"Do everything else that seems reasonable, but kill the shell game." That was the substance of Ronald Reagan's strategic decisions on Friday, and -- whatever else you think of them -- waffling this was not.

There won't, and shouldn't, be any significant opposition to the proposals to modernize our strategic communications -- this area has been neglected far too long in the eyes of strategic mavens of a wide variety of political persuasions. Nor should there be any real balking at the submarine force improvements. A highly accurate submarine-launched Trident II missile is only prudent, especially given the difficulty we seem to be having getting a land-based one deployed; and putting more cruise missiles on our general-purpose submarines is a relatively low-cost and effective way to complicate Soviet nuclear war planning. The thrust of the strategic defensive improvements also appears reasonable in light of the run-down condition of North American air defense -- some room for quibbling, perhaps.

The B1 decision will provoke a fight -- necessarily almost a wholly uninformed one, because the competing "stealth" bomber's characteristics, prospects and problems are highly classified. A close call.

The big decision, and the big surprise, was the complete cancellation of the MX shell game. To deploy a few MX's in very hard silos for a short time -- until these also become vulnerable in the late 1980s due to Soviet accuracy improvements -- is to provide this part of our ICBM force with something more than a fig leaf of survivability, but something considerably less than a full pair of pants. (The administration's written statement essentially concedes this point.) Further, unless there is a major breakthrough in ABM technology (in a field, as Harold Brown puts it, other than ABM public relations), the decision to forgo multiple shelters makes ABM defense of MX considerably more difficult. This is because the more shelters you deploy that the bad guys would have to shoot at, the less your ABM has to do. With just a few MXs deployed in silos, and no shell game, any ABM is going to have to be very sophisticated in order to do its job; it will have to be able to shoot down large numbers of Soviet warheads before being destroyed itself. A possibility by the late 1980s, but a real -- so to speak -- long shot.

So where does that leave us with the problem of ICBM vulnerability with which we began? Is it real? Yes, and the administration admirably has not followed one or two academics and a gaggle of journalists into giddily thinking that Soviet ICBMs are not accurate. What about the highly advertised mid-'80s window of ICBM vulnerability? Because of the other strategic force improvements one might argue that, in context, it is less serious -- but it is undeniably still there. The major problem that the "window" once posed, it still poses: namely, there is a need to have another ballistic missile force, in addition to the submarine, as a hedge over the long run against Soviet anti-submarine warfare improvements. Bombers are fine, but they require warning from vulnerable satellites or radars to escape from their bases. Prudence requires more in this, the nation's most important job -- deterring nuclear war.

The administration's program thus bets heavily on having one of two possible basing modes for the MX by the late 1980s. By then, its written statement says "we will have a better system." These are 1) deep underground basing -- burying missiles, e.g., in mesas in such a way that unrealistically massive attacks would be required to destroy them; or 2) continuous airborne patrol aircraft.

The latter particularly deserves careful and energetic study. The job of detecting, and tracking, and destroying dozens of such aircraft, flying anywhere over tens of millions of square miles of ocean, is not simple. The physics required is imaginable, but the engineering and electronics required to threaten such aircraft could be prohibitively expensive for the Soviets -- especially if we use clever countermeasures. In any case, the possibilities are great for aircraft that are so fuel-efficient (8 to 10 times better than current large jets -- as good as some types of ships) in roles other than carrying MX. For example, long-endurance aircraft may have utility as communications platforms, for ocean surveillance, or for military and commercial transport. (Perhaps not a bad way to reinforce Israel, or the Persian Gulf -- with that sort of fuel efficiency you can virtually get there and back without refueling.) So we ought to get on with them, and with the other R&D projects.

Indeed we have to. There's no going back now -- the shell game is dead without Ronald Reagan's support. But remember, such people as Sen. John Tower, the senior Carter administration officials, and -- most of all -- many officers of the U.S. Air Force are not crazy because they have long faces today. They supported the shell game because they felt a strong sense of duty to lead the country toward having a form of ICBM basing that -- however cumbersome -- had at least a chance to survive. With just the right combination of luck, sacrifice, effort, ABM protection or arms control limitations, it might have worked -- but the risks were great. Now, instead of taking that uncertain and difficult route -- or even maintaining it as a hedge -- we have headed down a path with different risks: we have made a high-stakes bet on our Yankee ingenuity.