It can happen,
Most likely one rotation earlier the ball was at dd over the number 1 (maybe 24) but the ball was a bit stronger. VB2 increases prediction with faster ball. That faster ball because of strong tilt does not stop on dd, 3, 6, 9 but it makes full additional rotation.
If you look at it as times of ball.
One spin 1100 ms ball may travel 6 rotations.
Next spin you may have 1070ms , this ball is a bit faster it will predicted a bit more but not so much as full ball rotation. So there are always braking points which are impossible to predict.
You may have even 1000 ms once make 6 once 7 ball rotations.
The ball is not 100% balanced, it is has marks, the ball track has dust, the ball doesn’t always cross same micro obstacles, sometimes it travels a bit easier way so it arrives to the last rotation a bit faster.
Leveled wheel may have 4 vertical diamonds where each theoretically takes 25% of hits.
Tilted wheel may have one dominant diamond, which groups wider range of ball speeds.
1000 may hit dd 12, 1050 dd 9, 1100 dd6, 1150 dd3
On tilted wheel you may have ball speeds
1001 to 1200 hitting dd12 after 7 rotations
Than 1201 to 1400 hitting dd 12 but after 6 rotations
It is logical to expect that ball speed ranges 1180-1220 are unstable and often will make different amount of rotations. It is impossible that the 1200 always makes 7 rotations and only 1 ms slower ball makes 6.
What goes in favour of VB and a bit more in favour of VB2 is that instead of 24 it predicted 1 a bit more. Often you have common ball jumps 10-15 pockets, but the ball which made additional rotation will arrive to dd very slow and hit it at bottom part. Such ball often doesn’t jump so even it dropped wrong it may stay in area you played.