There is a very interesting form of lightning called "ball lightning." Many scientists do not believe in this form of lightning because there is no physics theory that can satisfactorily explain it. Obviously, the Ball-of-Light Particle Model does explain ball lightning. Ball lightning are balls-of-light.
Ball lightning can be created when two strokes of lightning collide. Typically one stroke travels from cloud to ground and the other stroke travels from ground to cloud. When they meet, if polarized correctly, some of the energy can wrap around itself in a semi-harmonic pattern.
Graphic of lightning strokes creating ball lightning
Sometimes the two strokes are so highly polarized and have such highly matching wave patterns that more than one ball will be created. A string of balls is created. This is the so-called "bead lightning."
Graphic of bead lightning
Sometimes, cloud-to-cloud lightning will create ball lightning that is visible from the ground or from planes.
Ball lightning seems to come in certain sizes. One common size is about the size of a basketball. The limit to the size is based upon the energy in the strokes that create them.
Ball lightning appears to be inherently nonharmonic. For example:
Ball lightning decays in an explosive fashion sometimes because the nonharmonics can be influenced externally or simply start creating a chain-reaction type of greater, induced, nonharmonic fields. (If ball lightning was hit by a cosmic ray, then this could induce forces that quickly explode the particle.)
Ball lightning is observed to have different colors. Sometimes the light given off appears: red, white, yellow, or orange. The harmonics on the surface of the ball determine the wavelengths of the light that is given off.
If you try to pass an electrical current through a wire, you will discover it has a certain electrical resistance. If you try to pass an electrical current through a wire with a metal ball in the middle of it -- especially if the magnitude of the current accelerates -- then you will discover the wire has a greater resistance. At first this may seem backwards to some -- "Shouldn't the thicker part of the wire have less resistance?" What happens is the electrical current will induce magnetic fields in the metal ball which increase resistance -- nature is try to create a ball-of-light.
The metal ball on the top of flag poles increases the resistance of the pole. It is harder for lightning to travel down the pole. This is intuitive for many people: it is easier to create a static shock off of something with a sharp point than something with a rounder end.
While the ball on top of a flag pole might reduce the chance that lightning will hit, if lightning does strike a pole with a round end on it, it seems to me that this would increase the likelihood that ball lightning will be formed.
The Ball-of-Light Particle Model predicts it will be possible for scientists to create ball lightning on demand artificially either with: large electrical voltages across conducting metals spheres as in...
graphic of my spark experiment
Or, using current lightning experimental equipment, it is should be possible to shoot rockets into thunderhead clouds that trail a conducting wire in order to force lightning to jump a gap such as in this experiment...
Graphic of rocket, wire four balls with one ball-sized gap