Initially the calculation was based on the recent observations known for me, including observations by ISON-NM observatory (Feb. 27), Martin Mašek (Feb. 27) and Tomas Vorobjov (March 1). Estimation was made on a sample of 1,000 virtual particles (clones), calculated by the Monte-Carlo method and based on the nominal orbit solution. Calculation shown what only 2 of 1,000 clones will collide with Mars, i.e. 0.2%.
Late at night, we received information what found two another archival observations (October 4, 2012) by Pan-STARRS. Now, arc of observations increased to 148 days! Based on the new data, calculations was restarted again. The collision probability decreased in 2.5 times. Neither clone from a sample of 1,000 virtual particles not collided with Mars. Final calculation is based on a sample of 10,000 clones! It shown that only 8 virtual objects will be collide with the Red Planet, i.e. probability of this event fell from 0.2% to 0.08%, but still high enough for the events of such scale. Minimal distance of close approach, according nominal orbit solution is 0.00039 AU or ~58,000 km.