In this article, I will show how random errors, legal uncertainties, judicial bias, and application of multiple rules and doctrines affect the probabilities of PTAB invalidation of patents. I will use several simple models to show how those factors work against patent owners in a multiple-challenge environment.
1. Random Errors Work Against Patent Owners
A patent, which is assumed to have marginal merit, would be invalidated or survive by an equal chance of 0.5. Final results of the challenges would depend upon a large number of random errors such as attorney’s strategic errors, clerical errors, citation error, judge’s ruling errors, omission of required statements, missing deadline, and more. It is assumed that the administrative judges are impartial and all parties work within the bounds. I will estimate the probability for each of the challenges and the cumulative probabilities of invalidating the patent. In any single proceeding, those errors work in both ways: any of them may help the patent survive or kill the patent. However, the errors are aggregated in a total different way when there are multiple challenges running independently. To survive, the patent owner needs to beat the odds in each of the challenges. However, to kill the patent, the challenger only needs to make one single fatal-strike. The following table shows how probabilities look like:
Table 1. The cumulative effects of random errors in multiple challenges on the probabilities of invalidating a patent
Note: I use 0.5/0.5 probabilities solely for convenience, a result with similar treads could be shown by using 0.4/0.6 probabilities.
This table shows that errors in favor of invalidating the patent accumulate. The first challenge may be fair, but in the second, third, and Nth challenges infringers will eventually take the patent out. The bias caused by error accumulation is the reason why double jeopardy is prohibited under the U.S. Constitution. If we have a large set of data for patents with marginal merit, the data would show that patent owner might screw up in one or more of potentially thousands of things from filing a wrong paper or perhaps to mistake in fee payment. It is possible some patent owners just give up.
2. Legal Uncertainty Works Against Patent Owners
The following table shows how legal uncertainty works against patent owners. It is assumed that the legal theory has created a probability uncertainty. For a patent with marginal merit, it could be invalidated or saved. However, due to the desire to invalidate the patent, administrative judges may take advantage of the legal uncertainty in invalidating the patent. Table 2 below shows how this element might affect the probabilities of invalidating the patent in different challenges.
Table 2. Cumulative Effects Of Legal Uncertainty On Invalidating A Patent And The Probability Of Invalidating The Patent In Each Of Multiple Challenges
Note: R is the change in probability attributable to legal uncertainties. It is an arbitrary number, but 0.2 is a reasonable number to show legal uncertainty. Any value would be used to show the same trend.
P and Q are hypothetical probabilities for survival or invalidation. R is a change in the probability attributable to legal uncertainties (a reduction in probability to survive). A patent could be invalidated at a higher probability if judges resolve each legal uncertainty against the patent owner. To some extent, this is a subtle bias that is restricted by existing law and case facts, but the judges still work within the bounds of law. Patent owners cannot successfully challenge such bias. When all legal uncertainties in different proceedings are resolved against the patent owner, the probability of invalidating the patent will increase. The probability of invalidating the patent will rapidly increase with the number of challenges.
I made an assumption that all proceedings are not related. However, different proceedings may involve the use of same reference, same judges, and same case law. Despite the lack of independence, the table accurately predicts how this factor might work against the patent owner.
One problem is the abuse of broadest reasonable interpretation standard. The Patent Office does not provide general guideline on how to find equivalents for claimed elements. The PTAB can use one or two words in a claim limitation to find claimed elements. This gives them great margins to exercise discretion against patent owners.
3. Bias of Administrative Judges Works Against Patent Owners
It is known that the Patent Office selects administrative judges who are willing to invalidate patents. Since they can and do manipulate panels to achieve invalidating results, the impacts of their intentional conduct cannot be studied by any probability model. They can achieve whatever they want. However, I still use a simply model solely to show the boundaries of the impacts of their intentional conduct. Their conduct is limited by office rules and precedents which can prevent them from entering completely arbitrary decisions.
The real problem is the combination of using the broadest reasonable interpretation standard and worldwide prior art in a Section 103 challenge. This combination could kill most patents. Deciding patent validity under this combination is impossible in practice. No one can ever exhaust prior art if he or she is determined to find prior art. If a sufficiently large reward is offered, prior art reference could be found anywhere in the world in any language. Then, no body can prove or disprove the existence of such a reference.
Table 3. The Impacts of Biased Administrative Judges and Acts of Improper Conduct of Challengers on Probabilities of Invalidating a Patent
Note: S is a change of probability attributable to the bias of administrative judges or the improper conduct of the challenger.
I do not show cumulative invalidating probabilities because each challenge could have near unity for invalidation and a patent cannot be invalidated twice or multiple times. The product rule used in Table 3 is still applicable as the patent survival depends upon the product of survival probabilities. However, when bias is a factor, one challenge most probably would take the patent out. To see how multiple challenges affect the patent, imagine that the patent is challenged in multiple secret proceedings, after all proceedings are over, and then see results. Chances are that the patent might have been invalidated multiple times.
4. Application of Multiple Rules Against Patent Owners
Contrary to the trend to reject, disuse, abolish, and reform the common law adjudication model in the world, the patent office has made its rules more and more complex with endless technicalities. Its proceeding is run like a common law court. Therefore, I would like to examine the cumulative impacts of applying multiple motions, multiple doctrines, multiple rules, or their combinations in an invalidation determination.
The foundering fathers of the nation did NOT use a court as the agency for issuing patents even though court was then the body that issued patents in states such as Massachusetts. They must have anticipated that if a patent would be issued by a court, the patent system would not thrive. They knew that Britain had an equity movement to address the well known injustice that was often seen in common law courts. Despite such well known lessons, the Patent Office (at the behest of Congress) created the whole version of court-like proceedings for the PTAB.
To see the serious bias in PTAB procedural designs, one should note how the adjudication model was developed. The model was developed in England before the probability theory was known (The first recorded evidence of probability theory can be found in 1550 in the work of Cardan). The common law model in its current state is inherently bias against plaintiffs. This is not what the designer could understand at that time. Assuming that a plaintiff must pass four motion challenges, and, for each of the challenges, the plaintiff has a 0.5 chance to survive, the final chance for the plaintiff to survive all challenges would be 0.5*0.5*0.5*0.5=0.0625 (assuming that each challenge is based on procedural facts unrelated to other challenges). In other words, the chance of success for the plaintiff is much lower than a hypothetical probability based upon case merit. When a defendant keeps trying to throw out a case, it will hit a lottery at some point. When the system has this built-in bias against plaintiffs, the judge’s bias, however slight, can greatly affect case outcomes. While patent owners are defendants but not plaintiffs, the PTAB procedure actually puts them in a similar unfavorable position because the purpose is to invalidate challenged patents. The procedure is designed with all adverse actions for default, mistakes, failure to appear, etc. against patent owners.
One big problem in PTAB procedure is using multiple rules, doctrines, and holdings in invalidating the patent. The successive applications of multiple rules or doctrines in a proceeding result in a great bias. Application of multiple rules were frequently used in all common law courts. This practice is still common in courts in a few nations. For example, a rule or doctrine may decide what law is used, which precedent controls, what evidence is admissible, and how an issue should be decided. Those rules or doctrines were developed in ancient society where the ability to find truth was very limited. Many rules are accepted in a majority of jurisdictions because they are most PROBABLY right. Whenever, an issue is controlled by such a rule, the issue in a case would be decided by the rule directly. In other words, the outcome of the case would depend upon the analysis whether the rule should be applied or not. This analysis would hinge on a small number of factors unrelated to case merit.
Errors inherent in such rules impair the court’s ability to deliver justice. However, a much bigger problem occurs if multiple rules and doctrines are used in one ruling. All errors can accumulate. A plaintiff loses his or her case if he loses one, two, or all of the issues. Most plaintiffs would be denied of justice due to accumulated errors.
This error-accumulating problem is very serious in patent cases because the patent is supposed to have exclusive right for about 17 years (with about three years used as prosecution time). A challenged patent may be invalidated if any of the three issues decided under the rules or doctrines is against the patent owner. If each rule introduces 0.4 error probability, the accumulated error probability by three time applications would be 1- (1-0.4)^3=0.784. This can be viewed as the probability for the patent owner to lose due to accumulated errors in all three rules. The patent may be invalidated due to the errors inherent in those rules and doctrines. The PTAB procedure would have a similar bias.
5. Combination Impacts of Errors, Bias, Legal Uncertainty, and Rules
The PTAB procedure has all of those attributes that are discussed in the models. That is why the PTAB can invalidate 70% or more of challenged patents. Those four factors work in some kind of cumulative way. If a patent is challenged once, the patent owner suffers a great threat attributable to normal errors and errors inherent in all majority rules and doctrines. If judges are required to invalidate the patent by using whatever legal uncertain they can use, the chance for invalidating the patent will be further high. If administrative judges are bias, they can use incorrect precedents, rewrite case facts, re-invent prior art, call an apple an orange, and cite arbitrary citations. There is no chance for this patent to survive. So, many of challenged patents would be invalidated by random errors; and many of them are invalidated by legal uncertainty, and some of them are invalided due to the bias of administrative judges. While patents are invalidated only once, many of them could be invalided many times for difference reasons.
The model results unequivocally speak the truth about how errors and bias are aggregated in one way to hurt patent owners in multiple challenges the PTAB procedure most probably violates the Fifth Amendment of the U.S. Constitution.
The U.S. Constitution, Fifth Amendment reads: “No person shall be held to answer for a capital, or otherwise infamous crime, unless on a presentment or indictment of a Grand Jury, except in cases arising in the land or naval forces, or in the Militia, when in actual service in time of War or public danger; nor shall any person be subject for the same offence to be twice put in jeopardy of life or limb; nor shall be compelled in any criminal case to be a witness against himself, nor be deprived of life, liberty, or property, without due process of law; nor shall private property be taken for public use, without just compensation.”
The PTAB invalidation procedure violates due process because it drags patent owners into endless assaults by patent infringers and patent thieves. Due process requires compliance with standard notions of fair play and justice, which is lacking at the PTAB. The PTAB invalidation process does not afford due process because it allows an unlimited number of infringers and patent thieves to shoot down a patent in a never ending series of challenges. If one challenge misses the target, the second and the third challenge will hit it, or ultimately it will be hit by the Nth challenger. When patents can be attacked repeatedly in such a bizarre way, the patent reward of an exclusive right is a meaningless promise.
The same clause baring double jeopardy in Article III courts must make such unlimited jeopardy in PTAB proceedings unlawful.