The goal of grid friendly charging is to avoid additional load during peak use of the electricity grid, and to deliver a financial incentive to the use. In a Day Ahead tariff (DA), it is easy to schedule charging for the time of cheapest electricity. However, the full benefit can only be realized using Real Time Prices (RTP), because this allows the charging to respond to unpredicted changes in the balance of supply and demand. Because future prices are uncertain, finding the best charging period becomes a stochastic optimization problem, which has to take into account the risk of unexpected price changes. This paper explores the value of a predictor for future prices. Several factors are considered to influence the price: from the time of the day and the day of the week over current and predicted electricity prices. Because a full factorial design would be too difficult to tune, partial factors are identified for important interactions. The predictor is found using regression on historic data, and testing against a validation set. The aim of the predictor is to provide a reasonable estimate of future prices that help to make better decisions in the charging optimization. A case study is preformed based on historic data from the Illinois Electricity Grid prices. It demonstrates that the predictor helps to reduce charging costs over the stochastic optimization without prediction, but it is still short of optimal charging cost as identified in retrospect. This indicates that this is potential scope for further improvements to the predictor.