Have your production track, beeps track and playback track all routed to headphones and room speaker on diff (postfader) sends within the session.
Keep the fader for the production -> room speaker send visible in the session.
Set the beeps and production track to a low but audible level with the track faders. Leave the playback track at unity.
When recording a take, pull the production -> room speaker feed down during the beeps and leave it down for the duration of the take. After the take, option+click the send fader (not the track fader) to return it to unity. This will return the production playback level to the low but audible level you set before.
To play back a take, move the good take up into the playback track, cue and playback the line, and during the beeps pull down the production TRACK fader instead of the
speaker send. This will kill the production track everywhere.
If your actor has his headphones on, just leave the production track send to the room speaker all the way down and work the track fader during takes.
If you have someone in the room who doesn't want to hear production during takes, then you can do the headphone feed PREFADER and work the track fader in the room.
ADR cutting is often about making sure that everyone involved is hearing the appropriate thing at the appropriate time.
I once had a session where I had a headphone feed for the actor (low beeps, low production throughout), one for his voice coach who was on the stage next to him (low beeps, no production ever), one for the line producer who was wearing headphone in the control room behind me (moderate beeps, loud production, loud takes and playback-for checking sync), and a mix in the room for me and the film director (beeps, production lead in, no production during the take-so he could listen for performance).
Also this actor was the lead, this was a big budget animated feature where millions had already been spent on animations synced to the guide tracks, we had every one of his lines to do, this was a comedy so timing was key, and I had the actor for very limited amount of time. IOW, I couldn't waste time not knowing what was going on in every headphone feed I had running.
It was pretty intense, but very rewarding once I had finished. That's probably the most complex an ADR setup can become, but you have to be prepared for the varying levels of complexity between that and what we just discussed.