War games and war problems have not yet been accepted by some; for some regard them as games pure and simple and as academic, theoretical, and unpractical. It may be admitted that they are academic and theoretical; but so is the science of gunnery, and so is the science of navigation. In some ways, however, the lessons of the game-board are better guides to future work than "practical" and actual happenings of single battles: for in single battles everything is possible, and some things happen that were highly improbable and were really the result of accident. After nearly every recent war there has been a strong move made toward the adoption of some weapon, or some method, that has attained success in that war. For instance, after our Civil War, many monitors were built, and the spar torpedo was installed in all our ships; after the battle of Lissa, the ram was exploited as the great weapon of the future; the Japanese War established the heavily armed and armored battleships on a secure foundation; and the early days of the present war caused a great rush toward the submarine. Yet, in most cases, the success was a single success or a very few successes, and was a little like the throw of a die, in the sense that the result was caused in great measure by accident; that is, by causes beyond the control of man, or by conditions that would probably not recur. The game calls our attention to the influence of chance in war, and to the desirability of our recognizing that influence and endeavoring to eliminate it, when reasoning out the desirability or undesirability of a certain weapon or a certain method. Of course, every thoughtful person realizes that few effects in life are due to one cause only, and that most effects are due to a combination of many causes; so that, if any weapon or method succeeds or fails, it is illogical to infer from that one fact that the weapon or method is good or bad. A common illustration is the well-known fact that a marksman may hit the target when his aim is too high or too low, provided that he has erroneously set his sight enough too low or too high to compensate; whereas if he had made only one error instead of two, he would have missed. "Two wrongs cannot make a right," but two errors can compensate each other, and often do. The theory of the Probability of Errors recognizes this. In fact, if it were not true that some errors are plus and some minus, all errors in gunnery (in fact in everything) would be additive to each other, and we should live in a world of error.