Electroencephalography (EEG) is a crucial tool across neuroscience domains, including medical diagnostics, psychological research, and brain-computer-interfacing (BCI), due to its non-invasiveness, high temporal resolution, and cost-effectiveness. EEG data classification—the process of assigning predefined class labels to segments of EEG recordings based on patterns learned from training data—is challenging due to EEG’s high dimensionality, variability, and individual-specific differences. Recent research has shown that the time series classification algorithm HIVE-COTE v2.0 (HC2) is particularly effective at EEG classification, but that it is also orders of magnitude slower than algorithms from the EEG literature. We investigate ways of improving the run time of HC2 through channel selection and creation. We demonstrate that we can achieve accuracy that is not significantly different to full HC2 with up to 3 times faster runtime.