Of all the ideas that have the potential to cause controversy, this one seems to strike the most nerves. I have always been willing to conduct Magic: the Gathering matches by chess clock, and have never had any of the problems that are claimed to be associated with such use. However, it seems appropriate to address these concerns, and my experience with them, before going into the details of how to go about implementing chess clocks for Turnabout.
Too much priority passing
Whenever I have presented the idea, someone thinks that it is clever to point out that, under the most strict interpretations of the Magic Comprehensive Rules, there are numerous priority passes that take place, even on a basic turn where a person might simply play a land and do nothing else. This fact supposedly renders the idea of a chess clock meaningless.
The argument itself would only work if several factors were true that simply are neither true nor relevant. The first assumption is that I am talking about having people "send the clock" whenever a person would gain priority. No one does that now, so to think that it will happen just because there is a clock present is silly.
The second assumption is actually a challenge to my first response. It is that people will choose to waste time off of the other player's clock by insisting on every priority pass. The way such an event would happen is if a third assumption were true, namely that the clock in question is a simple chess clock with no frills.
The answer to both of these lies in realizing exactly what is being proposed. Most advanced clocks today have one or both of the following options: Bronstein timing, or delay timing. Bronstein timing allows for a "refund" of time spent, up to a certain limit chosen by the event organizer. Delay timing creates a timer that counts down between ending one person's time and beginning the next person's time. In both cases, the phenomenon of deliberately striking the clock to pass priority just to burn time off the opponent's clock fails because the opponent can simply pass it back within the time limit and not lose anything.
In addition, the arguments can be made just as strongly against the current shared time model, if not moreso. If someone actually tried to enforce every pass of priority during every turn, it is likely that a judge would issue a slow play warning. In a normal match, that is clearly taking away from the shared match time; in the clock situation, if any time is being wasted, it is the person's own time, and as such, I have no objection to people being idiots on their own time.
The clock shouldn't influence the game
Another argument against chess clocks is that they shouldn't influence how the game is played. A game where each side only has a certain amount of time to play will remove certain deck strategies from the environment unfairly.
If the goal is to allow all strategies to be playable, the match should be untimed. If there is to be any sort of time limit on the match, it will, by definition, harm some strategies over others. In such a case, the only question is one of whether to divide the match time evenly through chess clock technology or let each match divide the time based on what happens during the match.
DCI would never sanction such an event
Regardless of what the DCI would sanction, this is currently an unsanctioned format (except as Casual Non-Rated Constructed). If it's already unsanctioned, what harm is there in adding another element?
So now that these are out of the way, let's go about dealing with the implementation:
Length of Games
It is recommended that each player have 10 minutes and that the Bronstein or delay time be 3 seconds. The clock should measure game time, not match time. This gives 40 minutes of play time, plus 8 minutes of prep time (5 in the first game, 3 in the second). Together with the various time that will add due to the refunded or delay time clocks, this will put the projected end of the match at about the same time as normal.
For Bronstein clocks, the 3 seconds need to be set as the refund time, and also added onto each person's base time (so each player's time should be set to 10:03). For delay timers, this is not necessary; give each player 10:00 and set the delay timer for 3 seconds.
Player shortcuts and communication
There are already a number of shortcuts that players use to get through games without doing every priority pass. These are fine and should be encouraged. If a person wants to do something on another person's turn, he should ask for the clock.
Resolution of spells and abilities
The only question that remains over and above the basic is whose clock should run while resolving spells or abilities. The answer is that the person who controls the affected object should resolve things on his time. If objects controlled by both players are affected at once, the player controlling the spell or ability resolving should have the clock running for whatever time is needed to resolve his or her affected objects; then if the opponent still has things to resolve, the clock should pass to resolve them.