Ask your friend if he ever watches the Weather Channel. When there is a tropical storm or hurricane bearing down on the US from Africa they run a variety of very costly and complicated models to predict where and when it will hit land and how intense it will be. The predictions look like so much multicolored spaghetti with mostly general but no exact agreement between any of them. And this is with PhDs, supercomputers and real-time satellite and aircraft information feeding the models. How often is even ONE of those model correct? Rarely. Predictive modelling is a fancy name for educated guessing, and even the experts can't agree whose is more accurate because none of them are always accurate. ....
I think you are totally missing the purpose of modelling these sorts of things. It isn't about being "right", it's about trying to provide some perspective and proportion to a range of possibilities. "Exact agreement" should not be expected.
When my GPS tells me a trip will take 53 minutes, I don't expect it to be "right". But I can use that to estimate a reasonable range of expectations, which can still be thrown off by an accident and road closing. But I still use that info to help make a decision.
If we have a full understanding of something, and there is little randomness involved, then we no longer use modelling of this kind, we just calculate it. Modeling of this sort is for things we can't calculate, and therefore, by definition, will not be "exact".
... Same with financial predictions, but you add in factors such a human emotion, greed and corruption and it all becomes a crap shoot. So pick a few tools that you like that have decent track records, look for places they generally agree, then guess whether or not they are right and if you trust them enough to follow their advice. Nobody can predict the market long-term.
I'm also curious about what tools you think might have a "decent track record". And how does that jibe with "Nobody can predict the market long-term"? Can this info help make a decision?
Some people will be able to show a track record of market predictions. But we have no idea if that means they can do it again, or if they were just the lucky dart-throwing-monkey. I'd love to be shown how to tell the difference.
And if you have read the intro to FIRECalc, you'd know that it isn't a predictive tool. It wouldn't even call it modeling, it is producing a report of past activity (like looking at past weather events). That doesn't mean it can be predictive, but lacking predictive capabilities, it can be informative to understand the range of past scenarios. It's all we've got, so IMO, better to make use of what we know, than to go into it blind, or to try to find a guru.
-ERD50