As pointed out a few times in individual forum posts:
If you have an irregular shape in your antenna pattern you should pay attention to the parts that stick out or have a hole.
These can cause "nulls" in your coverage and unexpected coverage holes.
Take an azimuth chart where you have some parts of the signal around the 0dB line, that will represent your correct gain. But for example if you an area where the line crawls a bit inward, that may not show a very steep indent but in reality that is much bigger. If that line where -6dB then the actual covered distance is HALF! compared to where it hit the 0dB line.
A perfect example is the MR33 which has a shape where the signal goes much farther sideways.
I drew red arrows where on the polar char it goes slightly inwards and a green arrow where it goes a little outward.
On the actually predicted software where the numbers are in meters you can clearly see the effect that is exaggerated.
So beware if you take high gain dipole antennas on AP's mounted on high ceilings. The signal could travel far horizontally but have poor coverage on the ground.
The same is true for outdoor AP's with dipoles mounted on the ceiling instead of the wall, so you don't get much range horizontally but cover the floor above it 😜
Thanks @GIdenJoe for putting this out there for others to reference. It also reminded me of the days some time ago, when everyone started buying their home wifi routers, and there were also high gain omni antennas available separately. I remember on a couple of occasions, having the explain the "doughnut versus pancake" effect of the high gain omnis.
One friend had their wifi router in their basement office, and was getting a lousy signal in the 2nd floor bedrooms, so bought the high-gain omnis, and was complaining the signal upstairs got even worse. Once they realized the concept of passive gain and that those higher gain omnis actually sent less signal up and more signal into the earth through their basement walls, it made sense to them! They simply went back to the original antennas and found a better place to locate the router. 😁
Heh, that reminds me of the days as a cable technician.
We had alot of customers where the cable entered the house in the basement and the different coax cables to the outlets throughout the house also arrived there, so we had to put the modems there. And later we had these modem/routers with Wi-Fi.
Of course since most of the Belgian houses are concrete between floors and have stone walls Wi-Fi was really poor and we couldn't do much to help them unless they were smart enough to have some UTP's through the house where we could attach an access point. Later we also had these powerline homeplugs which went from 14Mbit PHY to 85 Mbit PHY second gen and later up to a few 100 Mbit. But you were still limited to the quality of the electrical installation and possible interfering electrical devices (usually led lighting with poor power supplies causing drops in the homeplug communication.. then you would go there and see that problem and try to find old telephone cables you could replace with utp cabling by trying to pull the telephone cable out while having the utp attached... those were the days 😉
That ISP in more recent times used a protocol TR069 to communicate with those secondary AP's or powerline homeplugs with wifi in them to get the same SSID config as the modem/router.