My criteria for RfAs

I thought I ought to enumerate these for my own reference; be warned that they're evolving.

I want to see both a breadth and depth of experience in the areas a candidate will be operating with the tools. Unless a candidate declares they will only be operating in one area of administrative responsibility, I have to evaluate on the assumption that they will operate in all areas.

If you have participated in RfA in the past, I will hold you to your own standards as well as my own.

What I'm voting for

I'm trying to stop Bad Things from happening:

Articles being deleted outside of criteria - this happens with almost no oversight
An uncivil editing environment (biting, warring, drama) on the assumption sysops police our norms
A hard life (I don't want to clean up after the janitorial staff, getting their incorrect actions reversed; I don't want to inspect their every action)

Criteria indicative of failure

https://toolserver.org/~tparis/editsummary/index.php?name=Josh+Parris&lang=en&wiki=wikipedia reports a number less than 100% for the last 150 edits, both major and minor (use the correct editor name)
It's no secret that edit summaries are important to some people. I'm one of them. If you're not prepared to prepare for an RfA by at least pretending that you think they are too, you aren't trying hard enough. And no-one says "uses edit summaries, consuming valuable server resources".
https://toolserver.org/~tparis/editsummary/index.php?name=Josh+Parris&lang=en&wiki=wikipedia reports a number less than 90% for all time edits, or less than 80% with a skew towards having recently "seen the light" (use the correct editor name)
What's the difference between "Bhwa ha ha! I deleted valuable content and replaced it with my own nonsense!" and no edit summary? There isn't one. No edit summaries make me work hard. I don't like that. This is indicative of poor communication.
An examination of your recent edit summaries shows summaries like "re", "+" and "cmt"
These are not edit summaries. They are what the page is for. These are blank edit summaries that also consume valuable server resources. But the big thing this says to me is: you can't summarize, you can't communicate - and these are vital in the duties of an administrator
https://toolserver.org/~snottywong/cgi-bin/afdstats.cgi?name=Josh+Parris&max=250 shows less than 50 appearances
Without sufficient samples to work from, the next metric fails. Also, practice makes perfect.
https://toolserver.org/~snottywong/cgi-bin/afdstats.cgi?name=Josh+Parris&max=250 shows a "didn't match result" above 15%
If you can't make these calls right yourself, how are you going to going to make them right when you work AfD? How can I trust you to judge CSD reasonably?
http://toolserver.org/~snottywong/cgi-bin/commentsearch.cgi?name=Josh+Parris&search=speedy+deletion+nomination&max=100&nosect=1&ns=none shows less than 20 CSDs
Insufficient data to make decisions on
Appearances on WP:AN/I - https://en.wikipedia.org/w/index.php?title=Special:Search&profile=advanced&search=%22Josh+Parris%22+intitle%3A%22administrators%27+noticeboard%22&fulltext=Search&ns4=1
Drama magnets do not make for good admins.
Editing without previewing
Frequent incremental edits to an article you're developing? Disruptive in the small - one article becomes a watchlist nightmare.
Incivility
Your talkpage will be read, its history examined to see what censorship has occurred.
Short participation in the project
I think it takes at least six months to get the hang of the place. Administrators are meant to understand what the community norms are.
Few edits
Automated edits aren't edits for the purposes of this criteria. I like to see more than 1000 edits where you've had to sit and think and construct comprehensible, compelling English prose. This is a requirement for demonstrating an ability to communicate clearly. This is not an editcount of 1000; it's substantially more.

Criteria indicative of success

High edit count, spread over many pages and namespaces
Note that inflated edit counts don't... count. If you can rack up a big edit count and not find yourself tripping any of the Criteria indicative of failure, you're doing something right.
Content creation
That's why we're here. Doing it helps you understand the value of it.
Admin practice
If you've been running around doing things in the admin space: xfD, thoughtful RfA contributions, helping out newbies experiencing shock and awe, and nobody has yelled at you, then that shows you're likely to be good at this stuff in the future.
Long history of service
You're still here? That's dedication. Wikibreaks not a problem, but at least six months since your last.
A good CSD ratio
The ratio of edits found by surviving delete requests compared to logged requests of 1:20 or lower is a dead giveaway that you know what you're doing (only look at recent edits, since Twinkle became the tools of choice for this; when dinosaurs roamed the Earth people had to do it by hand).