Active Device - selection criteria

The election of the active unit after fail over is dependent on the traffic groups configured failover method. The only method that leaves any ambiguity to the election is the load aware method as the HA Order and HA Group methods as well as the auto failback option make it clear which unit will be elected as active after failure. In the case of load aware failover we use the concepts of the current and next-active devices to determine the active member but the split brain scenario you described in your question certainly introduces ambiguity to those concepts. There's likely a tie breaker when all things are equal between the two F5's but I'm not privy to the particular mechanism. I'm sure there's someone out there that knows the particulars.

https://support.f5.com/kb/en-us/products/big-ip_ltm/manuals/product/bigip-device-service-clustering-admin-12-0-0/6.html

That being said though, if predicting the active unit is a concern in your load aware failover environment - use the load aware mechanisms (HA Load Factor setting, HA Capacity setting) to avoid an unpredictable situation like you've described. Alternatively use the Auto failback setting or the HA Order/HA Groups failover methods.