The Math of "bad lines"

[lurkerforages]

Sundevil, don`t take it personally but those of us who understand the math of sportsbetting are in the minority.
After reading the reply from Dave at CBS, the first thing that hit me was this guy doesn`t understand the math of sportsbetting.

Do you honestly think you will get more hits than a thread about how big some girls jugs are?

 

[FunkSoulBrother]

Sun Devil (or anyone else):

How do you determine the optimal money amounts to bet on each side for a scalp?

For example, in your example:

"Scalper A has $2000 to play with and finds one side of a game at +100 and the other at +120. Few would call one of these a "bad line" but Scalper A can play $1047.62 on the +100 side and the remaining $952.38 on the +120 side and guarantee himself a profit of $95.24 on the game."

How do you find the magic numbers of $1047.62 and $952.38? Is it the intersection of two lines or what? I would be most interested in a formula so I could better play these when I find them.

 

[SunDevil]

FSB, the "goal seek" tool in Excel is a wonderful thing. It's kinda tough to describe the entire spreadsheet involved verbally, but you'd want to have a cell with your starting bankroll and a cell where you input a number you plan to bet on side A. A simple subtraction formula could then be used to determine the amount remaining to be bet on side B. Once you have the amounts being bet on each side you can use simple formulae to determine the amount you win if side A wins less your amount you lost betting side B which equals your net profit if side A wins. Similarly you can apply the same idea to calculate your net profit if side B wins. The key to using the "goal seek" tool is to then create a cell whose formula is the net profit if side A wins minus the net profit if side B wins, then use goal seek to set this cell to value zero by changing the cell for the amount bet on side A.

Phew, like I said, it ain't easy to describe spreadsheets in words. Hope that helps. If not, maybe SPV can help since it's all his math anyway

 

[SunDevil]

Lurker,

Nah, I go for the boobies first myself, but I do have the simple hope to out-do a wannabe tout.

 

[SunDevil]

To all that supported me here I owe a partial apology. My numbers, while they are truthful, and ARE my own - despite SPV's claims - do not tell the whole story.

As I have now realized and written in another thread here, all this bankroll analysis assumes that the bettor wants a completely locked-in profit and is willing to bet accordingly even though he knows which of the two lines he's betting into is the one that's off the mark. This results in him laying down (especially as we inflate the size of the favorite in the "fair line") a very large percentage of his bankroll on the favorite (the "fair line")and only a minimal amount on the dog (the "bad line") - which SEVERELY diminishes his long-run equity if this is to be a large sample space of events. He's made much better off just by taking a position on the dog and letting it ride.

So, for a COMPLETELY RISK-AVERSE bettor my analysis holds, but otherwise it fails. Time for some flagellation...

 

[Jazz]

SD - I always assumed your example implied that it was for the 'completely risk-averse bettor'. One could always overbet whichever line one liked, too, so as to net zero if one was wrong - i.e., overbet the dog and if the dog loses you net zero, but win a good deal more than if you'd made a 'pure' scalp. Or as you said letting the doggy ride - all options depending on the circumstances.

 

[Jazz]

Funky - do you know how to code an Excel spreadsheet? If not, take some time to learn - search on 'formulas' and go through the tutorial. Then it's pretty simple. I assume others probably have better ways, but I've coded 8 different 'model' lines that can do the calculations for the 8 possible different situations for scalping or siding/middling. When a situation arises, I copy the model line down, then plop in the numbers (odds on both sides of the bet as well as one of the bet amounts) to get the proper second bet to place. The formulas are all designed to either maximize your profit (scalping) or minimize your cost (siding,middling).

Here's one of the model lines, used to calculate a scalp as in the example in this topic - a cell consists of the intersection of a row and column on the spreadsheet, and in this example the model line is located at row 21. This model line makes you enter the amount you're betting on the favorite line (the negative line) - I have another which makes you put in the dog amount ... do NOT enter negative signs with ANY of these values! You may have to change the column attribute to a 2 decimal number for precision, but that's easy ..

Row 21 on the spreadsheet:

col A - amount bet on the favorite
col B - dog odds (enter '605')
col C - favorite odds (enter '435')
col D - left blank for spacing
col E - col A + col F (total amt bet for reference)

col F - calculated amount to bet on the dog side ... formula on next line entered with '=':

= (A21 + (A21 / (C21/100))) /( (B21/100) + 1)

col I - dog win amount ... formula on next line entered with '=':

=F21*(B21/100)

col J - favorite win amount ... formula on next line entered with '=':

=A21/(C21/100)

col K - payout amount for a dog win (includes amount bet and amount won) ... formula on next line entered with '=':

=F21+I21

col L - payout amount for a favorite win (includes amount bet and amount won) ... formula on next line entered with '=':

=G21+J21

col M - maximum profit, based on the dog winning - this was arbitrary, since the profit is identical or almost exactly so when the favorite wins, and I was too lazy to calculate the average of the two which is certainly easy ... formula on next line entered with '=':

=((B21/100)*F21)-A21


If you enter this on some row other than 21, just make sure to change all references to 21 to the row you're entering it on.

Anyhow, once you've coded it, save your spreadsheet and then copy the model row to another row - never modify the model row unless you're doing it to change a calculation. That's because if you overtype a number in a cell with a calculation, the calculation usually no longer works, so in the example above, don't overtype the amount to bet on the dog in column F!

 

[raiders72001]

Jazz- Any possibility of you emailing your spreadsheets to me? I would like to try them out. If yes my email is raiders72001@yahoo.com.

 

[Sugarbear]

Brillant thread. Very few real good math debates/disscussions exist anymore in the archives that abound in the different places. This is the best one I have seen because it includes tutorials and no flaming.

 

[royalfan]

I cannot believe everyone is googa gaga over this post. It is flawed logic at its finest. The scalp on the heavy juice lay side has absoltely nothing to do with it. The 605 vs -435 is off WAY more than 120 vs 100. Some of you need to let simple common sense be your guide. Very amusing though to be sure!!

uppy:

 

[winbet]

I cannot believe everyone is googa gaga over this post. It is flawed logic at its finest. The scalp on the heavy juice lay side has absoltely nothing to do with it. The 605 vs -435 is off WAY more than 120 vs 100. Some of you need to let simple common sense be your guide. Very amusing though to be sure!!

uppy:

I think this proves my point on how everyone is sharper nowadays, its not got the same surprise as it had then, although you would have to worry what they were thinking then. :drink:

 

[kidslick]

Brillant thread. Very few real good math debates/disscussions exist anymore in the archives that abound in the different places. This is the best one I have seen because it includes tutorials and no flaming.

Fezzik has some posters that are top-shelf with math questions, SBR has a real good math guy ( Granchow, or something like that). Some places are good for specific guys. RX is good for quick opinions by the masses.

http://forums.lasvegasadvisor.com/

 

[Sugarbear]

I am interested in the old debates for a reason kidslick. Much different types of things lie in those old threads that were added by the non math people.


Plus the time line of the first 10 years of the offshores can be traced along some of the better stuff I have found. Very interesting how the offshores thought about things in there infancy and youth.