Chap 1: Introduction

Football betting is crazy. So much money is transacted each week that it is amazing how most punters continue to wager on football matches while not fully understanding the various factors that influence outcome. Many have made an effort to gain more knowledge on the subject but there is still relatively little literature out there to equip the betting public with an organized framework to see the game with a clearer perspective.

Like what I mention in previous post, there are softwares out that can predict 1x2, accurately 70% of the way but never 1 that is able to try predict asian handicap. True, the hit rate is good but any punters or user will tell yu most of the time, it will predict home win, but the payout is so miserable it can be wipe out after a series of 'off-day' betting. So at last maybe yu stand to gain 20% or so if yu are lucky that year, the problems lies with the weightage for such calculatons with tend to skewer to the home team , hoping that majority of the hot/home team will persists in their run, but the winnings are usually very little too significanly cover the loses. At last balancing to null.

The aim of my publish is to better equip the reader with predictions from a chart (formulas) that would provide a guilding light thru satistical analysis, in other words, over time, matches that have similar hadicap will tend to follow a certain trend or buck(reverse) a trend given all teams/players/pitch/weather are equal in a home/away circumstances. Naturally, the only effects are generated from results from previous weeks so as to statiscally visualize the next upcoming matches based on availability of asian handicap. My investigaton have the below assumption;

Teams all start off same in a year, no favourite & no underdogs, the champions or the just prometed have equal staus, previous reputation is zap.

Direct factors like players injuries or coach perferences matters nothing, we are not playing fantasy football but actual handicap prediction.

Indiect factors like pitch condictions/wheather is discounted.

Cup matches does not apply here, it is between a home or away team in a league over a season of home and away matches.

The statistical chart examines all the known factors that influence the outcome of football matches and simplify into a existing theory; all events will repeat itself given the same handicap over a period of time, if there are too many variable, the chart is able to pick it up and hightlight to the user, helping him/her in lesser a mistake when going all out for a bet.

I have generally classify the tools that people used to pick a winner to simply 3 major methods;

The home advantage chart

This chart generally factors in the home team with a initial handicap of 0.2, to accord for the 12th person which is the home crowd, tiz chart can also be call goals difference chart as a significant scoring in a match can pull a team up in points and improve their standing in the statiscal table, hence good for early prediction in the early weeks (4th-8th weeks) when form not set in.

The form chart

This chart generally factors 2 forecoming teams with their last 5 match form coming to play, i give a weightage to home win/home draw/away win/away draw/straight lost and compared the outcome in realtion to the asianhandicap.

The sequence chart

This chart can symbolises the runs/stops of overcoming the asianhandicap, I only have a thought but I have not maths out the statiscal capability of such a chart, but I know most khakis to have use its methods many a time in the betting battlefield. Most common guide, wah the f***king team won 3 in a row, time to lose lah, this is widely used in software prediction tools for 1x2 results, and is the most premitive tools i noticed in soccer prediction. However there are perks and cons in using such a method, i will mention further in future post.

Last but not list I will have 2 sets of predictions in each chart , 1 set is the nearest to the outcome from previous results, another is the nearest of 3 to the outcome from previous results, I do this as a localised sampling of results to achieve a more homogenious and non-bias prediction, I have before embarked on near 5/10 localised sample but it is too tedious and chaotic to maintain so down to 3.