Dark Review of a Wine Book — Rejected by the Journal of Wine Economics
By Roman Weil
I hope you enjoy this lark, which results from my having time on my hands and a reputation as an oenonomist — an economist who does empirical research about wine. The book review editor of the J. of Wine Economics (a Harvard guy, but forgive him that) asked me to review the 35th (!) edition of Kevin Zraly’s Windows on the World Complete Wine Course: Revised & Updated, Sterling Epicure; 2021; Hardcover: 464 pages, $35. ISBN-10: 1454942177; ISBN-13: 978-1454942177; First published 1985.
I had a ball writing this review for reasons you’ll see in a moment. The editor, after consulting with his co-editors rejected my first version because, they said, my review was more about what the book did not contain than what it did. Get rid of the stuff about its omissions and tell us its inclusions. So, I obeyed and did a boring review, which I’ll not give away anything about here. Instead I’ll show you what I wrote, which the editors rejected. All the following was in the first draft; none of this will appear in the Journal.
I’ll start by disabusing you of the notion this review is about Kevin Zraly’s book alone. You know him, sommelier at Windows on the World until its untimely demise on 9/11, but not me, so I’ll start with me.

Roman Weil
First topic: you are about to go to, if not already going to, business recruiting interviews, which almost always involve business meals, mostly lunches, which almost always involve wine. This was 1985, remember and in those days the BSD (see the Urban Dictionary) types might commandeer the wine folderol. In their version of the stress interview, however, they might pass you the wine list and ask you for a choice. Now, I ask, what do you do? That’s what this presentation is about: how not to be intimated by wine at business lunches.
Second topic: Keep in mind this is 1985 and I had not yet done my systematic research on amateurs, beginners and advanced, to distinguish between wine tastes, which I’ll discuss a bit later, with references. By then, I knew only that wines taste different from each other as a function of the grape from which they are made and where they are grown. We don’t have time to test both red and white wines, so I’ll focus on white wines because that’s more likely what your BSD hosts will think appropriate at lunch. So I constructed a 3 x 3 matrix of European, U.S., and Australia/New Zealand wines of chardonnay, sauvignon blanc, and Riesling. I explained that there is no right answer to preference among grape tastes. I, almost alone among winos I know, prefer Riesling and I explain why I think most people dislike them. Most sweet wines have insufficient acidity to balance the sweetness and I explain the difference between simple syrup and lemonade. Yada. Yada.
NBC’s Today Show got wind. “Stanford Business School teaches students how to deal with wine at business lunches. Migod. If you ever do that again, let us know and we’ll put you on TV.” Well, I’m no PR fool. Now, I’m back at Chicago from my visit and we put one on there, so it could be on the Today Show. Students had a ball, because I’d done one and knew how to trim it down so that more of the minutes counted.
But that just started the ball rolling. Every year, students said, “We heard what you did for last year’s students. We want one, too.” And the alumni affairs office said, “We think such seminars would be fantastic alumni events; do them for us.” By the late 1990s, I was doing several a year and finding it boring. What could I do? I read Frank Prial’s column in the NYTimes of Feb. 9, 2000 and had the idea of testing his hypothesis that the vintage chart was obsolete. I designed the tasting part of my seminars over several years to test so-called Appalling vintages against Good vintages from the same producer in triangle tests, where testers face three glasses. Two of the glasses have the same, identical wine. The third glass has a different vintage, but otherwise the same label. The testers say which of the three glasses differ from the other two and which they prefer. From several hundred individual tests of this nature (including some with Yale 62 classmates), I derived statistically significant results that 40% of the testers can distinguish the Good vintage from the Appalling but half of those prefer the Appalling.
I published that paper, but the requests for seminars didn’t stop. So I switched experimental design to test regular vs. reserve bottlings of the same wine from the same year from the same producer, whether European or U.S. After several hundred testers’ reported results, I found, again that 40% of the testers could distinguish the regular from the reserve and half preferred the regular to the reserve. The result appeared in Chance, 18, 2 (2005), 9-15.
Finally, I did a third round of triangle tests, reported in this journal [2, 2(2007), 136-144], wherein I tested whether wine reviewers’ words to describe wine tastes conveyed information to my testers. In a word, no.
Over 25 years, I conducted wine classes and seminars, with tests of students and alums, eager and some experienced amateurs and found that about 40% can distinguish one wine from another in tests where one-third would guess right at random, but half those 40% prefer the Appalling wine or the regular over the reserve. And no one can systematically match wine reviewers words with the wines themselves. As the JWE editor called the issue that published that paper, it’s Wine Bullshit.
So now you know my bias as I opened the 35th edition of the Windows on the World wine course, which has sold, the publisher says, over 3 million copies. I never visited the restaurant nor looked at any of the editions of the book before now. I acquired copies of the first edition, 1985, and 30th anniversary edition, 2014, as part of homework to do this review.
Bright idea. I asked economists I know, expert in wine, what questions they’d want a book, purporting to teach beginners and intermediates about wine, to answer. I have edited and compiled their questions in the attached Exhibit 1. I report whether the Windows book addressed these questions. I give my own comments and answers. In fairness to the book, I list questions it answers that none of us thought to ask. A critic of this review could well say, “That list of questions sounds much like economists compiled it.” Yes, but this is a journal (mostly) for economists.
None of us seems interested in the chemistry of wines, for example, but the Windows books spends 35 pages on the chemistry of wines, the physiology of tasting and smelling, and the mechanics of wine tasting. After examining those pages I cannot tell how reading them enhances my enjoyment of wines, other than the taxonomy and physiology of tasting, which I think useful. If I were at a gathering of colleagues from the Chemistry Department, then I might be able to word drop, but I’d be a fakir.
The book has grown from 191 pages in its first edition to about 465 pages in its 35th. The type fonts have grown as have the size of the illustrations, so I guess the content has not expanded as much as the growth in page count would indicate. Virtually all the information from the first edition about grapes and regions continues with more details and expanded coverage.
In comparing the 30th and 35th editions, I see little change except some updating of recommended vintages, which, given my biases from own controlled experiments, I think not valuable. The 35th contains a 40-page personal history not contained in the 30th.
If you want this book’s teachings without paying current prices, go to abebooks.com and buy a recent used copy. As I write this, I see copies of the 2016, 32nd, edition available for under $7.
Now, refer to Exhibit 1, where we get to the meat of the matter, the questions my colleagues and I posed before we looked at the book. You’ll see many go unanswered because we are economists and Mr. Zraly isn’t. I’m most disappointed that in spite of the fact that he recommends over twenty different so-called “guided” tastings/testings of wines, he nowhere suggests the triangle test format, which I think indispensable for judging wines. What do I know? He’s sold 3mm books and taught thousands of students to know and enjoy wines. I’ve tested fewer than one thousand subjects with several dozens of pairs of wines. I always get the same results: 40 percent of my subjects can distinguish which wine differs from the other and half of that 40% prefers the allegedly inferior wine—the lower ranked vintage or the regular, in contrast to the reserve, bottling. I come away from these tests believing that amateurs cannot tell the difference between wines that come from the same grape of the same age and from the same producer. In other tests, I find that the wine words used by wine critics convey zero—that is, no—information. That is words other than “sweet,” or “sweeter,” or “sweetest.”
The book contains overwhelming details about names of grapes, regions, wineries, distributors. I confess I poke fun at the book in my list of questions it does provide answers to. The book tells us nothing about how to conduct a wine tasting nor how to score and rank wines in one. It doesn’t help us evaluate wine critics’ ratings where there is a 100-point scale, but all scores seem to be 88 or above.
Here is the accompanying exhibit which gives the list of questions the oenonomists posed with answers from the book and my comments where the book is silent.
We welcome your comments below.
I neglected to say, in discussing the informativeness–or not–of wine words, that in addition to sweet/sweeter/sweetest, one can often learn to distinguish more alcoholic wines from less. I think it folly to do a triangle test on two wines whose alcohol contents significantly differ–say 10% v. 15%. California wines tend to be more alcoholic than French wines made from the same grapes, for example California Cabs v. French Bordeaux. This results from California temperatures during the growing season being hotter than France’s. The heat causes the grapes to grow with more sugar. The wine maker who doesn’t want to make a sweet wine, with as most winemakers, has to get rid of the sugar by turning it into alcohol. The sugar in grape juice can remain, making the wine taste sweeter, or be vinified into alcohol.
Roman
How about Hazlett red cat?