Footyforecast getting rid of unlikely draws
Here are the fundamental guidelines:
For each team work out the following;
1. The total number of points obtained for the last N games.
2. The maximum number of possible points for the last N games.
3. Divide the total number of points obtained by the maximum available and multiply by 100.
4. Calculate the forecast value.
In (1) and (2) above, N games might be all the home games for the home side and all the away games for the away side. N might be the last N games consisting of all home and away games for a team.
The forecast value is determined like this;
HOME POINTS = number of points for home team from last N games.
AWAY POINTS = number of points for away team from last N games.
HOME VAL = [HOME POINTS / (POINTS FOR WIN * N)] * 100.
AWAY VAL = [AWAY POINTS / (POINTS FOR WIN * N)] * 100.
PROJECTION = [HOME VAL + (100 - AWAY VAL)] / 2.
To compute the possible result of a match based on the Footyforecast technique, the value is compared with the following;
1. A projection value of 50 = a draw.
2. A value between 50 and 100 provides an increasing possibility of a home win the closer to 100.
3. A value in between 50 and 0 provides an increasing possibility of an away win the closer to 0.
There are a couple of variables to think about, for instance the number of matches to utilize and whether to utilize all matches or simply home for home side and away for away side. You might want to try out these values.
By outlining real resulting draws against the projection it is possible to produce 2 threshold values, one for away wins and one for home wins, any values in-between these thresholds are most likely draws.
As an example, with a value of 40 or less for away wins and a value of 60 or more for house wins. This means any match falling between 41 and 59 may therefore be draws.
What this technique does, with mindful tuning by the user is to remove lots of matches which will not be draws thereby providing you with a list to pick from.
Here is a worked example:
The values utilised are the points gained by the team for each game in a series of 4 current matches. Of course, you might pick more games to base your computations on.
West Ham
H4 = 3 (earliest match).
H3 = 1.
H2 = 1.
H1 = 0 (latest match).
Leeds Utd
A4 = 1 (earliest match).
A3 = 3.
A2 = 0.
A1 = 3 (latest match).
.
(NB: Utilizing just home games for home side and just away games for away side).
FFP Home = [(3 + 1 + 1 + 0) / 12] * 100 = 42.
FFP Away = [(1 + 3 + 0 + 3) / 12] * 100 = 59.
FFP Forecast = [42 + (100 - 59)] / 2 = 42.
If our threshold values are 40 and 60 then for this match the forecast lies on the anticipated draw area and at the lower end meaning that if it is not a draw the most likely other result would be an away win.
This might be translated as an X2 forecast, i.e. draw or away win, which most bookmakers accept as a bet.
Now its your turn.
Of course you may use different values to those shown here and by experimenting you may come up with better values to use. You might likewise decide to utilize all home and away games played by each team in your estimations rather than just simply home games for the home team and away games for the away team. You might likewise choose to outline real results against the Footyforecast system's forecasts to see how many real draws fall in the away win, draw, and home win forecast zones.
Good luck.