As astrologers we have the great advantage of learning how to predict events through retrospective research. There's no limit to the number of charts with accurate dates of past events that you can collect and study. Simply studying the significant events in your own life can be extremely valuable learning experiences. To see karma unfolding, even if it's a difficult karma in your own life, through the dashas and transits is fascinating. My standard suggestion to students, once they've learned the basic principles (which is no small task!), is to study their own charts and at least five others. The charts of close friends, family members, professional colleagues, or popular celebrities can all become rich research laboratories for studying Jyotish. I particularly like the Astrodatabank software program for this, which gives you the birth data and biographical histories of thousands of celebrities. However, if they're people that you are in regular contact with then you have the opportunity to see events as they happen and can practice making predictions. Of course, there's no better chart than you're own to work with since you know all the subtle details and can follow it on a daily basis.
The hardest thing for students to acquire is experience. If you're giving a reading and a client wants to know definitively whether they're going to get promoted at their job don't you think you should have studied at least 10 examples of job promotions? 100 would be even better. When you do a lot of retrospective research on events and you continually see the same patterns repeating it strengthens your predictive abilities. I like what Marc Boney says about this, "Do retrospective research enough with your own and the charts of other people, and it will give you that 'saturated experience' that enables you to eventually see events prospectively." For example, in my own research I've seen that people often move in Saturn-Moon or Moon-Saturn dashas.
This is especially true if transiting Saturn is also aspecting the natal Moon, like in the 4th house from the Moon. I've seen this enough times retrospectively, for instance, when clients give me a list of previous events, that it has now become useful for prediction.
For instance, this is a chart of a woman who recently moved to a foreign country to try it out while in her Moon-Saturn dasha. My explanation of why Saturn and the Moon are a common dasha sequence for moving is because the Moon is the karaka of the 4th house of home, while Saturn is the karaka of two dusthana houses - the 8th house of change, and the 12th house of loss. In this woman's chart Saturn also happens to be the 8th lord from the ascendant, Moon, Sun and dasha lagna Moon. Saturn also aspects her 4th lord Mercury. While these factors reinforce a move during her Saturn sub dasha, I don't believe that they're necessarily the cause of her move. Just being in a Saturn-Moon, or Moon-Saturn dasha could be enough.
A great way to do retrospective research is simply to ask a friend, who has an accurate birth time, for a list of the dates of important events in their life and then look up what dashas and transits were in effect at the time. Better yet would be to do this with a friend who's also studying Jyotish. You could quiz each other on events in each other's lives, or other people that one person knows well. "What's this person's main character traits? What's their profession? etc" Or pick a date of an event that you know about the person and ask, "Can you tell me what happened to them on ____?"
Of course there's the argument about retrospective research that hindsight is always 20/20, and that it just teaches you to justify past events however you see fit. What we need to avoid is what Andrew Foss, creator of Sri Jyoti Star software, calls, "The one a half effect." This is the creating of one's own theories based on just seeing something in their own chart (1x), and kind of seeing it in someone else's chart (1/2x). What you're left with is your own half-baked theory that may not work a majority of the time. This can be a problem, which is why it's always best to have some training with a tutor or at least cross check with someone who has more experience, in order to confirm that your interpretive logic is correct.