It's not just you. Sign stealing and pitch tipping was just gamesmanship back when you had to actually notice something on your own. Now, you can use machine learning to highlight what to look for as easy as pie.
All you need is a little machine learning magic, which is way, way easier in the last few years. Super accessible. All you'd need is to separate clips of as many pitches from the same angle as possible and someone to tell the machine it was x, y, or z pitch. Eventually, it will pick out any glaring differences for you.
You could also use it to predict pitch selection patterns. Kind of like this
rock paper scissors app, which is like a "my first ml algorithm" project. All you'd need there is a fairly basic app and a guy with a tablet entering the last pitch thrown throughout the game. The algorithm wouldn't guarantee prediction, but it'd give you an advantage.
This is a whole new era.