I have written and posted on how forecasting is generally more accurate than at least the media would have you believe — with the caveat the forecasts are often best part of a range of alternatives and intended to provoke thinking, and not necessarily to predict the future (since that is virtually impossible). So, we can get a good sense of the possibilities. But of course, futurists are stilled asked about their accuracy rates. So I’ve been on the lookout for these, having written about my own and my colleagues work, Futurist Editor Ed Cornish, and Ray Kurzweil. Me and my colleagues and Ed came out about 67% accurate; Kurzweil claimed 78% dead on and the rest being close, with just 3% wrong. So far, so good. But, alas, a little cold water was splashed in my face when I re-read former client and colleague Rich Albright‘s wonderful piece: What can past technology forecasts tell us about the future? Technological Forecasting & Social Change 69(2002) 443–464 445. He convened a panel of eight reviewers to evaluate 100 forecasts of of technical innovations made by Herman Kahn and Anthony Wiener [1] in their classic The Year 2000, A Framework for Speculation on the Next Thirty-Three Years published in 1967. The reviewers found found that fewer than half of the forecasted innovations occurred before the end of the century. I was curious as to why they seemed to find a much lower accuracy rate. So I went through the 100 forecasts myself, and put on my tough grading hat (don’t tell the students, but I’m kind of a softie on grading) and my results were not much different. They found that 40ish% and I netted out at 49% had clearly occurred by the turn of the century. Another 25% were close, but this was not horseshoes. As I discussed in “How Accurate Are Your Forecasts? there are many fine lines in terms of wording and interpretation. Nonetheless, the search continues. If anyone has a good lead on some forecasts that have been evaluated, please pass them along. Andy Hines
