I calculate value based on mileage depreciation, with assumptions of 150 to 200k lifetime mileage for my daily drivers (depending on make model). Estimate a residual value and then look at a linear depreciation curve, or in this case, line. Any vehicle with value above that line are overpriced, anything significantly below the line might be a good value.
This does not factor in added maintenance needed if buying used, or higher financing costs used vs. higher overall interest paid on a more expensive vehicle. Also does not take into account changing used market conditions. Still a valuable tool though.
This business-based decision model assumes that the total purchase price is irrelevant and maximizes the per-mile-driven economy. Also assumes a fairly constant rate of miles driven, and a fixed mileage point of sale at end-of-life. Obviously, if I'm able to keep a vehicle longer than the initial estimate, my costs will be lower.
Example in round numbers from my purchase back in 2016:
Mazda6: 20k purchase price in 2016
Drive to 200k, sell for 4k
16k over 200k miles, $0.08 per mile
Used w/40k: depreciation of $3,200
Any vehicle with 40k and less than 16,800 would be less expensive per mile driven to 200k.
Figure an opportunity costs of 2k for unforseen maintenance needed on a used vehicle and you'd be at 14,800.
That was numbers from 8 yrs ago but the logic still stands. When I ran the numbers for full size SUVs a few years ago, new was the logical choice given the crazy post Covid market.
Also have to consider the possibility of getting a buyback or otherwise "lemon" vehicle, especially in the 2 to 3 yr old used market and among those makes and models with high product infant mortality rates.