It’s easier if you look for the best actors with the least turkeys.
Logical way to do it is to calculate log-odds. Take the number of good movies by the actor and divide it by the number of shit movies. For example, someone who has made 50 good movies and 50 poor ones would be 50/50 = 1.00. Someone who has made 90 good movies and 10 poor ones would be 90/10 = 9.00.
Then take the natural log of that number. The natural log of 1.00 = 0.00. The natural log of 9.00 = 2.20. An actor with log odds of 0.00 means that it’s purely a coin toss whether a random movie will be good or bad. That’s Tom Hanks. He gets a D grade.
An actor with log odds of 1.00 will have about 73% good movies versus 27% poor ones, so a random movie will probably be ok, but there will be plenty of rubbish ones in there too. That’s Tom Cruise. He gets a C grade.
An actor with log odds of 2.00 will have about 88% good movies versus 12% bad ones. That’s probably someone like Dustin Hoffman – very solid, but occasionally has a misfire. He gets a B grade.
When you get to log odds of 3, then you are talking about 95% being good. That’s someone like Daniel Day-Lewis, who is as close to a sure bet as you’ll get. He gets an A grade.