Nerf....

Something is better than something else, quick, nerf it....

Seems to be that any perceived imbalance gets the nerf. Stop it!!!

Slow down! There are a few reasons; I am going to consider three:

1.       Management practice (Deming)

2.       Squeaky wheels (Drake versus Hurricane)

3.       Decision making under stress (examples from military studies)

The current nerf is jump bridges, which pretty much has been roundly condemned in a number of forums. It has some positive aspects, but fundamentally will most likely not achieve its intended purpose. What is the purpose? People surmise (CCP hasn’t explicitly stated) that its do discourage large super-cp blobs and encourage small scale PVP.

The next nerf is agents, due out in about 48 hours. I think this change is very mellow, it tidies up an existing system, hopefully will add a little clarity. However what resources at CCP were required to do this to what in game effect? After all the system works as it currently is, so why ‘fix’ it? So it raises to me what is CCP’s vision? as per an earlier post.  In what way does this change work towards CCP’s goals. I would love to see a list of the intended changes, when, why, how, instead of, justifications and the likes.

Part 1, Management practice:

Edward Deming got the job of helping rebuild Japan’s economy after WW2. His principles later became known as ITIL and Microsoft has their own take on them (with MOF4.0) http://en.wikipedia.org/wiki/W._Edwards_Deming

 Some of his key principles can be summarised as quality check everything or nothing. Quality leads to lower costs, profits are generated by loyal customers. Making a system change needs to be carefully considered and managed and based on testing and knowledge – PLAN, DO, ACT, CHECK.

Deming used a funnel suspended over the ground and dropped a ball at a specific spot through the funnel. Then corrections are made in different ways and the results compared. Here’s a simulator and some graphs:
http://www.symphonytech.com/funnelexp.htm

What it shows is the adverse effects of adjusting a process when not required.

So what’s my point? CCP keeps nerfing things, making changes, which while tested on SISI are implemented with a major impact on the game, often it seems not predicted by CCP but by some bittervets (Ripard Teg and Mord Fiddle for example). It seems there is a disconnect between the testing and reality, but how can this be? Seems the testing on SISI is either not relevant to Tranquillity (such as you test PVP but not PVE) or simply dismissed as a blip, a statistical anomaly. Regression testing is about the integration of the code, rather than the impact of that code on the game play. Given the number of patches that follow a patch (seems that there are 3 minor ones that follow one major one) it raises some very serious concerns about the evaluation, testing and implementation of patches. It is extremely hard to manage the code as it grows, and the impact of all the changes one makes. So it is a very hard process. But given the negative impact of a number of patches (nerfs), the number of sub-sequent patches for releases and the screams on player forums Plan, do act, check is not working. The changes seem too radical too often, more on that in part three.

CCP needs to address a significant break down in the change process.

Part 2, Squeaky wheels

Nerf the drake! Who hasn’t heard that? But if you look at the numbers the question I pose about the drake is so what? If you look at the statistics, these are from the ‘Quarterly Economic Newsletter Q3 2010’, the drake is imply another ship in game.

The number of ships in use:

The drake is a level 3 mission runner, and is up there with the level 1 mission runner the kestrel. Given the Caldari race is the most popular it would fit that the predominance of Caldari ships would follow (the raven makes the top 10 as well, but not the caracal interestingly). Given the drake is also good with 7 weapons slots it makes for a good cheap PVP ship too. Like the Hurricane with its BC size and 7 weapon high slots. A BC can easily be reached by players and doesn’t break the bank to lose it, can be insured for full replacement (unlike T2 ships) and is not as skill intensive to reach as many other ships. All in all a BC of any race makes a good cheap PVP ship (Brutix is probably bottom of the list but not bad, and the Harbringer is certainly a good ship).

As for PVP the Hurricane killed 5000 more ships than the Drake, and the Drake died 4000 times more than the Hurricane. On that basis it should be the Hurricane that is nerfed not the Drake. I think the issue around the Drake for CCP is that it fires missiles, which add another computational calculation in PVP, potentially adding to lag. The ship itself, statically speaking, is fine.

The message, all too often it seems a nerf is based on an emotional response from the player base and not a statically supported one. The jump bridge nerf fits in here.

The squeaky wheel gets the oil.

Part 3, decision making under stress

First read the Psychology of Military Incompetence by Norman Dixon (http://www.amazon.com/Psychology-Military-Incompetence-Norman-Dixon/dp/0708814824/ref=sr_1_2?s=books&ie=UTF8&qid=1305626347&sr=1-2) and one that helps illustrate this point, the battle of Khafji with ‘Storm on the Horizon’ (http://www.amazon.com/Storm-Horizon-Khafji-Battle-Changed/dp/0345481534). The battle of Khafji can also be read on Wikipedia at: http://en.wikipedia.org/wiki/Battle_of_Khafji

OK, so what do they say that relevant? Slow down, assess, respond, re-evaluate. When the brown stuff hits the whirly thing people, even very well trained (Khafji with the Observation Post 4) and highly decorated (General Haig) can quickly find themselves overwhelmed by information and the need to make a decision and to make it fast. The Battle of Khafji also showed that if cool heads prevail, that time is taken to ASSESS and then respond a significantly different outcome results, a more positive one.

The message in this section is when you need to make a decision, work can be a stressful place, so slow down (get a grip and calm down, to help you think clearly), assess the situation then make a calculated response and measure its effect.

So what does this all add up to?

CCP needs to ignore the squeaky wheel and look at the statistics (assess). It needs to publicise its findings and what it thinks it MAY do about them. CCP positions then can be debated (evaluated) in the forums and reported back through the CSM. The nerf is then tested (checked) on Sisi and enacted on Tranquillity if it reaches the intended solution. Then it needs to be checked (look at the statistics and further feedback via the forums through the CSM) on Tranquillity to see if it in fact works on the live environment.

Currently smacking things with a big hammer is breaking fingers and still not fitting the round peg in the square hole. The nerf gun doesn’t seem to be the best way to hit the bullseye.