I think it might be better to have the suggested list from the qualifying individuals be randomized for one. I agree that having new users do this might not be the best idea, which is why I think randomizing it might help so lazy people just clicking the first few names in front of them don't pick the same users all the time. At least in this manner users can work to earn a trust level to be included in the randomized default choice pool.
I also don't like the idea of a higher trust ranked member being able to exclude completely the trust of a lower ranked members. This is in effect no different than burning a users account with a negative rating via the default trust for nothing more than a personal choice not to trust a user. The trust system shouldn't be a tool of moderation of trust ratings, individuals should be making these distinctions via ratings and trust settings. If someone is out of line they should be appealing to the public of the forum to take action, not using their superior position of trust to negate the trust of others UNLESS there is sufficient evidence they are engaged in fraudulent trade activity and done so in the form of a negative trust rating. The use of exclusions IMO should be averaged out between the 2 users ratings at least, but not completely negated otherwise all that will happen is those with the highest ranks will use exclusions instead of negatives to take non-trade related retribution upon users.
As far as I know there are no standards for exclusions anywhere, so users could therefore make them for any reason and have a significant negative impact on a lower ranked account regardless of how hard they work or how many successful trades they make, they will never be able to recover from it.
Yes, but my point is that if those are the only 30 displayed they will be the ones getting funneled all of the new user trust (increasing their trust rating) and it will be an endless cycle of more people trusting them producing a huge rift, and basically just reproducing the default trust list in a slightly modified form.
That's why I suggest to show not the 30 best, but 30 among the X best.
This sounds like a more reasonable solution to me. Furthermore making the higher weighted users have more likelihood of being displayed is another feedback mechanism that favors higher ranked users over anyone within the acceptable parameters. I think a randomly generated list selected from the users within the acceptable rating, then explained to the user might be more efficient instead of forcing new users to pick from lists of people they likely don't know. They can always update it to include or exclude users later.