It's not too difficult to test the accuracy under more rigorous conditions. Make sure you record the predictions ahead of time. Decide ahead of time what woul

It's not too difficult to test the accuracy under more rigorous conditions. Make sure you record the predictions ahead of time. Decide ahead of time what would need to happen in order to consider the predictions 'accurate.' Wait, and then record which ones turned out to meet their 'accurate' criteria. Deciding ahead of time what would be considered accurate or not is key, because otherwise practically anything can be justified with post-hoc reasoning. The other thing that happens is that the predictions are not very specific. Here is an example. Taken from a horoscope for me today: "The answer will be right there waiting for you." There is no indication what the question is. Just that there is an answer. What's the utility of this? Well there are millions of ways this "prediction" can come true. The chances of something happening any day that fits that criteria is high, and I think most days I'm dealing with an answer to something. So this prediction is meaningless. Here are some levels of specificity: You are going to have a misunderstanding with a friend today. You are going to have a misunderstanding with your closest friend today. You are going to have a misunderstanding about food with your closest friend today. You are going to have a misunderstanding about peanuts with your closest friend today. You are accidentally going to give peanuts to your closest friend today, who will have an allergic reaction. Keep an epi-pen near. I think it's pretty clear that astrology will never deliver a prediction at the level of that last statement, which also happens to be the most falsifiable and most useful. Everything about astrology is really a tool to try to shift your perspective. This can be very useful, which it why it maintains it's popularity. But it has nothing to do with actually telling the future.