Page 1 of 1

Knight's Move

PostPosted: Sun Oct 30, 2016 3:38 pm
by dreamweaver
Hi Stan,


I am not sure if there is an anomaly in the following filter or I have misinterpreted and applied the workings of the
following Filter : KNIGHT'S MOVE.

From my understanding the Knight's move filter mimics the same movement pattern that a Knight chess piece would make
being L-Shaped. Also according also to the documentation the

" Knight's Move statistics compare the winning numbers of analyzed draw/ticket with the winning numbers from the previous draw. The statistics then show how many numbers from the previous draw made chess-like knight's move into the analyzed draw."

I have applied this filter to predict the number of Knight's Move to the Australian Powerball Draw 15/9/2016:

The statistical results show there was (3) such Knight's movements that occurred between the 8/9/2016 and the 15/9/2016.

Draw 15/9/2016 : Winning Numbers : 2, 12, 13, 27, 29, 32
Draw 8/9/2016 : Winning Numbers : 9, 10, 17, 25, 28, 31

I have calculated that there was (4) such movements. Please see Figure 1 :

Figure_1.JPG


I have also attached the number GRID to graphically show the where those movements occurred.

Figure_2.JPG


As you can see:

Number 17 can make L-shaped knight's move to numbers 2 and 27 respectively.
Number 28 can jump to number 13
Number 10 can jump to 27

Giving a total knight's movement count of (4). But the statistical properties reports a Knight's Movements
of (3) as per figure below:

Figure_3.JPG



If you can help clarify why I get a different result that would be greatly appreciated.


Thanks Stan..

Re: Knight's Move

PostPosted: Wed Nov 02, 2016 10:45 pm
by stan
The knights move stats and filter are actually based on a grid with 10 columns (see here http://www.expertlotto.com/en/help/en/s ... _move.html). So you can't check the knights move on your ticket panel grid.

Re: Knight's Move

PostPosted: Thu Nov 03, 2016 10:42 am
by dreamweaver
Hi Stan,


Thank you for the reply much appreciated but the grid for a Powerball coupon is design with 5 rows x 8 columns
which is different from the grid you have referenced. That grid is for a 49 number lotto game, the Powerball
game has a pool size of 40 numbers. Would it not be more appropriate and more significant that the grid designed
for a particular lotto game be used in the Knight's Move analysis. The grid you have referenced can possibly give
a whole set of different characteristics and outcomes.

I have also noted the L-shaped movements for the two draws that I used in my example and performed a Knight's
Move analysis based on the Reference grid you have provided, I end up with (4) such events and not the
(3) that statistical properties reports. They are as below.

1. 10 jumps to 29
2. 17 jumps to 29
3. 25 jumps to 13
4. 31 jumps to 12

So why are we still not getting the result as the statistical properties reports ?


Thanking You

Re: Knight's Move

PostPosted: Sat Nov 12, 2016 8:45 pm
by stan
Ok, I looked at the source code the logic is actually reversed - the stats show how many number in the current draw can move to numbers of previous draw. In you example there's number 29 listed twice but it counts as a single hit. It's a subtle difference but it counts if two or more numbers make jump to the same number. I'll update the documentation so it matches the actual implementation.

Re: Knight's Move

PostPosted: Sun Nov 13, 2016 8:09 am
by Meisterkeks
Hello,

In Germany the tip has 7 rows and 7 columns. I have the two draws
22.10.2016
26.10.2016
Compared and come to the same problem:

Image

Image
Image

Can you also correct?