) that might exists.

Originally Posted by Ganchrow

Next up you may want to try implementing a multiple simultaneous bet Kelly Calculator.

On my home computer I think I got the multiple kelly bets in a python implementation which I copied from Steven Brechers "Beating the races with a computer". I will be back on this.

[Ganchrow]

Originally Posted by betso

Thanks, this is basically the first VBA code I write, so Option Explicit is new to me. Seems like a sensible thing to do though. Thankful for any other comments (and possible bugs) that might exists.



On my home computer I think I got the multiple kelly bets in a python implementation which I copied from Steven Brechers "Beating the races with a computer". I will be back on this.

I wouldn't mind checking that out. Could you PM the source code? But just to be clear, is this an implementation of simultaneous pooled bets on the same outcome (as one might expect from horse racing), or of simultaneous fixed odds bets on a specified number outcomes. The latter is significantly more useful in sports betting.

[Poventino]

Handy tools when you wager International games. Thanks.