The subtitle of the July 2013 edition of “The Scrum Guide” is “The Rules of the Game.”1 This is an ironic choice. The Rules of the Game is also the title of Andrew Gordon’s in-depth analysis of the Royal Navy’s performance during the Battle of Jutland, a performance that failed to meet expectations and led to bitter recriminations. It is not the kind of performance software teams would wish to emulate.
Jutland was the great naval battle of World War One. In the late afternoon of 31 May 1916, the main battle fleets of Great Britain and Imperial Germany found each other in the North Sea. They fought on and off through the fading light and darkness for the rest of the day and into the night.
For the Royal Navy, the battle offered great promise. Victory over the German fleet would have opened communications with Russia through the Baltic, and permitted offensive action against the German coast. Together, these might have shortened the war.2 And victory was expected. Since Admiral Horatio Nelson’s victory at Trafalgar in 1805, the Royal Navy had enjoyed a preeminent position; no other naval force could compare in size and power.
The promise of victory grew more certain during Jutland’s opening moves. Signals intelligence gave the Royal Navy early warning of German movements, allowing the British to concentrate overwhelming force at the anticipated contact point. British scouting forces successfully located the German battle fleet, and led it toward the Royal Navy’s battle line. The Germans soon came under the largest concentration of naval gunfire in history, far away from their bases, outnumbered, and outgunned. Defeat seemed certain. But the promise was not fulfilled; the German fleet not only survived, but managed to inflict more punishment than it received.3
The failure of the Royal Navy to win a decisive victory is the dominant theme of Jutland. Most assign blame to the fleet commander, Admiral John R. Jellicoe, or his chief subordinate, Admiral David R. Beatty. Gordon’s analysis goes beyond personal explanations and examines the Royal Navy’s system of command. Gordon illustrates how the Royal Navy’s command mechanisms—the “rules” that had been established to guide the behavior of officers in battle—hindered rapid decision-making, crippled individual initiative, and thwarted success at this most critical juncture.4
The primary problem was an overreliance on orders and instructions from above; this created an environment where subordinates were hesitant to act on their own initiative, even in situations where such behavior endangered their forces or their mission.5 Both Beatty and Jellicoe were forced to assume the burden of commanding the bulk of their forces directly. They shouldered this responsibility quite well, but the challenge of attempting to coordinate the movements of a large battle fleet, in fading light and darkness, while maneuvering to intercept a fleeing enemy was too great for any one person, or even a small group. Jellicoe and Beatty needed greater initiative from their subordinates in order to deliver on Jutland’s promise.
This was not something the Royal Navy was prepared to deliver. The limited initiative displayed by subordinates was an unintentional—but wholly predictable—consequence of the system of rules that governed their behavior. The rules took the place of intelligent action. Instead of focusing on using every available means to defeat the enemy, the Royal Navy adhered to the “rules of the game.”
The Scrum Guidance, by creating a similar system of rules, risks creating nearly identical, unintended side effects. Scrum teams often will hesitate when confronted with situations that are not anticipated or accounted for by the rules, rather than addressing the problem creatively on their own initiative. This is common, for example, when access to the Product Owner is limited. With no one to groom or prioritize the backlog, the influx of work slows, and progress begins to stall.
A more insidious problem is that rules can frequently hinder learning, particularly when situations that contradict the rules are encountered. Because the rules provide a context for framing the problem, the most common response is to conclude that the rules have not been implemented properly. The team convinces itself that if they could only be “good enough” the problem would be solved. This view can blind a team to alternative approaches and can hinder the customization of Scrum for their own context.
If problems do arise, wasteful arguments about the correct interpretation and enforcement of the rules are likely, particularly in stressful situations or where failure has occurred. This can easily divide the team and shift focus away from the main goal of delivering software.
Gordon’s analysis illustrates all three of these negative outcomes. Limited individual initiative was a key component of the Royal Navy’s failure to decisively defeat the Germans at Jutland. In the years before the battle, alternative approaches to command were evaluated and discarded; their value was missed because the existing framework—the existing system of rules—prevented a fair assessment of them. And, most visibly, the aftermath of the battle saw a split between Beatty and Jellicoe, which led to a “Jutland controversy,” centered on their different approaches to leadership and their interpretation of the “rules.”6
Rules are necessary to help guide behaviors and align the work of teams. The performance of the Royal Navy at Jutland offers a salient example of the problems that can develop when too much emphasis is placed on adhering to rules. This is relevant for software teams, because software teams—like navies—make it their business to capitalize on dynamic and changing environments. Success in such circumstances requires individual initiative and low-level decision-making. The Scrum Guidance, by emphasizing “rules of the game” risks hindering the ability of teams to capitalize on the initiative of their members and learn from unanticipated circumstances, both of which are goals of the Scrum Framework.
5. The best examples of this are the handling of the 5th Battle Squadron early in the battle (Gordon, p. 81-101) and the failure of the destroyer flotillas to report encounters with the Germans during the night (Gordon, p. 472-499)↩
6. Gordon, p. 537-561; Yates, p. 257-275↩