The hit rate of space weather forecasts issued by the Japanese forecast center in the National Institute of Information and Communications Technology (NICT) between June 2014 and March 2015 are compared with that by the persistence method. It is shown that the hit rate of the forecasts by the Japanese center is better than that by the persistence method. Several problematic events on the space weather forecasts during the same period are analyzed. Those events are (1) geomagnetic storms associated with coronal mass ejections (CMEs) on 9 September 2014 and on 15 March 2015 with different durations of southward interplanetary magnetic field (IMF),(2) a large active region, AR 12192 without CMEs, solar energetic particle events, and geomagnetic storms,(3) a geomagnetic storm on 7 January 2015 caused by a faint CME, and (4) disagreement between the in-situ observation at 1 AU and the prediction of the Potential Field Source Surface (PFSS) model on timing of sector crossing in January 2015.