bed.g  XY points well outside the delta triangle

Hello
The Duet calibration instructions for delta printers https://duet3d.dozuki.com/Wiki/Calibrating_a_delta_printer say the following about the bed.g file:
"you need probe points that are well outside the triangle formed by the towers"
My Anycubic Kossel Linear Plus frame (680 mm tall towers, triangles with 300 mm sides, 240 mm diameter bed) looks like this seen from above:
Where can I find those probe points well outside the triangle ? To make things worse, I have a borosilicate glass over the original aluminium bed and I think the glass is not "flat" at the edge, but it's tapered, so I actually have less than 240 mm in printable areas's diameter.
Thanks in advance,
Mihai 
What it actually says is:
For typical delta printer builds, 6factor is best. You should be able to measure the diagonal rod length accurately enough, and for auto calibration to be able to determine it accurately, you need probe points that are well outside the triangle formed by the towers. Likewise, without points well outside that triangle, auto calibration cannot accurately distinguish the effect of bed tilt from misplaced towers.
I guess it's not as clear as it might be. What I meant to say is that if you want to use 7 or 9 factor calibration (which adjust the rod length) then you need probe points well outside that triangle. So use 6factor auto calibration.

Hello Mr DC42,
I understand. I have Hayden Magball rods with stickers saying each rod's length.
Also I followed your recommendation from topic
https://forum.duet3d.com/topic/11243/deltaxandyheadpositionchangeduringzautocalibration/5
and I specified one single rod length in the M665 L parameter (the average length as resulted from the equation 4288.14+2288.15=288.143), waiting for the 2.04RC1 firmware which can handle correctly three rod lenghts in M665 L parameter.
And I chose a 6factor in the bed.g file, since the manufacturer of my Hayden rods gave me the rods with stickers on it. I hope this is the correct way to do things, please can you just give me a quick "yes" or "no" on this.
Thank you,
Mihai 
@mihaitintea said in bed.g  XY points well outside the delta triangle:
Hello Mr DC42,
I understand. I have Hayden Magball rods with stickers saying each rod's length.
Also I followed your recommendation from topic
https://forum.duet3d.com/topic/11243/deltaxandyheadpositionchangeduringzautocalibration/5
and I specified one single rod length in the M665 L parameter (the average length as resulted from the equation 4288.14+2288.15=288.143), waiting for the 2.04RC1 firmware which can handle correctly three rod lenghts in M665 L parameter.
And I chose a 6factor in the bed.g file, since the manufacturer of my Hayden rods gave me the rods with stickers on it. I hope this is the correct way to do things, please can you just give me a quick "yes" or "no" on this.
Thank you,
MihaiYes.