U.S. Joint Forces Command was a major actor in the area of joint wargaming during this period (it would not be deactivated until 4 August 2011). USJFCOM was already the “transformation laboratory” of the U.S. military; much of its experimentation was done through wargaming. JFCOM’s wargames were primarily supported by two subordinate organizations: the Joint Warfighting Center, which managed the games, and the Joint Warfare Analysis Center, which endeavors to improve their accuracy, particularly in the area of systems effects. Most JFCOM wargames were conducted at the Joint Training, Analysis, and Simulation Center, in Suffolk, Virginia. JFCOM was formally dissolved in August 2011.

Millennium Challenge 2002 (MC02) was the first big Title 10–like wargame run by JFCOM. MC02 was big in every way: cost ($250 million), personnel (13,500), and area (twenty-five locations across the United States). It was also very complex, consisting of twenty-three workshops and sixteen experiments over and above the main game, which itself was adjudicated by a new “federation” of forty-two wargame engines and simula- tions. MC02 was also ambitious, in that it attempted to explore eleven joint-service con- cepts, twenty-seven joint initiatives, forty-six service initiatives, and twenty-two issues from the regional combatant commands.

Unofficial sources have called MC02 a big success, but most of the publicity went to a public exchange between JFCOM and one critic. The incident seems a bit humorous in hindsight, but it illustrates an important point. Shortly after the exercise, a story broke claiming that Millennium Challenge had been fixed. An e-mail from the chief of the Red team, one Lt. Gen. Paul Van Riper, USMC (Ret.), to a friend was cited as proof. The e-mail complained that, at times, the Red team had been restrained inappropriately, allowing a Blue technology or concept to perform better than it would against a real adversary. JFCOM’s initial response was that the e-mail must be a fake or at least taken out of context. General Van Riper fired back with a second e-mail (this one got into the press!) reaffirming the original report. JFCOM then replied that the general must be confused about the difference between a wargame and an experiment, that Millennium Challenge involved both, and that he had not been constrained during the game phases, as such, but that “free play would have been inappropriate during carefully controlled experiments.” General Van Riper insisted that he knew very well the difference between a wargame and an experiment but that while Red play was conducted appropriately during most of Millennium Challenge’s games and experiments, some of the events had been conducted as experiments (including Red play) but had been called “wargames.” The general argued that this was more than semantics, as a wargame was a more rigorous test than an experiment. That is, calling an experiment a wargame gives the outcome more credibility than it deserves.

Following MC02, JFCOM recognized the danger of the “politicizing” problem that the Air Force had spotted a few years earlier: slanting a game to get the (leaders’) desired result. Games set in the relatively near term were needed to “prototype” concepts that could be implemented fairly soon, while wargames set in the more distant future were needed to identify technologies and concepts that would require significant enabling technology to be developed. Instead of establishing the alternating-game system adopted by the Air Force, JFCOM’s director of joint experimentation, Maj. Gen. James Dubik, USA, began cosponsoring the services’ near-term Title 10 games.

And a footnote that does a better job of explaining the lesson they learned:

Many, many e-mails over the DoD wargamer LISTSERV confirm comments made by Barney Rubel during his review of this manuscript: “Amen. I was with Van Riper during MC02 play and observed firsthand the problem with trying to combine a wargame with a field exercise. As far as I am concerned, Van Riper was right on. I interacted quite a bit with JFCOM J9 at the time and at the GO [general officer] level, the whole operation was politicized and the rank and file consisted of a bunch of retired Army colonels who all seemed to have personal agendas of one kind214 or another. J9 grew explosively, with way more money being thrown at it than there were ideas for how to use it.”

Not that any of that answers your question. My best guess is they kept strategic wargames on computers, while using actual troops/assets is for wargaming tactics.