When debating others, be that on the internet or face-to-face, there will inevitably be times when one or more other individuals in the debate are being dishonest about their beliefs (either pretending to believe things they don't actually believe, or else being dishonest with themselves about the reasons for their beliefs). This is generally described as arguing "in bad faith".
When this happens, it is important to be able to move them to a lower level of trust so that they can be excluded from discussions where more trust is necessary.
It is also important not to mistakenly assume bad faith where someone sincerely holds and expresses beliefs which don't appear to make sense because they are based on sincerely-believed premises which they had not yet thought to challenge.
More generally: it is important to be able to assign debate participants to a level in the hierarchy of trust (HoT) with a reasonable degree of accuracy.
The most difficult levels in the HoT are typically "reasonable disagreement" vs. "unreasonable disagreement"; in both cases, we disagree with what they are saying, but need to determine whether what they are saying is at least reasonable.
Heuristics which may prove useful towards determining this include:
- Do they assume good faith from others? Someone who comes into a discussion with the assumption that anyone who disagrees with them (or holds certain beliefs) is being dishonest pretty much makes it impossible to engage them in reasonable discussion. They have established that they may assume anything you say is either self-delusion or an outright lie, and at best you will spend significant time trying to re-establish trust with someone who seems disinterested in supporting that process.
- Do they take in new information? If an argument someone has made has been refuted, and they have not answered the refutation, they cannot later make the same argument as if it was generally-accepted truth. (Critique responsiveness seems to be a good phrase for this.)