Wikipedia:Follow consensus, not policy

There are three stages to learning how to both use and contribute to Wikipedia.
 * 1) When you start out, it is safe to Ignore all rules, since nothing you do can permanently hurt the encyclopedia.
 * 2) Over time, you can start to learn how other people have thought about things in the past. In this period of time process is important.
 * 3) After about 2–3 months, you start to get an idea about how people interact, and how they think or expect things to be done. You can then try to follow the actual wiki way, and try to stick as close to Consensus as possible instead. Traditionally, this is when you got noticed and got nominated for administrator, though that process seems to have changed somewhat, recently.

Project namespace is as reliable as Wikipedia itself
Note that at any point in time, we first establish a consensus, and then if some volunteer takes the time, only then do they write down the established consensus on some page in the project namespace. Some people also write down how they think people should act. Sometimes a small number of people hold discussions or votes to try to tell the rest of Wikipedia what to do. Because of this, you should be wary of what you read in the project namespace. It will lag behind, give bad advice, or even be downright wrong. As is true for Wikipedia itself, different pages are of differing quality.

How to start
It's ok to ignore any kind of guidelines until you get around to actually having some time. Nothing you can normally do in good faith can permanently damage the encyclopedia (see:Page_history), so feel free to experiment! Who knows, you might learn new and better ways of doing things, and though that's not Wikipedia's primary mission, it is a part of what we do. :-) Playing is always the best way to learn. So do play!

After a while, you might want to learn more about what other people have already learned about how wikis are run. Try start out by reading pages marked as policy, then try guidelines, then read essays. Note that a page marked policy isn't necessarily right or as important as it seems, but people have tried to be as accurate as possible, so it's a good start. Just be wary of quoting policy at an experienced Wikipedian, as if it is written in stone.

Finally, after a while, you'll start to grasp that all these things called policy, guidelines and essays, and the mediawiki code, et al., actually just have one goal. They describe what a wikipedia is. Or at least, they form a representation of what people thought a Wikipedia was, up to a couple of months ago. Keep your eyes open. Do you see people deploying new initiatives? Do you notice the slight changes and improvements that haven't been written down yet? Do you notice how you can do things somewhat smoother yourself? So like go ahead, be BOLD and sofixit!

No wait! Just a moment, are you sure that you will have consensus for what you're doing? Maybe it might be a good idea to hold back for a moment, what's this consensus thing. Did you learn about consensus in school? Chances are you didn't. Some new people on Wikipedia will tell you that consensus is a kind of voting. Well, voting can be part of a consensus formation process, but it isn't consensus itself. What consensus is and how it has been formed on Wikipedia has always been taken for granted, so people never really write it down. At some point they just get it, go "Duh", and then promptly decide it must be equally obvious to everyone else (despite it having just taken them several months ;-))

Having consensus
There are several ways to know you have consensus:


 * You can read old guideline pages, that's a good approximation. That's why "following policy" seems to work so well. But don't think this is the be all and end all. People have been blocked for simply following policy with their brains turned off. Anytime you are going to apply something that's already written down, be sure to check whether it is actually a smart thing to do, and be sure the thing has consensus
 * You could ask first. This is the most hated option, since it means it'll take you ages before you can finally do anything. Others might get restless too. Learn to know when to ask, and when not to ask.
 * You are doing something new. If you do something new that doesn't really touch anyone else's work, no one is likely to complain much, which is good. If you are doing something that helps people out of a bind or that solves problems effectively, probably people will accept it gladly. Beware of a big trap though. Simply making stuff up won't cut it. You may often need to do quite thorough analysis of existing systems, and decide on factors like scalability, resistance to gaming, complexity, time needed to learn, time needed to apply each time, etc. Techniques applied have included game theory, systems analysis, and statistics. Practice with using a wiki helps too.
 * Easier to apologize than to ask permission. This is a risky strategy, it works okish for small things though. Note that it's always going to be mildly controversial. Certain situations where people do this is very common, and is written down as the snowball clause.
 * Read/participate in discussions. This is a great way to get a feel for consensus on particular issues.
 * Test the waters, try to run a number of small tests, and see what methods work more easily, and/or meet the least resistance from others.
 * Be psychic. Ok, so you can't really be psychic, but if you are very experienced, you can intuitively try to predict what the consensus on a particular issue might be. It's very easy to overestimate your skill at this. If you misestimate and you do something big, you may possibly get kicked off of the wiki. It helps to try to test the waters first. On the other hand, if you practice this skill a lot with smaller issues, you'll find that things go a lot easier for you.

How good are you?
If you can just go about your business, without having to worry about what's exactly written in the guidelines, you may well have come a full circle. You are now ignoring all the rules, and Following consensus, not policy.

How good are others?
Some people try to cover their mistakes in retrospect, by quoting "I just ignored all rules". But that's not good enough. Point out that perhaps they did ignore all rules, but at the same time, they hurt your feelings, or hurt the project you were working on. Apparently they misjudged consensus. So tell them so.

But they probably did want to do the right thing. Just because they stumbled, doesn't mean that they are necessarily evil or trying to hurt you. Instead, try to help them reach the smoothest possible conclusion. Try to follow the guidelines given on this page, and help them as best you can.

Most importantly, talk with people
Don't assume someone is wrong upfront. Always ask them for their reasoning. Why did they think this approach was best? Perhaps with minor adjustments, their approach really is the best way forward. Months later, you may well find yourself applying their method, and having to explain it to some other person. Now where did that guideline go? Oh dear, it was never written down! Perhaps you should take the time to do so now. :-)