Success to the successful archetype represents two reinforcing structures which may be in a delicate balance though as soon as one gains a small advantage the resource allocation favors the more successful and the result is then rapidly skewed in the direction of the more successful. There is also a video
for this model with is a component of the Systems Archetypes Course
This structure consists of two growth or reinforcing loops which are linked though the allocation of resources based on the comparative success of two activities.
When you first encounter this situation there may be a steady level of progress for both activities, though what's more likely the case is that there is a greater allocation of resources to one than the other causing one to demonstrate far more success than the other. And as such, continuing to be allocated more resources than the other.
The simulation structure makes the allocation variance a bit more obvious as [alloc var] may be positive or negative and [factor] is simply an adjustment variable which can have values between zero and one. [pulse] and [when] are used to shock the structure.
With [factor] = 0.4 and [pulse] = 0 resources are equally allocated which produces a steady accumulation of success for both a and b.
If at [when] = 4 we give b just a little nudge with [pulse] = 0.1 we see a very rapid deviation in the results of a and b. This is the nature of a growth structure in action. Just a little difference is quickly amplified.
If at [when] = 4 we give b a larger nudge with [pulse] = 0.4 we see an even more rapid deviation of the success of a and b.
Strategy: Seldom is this recognized before the fact, or after the fact. It's typically just assumed that either a or b is a poor performer and is a problem to be dealt with. What's appropriate is to look deeper and understand what's happening and correct the allocation it's not too late to correct. Once you go over the falls in the barrel it's not time to ask if it was a good idea.
Change the sliders to run the model with various parameters to see how they impact the behavior of the model. Also change the Simulation Time Step in Settings to .5, .25, and .125 to see how this affects the nature of the graphs.