Espionage aficionados - regardless of if you prefer the spaceship or real-world variety - should immediately read Malcolm Gladwell's article in this week's New Yorker about the WWII British caper 'Operation Mincemeat'. Mincemeat was judged by its masterminds to be an astonishing success, but as Gladwell points out, it is impossible to prove if Mincemeat made any particular impact upon the German distribution of forces due to the regressive nature of the 'expression game'. The expression game will be instantly familiar to anyone involved in EVE espionage, and thankfully for us the expression game doesn't have ugly real-world consequences attached to it when we screw it up. 

style="margin: 10px; border-collapse: collapse; float: right; width: 250px; height: 185px;"
border="1">

href="http://www.tentonhammer.com/node/84294"
title=""> src="http://www.tentonhammer.com/image/view/84294" alt="EVE Online" width="250" height="185"
style="border: 0px solid ; width: 250px; height: 185px;" />

style="font-style: italic;">Spymasters try to get inside their enemy's head.

Mincemeat (greatly simplified) was based around planting false invasion plans on the dead body of a bum dressed up like a British agent, dumped in the waters off the coast of Spain. The idea was to have the Spanish leak the plans to the Germans to give them an air of authenticity, and induce Hitler to redeploy some of his divisions away from Italy and into Greece. The Spanish did behave as the British had hoped, and the Germans did actually redeploy divisions to Greece. But did Mincemeat actually pave the way for the successful Allied invasion of Italy, as the spymasters claimed? Here we run across the nasty causality problems at the heart of espionage itself. 

Did the Germans assign the planted information any credibility? Did they move to defend Greece with additional divisions because of the alleged invasion, or because of its obvious lack of military defenses? When confronted with the invasion plans, did the Germans assume it was a legitimate document and move the divisions to Greece because of it, or did they assume it was an illegitimate plant designed to be obviously fake, to draw the divisions away from Greece, and thus implying that Greece was actually the target all along, and needed defense? This murky rabbit-hole is the expression game. 

In EVE, we are confronted with the expression game every time an agent forwards a piece of information about an alliance-level plan or operation. Gladwell points out that in the real world, critics can make the case that the expression game renders espionage itself almost pointless, if not a deadly waste of life (an argument with some teeth, for those who've read Legacy of Ashes or The Sword and the Shield) Yet in the context of an online spaceship game where no one can really get hurt, the endless twists of the expression game - trying to parse the intentions of your foe, navigating blinds, double-blinds, and one's own illusions thereof - is one of the most rare and satisfying types of gameplay available. 

An example. An agent reports that in a matter of days, your enemy alliance will be invading a neighbor - but the target could well be you. You have several alliance mails, but they are all opaque - or, depending, they explicitly mention a target that isn't you. Here we dive into the woods; surprise attacks are a crucial part of an alliance leader's arsenal, and it is all too often that the leadership announces a hit on target X only to hit Y. As a spymaster, your job is to somehow determine whether the announcement of the invasion is real. You try to get inside your enemy's head, compiling what you can about the leader's personality, the leadership structure of the alliance, the gossip of the line troops, everything. If you successfully warn of an impending attack, you may have saved the day; if you warn your troops and the attack didn't occur, is it because the enemy noticed your readiness, and attacked a different target instead? At each stage of the analysis the uncertainties multiply. 

style="margin: 10px; border-collapse: collapse; float: left; width: 250px; height: 160px;"
border="1">

href="http://www.tentonhammer.com/node/84475"
title=""> src="http://www.tentonhammer.com/image/view/84475" alt="EVE Spy Game" width="250" height="160"
style="border: 0px solid ; width: 250px; height: 160px;" />

style="font-style: italic;">Reporting Agents were one of the most critical kinds of spies before Dominion.

This brings us to one of the more unfortunate aspects of the state of the spy game today: the near-removal expression games from the realm of reporting intelligence. 

Before the release of Dominion, the reporting agent was perhaps the most critical type of spy in a nullsec alliance's arsenal. These sleepers functioned as information siphons, relaying operations announcements and hostile communications to their handlers. This data was absolutely critical because of the nature and risks of the pre-Dominion sovereignty system. Since sov was held by control towers, and control towers had to be stront-timed correctly or else territory could be lost, the reporting agent was essentially the only defense against a catastrophic first strike. Without foreknowledge from an agent and coordination with a logistics team, an enemy fleet could steamroll an outpost system using a massive capital fleet and ensure that the towers there were either kited into a vulnerable hour, or hit into reinforced before they had a chance to be timed. 

This system of territorial control was widely criticized (and I added to this chorus vehemently) due to the possibility of having a system of region lost to an attack outside of one's normal timezone, while the members of an alliance were asleep or trapped at their day jobs unable to log in. I doubt that anyone will truly mourn the passing of the pre-Dominion sov system, yet at the time of that patch it wasn't obvious what kind of consequences it would have on other types of gameplay.

In the past six months, we have seen the reporting agent reduced in status to that of a useful accessory. Because the new system is based around fixed timers which only have to be set once by the alliance leadership, many of the 'early warning' functions of reporting agents are carried out by covert ops pilots. Any invasion is predicated by a minimum of three hours of warning time by onlining SBU's, followed by multiple 24 and 48 hour timers on infrastructure hubs and outposts. There's no longer any need for spies to liaise with logisticians, save for the odd mining tower or CSAA. Because of the removal of the risk of first strikes, there's no longer any need to bother with the 'maze of mirrors' which characterizes the expression game. 

There are still some expression games to play in the realm of counterintelligence; spyhunting remains a place of paranoia, conjecture, and attempting to outwit and uncover hostile spies. But it's a tragic loss to see reporting intelligence minimized so, when it was once the most preeminent and complicated field of espionage gameplay in EVE.  Should it somehow return, it would add much needed vibrancy to the visibly stagnating nullsec metagame.