A predial algorithm looks at the call completion phase for an agent, and dials another number in anticipation of the agent becoming available to take a resulting live call. Terms used to describe this process include 'progressive' and 'intelligent' dialing. The predial time, which we call the predial interval, is described in Predial Setting in Oceanic®. In all discussion of predial, it is assumed that as live calls are made they may be allocated to any waiting agent.
This algorithm is used extensively by telemarketing and scripting products which don't have access to the more sophisticated overdial algorithms of a predictive dialer. In a world of fast call setup and network response, the maximum reductions in wait times that occur will normally be several seconds only. And if the aim is to avoid any abandoned calls at all as a result of predial, then even this is probably ambitious (see especially Sober Thoughts On Predial).
Two agents on a campaign, Jake and Louise, are about to finish a call, and are both pooled for predial purposes. Predial is initiated for Jake first, and then Louise. Jake finishes first as expected, but the first connect is on the line dialed for Louise. Oceanic® assumes that Jake will be connected to this number. See also Progressive Dialing.
Overdial and predial algorithms may be used in combination; see Combining Overdial and Predial for how to do this in Oceanic®. When this happens in Oceanic®, the number of lines (pre)dialed, is determined by the overdial algorithm, and hence may be considerably more than in progressive dialing, where typically just one line is predialed per agent.
In a world where ACDs no longer take seconds to get a call onto the network, the scope for predial is really quite limited. If we have given it a lot of attention in Oceanic®, then it's simply because it is a way of dialing much talked about, (but perhaps not as widely used?).
If you want to explore predial in more depth, be sure to start with Sober Thoughts On Predial .