For free backtesting, you have to know how to code. Quantconnect and Quantopian offer free access to their data packages if you use their cloud to program your script and test (they can see the results, btw).
There's a few standalone programs that do it, but they're expensive, and don't include the data (I think one is called Arbiter or something like that). Quality data for a lot of stocks is expensive (I think 10 years of S&P500 symbols at 1 second resolution is like $20k).
If you have the data, can program, and want to set up something local there's a few great Backtesting programs written in Python on GIThub.
I wouldn't bother with TradeIdeas, their backtesting only goes back to 90days last I checked. It's way too easy to overfit and the small sample size of only 90days will make the algorithm very susceptible for erratic performance. You generally want to optimize for years of data and then test for another set of years the algorithm/strategy has never seen.
DAS Replay is a great mode for visually / manually backtesting a strategy, but you can easily introduce various biases in doing so. They have data going back to Oct/Nov 2018 if I recall.