The published verification stats for ECM may look better than those for GFS, but when it comes to spotting future pattern changes, I trust GFS a thousand times more than ECM, which always appears to miss them until the last moment.
Perhaps ECM is better at the finer detail up the rhe end of reliable, but is useless when it comes to broad brush issues further out, where GFS appears to be much better.
Originally Posted by: Maunder Minimum