Today’s installment of Book Club looks at Chapter 5 of “Men Against Fire” (MAF), which introduces the metric at the heart of Marshall’s argument: the eponymous ratio of fire, or the fraction of men engaged in combat who fire their weapons at the enemy. Marshall reports that this ratio was astonishingly low in the WWII engagements he investigated, never surpassing 25%. He begins to draw certain conclusions from this, and I think it has interesting implications for other areas of human endeavor.
Marshall conducted “post-combat mass interviews with approximately four hundred infantry companies in the Central Pacific and European theatres”. He found that “on average not more that 15 per cent of the men had actually fired at the enemy positions or personnel … the figure did not rise above 20 to 25 per cent of the total for any action”. Marshall notes that most of these engagements “had been decisive local actions in which the operations of a company had had critical effect upon the fortunes of some larger body and in which the company itself had been hard-pressed”.
Writing of those men who do fight, Marshall observed:
- In the engagement which brought the low ratio of fire to his attention (an attack by the Japanese on a battalion of the 165th Infantry Regiment on Makin Island, in November 1943), the majority of those who fired were “heavy weapons men. The really active firers were usually in small groups working together.”
- “[I]n the main the same men were carrying the fire fight for each company day after day. … You could pick out your man who would probably keep going until he was dead. Or for that matter, after a few trial rounds, you could spot the man who would probably never get going though his chances of dying were relatively good. … For it must be said in favor of some who did not use their weapons that they did not shirk the final risk of battle. They were not malingerers.”
- It is not possible to determine, prior to combat, which men will take the fight to the enemy. The notion that discipline correlates with combat effectiveness is misguided; many of the most effective fighters were poor soldiers, but a larger number were not.
- It was commonplace to find individual soldiers of previously undistinguished service who “became pivots of strength to the entire line when fire and movement were needed … the sustained action of these men … rallied others around them.” Such men were not always appreciated by their commanders, who felt their authority threatened by them.
Marshall’s results have been disputed, but let’s take them at face value, and see what they might suggest for those of us who enjoy lives of peace.
The obvious place to begin is with one of Marshall’s own parallels:
In the workshop or office, or elsewhere in society, a minority of men and women carry the load of work and accept the risks and responsibilities which attach to progress; the majority in any group seek lives of minimum risk and expenditure of effort, plagued by doubts of themselves and by fears for their personal security.
As on the battlefield (although Marshall intended the analogy to work the other way) 80% of the progress is accomplished by 20% of the people. Marshall was concerned with increasing the number of men who contributed to the battle, so as to arrange matters to secure victory for his side more quickly and at less cost. My aim is more modest; to enquire how, in an organization, more of the men within it may usefully advance its aims, for the profit of all.
Ratio of What?
Marshall zeros in on a very simple metric in the “ratio of fire”. This metric has several nice properties:
- It’s easy to measure: A soldier knows whether or not he fired his weapon in an engagement, so a quick survey will yield a ratio.
- It’s relevant: Some people argue that all combat is just fire and motion, so the question of whether or not fire is happening is pretty clearly to the point.
For most other organizations and enterprises, however, the metric is murkier. What do you measure for software development? Don’t, for the love of all that’s holy, measure lines of code, or pages of documentation. Don’t measure an organization on the CMM(I) scale. What those three things get you, respectively, is:
- Excessively verbose, non- or barely- functional code
- Redundant, impenetrable, legalistic documents
- The process as an end in itself
I don’t know what you measure for software. I think the real point of software is to understand a problem, and then to explain it to a computer. If you were going to measure something (i.e. an activity, not an output), you’d want to ask your programmers: “Did you try to understand something today”? Unfortunately, unlike firing a weapon, there’s a lot of room for interpretation with this question, given the considerable gulf between “sort of” understanding something, and “really” understanding something. (Also problematic: How deep an understanding is appropriate for the problem at hand?)
It is a puzzlement.
The most interesting observation in this chapter is that the most active men tended to fight in small teams, and that the heavy weapons men (i.e. those on crew-served weapons) tended to acquit themselves well. This suggests that the XP practice of pair programming may not be entirely wrong-headed; a pair of men actively engaged in the work of an organization may motivate one another to aggressively press that work forward.