Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.2 9/18/84; site uw-beaver Path: utzoo!watmath!clyde!cbosgd!ihnp4!houxm!vax135!cornell!uw-beaver!laser-lovers From: laser-lovers@uw-beaver Newsgroups: fa.laser-lovers Subject: impartiality in blind comparisons Message-ID: <1186@uw-beaver> Date: Sun, 19-May-85 17:40:05 EDT Article-I.D.: uw-beave.1186 Posted: Sun May 19 17:40:05 1985 Date-Received: Tue, 21-May-85 05:42:20 EDT Sender: daemon@uw-beaver Organization: U of Washington Computer Science Lines: 44 From: Brian Reid I've had a wine tasting at my house every Monday night for several years. One of the things that I have noticed is that people can often recognize particular wines. From time to time one of the winemakers shows up, and they always dazzle us with their ability not only to recognize their own wines, but to tell one year from another. One person owns many hundreds of bottles of wine from a certain winery, and he can almost always detect that wine in these tastings. Whether he *prefers* it is a separate issue from whether he can *recognize* it. Similarly, I suspect that I could recognize, by brand name, the output of this or that laser printer, and this or that text formatter, for a reasonably large number of printers and text formatters. I'm not as scholarly as Chuck Bigelow on these matters--I could never manage to tell who the 16th-century Italian designer of some font might have been--but I can normally spot Xerox or Imagen or Apple/Adobe or Talaris/QMS output, even if I don't know whether I am looking at Times New Roman or Optima. I would guess that the people in Les Earnest's group can do this too. Some of them probably have pretty good eyes for this kind of detail. And therein lies the problem. Given that most experts in this field can *recognize* the output of their own systems, and thereby pretend to *prefer* it if they so choose, and given the nature of the accusations Les has been making, it is very important to have comparisons like this made by disinterested third parties. Even if the people doing this "blind comparison" have no financial interest in any of the products being tested, there is a tremendous amount of interaction between people at Stanford and the various laser printer companies, and there are going to be plenty of folks who are quite partial even without owning stock in Imagen or Adobe or Apple or QMS or Xerox. It is my understanding (I'm sure that Les will correct me if I am wrong) that Imagen was founded by a bunch of people who left Stanford (primarily from the TEX group) to sell commercially a laser printer whose initial prototypes were made in the basement of the Stanford CS department by folks in that research group. If that is true, then I claim that a "blind tasting" by members of that group is not an impartial tasting, and that Les therefore did exactly the right thing by keeping the results private. -- Brian Reid decwrl!glacier!reid Stanford reid@SU-Glacier.ARPA