Tags

In the previous post, I outlined a comprehensive plan for evening out the gap between PVP-low and PVP-high areas, so that transitioning between them would not look like jumping off a cliff into a pool of sharks with lasers on their heads. A plan that, while it stands very little chance of ever being implemented due to both the public opposition and risks involved, is at least interesting to think about.

Well, this blog has always been about thinking, so an intellectual exercise is par for the course.

I’ve had some new clever ideas since, first few on that front in quite a while. This post is by necessity sort of a mishmash of them.

Crime and Punishment

— Raskolnikov! How could you?! Killing a granny over 20 kopecks!?!
— Yeah? Well, five of those make a rouble!1

First, some observations. In other contexts where the concept of crime comes up, the goal of criminal justice system is to prevent crime, as defined by law, and protect the victims of same. At the same time, all the tools available for the purpose are limited — they typically cannot be employed until the crime is at the very least in progress, for the very same reasons.2 Nevertheless, the idea of crime prevention is not abandoned, and the law’s stated purpose is never revenge for the crime. While it is impossible to determine if a crime was actually prevented or not, assuming that if no preventive measures occurred, the crime rate would be more or less even, allows one to indirectly see, by the reducing rates of crime, the effects of preventive measures.

There are multiple avenues through which law enforcement considers itself working for preventing crime. However, for our purposes two are the most important:

  1. The inevitability of punishment is seen as a deterrent, that is, it is believed that if the criminal justice system is sufficiently effective in punishing criminals, crime rate decreases. While not necessarily correct, this logic is usually acceptable.
  2. Statistically, the most common source of crime are repeat offenders.3 While jailhouse environment does not really work to rehabilitate people and prevent them from reoffending,4 long prison terms serve as a measure to prevent, or at least significantly delay crime, by isolating the incarcerated where they can’t do much harm, except maybe to each other.

Every criminal justice system is only aware of how well are they doing at inevitability to a certain degree, as there’s always an unknown number of unreported5 and undiscovered crime, unsolved crime, and other nonsense, but every prisoner behind bars is a clear indication that he isn’t actually committing a crime right now, rehabilitation be damned. This is why the practice is actually perpetuated, no matter how imperfect it is — it works, for suitably small quantities of ‘works’.

That "suitably small" is generally sufficient to provide an environment where law-abiding citizens can go about their law-abiding business without cushioning for risks too much, which was the whole point. What Eve needs is a smooth progression between such an environment and total lawlessness.

Meanwhile, in New Eden

In Eve, both of these avenues of crime prevention are actually implemented, that is, have mechanics meant for them. The discourse on the topic, however, very rarely mentions anything except CONCORD, which is the #1 avenue of crime prevention — inevitability of punishment. CONCORD has the inevitability part down pat, it is inevitable like death and taxes, where it is in effect, it will always happen. If you somehow make it fail to happen, you’re screwed even more because you’re engaging in an exploit, committing an affront directly against God. But what, exactly, is CONCORD protection actually doing?

  1. It destroys the offender’s ship. It does not destroy their pod.
  2. It effects a time limit on how long the crime can be in progress — the gap between the first shot and the moment CONCORD arrives.
  3. It reduces the offender’s security status, which is currently mostly a nonsense number. (But more on that below.)

#1 on that list is essentially a tax on committing crime, both in ISK required to purchase a new ship and the time the player requires to reship. #2 is a minimum limit on how much DPS one needs to bring to commit their crime successfully. Indirectly, that creates, for some of the more typical crime like suicide ganking miners and haulers, the minimum amount of profit one needs to make the crime pay off.

Structurally, #1+#2 are completely equivalent to being fined the total cost of the ship (barring the mess with the loot fairy) and having weapons deactivated after a certain delay (in seconds, until CONCORD arrives) for a certain period (minutes until you can reship). But despite that structural equivalence, there’s one thing that is different in application possibilities: CONCORD "protection" that is expressed as a fleet jumping in to shoot the offender dead can only be toggled on and off. The only variance that can be had is in the time domain — i.e. time before arrival, and possibly, time when it is and when it is not available. But expressing it as a monetary tax allows variance in amount of said tax instead, which can be as granular as needed.

What’s essentially missing in Eve is #3, though — the long term effect of security status which would actually prevent a highly criminal character from committing further crime in protected areas.6 I.e. the equivalent of incarceration’s effect on crime prevention. Nevertheless, it does have a mechanic for it, it’s just that it’s not working terribly well: Faction navy.

Technically, the role of faction navy is to prevent you from entering space if your security status is too low to be in there. However, seeing as it how never actually scrams anyone,7 that never really happens, and it does not prevent you from being active in that space once you have entered.

Making Slightly More Sense

Once you consider this, the obvious outcome is that a roaming DED fleet mechanic is actually unnecessary to get a smooth scale of experienced safety, just like CONCORD itself is unnecessary, and something much simpler and more predictable can be done. Here’s how it could work in the context of Plan 9:

  1. When a character acquires a criminal flag anywhere in space claimed by an entity that is part of the CONCORD agreement, regardless of system security level, the following things happen:
    • The character immediately pays a fine. In a 1.0 system, the fine is equal to X*M, where M is the total cost of the offender’s ship and cargo, and X is an appropriate multiplier based on system security level. It is reduced proportionally to system security status, so that it becomes equal to 0 somewhere around 0.1 security status or thereabout.8 If the player’s wallet is insufficient to pay the fine, however small it might be, a CONCORD fleet actually spawns immediately, and destroys their ship just like it would do now.
    • If you want to be nice, part of the the fine collected goes to the wallet of the injured party, while the remainder becomes a bounty on the offender’s head, split in any desired proportion. More bounties on people is good.
    • At this moment, the criminal’s faction standings are negatively modified as described previously, for the act of committing a crime in faction space. Positive standings are modified further than negative ones.
    • For a while, nothing happens, and the crime continues. After a delay dependent on the security status of the system, the criminal’s drone bandwidth is reduced to 0, weapons are deactivated, number of lockable targets is reduced to 0. I.e. the battle is momentarily interrupted. After a further delay dependent on system security status these effects go away, so the battle can continue if the victim is still here, but the safety selector is forced to green.9 The selector remains forced to green for a further period of minutes, depending on the security status of the system the crime occurred in, regardless of where the criminal goes afterwards.10

    That’s the "inevitable punishment" part.

  2. The duty of actually preventing crime, by preventing potential repeat offenders from entering until they are rehabilitated, i.e. raise faction standing, remains with the suitably boosted faction navy, which treats people with negative faction standings11 just like it treats enemy faction warfare players — i.e. scrams them, follows them around safe spots, and spawns in quantities dependent on both the system security status and the faction standings of the offender.12

Looks like such a mechanic would work a lot smoother than randomly roaming fleets.

Further Variations

There’s a few ways Plan 9 can be tweaked which I did not previously detail.

  1. The list of player activities that can be used to affect system security that I can think of is:
    • Units/costs of materials extracted from the environment.
    • NPC ships destroyed and/or NPC bounties earned.
    • Units of materials used to manufacture items or units of items manufactured and their costs.
    • Units of ME and PE researched.
    • Units or prices of items bought and sold.
    • Jumps through gates.

    Of these, the ones actually too easy to game are jumps and items bought and sold. However, both can be rephrased to count not unit jumps or units/ISK of items, but number of individual characters participating in either, regardless of how much individual jumps or individual purchases they make. That would require much more effort to deliberately game. Using everything on that list to affect system security would make for a considerably better reflection of actual usage as well as reduce the effects on long-distance trade routes.

  2. Criminal acts committed in a system could also reduce it’s security status over time, thus making the pendulum swing both ways and reducing the concerns about highsec systems being permanently stuck at high.13
  3. The constant sec drain applied to a system does not have to be the same for every system. In fact, in one of the comments on Plan 9, NeVrain suggested a very clever variation based on heat transfer equations and NPC pirate faction ‘population’ centers, which is probably the most advanced way to keep the world highly dynamic in detail but generally predictable as a whole.
  4. With the fine+faction navy mechanic, experienced security can be smoothly (and computationally cheaply) varied within a system, thus making certain places in it less secure than others — by distance from key spots like stations or gates, for example, which is harder to do with a DED fleet mechanic.

Oh, and one thing I actually noticed only much later: Implementing Plan 9 in this fashion would allow the reuse of the ideas from the A-Life based design I had for a completely different project that failed to materialise. I already mentioned that an artificial life engine has been a major intended component of Ultima Online, which was removed because players just killed everything too fast. But that’s a failure of the implementation, certainly not the concept itself. An ecosystem properly adjusted for the biggest predator, i.e. the human players, would not have such a problem.

Not to mention that I suspect this might be the only way towards practical procedural story generation. :)


  1. A joke from the 90s. Seeing as how that’s the only joke mentioning Dostoyevsky’s book that I remember, you can see that the book’s not terribly popular, at least compared to "17 moments of spring".

  2. Unless you live in a universe where reliable prediction of crime exists, or where the ideas of law derived from Roman Empire have had no influence. Thankfully, either is rare outside fiction these days.

  3. In the US, 53% of all males convicted go on to repeat offense.

  4. Amusingly, if it did, the rate of repeat offenses would go down, so there would be less point to incarcerate people so… It would stabilise somewhere again.

  5. Certain types of crime are particularly likely to go unreported for cultural reasons.

  6. What it instead prevents a criminal character from is doing non-criminal activities, like shopping. Shooting players is not hindered much.

  7. Unless you’re an enemy Faction Warfare player. Then they don’t just scram you eventually, but also follow you around, which is far more effective. That still didn’t stop GalMil from raiding war targets in Jita, though…

  8. Thus addressing concerns that this would completely destroy the Fight Club. Reducing the amount of space reserved for pure Fight Club activity would actually make encounters in it much more frequent, by the way.

  9. Limited engagement flag would actually remain — so that if the victim actually fought back, the battle could be legally concluded. But with the safety selector at green, no further crime can be committed until it is released, which is the point.

  10. Lorewise, that would mean that forcing everyone to have safety selectors on their ships actually gave CONCORD an opportunity to police space far cheaper than it previously could manage.

  11. Incurred by committing crimes within faction space more likely than not.

  12. Severely negative faction standings could also deny docking in the faction’s space, by the way. The equations setting how much faction navy a single persona non grata merits where, depending on their faction standing, how hard it is to run away from them, and how much faction navy needs to die before the NPCs back down and leave them free to operate in a particular system need a lot of thinking with actual data at hand, which I obviously don’t have. These equations would need to be adjusted to produce a smooth scale of experienced security as derived from a smooth scale of opportunity to profitably commit crime. Just how much opportunity is acceptable where is a matter of much debate, far beyond the scope of this post — I only aim to show that sufficient flexibility is possible with such a mechanic to accommodate any play style in any desired quantity.

  13. Imagine the Burn Jita operation actually succeeding in driving Jita to 0.5 for a few days. Then imagine a sufficiently big group dedicating time and effort to disrupt a highsec system on the Jita-Amarr route, otherwise pushed upwards by constant player traffic…

About these ads