Advice projects must not ignore (but often do) . . .

A Must for the Project Manager
(and his or her boss)

Tom DeMarco & Tim Lister: Waltzing with Bears -- Managing Risk on Software Projects
2003, Dorset House, ISBN 0-932633-60-9, 196 pages
reviewed by Conrad Weisert, June, 2003, ©Information Disciplines, Inc.

Scope is broader than title indicates

The misleading subtitle may lead some prospective buyers to assume that this book has a narrower scope than it really has. In particular:

It is about system development projects in general, not just "software projects". Much of the excellent advice applies just as much to projects that purchase and install packaged application software products as to projects that develop custom software. This is especially important in today's environment, where the most popular fad methodologies, such as extreme programming and the "Unified Process"¹, don't support such projects at all.
"Risk" is not something separate from planning and estimating. Although the authors emphasize risk factors, the book offers practical general guidance on realistic project planning and estimating.

Political and organizational obstacles

The central point of Waltzing with Bears is the unending struggle between rational project estimating and management by wishful thinking. This struggle has been explored many times before in books, articles, and courses, including:

Ed Yourdon's well known book on Death March projects,
My own 1998 article "The Burden of Proof in Project Estimating"

Nevertheless, project managers under pressure continue to capitulate and tell their boss what he wants to hear. There is always something special about this project, but we'll surely do the next project right. (I've watched some organizations go for several years without undertaking a normal project!)

Now one more enlighened voice gives support to the poor project planner. DeMarco and Lister's examples and anecdotes are both entertaining and persuasive.

A minor quibble:
Unnecessary ethnospecificity

In today's milieu of international applications and worldwide trade, authors ought to avoid Americanisms that will baffle or irritate readers whose first language is not English or who are unaquainted with U.S. culture. Unfortunately, DeMarco and Lister have liberally sprinkled their text with slang expressions, non-existent words, and obsolete² English units of measure.

If I were to use Waltzing with Bears as a text for a course in another country, I would have to provide explanatory notes. I wouldn't mind that, if there were some trade-off, some clear advantage for a U.S. audience, but there isn't. Nothing at all would have been lost if the authors had used plain English instead of, for example, "slam-dunk"³ (p. 9), "daisy chain" (p. 25), "hairy edge" (p. 31), "gotta", and "gonna"⁴ (p. 147).

I myself was bewildered by "mojo" (p. 116), a word I was unacquainted with. After consulting the publisher and some dictionaries newer than mine, I concede that it's a legitimate, though recent, English word meaning "charm" or "magic spell", but it's still unclear how it contributes to the authors' point,

What ever happened to the system development life-cycle?

The authors avoid confronting issues about project phases. In fact the word doesn't even appear in the index. That's surprising, since a fundamental purpose of the phased life-cycle concept is to limit risk.

On the other hand that may be a sensible choice, given the controversies raging not only about specific life-cycle methodologies but even about the very concept of the phase-limited commitment. The reader may miss the context that many project planners like to lean on. It's as if we expect do the entire project in a single phase. Waltzing with Bears, therefore, must be considered another adjunct to project-management methodology rather than a source of stand-alone guidance.

If the book has a weak point it's Chapter 16, "Incrementalism for Risk Mitigation". The arguments there are unconvincing and the discussion seems unrelated to the rest of the book.

Bonus features

Appendix A is a reprint of part 1 of William Klingdon Clifford's 1876 paper "The Ethics of Belief", still relevant reading for managers.

An 8-page bibliography directs the reader to some useful references, including the 1997 IEEE software special issue on "managing risk".

Highly recommended

¹ -- A system development life-cycle descended from Ivar Jacobson's "Objectory"
² -- Obsolete for most of world, including now even Canada
³ -- Basketball may be an international game, but not everyone is a fan. (I asked 8 acquaintances (1 Russian, 2 Indian, 1 British, 5 American) to explain the term. Two of them had no idea.)
⁴ -- Like the now-ubiquitous "wanna" these words were not listed in reputable 1970 dictionaries. More recent dictionaries identify them as "non-standard" or "colloq., esp U.S.". The respected brand new Cambridge Advanced Learning Dictionary continues to ignore them, as do prudent writers.

Return to book reviews.
Return to table of contents.