The "reputation" criteria for indicators

[DaveSkender]

In our guiding principles and contributing guidelines, we briefly mention that we only implement reputable indicators in the library. Since we don't elaborate on what that means, I thought I'd do that here and provide more context.

the criteria

what we usually say:

An indicator is a good candidate for our library if it can be found with a mathematical formula written by a reputable and recognized industry expert. The author provides a thoughtful and scientific explanation in a popular book or magazine published by a well-known media outlet. An author's reputation is not based on how many community-contributed indicators they've created on open platforms.

our sniff test

• Created by a recognized industry expert with a mathematical formula (not code) and scientific explanation.
• Sourced from a respected and popular book or periodical, through a recognized publisher.
• Can be found in most finance or investing textbooks or closed commercial charting platforms (not community contributed).
• Is intentionally published by the author to the public. It is not a proprietary or license-only formula.
• Is time-tested and continues to be popular after many years.

examples of sources that ARE reputable:

• Found in most finance textbooks (e.g. correlation, beta, linear regression)
• Found in Wikipedia and is also used in most closed commercial systems (e.g. exponential moving average)
• Published by Welles Wilder in his popular book "New Concepts in Technical Trading Systems" (e.g. Parabolic SAR)

examples of sources that are NOT reputable or time-tested:

• A community-contributed indicator on TradingView, MetaTrader, NinjaTrader, etc.
• Something random a person came up with.
• Raw code from another platform that cannot be tied back to the above reputation criteria.

rationale for having criteria

• More indicators create choice confusion for our users. We add value by curating the options to exclude low-quality untested indicators.
• Simple prioritization. We can't do them all, so we focus on fewer, higher-quality choices.
• It's too expensive to develop and maintain indicators that few people want or need.
• You can create custom indicators to extend the library for your unique needs.
• Most community-contributed indicators have questionable value, despite the sales pitch from authors.

With all that said, we acknowledge that there are 1-2 indicators in the library that are not from reputable authors or sources. They are not examples of precedence. We used poor judgment when adding them and may remove them at some point. They only persist to avoid breaking anyone's usage.