There has to be a balance between the ability of newbies to ask questions of gurus and the problem of newbie accounts being used to spaminate SO.
On one hand, we see that experienced users frequently ignore newcomers by looking at reputation and accept rate. If this happens often enough to a question, the majority of users will just skip the question. There are exceptions, to be sure, but the fact that they are exceptions and not the rule is sad.
On another hand, experienced users are afraid of poorly formatted or poorly formulated questions — which are typical among newbies — and rightfully so, because those questions interfere with users' ability to readily find and understand answers among tons of questions.
To find a balance between these two points, I propose the following feature:
- Each question from any user with reputation less than 100 should be marked as "sandboxed"
- "Sandboxed" questions are searchable only under special conditions
- "Sandboxed" questions can't be down voted, but can be upvoted
- Each user with reputation more than 1000 can remove any "sandboxed" mark, making a post an ordinary question that may be voted on in the usual way
- "Sandboxed" questions with two or more upvotes automatically become regular ones
- Optionally, introduce a new silver badge
Sandbox Savior
for removing more than 100 "sandboxed" marks
The general idea is to have a sandbox for newbie questions, so that gurus can't downvote them — encouraging newbies to ask more — and so that spammy questions won't interfere with SO activities. This will also provide some reward for gurus who help newbies to become more informed and useful.
ADD Primary goal is that SO should be recognized as a source for good ideas: questions/remarks/answers and so on, rather than tons of spam and somewhere among them there're brilliants. So we need establish some kind of filtration or as in proposal sandboxing.
We see that experienced users frequently used to ignore newcomers
I don't think this is true at all. Is this your subjective impression or do you have some hard data for it?