Jump to content
Announcements
*** Punchestown Festival Competition: Well done to 1st. Craig Bluenose, 2nd. Lee Grays & 3rd. Carole Dawney ***
** April Poker League Result : 1st kevsul, 2nd Rob Valk, 3rd McG **
** Last Man Standing Results - glavintoby & Redno2009 both win £125 **
** April Naps Competition Result: 1st Sugardaddyken, 2nd TRAINMAD091, 3rd Harry Faint99, 4th Gary66. KO Cup Winner Waggy, Most Winners: Xtc12 **

Value Betting, a new approach


Recommended Posts

Do you feel lucky? Tell me! Do you feel lucky?

If the answer is yes, you can play Lotto, Keno, Joker, or you can place a bet. But beware, only ONE bet, if you are counting on luck.

If you want to place more bets in the long run, luck does not play a significant role, betting is a game of patience and ability. To gain profit, through time, the only thing you need is to make better estimates than the betting companies. Every reasonable person may wonder how. How can a simple betting player estimate the odds better than the giants who possess unlimited resources and the most capable human resources?

This is a hard task, but not an impossible one. Many punters try to predict value bets by using the “famous”, in the gambling world, Poisson distribution. Well, it’s not working, at least efficiently. Maybe it can predict possible value existence in a game but rarely produces viable results, mainly because it can’t “see” the non-linear nature of football games.

After years of trying, fail, and trying again, I come up with a fresh approach to help me pre-select the games I must deal with.

First, we must see the football divisions we care to bet on as autonomous entities. There are big differences between England League One and Bundesliga, for example, so I like to use a fresh approach to estimate the fair price of football outcomes.

Initially, I calculate a set of variables (predictors) for every game based on past results, special characteristics of every football division, etc.

Then for every predictor (there are 13 of them) and for every division I pick optimized expected goals equations to predict the percentage of a team, to score a specific number of goals in a specific game. These calculations take place for the home and guest team for every game.

These percentages are not in a distribution form. I’m sure you can remember a lot of games expected to end with 0-1 goal based on Poisson predictions and end with a score like 3-3 or 5-2, etc. The expected goals equations can represent this kind of behavior, unlike distribution-like or linear type predictions.

There is no automated system to make you rich, your participation is essential in every decision, BUT with this method, you can have a credible assessment of every aspect of a football game.

You can see how the fair odds “move” on various scenarios in a single game. In my new book “MATH to WIN: Football”, available on Amazon bookstore, I present the optimized equations for every division, I explain how to build game scenarios and how to apply educated reasoning with the help of tens of real-life examples. 

If you are a casual punter and you don't want to spend much time in analysis and other boring stuff, the book can help you by providing a series of informational tables for every division. These tables will be your sidekick in order to select games with the positive or near-zero value, so you avoid the bookies' traps.

If you will spend more time on analysis but you are not doing so well with computers you can apply the simplest form of equations. You only need an average familiarity with MS Excel or a similar program. The simplest equations dealing only with the game's outcome (1, X, 2) and are quite easy to implement.

Now, if you want to get the most of this method, you will use the expected goals equations. You must have at least average computer skills, especially in MS Excel or a similar program.  If you are familiar with programming the things get more interesting because you can try to build your personal models.

On my website, www.mathtowin.com, under the Downloads section I will provide, completely free, the "raw material ", the predictors' values. With these values, you can implement fully my method, but also you can use them to build your own models.

I would like to have your comments and feedback, the participation makes us better!

 

Edited by johnkam
Link to post
Share on other sites
  • 2 months later...

Hi,

 

I downloaded the sample of the book and thought it was very good so went and bought it. However on page 20 you break down prices by odds bands, ok but not sure what the acronyms VS-Hw etc. are meant to mean? Be easier just to show the odds bands no?

Also on page 84 you then start introducing things like homeSCFT, homeSCPprob etc. and these are used as the basis for the equations. Now you hint at what they are, not in the book but on the website but no full explanation anywhere. The R notebook simply has values pre-done. I think it would be of use to anyone of a technical nature to be able to get how these figures are calculated. I code in Python and I've done stuff like attacking/defensive strength for things like Poisson and I already pull data from football-data completely automatically using a VPS. 

Seems a shame you don't provide all the calculations and it means people are then reliant on you to keep updating those numbers in the spreadsheet on a daily basis.

 

 

Gary

Edited by garyk1968
Link to post
Share on other sites
  • 3 weeks later...

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...