Reverse Bifurcation in Practice

By Jason Rantanen

Patent litigation can be particularly frustrating and inefficient when litigation costs far exceed the potential damages resulting from an infringement.  This issue is particularly problematic when the patent holder's expected litigation costs are significantly lower than those of the accused infringer, such as in the instance of the prototypical "patent troll," because under such circumstances the settlement value is largely driven by the expected cost of litigation to the accused infringer as opposed to the actual damages from the infringement, resulting in a windfall for the patent holder.

One strategy for addressing this issue – and for streamlining patent litigation generally – is to decide cheaper and easier issues early in the litigation process, a concept discussed by Dennis Crouch and Robert P. Merges in their article Operating Efficiently Post-Bilski by Ordering Patent Doctrine Decision-Making, 25 Berkeley Tech. L.J. 1673 (2010).  An example of the strategy is "reverse bifurcation," in which the damages phase is moved to the front of the litigation sequence with the goal of figuring out how much is really at stake.  This procedural move may be particularly appropriate in the case of patent holders who sue a large number of parties, including parties with relatively small potential damages pools, in the hopes of extracting the aforementioned undeserved rents. 

This approach was recently employed by Otter Products LLC in its suit with Gellyfish Technology of Texas.  Earlier this year, Gellyfish sued Otter and 36 other defendants for infringement of Patent No. 6,847,310.  After being served with the Complaint, Otter attempted to identify the products accused of infringing the patent, finding that only two of the products named in the Complaint were ever sold by Otter, and those products (which have since been discontinued) had trival sales with potential damages estimated in the $800-$1200 range. 

When Otter's settlement offer was rejected, it filed a motion seeking to sever itself from the existing case and requesting a "reverse bifurcation" procedure in the new matter so that damages discovery and trial would precede liability discovery and trial.  A copy of that motion is available here:  Download Gellyfish v Alltel Motion.  The motion appears to have had some effect: shortly after it was filed, Otter withdrew the motion citing additional discussions between the parties that suggested an amicable resolution.

While no order issued in response to Otter's motion, Otter's approach makes considerable sense.  From a judicial and party efficiency standpoint, forcing patent holders to identify the accused products early on during litigation may greatly simplify discovery and perhaps eliminate the need for a merits determination, which itself can be extremely complex.  It may also help to discourage patent holders from bringing suits in instances where the potential damages are minimal compared to litigation costs.  And forcing patent holders to think about a realistic universe of damages (as opposed to pipe-dreams of every product sold by the accused infringer) may encourage earlier settlement – as perhaps occurred here.

Note: Thanks to Robert Merges for pointing me to the Otter case and the EDTexweblog.com for background on the case

22 thoughts on “Reverse Bifurcation in Practice

  1. 22

    Ben,

    You are making an assumption and a value judgment.

    The true comparison would be the opportunity cost of the time for those incourt that otherwise would be generating revenue for the company. You merely assume no adverse affect for an NPE employee to be away from his other duties.

    There is no blanket generalizations that can be made concerning the profit ratios (return per unit cost) for those employees of NPEs and PEs, no matter the role of the employee.

    Remember – the patent system does not guarantee a profit – it does not even give you the right to make anything – all it does is give you the right to stop others.

  2. 21

    the litigation costs for the patentee would be roughly equivalent to the accused infringer

    Would it? In the case of a patent troll, then the accuser would have no one in the company that would be adversely affected by showing up in court; nor would the company. Comparatively, the defendant would be adversely affected as their management, engineers, and potentially other employees would have to show up in court or to depositions, etc thereby unable to work on helping the company innovate and create new products (or fix old ones).

    Sure the court costs would roughly be the same, but would the actual costs to be the same? Not likely.

  3. 20

    “Completely meritless, untenable cases get filed all of the time and should be dealt with swiftly and harshly”

    Can you explain, exactly, what a completely meritless patent case would be? If there is a valid patent at issue, it is presumptively NOT a meritless case. It might be a poor case, but then it should be easy to deal with “swiftly and harshly”

    Everyone’s always looking for an easy way out.

    “I also have experienced litigation misconduct a shocking numer of times”

    You mean personally? Which end were you on?

  4. 19

    You could pretend to be a disinterested party, start an annonymous blog smearing your opponents and calling them names like “troll” – oh yeah, that is soooo much more honorable than whatever the so-called “troll” might be doing…

  5. 18

    Since “damages discovery” is so completely intertwined with product-details-discovery for infringement determinations or speculations, plus disputed claim scope, plus sales information, that it can represent a major part of the discovery burden on defendants, I also have difficulty seeing how this is going to solve troll settlement-shakedown problems?
    [But requiring more information in advance from patent owners as to which of a defendants numerous products they think are infringed might help in some cases, assuming the patent owner actually did any serious infringement investigations before suit.]
    Would not more, faster and better summary judgment decisions [BEFORE large defense costs force lucrative settlements] be a better solution? Could not prompt consideration of an S.J. motion for non-infringement also in effect cause an early Markman decision on a dispositive claim element?]
    Are not courts putting off consideration of any S.J. motions until nearly the trial date defeating the whole purpose of S.J.? Especially S.J. motions based on anticipatory prior art, bar dates, or the like?

  6. 16

    What we have here is a failure to communicate…. (love Cool Hand Luke).

    I do not believe that all entities have a “right to a trial.” Completely meritless, untenable cases get filed all of the time and should be dealt with swiftly and harshly. I also have experienced litigation misconduct a shocking numer of times.

    And it is up to us, the IP/Patent Bar, to find solutions…

  7. 15

    in most cases obvious litigation misconduct is not going to happen.

    LOLOLOLOLOLOLOLOLOL.

    Wake up.

  8. 14

    As evidence, the district court cited Eon-Net’s interrogatory response produced in another case involving the Patent Portfolio which snidely stated that “the skill in the art required is that sufficient to converse meaning-fully with Mitchell Media

    Disgusting.

  9. 13

    Agree Rachael, the troll has the right to trial and in most cases obvious litigation misconduct is not going to happen.

    The judge needs to be the one to stop this, that said many patent claims in many fields are grossly overbroad and can lead to great damages – if they were valid and this does nothing for all those.

  10. 12

    Reverse bifurcation will not work in all instances, of course. In this case the strategy was to highlight that there was no real money on the table, absent extortion. (The point re: the need for a Markman is well-taken though in many cases, but not all, counsel for both sides knows how that will likely play out.)

    The tactic is but one arrow in the quiver. We need to creatively, courageously, help judges to find ways to manage these cases and even “level the playing field” as Judge Davis says.

    Hopefully, The Merges/Crouch paper gives us more strategies, and in suspect it does (pulling it next).

    We cannot lament “trolls” or the current label for any person engaged in disfavored litigation conduct. We must outsmart them.

    (gets off soapbox)

  11. 11

    Wow, case right on point. A troll bites it in link to cafc.uscourts.gov

    EON-NET v. Flagstar Bankcorp.

    The troll demanded $25-50k settlements (in nearly 100 cases) after filing suit. Flagstor litigated through claim construction, expending $600k. The court opinined that the lawsuit/settlement strategy was in bad faith.

    There was also litigation misconduct. “As evidence, the district court cited Eon-Net’s interrogatory response produced in another case involving the Patent Portfolio which snidely stated that “the skill in the art required is that sufficient to converse meaning-fully with Mitchell Medina.” Medina was the client.

  12. 10

    me, if you take the patentee’s claim construction as his best case, concede it (risk), and then concede a very limited amount of damages, I don’t know why the case cannot be resolved on summary judgment. The patentee will demand discovery on damages, which will be granted. But if there is nothing to discover, case over.

  13. 9

    But how often can you get a client to do that? And particularly in biotech, where claim construction is often essentially a crapshoot.

    Maybe after each side puts in their proposed claim construction and definitions for disputed terms – but then you are in a seven figures in most cases already.

    I am not saying this is a bad idea, but I think both sides really do have an idea of the actual likely damages – it is just that the troll does not care.

  14. 7

    I think that Otter, for argument’s sake, assumed a worst-case claim construction in its own damages calculation. See footnote 1, page 2 of the Otter Motion to Sever.

    This or course, may be contested, but Otter also uses the fact that Gelleyfish made no substantive reply as almost a de facto admission that the damages calculation was uncontested.

  15. 6

    Exactly, how can one determine damages prior to the Markman? I guess it could work in a simple case, but in any case where claim construction matters I cannot see how this could work.

    That said, we need to find something that works so we have fewer patent trolls, maybe a simpler Markman process where the schedule does not drag out for a year and have several experts involved…

  16. 4

    This issue is particularly problematic when the patent holder’s expected litigation costs are significantly lower than those of the accused infringer, such as in the instance of the prototypical “patent troll,”(whatever that is)

    Since, nowithstanding any damages, the litigation costs for the patentee would be roughly equivalent to the accused infringer, what you really mean is the patent holder’s expectation of settlement is high compared to the accused infringer’s expected costs of litigation.

    Naturally, the imbalance is precisely why the case is brought. However, since the patentee is the holder of a valuable right that has been presumptively infringed, comparison of relative costs of litigation is irrelevant and meaningless.

    “…because under such circumstances the settlement value is largely driven by the expected cost of litigation to the accused infringer as opposed to the actual damages from the infringement, resulting in a windfall for the patent holder

    Not really sure what the point is since the above calculus would apply in any circumstance. Besides, the article does not take into account the collateral value of a litigation, even when individual damages in a particular case are small. If the patentee is victorious, then the judgement can be used to expand the scope of enforcement, usually without the need for additional litigation.

  17. 2

    I think the approach does make sense, but a few complications:

    1. I take it that it won’t work if there is non-discontinued product that may be subject to an injunction.

    2. Requiring patentees to identify all infringing products in the complaint, without the benefit of discovery, is basically applying a Twombly-style requirement of specificity. This (a) has the problem that it runs against the grain of Form 18, which seems to allow the accused product to be described very generally, and (b) will elicit howls of protest from the people who have been complaining about Twombly (which I think is pretty much the entire plaintiff’s bar plus most civil procedure professors).

Comments are closed.