I'm still learning about Wikipedia; consequently my philosophy is still evolving. My primary philosophy is to try to adhere to Wikipedia rules and guidelines. Of course with guidelines like WP:IAR and WP:BOLD, that can be tricky at times.

Something I see from both people I agree with and people I disagree with far too much: deleting what you don't agree with, without at least putting an explanation on the Talk page.

1. If a link is broken, don't just remove the link just because you didn't like it anyway. If it bothers you enough to delete it, it should bother you enough to try to figure out how to fix it.
2. If you think a sentence doesn't belong somewhere, try to figure out where it might belong — even if you don't like what the sentence is saying.
3. If you think a sentence or section doesn't belong anywhere in the article, really consider taking it to the talk page first — especially if you don't like what it is saying.

That said, vandalism should be quickly deleted, and sometimes people make mistakes about whether something is vandalism or not.

I will make arguments that support a position I don't agree with. If I can think of a reason why my position might be flawed, I'll point it out. If someone has made a valid argument against my position, even if I think that argument is mainly irrelevant (possibly because it is outweighed by arguments in favor of my position), I will acknowledge it and frequently defend it from people who share the same position as me. I try very much to be a peacemaker in discussions.

When judging edits, I think a necessary part of WP:AGF is that when you read the edits, you should try to forget who is making the edits. Do not try to consider where they might be going with the particular edit. Either the edit is good or it is not.

Now I will admit to a few caveats in how I implement this. Specifically, if the editor is an IP address and/or has a history of vandalism, I consider more carefully whether a statement is true or not. I.e., whereas with most signed-in editors I will assume that a statement they make is true unless I have reason to believe otherwise, I do not make that assumption with people who have a history of vandalism or who edit under their IP address. In those cases, I usually go the extra step of validating the new "fact". If they do not provide a reference for a "fact" that is non-obvious, that is one instance where I will delete first and ask questions later (i.e., on the comment line and/or on the talk page).

So my rule of thumb is that in judging edits, you should judge the edits, and not the editor.

I have had several first-hand experiences where other editors have assumed bad faith of me. Some of the earliest ones came up when I was trying to suggest they assume good faith towards another editor. It seems to me that because of this assumption of bad faith, any edits or suggestions I make are likewise assumed to be bad. Furthermore, since these edits are assumed to be bad, I must obviously be acting in bad faith. And so on…

The most striking case was when I was trying to be nice to an editor with which I had recently had some disagreements. In trying to fix something to the style they liked, I made a comment in the edit history that I was trying to help them out. This was taken by yet another editor who didn't like me to mean that I thought the first editor needed help due to incompetence — thus reinforcing the bad faith assumption of me.

I'm not holding a grudge on this, but multiple editors have talked about how assuming good faith is not a "suicide pact" (as an argument against such an assumption), and it really just struck me as funny, in a sad kind of way. No, assuming good faith is not a suicide pact, but assuming bad faith often is.

Finally, another argument I've heard is that an editor assumes "neutral faith". The problem with this is that it is an unstable equilibrium point — one incorrect interpretation and it's potentially all downhill from there. If you could correctly size up everybody's statements, then there would be no need for the assume good faith guideline in the first place. The written medium is an imperfect one, especially when it comes to conveying intentions.

In my limited experience, one of the clearest indicators of assumption of bad faith are when editors use the "duck test" or reference WP:SPADE in an effort to deflect requests for specific evidence of bad faith actions. Another variation on this is "do your own homework", which often comes from editors who usually understand the need for those making the claims to provide the supporting evidence. If you see these used, ask yourself, have the editors actually contributed any useful information or are they merely trying to avoid answering a question that would expose their assumption of bad faith?