This one's only counting active mass shooters. When it's still a lesser shooting with under 4 victims, the odds of a vigilante rando with a gun - that is, a citizen packin' heat and not a cop off the clock - stopping the violence is about 1 in 7000.
There are two different categories: "active shooter" and "mass shooting".
An "active shooter" has strict definitions and is tracked by the FBI. These are the events depicted in this graph. An active shooter is someone trying to kill people at random in a public place. The number of casualties is irrelevant. A few years back a guy tried to attack a courthouse in Texas and was killed by a cop before he even got a shot off. That still counts as an active shooter.
A "mass shooting" has no single definition, and media and government organizations that use the term set their own parameters. Many of them define it as "four or more people killed or injured", regardless of circumstances.
The problem with the term "mass shooter" (and the reason why the FBI doesn't use it) is that it's overly broad. Guy goes nuts and kills his family before offing himself? Mass shooting. Robocop shoots four guys in the dick? Mass shooting.
EDIT: It's worth noting that the linked source clarifies that the graph shows all active shooter incidents between the year 2000 and 2021. This throws off your calculation significantly.
No it's not, dgu's happen all the damn time. Hell there is a subreddit that tracks the ones that are found. There are countless videos of people being attacked, and pulling a firearm and the violence magically stops. That's a DGU, even though no round was fired. So it doesn't show up on lists like these, which have an agenda.