In this study, we examine the accuracy of global geospace simulations by analyzing the relationship between the solar wind and its propagation parameters and the errors in auroral electrojet index AU and AL, ring current index SYM-H and the cross-polar cap potential (CPCP) in simulations. We show that generally the error distributions are wider for higher level of solar wind driving. Our results also show that the observing solar wind monitor distance from the Sun-Earth line and the phase front normal angle produce only minor effects on the error distributions, however, for oblique angles (<0.4) of the phase front normal there are noticable effects on the error distributions. Furthermore, we show that the results hold true also when using two magnetometer station recordings, one at subauroral and the other at auroral latitudes, which speak to the similarity of the error sources in local and global activity measures. These results are important elements in assessing the accuracy of the timing and magnitude of space weather events recorded on ground.