After reviewing Army "combat developments," with special reference to the 1962 U.S. Army Tactical Mobility Requirements Board ("Howze Board"), this report argues that the Army could improve combat developments by closer connection and interaction between its model building and testing activities. Presently, models and their simulations are uncritically used with little attention given to whether the model is empirically validated. This will be troublesome for the Battle Labs when they use simulations to carry out their work. Another problem is that many data or numerical inputs used in models may be of questionable quality, often because they are the output of some other invalidated model. These conditions suggest that the Army's system should have a mechanism that (1) tries to lay out programs of models and assertions about tactics and operational performance that can be empirically validated and (2) sequentially field-tests those assertions. It may even be necessary for the Battle Labs to take on or acquire this function to carry out their objective. Otherwise, imperfect expedients like the Howze Board will continue to be used.