Diluting spark-ignited (SI) stoichiometric combustion engines with excess residual gas improves thermal efficiency and allows the spark to be advanced toward maximum brake torque (MBT) timing. However, flame propagation rates decrease and misfires can occur at high exhaust gas recirculation (EGR) conditions and advanced spark, limiting the maximum level of charge dilution and its benefits. The misfire limits are often determined for a specific engine from extensive experiments covering a large range of speed, torque, and actuator settings. To extend the benefits of dilute combustion while at the misfire limit, it is essential to define a parameterizable, physics-based model capable of predicting the misfire limits, with cycle to cycle varied flame burning velocity as operating conditions change based on the driver demand. A cycle-averaged model is the first step in this process. The current work describes a model of cycle-averaged laminar flame burning velocity within the early flame development period of 0–3% mass fraction burned. A flame curvature correction method is used to account for both the effect of flame stretch and ignition characteristics, in a variable volume engine system. Comparison of the predicted and the measured flame velocity was performed using a spark plug with fiber optical access. The comparison at a small set of spark and EGR settings at fixed load and speed, shows an agreement within 30% of uncertainty, while 20% uncertainty equals ± one standard deviation over 2000 cycles.