Wow, a lot of typos; that's what I get for typing while watching This Old House.

Let's use this thread going forward.

Just finished Chapter 1. It is quite a good read. Still plenty of time for others to jump in and participate.

Quite the interesting read.

The biggest thing I was surprised by was how wrong people could be (even so-called "experts" are barely better than 50/50) without really any consequence. Is it because the public has such a hard time staying interested in a topic long enough to realize? Have we just come expect it is acceptable to be so wrong? Or, does the public demand predictions in areas the simply can't reliably be had yet?

I found the discussion on foxes and hedgehogs interesting. It reminded me a lot of this forum (I'll leave you to decide who would be in what category).

I'm not a fan of sports, so the baseball chapter was a bit rough to get through, although I did appreciate the vast amount of data available, making it an excellent (and profitable) arena for testing models.

I wrote down some thoughts. Will post them once I have a keyboard.

I read this book over the holidays, but I'd be happy to post my thoughts according to the reading schedule. I do a fair bit of forecasting for work (in economics) so I have a lot of thoughts on Silver's take.

Ok. Here are a few of the points I was most attracted to in these first few chapters.

1) The best and most surprising argument, to me at least, was Silver's emphasis on how predictions often use incorrect sample sizes. It seems like such an obvious thing to do, but he points out how S&P and Moody's previous reliability could not be used with the new mortgage backed securities because they were a knew unknown. Past experience was not relevant because it was for a different thing. I thought this was a very damning argument.

2) I don't know if I buy into the whole discussion of hedgehogs and foxes though. To an extent I want to believe it, but I find that argument only works real well with television personalites who employ the hedgehog mentality to get better viewership. I don't know how many hedgehogs are experts in their fields..

3) Finally, I am a pretty big sports fan and also found Chp. 3 a bit of a slog. I'm not entirely sure what it was, but it just didn't seem to hit home as much for me. Maybe because I never really saw issues in baseball prediction, even if such issues did exist.

Overall, I'm liking it so far. Really looking forward to the poker section, since that is a great hobby of mine.

1) I agree. I was shocked when he made such a big deal of being out-of-sample. Seriously, isn't that Stats 101? I'm sure banks weren't making mistakes like that. They must have known and just hoped they wouldn't get caught.

2) Like any personality profile, it has serious flaws, I'll agree with that. Having no experience with such fields, I really don't know either.

Oh, and Stackel, feel free to jump in if you want.

Regarding your questions, I had previously thought that maybe America had wanted predictions in fields that predictions weren't going to be reliable, but I think my sense is shifting some. The reason I think lies in his argument that people don't really think using probability. His whole argument that people tend to argue that 'this is for sure going to happen' etc., and the fact that people tend to overvalue (or even completely don't understand) percentages.

I like when he said that if a forecaster says it will happen 90% of the time, this means that 10% of the time it won't happen. When that 10% happens, people will always say that the forecaster was wrong, but in fact, he was right (so long as his percentages are true in the long run).

"When that 10% happens, people will always say that the forecaster was wrong, but in fact, he was right (so long as his percentages are true in the long run)."

Excellent point. A great example of how most people fundamentally don't understand stats.