While visual and laser-based simultaneous localization and mapping (SLAM) techniques have gained significant attention, radar SLAM remains a robust option for challenging conditions. This paper aims to improve the performance of radar SLAM by modeling point uncertainty. The basic SLAM system is a radar-inertial odometry (RIO) system that leverages velocity-aided radar points and high-frequency inertial measurements. We first propose to model the uncertainty of radar points in polar coordinates by considering the nature of radar sensing. Then in the SLAM system, the uncertainty model is designed into the data association module and is incorporated to weight the motion estimation. Real-world experiments on public and self-collected datasets validate the effectiveness of the proposed models and approaches. The findings highlight the potential of incorporating radar point uncertainty modeling to improve the radar SLAM system in adverse environments.