Hi edymurph, I’ve attached 2 spreadsheets. One is on the forecast function in Excel 2016 and the other is on Pivot Point Calculations. My prediction ideas are explained below:
Trying to predict the next figure in a number series.If you look at the average or median of 1820 Lotofacil draws for level 9, 586 is the average or median figure.
Step 1: DirectionIn trying to determine if the next result would go up or down, there seemed to be a pattern (most of the time but not always 100% accurate) that if the latest result was above 586, the next number would be lower that the latest. If it was below 586, the next one would be higher than the latest (see the sheet “directions test”)
Step 2: Input figurePivot points. These are mathematically calculated levels used in stock markets and financial trading to determine where the market is next likely to move to. On the spreadsheet attached the first main three types called Floor, Camarilla and Fibonacci are often used by traders.
Regarding the level 9 calculations, if you use the last 3-4 results and convert them into High, Low and Close, they will give pivot calculations. You could also use just the last 2 figures if you want, (this is something to experiment with).
So for example, if the latest 4 figures were 571,578, 592, 576
This would be High 592 Low 571 Close 576
Close will
always be the latest figure you are using
If you were just using the last 2 figures, it would be High 592, Low 576, Close 576
After entering these into the pivot points spreadsheet, you have to decide where you think the next figure is likely to go. If you try some pivot calculations, they can sometimes be rather accurate, but then sometimes they are off a bit from the next figure. And also you have to try and select the best level to use.
With the Match Winning Numbers Filter, it needs 3 levels predicted with 3 correct answers to find the winning ticket. Since the direction may be able to be guessed, the pivot points are there to try and choose a good input figure to use in the Filter.
Excel 2016 Forecast SheetI also tried another idea that requires Excel 2016 installed. In the Data tab there is a feature called Forecast Sheet.
You could put in the last 50 or 100 results (or even a shorter number) for level 9 in column B. In column A, you need to put a dummy date in so the Excel forecast feature will work. Eg type in 1/1/2017 and drag it down the cells in column A.
Highlight all the data in columns A & B and click on forecast sheet in the data tab.
Click on “options” at the bottom, set the forecast start date back say 7 days before the latest. Make the forecast end date 3 days ahead of the latest. You could try confidence interval 80% (this figure makes the upper and lower ranges narrower, so you can experiment with this figure also), and for seasonality, select “set manually”.
In the manual seasonality box, type in 1, and look at the graph and see if the orange line is matching up to the blue line actual results. Then try 2, look at the lines etc. Keep going until you find the best orange forecast line that best matches the blue line. When you select the one you like, Excel then makes a new page with a graph & it’s predictions.
When I tried this data on Lotofacil level 9, numbers in multiples of 4 seemed to look ok, such as 4, 8, 12, 16, 24 and 32. These gave the best looking graph results.
You can also try different forecast start periods such as 6, days, 4 days before the latest etc. 7 days seemed like a good start date to try and find a good orange forecast line to match up to the blue line data. And also the forecast end date could be larger than 3 (and also less at just 1), but projecting further ahead, how accurate would the forecast be??
The actual predicted or forecast figures from Excel may or may not be completely accurate to use, but it is interesting to note if they are predicting the next figure will be higher or lower than the latest one. And also if this is a reliable prediction figure that it gives.
This is the link to the Microsoft instructions on this feature.
https://support.office.com/en-us/articl ... a7062329fdHope the above ideas help a bit