As an advise for amateurs we would like to underline: if it's possible, try to observe 3 asteroid positions per night and do not neglect to observe the Numbered Minor Planets in your telescope field of view as a "by product" of your observational programs. They are a good test for everynight estimation of system "Atmosphere + Telescope + Matrix + Catalogue + Processing" altogether. The first control of obtained asteroid positions can be made also with linear approximation for each coordinate set of RA and DEC observed during one night - without any orbit.
We think such estimations ought to be produced every morning after CCD observations of the moving celestial objects by observer himself. Our analysis can find these possible errors, as crude as systematical, too late. Obviously it is desirable to test each asteroid's position before sending to the MPC. May be, our EPOS Software can be useful for our colleagues, namely for amateur astronomers.