Likelihood of Success in Percentage Terms; What Does It Really Mean?

Anyone who practices trademark prosecution has likely faced that most dreaded of questions: "what is the percentage likelihood that you can overcome the examiner's objection". Indeed, some of you may have put that question (or have been made to do so by your client) to your associate. I have thought long and hard about this, both from the vantage of being asked it as well as seeking the information from others. Here are some of my thoughts.

First and foremost, there is the issue of applying numerical percentages to an inherently uncertain activity. Frank Knight, the famous economist from the University of Chicago, taught us nearly 90 years ago that there is a fundamental difference between uncertainty and risk. In his own words:
"Uncertainty must be taken in a sense radically distinct from the familiar notion of Risk, from which it has never been properly separated.... The essential fact is that 'risk' means in some cases a quantity susceptible of measurement, while at other times it is something distinctly not of this character; and there are far-reaching and crucial differences in the bearings of the phenomena depending on which of the two is really present and operating.... It will appear that a measurable uncertainty, or 'risk' proper, as we shall use the term, is so far different from an unmeasurable one that it is not in effect an uncertainty at all."
In other words, uncertainty is characterized by being incalculable, as opposed to risk, for a which a meaningful calculation is possible (à la probability). In truth, the decade's most prominent curmudgeon, Nassim Nicholas Taleb, expressed skepticism over whether any risk can really be calculated. Thus, in The Black Swan, Taleb famously questioned Knightian uncertainty in the following words:
"In real life you do not know the odds; you need to discover them, and the sources of uncertainty are not defined. Economists, who do not consider what was found by noneconomists worthwhile, draw an artificial distinction between Knightian risk (which you can compute) and Knightian uncertainty (which you cannot compute), after one Frank Knight, who rediscovered the notion of unknown uncertainty and did a lot of thinking but perhaps never took risks, or perhaps lived in the vicinity of a casino. Had he taken financial or economic risk he would have realized that these "computable" risks are largely absent from real life! They are laboratory contraptions."
Whether or not one accepts Knight's or Taleb's view of things on this point, the

upshot for us is the same. There is something deeply troubling about being asked to give a percentage to the likelihood of success of a trademark application, an activity that certainly lies within the broad field of uncertainty.

Rare is the slam dunk rejection ("Coca Cola" for soft drinks; "table" for tables) or slam dunk acceptance ("UBUGUGU for any goods or services). Not infrequently, the tendency of a trademark owner is to select a mark that, at least obliquely, seeks to relate to the goods or services to which the mark is being used. As such, the ultimate fate of such a mark upon examination or the like (depending upon whether the jurisdiction addresses relative grounds only upon a challenge by a third party) will usually fall within the parameters of Knightian uncertainty. And yet--there I am, being asked (or asking) for a percentage view of likeklihood of success.

For students of psychological measurement, it would be a fascinating study to map out the parameters by which we trademark practitioners give a percentage estimate for likelihood of registration success. Some of these considerations may be the following:

  1. How broad a band are we setting when we use the ubiqitous 50/50 as the anchor?
  2. Do we err on the side of overstating or understating the likelihood of success?
  3. If so, what are the factors that lead us to overstate or understate?
  4. If so, how does that tendency affect the percentage that we give?
  5. Are the percentages that we give linear or non-linear? If so, in what maner?

From the point of view of the person seeking the percentage, the focus is on how such person uses such information. To wit:

  1. Does it affect a decision to instruct the associate to proceed with prosecution?
  2. If so, how is the information used to reach the decision?
  3. Does it matter if the percentage given by the associate is accompanied by a narrative that further describes the factors underlying the likelihood of sucess?
  4. Does the decision-making of the instructing party differ, depending upon whether the request for a percentage estimate comes from the instructing party or the client?

Questions galore--any thoughts or guidance would be most welcome.

Questions, questions, questions
Likelihood of Success in Percentage Terms; What Does It Really Mean? Likelihood of Success in Percentage Terms; What Does It Really Mean? Reviewed by Neil Wilkof on Friday, September 24, 2010 Rating: 5


  1. Hi Neil

    Once again we align on an IP issue. This is getting weird. Recently, I realized that the percentages related to patent litigation success probability batted about by the Chief IP Counsel of my Fortune 500 client were not only erroneous, but also unhelpful to my senior VP of Innovation. We lawyers all-too-frequently respond to substantive inquiries regarding "can we win" by doing hand waving about percentages that are, frankly, made up out of thin air. This is not only wrong, it is unfair to business people who are looking for real insights about whether they should take a legal risk or not. Unfortunately for the business people we represent, our perception of how hard or difficult it will be to prevail on an IP legal issue is often manifested in the form of bogus percentages. It is a great big Fail from us, in my humble opinion.

  2. Neil, in a perfect market you could form a good assessment of the perception of the percentage chance of success of any event by looking at the odds that are decided at arm's length between punter and bookmaker. In this context, it is hard to see who is going to run the book but the same result is achieved by comparing the price for a flat fee agreement (which provides the expectation level) with the price for a fee wholly conditional on success. The ratio provides an estimate of the advisor's perception of probability of success. In fact, it must underestimate the probability of success because the conditional fee will be at a premium to price in the advisor's willingness to accept risk and the cost of money (as payment would normally be postponed until success or failure becomes clear). None of this measures actual probability of success but that probably doesn't matter.

    An interesting question for advisors and probability that often troubles me is that there is a big difference between the perception of where the balance lies, and the chance of succeeding. The former requires an assessment of the case as a whole, whereas the latter requires an assessment of the likely variation in assessments of the case from one potential decision maker to another. The assessment of a case as "55/45" can often mean that the advisor sees the scales as tipped slightly to one side. However, if every advisor comes to the same conclusion, that means the outcome is 100/0. The key question is to estimate what percentage of the target decision-making audience will disagree with the advisor's perception. How far the end result goes from 55/45 depends on whether the advisor can correctly judge the majority view, and how consistent the decision makers are among themselves.

  3. I looked up "ubugugu", and it would appear that this word means "greed" in Kinyarwanda, so it may not be suitable for branding the Gordon Gekko Financial Services.

  4. "any thoughts or guidance would be most welcome"

    See Levicom v Linklaters, earlier this year (Decision at Appeal, 11 May 2010).

    As regards the likelihood of success in percentage terms, it certainly provides food -and pause- for thought.

  5. Generally, both sides will be overly optimistic about the chances of winning (whether expressed in terms of percentages or not; note that probability is usually expressed on a scale of 0-1), see This is one reason why cases go to trial at all, according to classic law and economics.

  6. I'm surprised by Frank Knight's comments (my personal view is that they are profoundly misleading and unhelpful).

    The founders of modern probability theory - people like James Bernoulli and the Marquis de Laplace - though of probability in terms of a degree of belief: what would now be known as "subjective probability". An assessment of probability is always a statement of one's knowledge or lack of it. There is no meaningful division between uncertainty and risk.

    While there was a long period when subjective probability was out of fashion, it has now become much more respectable again. It underpins most modern decision theory (which is after all our business, though we tend to use far less formal methods than my engineering friends might) and the use of Bayesian methods is much more common in science than it was when I worked as a statistician. So much more so, that it is becoming the new orthodoxy.

    So, a good start in answering your question is to try to approach it from a subjectivist/decision theory perspective.

    A second useful pointer is to realise that what you (subjectively) estimate is a probability distribution of any particular outcome (in some of my advices there may be a range of outcomes rather than a binary one, but there is still a distribution). When you summarise that distribution as a single figure in the percentage chance of success, you engage the same questions as those faced by anyone presenting summary statistics. You may not be aware of using a particular summary statistic, but that is implicit in what you do. For example do you use a mean, median or something else?

    So a client may not really be interested in knowing their expected outcome (factoring in litigation costs and so on). If we only cared about expectation, then insurance companies would make no money of course.

    Whether all this reflection really helps I don't know for sure, but it is worth being clear in one's mind what one is doing. I am sure that most lawyers intuitively average over their client's utility function (or at least try to) but maybe a sophisticated client could be given more information than a single figure.

    Its an interesting point to raise. Anyone working on a CFA is having to make these evaluations for themselves.

  7. Percentage chances of success serve as important guidelines for quantification of damages in professional negligence actions (Kitchen v RAF; Allied Maples Group v Simmons & Simmons). What happens in the event of a client losing an opportunity of a patent or TM registration because of his attorney's negligence? If it can be proved that the application would have had, say, a 60 % chance of success, the claimant's damages would be reduced by 40 %. If the application would in any event have had no chance of success, the damages would (notwithstanding the admitted negligence) be zero. Negligible chances of success (the normal benchmark being 15 - 20 %) would also reduce damages to zero.

  8. I am uncertin if this applies in Trade Mark prosection but in IP litigation I use the following:

    You are unlikely to be able to be more precise than 25% increments.

    Therefore there are only 4 answers:

    1) 0-25% chance of claimant winning - this lowest estimate is rarely given if you are the claimant as it suggests you never should have begun the action.

    2) 25-50% chance - this can be given where the claimant has been advised they may lose the litigation but has resolved to go ahead anyway.

    3) 50-75% chance - can be given where the claimant has been advised they have a fair chance of success and it includes a proper allowance for litigation uncertainty.

    4) 75-100% chance - this etaimate is uncommmon for the same reasons as 0-25% band estimate but in reverse.

  9. Don't know why you all see it as so uncertain. There are two main areas where value judgments come into play - surplus (descriptive marks with a hint of distinctiveness); and slogans. Harder to predict oppositions mind....

  10. It's simply a fudge. Claim a 50-50 chance of success and you'll sound thoughtful and measured, and yet never be wrong. But the same is true if you quote 25%, 50%, 90% or 10% chances of sucesss.


All comments must be moderated by a member of the IPKat team before they appear on the blog. Comments will not be allowed if the contravene the IPKat policy that readers' comments should not be obscene or defamatory; they should not consist of ad hominem attacks on members of the blog team or other comment-posters and they should make a constructive contribution to the discussion of the post on which they purport to comment.

It is also the IPKat policy that comments should not be made completely anonymously, and users should use a consistent name or pseudonym (which should not itself be defamatory or obscene, or that of another real person), either in the "identity" field, or at the beginning of the comment. Current practice is to, however, allow a limited number of comments that contravene this policy, provided that the comment has a high degree of relevance and the comment chain does not become too difficult to follow.

Learn more here:

Powered by Blogger.