The Accidental Adversaries structure represents a situation where two interacting entities which should produce growth unwittingly limit the growth of each other because of their actions to promote their own growth. There is also a video for this insight which is a component of the the Systems Archetypes Course.
This structure begins with two entities A and B whose activity toward each other should create a reinforcing structure R1 promoting the growth of each.
B sensing there are impediments to their success implement fixes which further promotes the success of their business.
In all likelihood A is doing the same thing, which would be OK, except...
What makes this an Accidental Adversaries structure is that [A Fixes] create [A Obstruction] which tends to diminish [B Success] and [B Fixes] create [B Obstruction] which tends to diminish [A Success]. The resultant R4 Vicious Reinforcing Loop ends up amplifying the initial degradation of success.
It should be easy to tell that the stock & flow version of this model mirrors the causal loop.
There are a whole bunch of assumptions built into the relationships in terms of the magnitude of the feedback paths and you can take a look at them when you interact with the model. Suffice it to say that when the model is run with no fixes in the mix both A and B experience success.
With A and B applying fixes to correct how they believe their businesses should be performing results in a dismal situation for both of them.
Strategy: There are essentially a few options here. 1) All parties involved need to stop defeating each others success and rely on the collaboration; 2) Stop pretending to be collaborators and just compete with each other; 3) Figure out how to do fixes that don't derail your collaborators.
I encourage you to look at the underling assumptions built into the model in the various variables and flows and run the model with different parameters until you're comfortable with what's happening here.