Case No. 3:10-cv-03561-WHA
Judge: Hon. William Alsup
Date Comp. Filed: October 27, 2010
II. ARGUMENT ...... 2
A. The Court should strike Dr. Cockburn’s entire reasonable royalty
analysis, because he admittedly has no information about the rights
Google would have received under the proposed Sun-Google
partnership, which is the basis of his calculation ...... 2
B. The Court should strike Dr. Cockburn’s upward adjustments to his
patent-damages calculation, because Dr, Cockburn admits it is a proxy
for Oracle’s lost profits and that it lacks a reasonable factual
foundation. ...... 6
C. The Court should strike Dr. Cockburn’s opinion regarding the value of
a lost copyright license with Google, because it ignores the governing
legal standard and has no factual basis. ...... 11
D. Dr. Cockburn’s opinion fails to separate out patent damages on a claim
by claim basis or copyright damages among the different categories of
allegedly infringed copyrighted material. ...... 14
E. Cockburn should not be permitted to prejudice the jury by presenting
data about licenses for noncomparable technologies or settlements of
noncomparable litigation. ...... 16
III. CONCLUSION ...... 17
TABLE OF AUTHORITIES
Cream Records, Inc. v. Jos. Schlitz Brewing Co.
754 F.2d 826 (9th Cir. 1985) ...... 13, 14
Daubert v. Merrell Dow Pharms., Inc.
509 U.S. 579 (1993) ...... 1, 9, 10, 16
Georgia-Pacific Corp. v. United States Plywood Corp.
318 F. Supp. 1116 (S.D.N.Y. 1970) ...... 7, 9
[ID Security Sys. Canada, ]Inc. v. Checkpoint Sys., Inc.
249 F. Supp. 2d 622 (E.D. Pa. 2003), amended, 268 F. Supp. 2d 448 ...... 11
Jarvis v. K2, Inc.
486 F.3d 526 (9th Cir. 2007) ...... 11, 12, 13
Lucent Techs., Inc. v. Gateway, Inc.
580 F.3d 1301 (Fed. Cir. 2009) ...... 16
Mackie v. Rieser
296 F.3d 909 (9th Cir. 2002) ...... 12, 13
Medtronic, Inc. v. Boston Scientific Corp.
Civ. No. 99-1035 RHK/FLN, 2002 WL 34447587 (D. Minn. Aug. 8,
2002) ...... 2, 5, 6
Oracle USA, Inc. v. SAP AG
No. C07-1658 PJH, 2011 WL 3862074 (N.D. Cal. Sept. 1, 2011) ...... 12, 13
Panduit Corp. v. Stahlin Bros. Fibre Works, Inc.
575 F.2d 1152 (6th Cir. 1978) ...... 7, 9
Polar Bear Prods., Inc. v. Timex Corp.
384 F.3d 700 (9th Cir. 2004) ...... 12, 13
ResQNet.com, Inc. v. Lansa, Inc.
594 F.3d 860 (Fed. Cir. 2010) ...... 16
Rite-Hite Corp. v. Kelley Co., Inc.
56 F.3d 1538 (Fed. Cir. 1995) ...... 7
Sid & Marty Krofft Television Prods., Inc. v. McDonald’s Corp.
562 F.2d 1157 (9th Cir. 1977) ...... 11
TAS Distrib. Co. v. Cummins Engine Co.
491 F.3d 625 (7th Cir. 2007) ...... 11
TK-7 Corp. v. Estate of Barbouti
993 F.2d 722 (10th Cir. 1993) ...... 10, 11
Trademark Research Corp. v. Maxwell Online, Inc.
995 F.2d 326 (2d Cir. 1993) ...... 11
Zenith Elecs. Corp. v. WH-TV Broad. Corp.
395 F.3d 416 (7th Cir. 2005) ...... 10
As the Court directed in its September 27, 2011 order, Google submits this supplemental
brief in support of its pending motion in limine to strike portions of the opinions that Oracle’s
damages expert Dr. Iain Cockburn seeks to offer, based on Dr. Cockburn’s October 10, 2011
reply reports and his testimony at his October 17, 2011 deposition.
The defects with Dr. Cockburn’s analysis are even clearer in light of his reply reports and
deposition. First, his reasonable royalty analysis, both as to Oracle’s patent and copyright
claims, depends on his opinion that the patents-in-suit are worth 30% and the copyrights worth
15% of the value of the proposed Sun-Google partnership in 2006. Dr. Cockburn concedes that
the partnership would have given Google many other items of great value—thousands of patents,
numerous copyrights including source code, the JAVA trademark, and other technical knowhow—
but he has never made any effort to value any other component of the partnership and
admits he couldn’t do so if he had to. Having no sense of the whole, he has no basis to opine on
the value of any of the parts. Second, Dr. Cockburn confirmed that his upward adjustment of his
baseline patent royalty tacks on Sun’s projected lost profits (which is legal error) and is based on
a single, speculative internal Sun projection (which is not a sufficient factual basis for any
opinion). Third, Dr. Cockburn conceded Oracle cannot seek a hypothetical copyright license,
because Sun never would have licensed Google to use Sun’s copyrighted material in the manner
Sun alleges it has done here—to develop a purportedly competing, incompatible version of Sun’s
Java platform. Fourth, Dr. Cockburn has ignored the Court’s directive to calculate separate
damage amounts for the various patent claims or copyrighted materials. That makes his opinion
useless to a jury that may find liability on some claims but not others. His revised analysis still is
generalized, cursory, and geared only to putting the largest possible number before the jury.
Oracle devoted most of its opposition to Google’s motion to erecting strawmen, working
to justify positions taken by Dr. Cockburn that, although they are wrong, were not the bases of
Google’s Daubert motion—such as Dr. Cockburn’s insistence on using Sun’s opening
negotiating bid of $100 million as his baseline, rather than Sun’s subsequent offer of $28 million.
But none of Oracle’s responses to Google’s actual challenges has any merit.
A. The Court should strike Dr. Cockburn’s entire reasonable royalty analysis, because
he admittedly has no information about the rights Google would have received
under the proposed Sun-Google partnership, which is the basis of his calculation.
The foundation of Dr. Cockburn’s reasonable royalty analysis—as to both the patents-insuit
and the asserted copyrights—is the 2006 negotiations between Sun and Google for a
technology partnership to “distribute a Linux Java mobile stack under an Open Source license.”
Declaration of Daniel Purcell in Support of Google’s Supplemental Brief (“Purcell Decl.”) Ex. A
(Cockburn Dep.) at 77:20-78:3. Dr. Cockburn wrongly uses Sun’s opening demand in February
2006 of about $100 million as his baseline, ignoring the parties’ real-world negotiations over the
following two months that, before negotiations broke down at the end of April, resulted in a final
offer from Sun of $28 million. After setting the $100 million baseline, he assigns 30% of that
total amount to the six patents-in-suit and a further 15% to the copyrights.
But for purposes of this motion, the problem is much more fundamental than whether the
right starting point is $100 million, $28 million, or something else. Dr. Cockburn admits Google
would have received many other things in exchange for that money beside the patents and
copyrights at issue here—including a full license to thousands of Sun patents, the right to use
numerous Sun copyrights (including Sun’s copyrighted source code to its Java ME platform), the
ability to stamp its products with Sun’s JAVA trademark, and access to Sun’s engineers and their
many years of expertise in developing Java-based technologies. But Dr. Cockburn admits he has
no specific information about anything Google would have received under the partnership other
than the patents and copyrights at issue, and has made no attempt to value any of those other
components. Because he has no sense of those other components or their values, he cannot
“identify a reliable factual basis” for calculating the percentage of the whole attributable to the
patents and copyrights. Medtronic, Inc. v. Boston Scientific Corp., Civ. No. 99-1035 RHK/FLN,
2002 WL 34447587, *12 (D. Minn. Aug. 8, 2002); see also id. at *11 (striking an expert’s
opinion that two patents-in-suit “accounted for virtually all of the $25 million” paid for a license
to an intellectual property portfolio, despite the fact that he never “tried to determine the value to
[the licensee] of any of the items in that portfolio other than the patents-in-suit”).
Dr. Cockburn admitted each of these methodological failures in his deposition. First, he
admitted that Google would have received, among other things, a broad intellectual property
license in exchange for his $100 million starting point:
Q. Part of that component would have been a license to a portfolio of Sun
Purcell Decl. Ex. A (Cockburn Dep.) at 78:4-13. But Dr. Cockburn conceded he knew nothing
about the Sun patents in the portfolio—how many there were, what they covered, or their
potential value either to Sun or Google.
A. That’s correct.
Q. And that intellectual property portfolio included patents, correct?
A. Yes. Again, I’d have to refresh my recollection as to the precise terms
in the contract, but I understand the payment under such circumstances to be a
payment for a bundle of intellectual property, including patents owned by Sun.
Q. Do you know how many Sun-owned patents were contained in that
bundle of intellectual property?
Id. at 78:14-23, 78:25-79:9.
A. I don’t know.
Q. Have you looked at any of the Sun-owned patents contained in that
bundle of intellectual property other than the patents asserted in this lawsuit?
Q. Do you know anything about the other Sun-owned patents in that
bundle, as far as what they covered, what functionality they covered? ...
A. I’m not aware of any list of—of such patents. I’ve seen some numbers
thrown around. But I have not looked specifically at any patents other than the
ones that are in suit in this matter.
Q. Have you made any attempt to calculate the value to Google in the
context of the—the April 2006 negotiations of the other Sun-owned patents not at
issue in this case?
A. I have not isolated the value of—of those other patents, no.
Second, Dr. Cockburn’s answer was the same regarding the Sun copyrighted material that
Google would have been permitted to use under the contemplated partnership. He understood
Google would have received a broad copyright license, including to source code, but hadn’t
made any efforts to understand the scope of the copyrighted material or calculate its value.
Q. … With respect to the bundle of intellectual property rights that was
part of the subject of the negotiation between Sun and Google, did that bundle
include the rights to copyrighted Sun material?
Purcell Decl. Ex. A (Cockburn Dep.) at 79:14-80:13, 80:15. Dr. Cockburn specifically conceded
that access to Sun’s proprietary, copyrighted source code may have had substantial economic
value to Google. Id. at 80:16-18, 80:20-22.
A. It’s my understanding that it does, yes.
Q. And do you have a sense of the scope of the Sun copyrighted material
at issue with respect to that bundle?
A. Amongst other things, I think it includes copyrighted source code. I
understand that Sun claims a copyright in the APIs. You know, there may be—
there may be other copyrighted material encapsulated by the—the terms of the
Q. Have you attempted to value the other copyrighted material included
within the bundle that is not asserted in this case.
A. No. My analysis supports apportioning the value of—of the entire
bundle specifically into—into two components—one of which is—is that
associated with the patents-in-suit. Another reflects the value of the copyrights,
but I have not specifically isolated the value of the other intellectual property.
Q. Are you sufficiently familiar with the content of that other intellectual
property that you could estimate its value? …
A. Not as I sit here today.
Third, Dr. Cockburn likewise knew that a contemplated partnership with Sun would have
given Google rights to use Sun trademarks, including the JAVA mark—but again, he made no
effort to understand the scope of the rights on the table or calculate the value of those rights.
Q. The bundle of intellectual property rights that was part of the
negotiation between Sun and Google, in addition to patents and copyrights, it also
included the rights to use Sun’s Java trademark, correct?
Purcell Decl. Ex. A (Cockburn Dep.) at 81:21-82:5.
Q. Have you attempted to calculate the value of the use of Sun’s Java
trademark with respect to the overall bundle of rights?
Fourth, Dr. Cockburn admitted that his baseline—the never-consummated deal between
Sun and Google—would have delivered additional value to Google beyond just an intellectual
property license. Yet once again, he admitted that he did not fully understand what else Google
would have gotten, and had not tried to value any non-license benefits or otherwise taken those
benefits into account in his analysis.
Q. The deal between Google and Sun under negotiating—negotiation in
April of 2006 would have provided Google with additional value beyond just a
license to a bundle of intellectual property, correct?
Purcell Decl. Ex. A (Cockburn Dep.) at 84:7-85:5.
A. Well, the value to Google is, as you have suggested, at least potentially
made up of a number of—of factors, one is potentially the right to—to access
certain source code or deploy certain Sun technology, the ability to access the base
of developers writing Java code. To the extent that they saw some value in the
trademark, they might have been—viewed that as part of the considerations under
I think there is likely to be some value to them in terms of being able to
capitalize upon Sun’s expertise and—and many years of investment in developing
Java and fine-tuning that technology to work in the mobile environment. There
may well be other benefits that they would receive in exchange for—for the
payments associated with this agreement.
Q. In your analysis, do you attempt to value the other benefits that Google
would have received apart from the intellectual property license?
A. Independently of—of the negotiations around this agreement, no.
The bottom line is Dr. Cockburn simply has no idea what is included in his $100 million
starting point. Despite knowing that his starting point included myriad items of value, he made
no attempt to identify or account for any of those items. Having no knowledge of the whole, he
lacks any logical or legal basis for calculating the value of any of the parts.
Oracle undoubtedly will protest that, unlike the expert in Medtronic, whose valuation of
the patents in that case was based on “a gut feeling,” Medtronic, 2002 WL 34447587, *7-*8, Dr.
Cockburn purports to support his 30% valuation of the patents-in-suit and 15% apportionment of
the copyrights with three performance-based studies. Google explained in detail in its trial brief
why each of those studies is rigged and unreliable. Google Trial Br. [Dkt. No. 534] at 18-19.1
1 Briefly, Oracle’s employee-conducted, made-for-litigation benchmarking studies disabled far
more than just the allegedly infringing functionality, leading to massive overstatement of the
importance of that functionality to the performance of the Android software. Cockburn’s own
econometric study of second-hand eBay smartphone purchases evaluated an entirely different
product market and ignores basic econometric concepts by failing to control for variable bias.
And Oracle’s commissioned consumer-preference study tested a cherry-picked list of smartphone features without first bothering to establish those features’ importance to consumers,
or whether there were other, more important features that also should be included—then simply
assumed a one-to-one relationship between the relative importance of the tested features to one
another and Android’s market share. Google will show at trial that none of this is science.
But for purposes of this motion, and the logic of Medtronic, the studies make no difference. The
studies purport to measure only the relevance of the allegedly patented and copyrighted features
to device performance and consumer preferences. They have nothing to say about any of the
other benefits that would have been conferred by the other aspects of a Sun-Google partnership.
They do not address the value of the other Sun patents, Sun’s proprietary and copyrighted source
code, the Java trademark, or access to Sun engineers and their Java ME experience. They are not
logical substitutes for the critical questions that Dr. Cockburn has failed even to ask.
With no means for calculating the percentage of his $100 million starting point that ought
to be attributed to the patents and copyrights, Dr. Cockburn’s entire reasonable royalty analysis,
as to both the patents and copyrights, falls apart. Accordingly, this Court should exclude Dr.
Cockburn’s opinion as to the proper amount of a reasonable patent or copyright royalty.
B. The Court should strike Dr. Cockburn’s upward adjustments to his patent-damages
calculation, because Dr, Cockburn admits it is a proxy for Oracle’s lost profits and
that it lacks a reasonable factual foundation.
This Court should strike Dr. Cockburn’s upward adjustment of Oracle’s purported patent
infringement royalty from $29.6 million (30% of his $100 million starting point) to $176
for two reasons. First, it violated basic legal rules governing patent damages by
smuggling Sun’s purported lost profits into what is concededly a reasonable royalty calculation.
Second, it lacked any reasonable factual foundation, because it was based on a single,
speculative, internal Sun projection of revenue for a business line that did not exist. Dr.
Cockburn did nothing to verify the foundation of the Sun projection, to the extent it had any. At
his deposition, Dr. Cockburn admitted all of this.
First, Dr. Cockburn admitted the sole source of his nearly six-fold upward adjustment of
patent damages is profits Sun allegedly lost the chance to earn because Google deployed the
Dr. Cockburn’s original report asserted an upward adjustment to $201.8 million, Oracle Opp’n
at 3 (citing Cockburn Report ¶ 47), but in his reply to the report of Google expert Dr. Gregory
Leonard, he has revised the amount of patent damages downward to $176 million. Purcell
Decl.”) Ex. B (Cockburn Reply to Leonard) at Ex. 6.
Android platform. This is improper as a matter of law. There is a well-understood standard for
seeking lost profits on a patent infringement claim, first adopted by the Sixth Circuit in Panduit
Corp. v. Stahlin Bros. Fibre Works, Inc., 575 F.2d 1152 (6th Cir. 1978), and adopted by the
Federal Circuit thereafter. See, e.g., Rite-Hite Corp. v. Kelley Co., Inc., 56 F.3d 1538, 1545 (Fed.
Cir. 1995) (citing Panduit, 575 F.2d at 1156) (noting that the Federal Circuit follows Panduit,
which “requires that a patentee establish: (1) demand for the patented product; (2) absence of
acceptable non-infringing substitutes; (3) manufacturing and marketing capability to exploit the
demand; and (4) the amount of the profit it would have made.”). Oracle hasn’t pleaded a lost-profits
claim here. Even if it had, Dr. Cockburn makes no attempt to address, much less satisfy,
the four-part Panduit test. He admittedly calculates only a reasonable royalty. Although a
patentee’s likely lost profits is one factor that can be used to justify a reasonable royalty
calculation, see Georgia-Pacific Corp. v. United States Plywood Corp., 318 F. Supp. 1116, 1120
(S.D.N.Y. 1970), Oracle may not simply boost the baseline royalty upward by the amount of its
purported lost profits. Permitting that would make the Panduit requirements meaningless.
There can be no dispute that Dr. Cockburn increased his estimated reasonable royalty by
the amount of Sun’s purported lost profits. At deposition, Dr. Cockburn testified that, in
adjusting his patent royalty upward from $29.6 million to $176 million, he relied on only three
factors: “the litigation premium, fragmentation, and … the value to Sun of capturing
compatibility and control.” Purcell Decl. Ex. A (Cockburn Dep.) at 136:4-12. Dr. Cockburn
admitted at deposition that he was unable to quantify the effect of either the litigation premium
or fragmentation.3 Id. at 136:13-16, 137:20-138:16.
The vagueness of Dr. Cockburn’s opinions regarding “fragmentation” deserves close attention,
particularly given Oracle’s stated intention to use “fragmentation” to seek extraordinary
injunctive relief effectively turning over the entire Android platform to Oracle. When asked how
one could know whether fragmentation had caused economic harm to Oracle, Dr. Cockburn gave
no indication that Oracle had any way of proving such harm at trial, other than through its usual
alarmist rhetoric. He opined that “we might in 20 years’ time be able to conduct a retrospective
study—I imagine it would make a very good Ph.D. thesis—to go back and look at what—you
know, what will have happened between 2008 and at some point in the future.” Purcell Decl.
Ex. A (Cockburn Dep.) at 139:5-18 (emphasis added). The only actual data Dr. Cockburn relied
on to support his claim that fragmentation had harmed Oracle had nothing to do with Oracle
financials—it was a calculation showing that Java developer message boards have recently been
less active, in terms of new threads started, than Android developer message boards. Id. at 143:6-16 (admitting that he did not “rely on any other piece of data specifically” other than the
figures regarding activity on Java and Android developer message boards). Finally, when
directly asked to identify specific evidence of harm to Oracle as a result of fragmentation, Dr.
Cockburn had nothing to offer beyond Oracle’s own ipse dixit.
Q. How can you be sure that there even is any harm from fragmentation?
Id. at 156:18-157-10.
A. Listen to the participants in this case. Oracle are very concerned, and I
think have a very reasonable basis for being concerned, about the impact of Java
on their business. They’re seeing it in their customers jumping ship. There’s—
you know, the record supports, you know, a substantial and ongoing and likely
accelerating impact of Android on Oracle’s Java business.
Q. Is there anything in particular you can point me to? I mean, I’m
familiar with Oracle’s assertions in the case, but is there anything in particular you
can point me to as evidence of harm to Oracle currently from fragmentation?
A. Well, I cite a number of—of—I don’t recall the specific references
here. I think the—the I wish I can add to my previous answers to your questions
in this line.
Accordingly, Dr. Cockburn’s “only basis for quantifying the upward adjustment of patent
royalties [was] the third upward adjustment—the value to Sun of compatibility and control.” Id.
at 176:12-16. But his method for calculating “the value to Sun of compatibility and control” was
simply to add the amount Sun projected it would earn in profits from a partnership with Google.
Dr. Cockburn admitted that the only basis for the bulk of his upward adjustment was a single
internal Sun presentation projecting Sun revenue from “Project Armstrong,” Sun’s internal code
name for its unconsummated partnership with Google. Id. Ex. C (Project Armstrong proposal).
Q. [B]ut looking at [Deposition Exhibit] 510, the presentation, is this
document the basis for your quantification of the upward adjustment for the value
of control and compatibility to Sun?
A. It’s the basis for that part of my adjustment, which, if you like, is a
lump sum. I also account for compatibility control as regards of lifting a cap on—
on the revenue sharing part of the agreement. This is the principal part, yes.4
4 In his original report, Dr. Cockburn explained that, of the total $175.6 million upward royalty
adjustment, $167.2 million (approximately 95.2% of the total adjustment) was based on Sun’s
lost profits, while a further $8.4 million (the remaining 4.8% of the adjustment) was based on the
removal of the revenue-sharing cap. Oracle Opp’n at 3 (table) (citing Cockburn Report ¶ 47).
In his reply report, Dr. Cockburn reduces the upward adjustment to $146.4 million. Purcell Decl.
Ex. B (Cockburn Reply to Leonard) at Ex. 6 (giving total patent damages figure of $176 million,
which is $146.4 million more than his $29.6 starting point). Dr. Cockburn no longer explains
what portion of his revised upward adjustment is attributable to lost profits, but applying the
same percentages as in his previous report, 95.2% of $146.4 million calculates to about $139.37
million in alleged lost profits tacked onto Dr. Cockburn’s current royalty calculation.
Q. Is there any other basis for that principal part of your upward
adjustment other than this document?
Purcell Decl. Ex. A (Cockburn Dep.) at 177:6-20.
A. No. Other than—other than my locating it in the general context of the
record and the other evidence I considered, this is the specific document.
The Project Armstrong presentation does not discuss, or place a value on, “compatibility
and control.” It projects Sun’s revenues from a partnership with Google. Purcell Decl. Ex. C
(Project Armstrong proposal). And Dr. Cockburn admitted in his initial report that the purported
“compatibility and control” “adjustment warrants an increase in the annual fee component to
reflect harm to Sun each year from lost sales it had otherwise anticipated to make.” Cockburn
Report ¶ 39 (emphasis added). That is lost profits, plain and simple. Cockburn confirmed at
deposition that he used the projected lost Project Armstrong profits “as a proxy” for “the value of
compatibility and control to Sun.” Purcell Decl. Ex. A (Cockburn Dep.) at 194:1-8. It does not
matter by what name Dr. Cockburn calls this adjustment; it is a straight importation of lost
profits. That is barred by Panduit.
Oracle offers (and could offer) nothing in its opposition to Google’s motion in limine to
rehabilitate Dr. Cockburn’s analysis. It claims Dr. Cockburn has just used Sun’s lost profits as
one of the Georgia-Pacific factors, but that is plainly wrong. Dr. Cockburn did not weigh the
projected lost profits against other factors and conclude that, because Sun might have made
profits from a partnership with Google, there should be upward pressure on the royalty; he
increased the royalty by the amount of purported lost profits. Although Oracle argues that Dr.
Cockburn also found other Georgia-Pacific factors would support an increase in his $29.6
million baseline, it cannot (and does not) argue that Dr. Cockburn used any of those factors as a
basis for quantifying his upward adjustment. The sole basis of that quantification remains Sun’s
alleged lost profits. That is disqualifying legal error under Daubert.
Second, even apart from that legal error, Cockburn’s patent royalty adjustment should be
stricken because he lacks a reasonable factual basis for it. As just discussed, his only basis for
quantifying at least $139.37 million of that adjustment is the Project Armstrong presentation.
That document, created in late February or early March 2006, offers a tentative projection of
revenues Sun might have made, had it partnered with Google. Not only are there no underlying
financial data backing up the financial projections in the presentation, the author of the
presentation, in her cover email, acknowledged that “not everything is spelled out” in the slides,
and moreover that “the numbers will move as the business model is more fully developed” and
“these numbers have not been vetted bottoms up (by customer).” Purcell Decl. Ex. D (Knopoff
email) at OAGOOGLE0100166873 (emphases added).
Cockburn made several critical concessions at his deposition confirming how speculative
this single presentation truly was. He conceded that Project Armstrong would have been a
completely new line of business for Sun, which had never before worked on a full stack
operating platform; its Java ME business was confined to the middleware layer of a full stack.
Purcell Decl. Ex. A (Cockburn Dep.) at 152:13-23; see also id. at 176:8-11 (“[F]or example, Java
ME runs on BlackBerry. There’s a big distinction between that and Android, which is a—what
we were describing earlier as a full stack.”). He conceded he did not discuss the Project
Armstrong projections with anyone at Oracle to assure himself they were reliable. Id. at 177:21-
25. He admitted he had never seen any other document confirming that the aggressive time-tomarket
assumptions in the Project Armstrong document were feasible. Id. at 185:7-13. And he
conceded that—even though Sun made clear its projections would “move as the business model
was more fully developed” and “have not been vetted”—he never saw any other document that
confirmed the projections in light of further due diligence. Id. at 187:23-188:19.
Daubert makes clear that, to be admissible, expert testimony must “connote[ ] more than
subjective belief or unsupported speculation.” Daubert v. Merrell Dow Pharms., Inc., 509 U.S.
579, 590 (1993). The single Project Armstrong projection essentially defines “subjective belief
and unsupported speculation.” Courts regularly exclude expert testimony based on unverified
revenue projections, particularly a litigant’s self-serving, internal projections. See, e.g., Zenith
Elecs. Corp. v. WH-TV Broad. Corp., 395 F.3d 416, 420 (7th Cir. 2005) (rejecting as basis for
damages claim a party’s “internal projections, which rest on its say-so rather than statistical
analysis” and “represent hopes rather than the results of scientific analysis”); TK-7 Corp. v.
Estate of Barbouti, 993 F.2d 722, 732-33 (10th Cir. 1993) (rejecting damages expert opinion that
“failed to demonstrate any basis for concluding that another individual’s opinion on a subjective
financial prediction was reliable”); ID Sec. Sys. Canada, Inc. v. Checkpoint Sys., Inc., 249 F.
Supp. 2d 622, 695 (E.D. Pa. 2003), amended, 268 F. Supp. 2d 448 (rejecting damages expert’s
reliance on projection by plaintiff, who had incentive to inflate predicted financial success). And
it is a basic principle of damages law that recovery of purported lost profits for new, unproven
lines of business—like the development of a full stack would have been for Sun—is disfavored.
See TAS Distrib. Co. v. Cummins Engine Co., 491 F.3d 625, 633 (7th Cir. 2007) (“as a general
rule, expected profits of a new commercial business are considered too uncertain, specific and
remote to permit recovery”); Trademark Research Corp. v. Maxwell Online, Inc., 995 F.2d 326,
332 (2d Cir. 1993) (where new business seeks lost profits, “a stricter standard is imposed for the
obvious reason that there does not exist a reasonable basis of experience upon which to estimate
lost profits with the requisite degree of reasonable certainty.”).
Even if it were legally permissible for Oracle to recover lost profits through a reasonable
royalty (and it is not), Sun’s revenue projections are not a reasonable basis for any expert
opinion. Because Dr. Cockburn relies on nothing else, his opinion is inadmissible.
C. The Court should strike Dr. Cockburn’s opinion regarding the value of a lost
copyright license with Google, because it ignores the governing legal standard and
has no factual basis.
As Google previously argued, this Court should strike Cockburn’s opinion regarding the
value of a fair-market copyright license between Sun and Google for two reasons: his opinion
ignores the legal standard for calculating the value of such a license and it has no basis in the
record. In his reply report, Cockburn revised downward his estimated copyright-license amount
from $102.6 million to “$89.6 million through the end of 2011,” or alternatively $61.7 million
through the end of September 2011. Purcell Decl. Ex. E (Cockburn Reply to Cox) ¶ 80.
First, as the Ninth Circuit has made clear, “where the infringer could have bargained with
the copyright owner to purchase the right to use the work,” a hypothetical license measurement
looks at “what a willing buyer would have been reasonably required to pay to a willing seller for
plaintiff's work.” Jarvis v. K2, Inc., 486 F.3d 526, 533 (9th Cir. 2007) (quoting Sid & Marty
Krofft Television Prods., Inc. v. McDonald’s Corp., 562 F.2d 1157, 1174 (9th Cir. 1977)). To
prove its entitlement to hypothetical-license damages, a plaintiff is “required to show that, but for
infringement, the parties would have agreed to license the use of the copyrighted works at
issue.” Oracle USA, Inc. v. SAP AG, No. C07-1658 PJH, 2011 WL 3862074, at *7 (N.D. Cal.
Sept. 1, 2011) (emphasis added) (citing Polar Bear Prods., Inc. v. Timex Corp., 384 F.3d 700,
711 (9th Cir. 2004)). Critically, in this case, where Google did in fact bargain with Sun over a
license, the hypothetical license for “the use of the copyrighted works at issue” means a license
for Google to use Oracle’s copyrighted material as it allegedly did, by incorporating that material
into a competing product in a manner that is incompatible with Java platform standards.
Oracle cannot satisfy this test here, because Dr. Cockburn cites no evidence (and no such
evidence exists) that Oracle ever would have granted a license to a competitor for an allegedly
incompatible implementation of Java. Determining the price of a hypothetical copyright license
is an “objective, not a subjective” analysis, and courts must reject “[e]xcessively speculative”
claims. See Jarvis, 486 F.3d at 534; Polar Bear, 384 F.3d at 709; Mackie v. Rieser, 296 F.3d
909, 917 (9th Cir. 2002). Where it is clear the parties “would never have agreed to a license” in
the real world, there is no basis for awarding actual damages based on the value of a hypothetical
license. Oracle, 2011 WL 3862074, at *7.
During his deposition, Dr. Cockburn admitted that, during their real-world negotiations in
2006, neither Sun nor Google ever raised the possibility of licensing an incompatible
implementation of Java, much less agreed to it. Dr. Cockburn also conceded, as he had to, that
he was unaware of any instance where Sun had ever licensed an incompatible implementation of
Java to a competitor, as Oracle is contending Google to be in this case.
Q. Do you have an opinion as to whether Sun ever would have granted a
license to Google for an incompatible version of Android? …
A. I can see circumstances under which they would have been willing
to—to license an incompatible implementation of Java. You know, if the price is
So I wouldn’t say never. I think they would have been reluctant to do so.
But it would depend very much on the terms that were offered.
Q. None of the actual negotiations between Google and Sun in 2006 were
for an incompatible implementation of Java, correct?
A. The negotiations, as I understand them, contemplated a compatible
Purcell Decl. Ex. A (Cockburn Dep.) at 237:24-238:1, 238:3-18, 240:2-5 & 7-12. That ends the
Court’s inquiry. Sun never would have granted the hypothetical license at issue here.
Q. Neither party ever put an incompatible implementation on the table in
those negotiations, correct?
Q. … My question is: Are you aware of any instances where Sun has ever
offered to license or licensed an incompatible implementation of Java to a
A. What do you mean by “competitor”?
Q. Somebody who’s competing with Sun for customers, as opposed to a
partner like DoCoMo [a wireless carrier].
A. I can’t think of one.
In its opposition, Oracle argues around this issue, pointing out that a subsequent opinion
in the Oracle v. SAP case clarified that it is not necessarily the case, as a matter of law, that a
plaintiff must present evidence of “actual licenses it entered into or would have entered into for
the infringed works, and/or actual ‘benchmark’ licenses entered into by any party for comparable
use of the infringed or comparable works.” Oracle Opp’n at 8-9 (citing Oracle USA, Inc. v. SAP
AG, No. C07-1658 PJH, at Dkt. 1088). This changes nothing. Oracle concededly has no
evidence of comparable licenses, but the broader point is that it has no evidence of any kind that
it ever would have granted Google the hypothetical license Cockburn envisions.
As the Oracle court also explained—and Oracle does not dispute here—the Ninth Circuit
“has never upheld a hypothetical license award” without “actual proof that the plaintiff would
have licensed the infringed work to the defendant or a third party for the specific use at issue and
proof that the infringement caused the loss of that opportunity.” Oracle, 2011 WL 3862074, at
*8 (citing Polar Bear, 384 F.3d at 704, 709; Jarvis, 486 F.3d at 528, 533-34; Mackie, 296 F.3d at
913, 917; Cream Records, Inc. v. Jos. Schlitz Brewing Co., 754 F.2d 826, 827-28 (9th Cir.
1985)). Neither has the Ninth Circuit ever affirmed hypothetical-license damages in a case
involving competitors (who do not commonly license each other), as opposed to partners. See id.
at *8 & n.2. There is no reason for this Court to do what the Ninth Circuit has never done.
Oracle also characterizes Google’s argument as “baffling” because “Sun did in fact offer
Google a license.” Oracle Opp’n at 9 (emphasis in original). Oracle misunderstands Google’s
argument. Oracle must concede that Sun offered to license Google a platform that would have
been fully compatible with Sun’s Java standards and jointly controlled by Sun. That is 180
degrees from the hypothetical license here—an allegedly incompatible Java implementation that
is fully controlled by Google. Sun never negotiated for that, and never would have done so.
Second, Dr. Cockburn’s calculation of the hypothetical copyright license is inadmissible
because it has no factual foundation. Like Dr. Cockburn’s patent royalty adjustment, it relies
only on the same speculative Sun projection of Project Armstrong revenue. Purcell Decl. Ex. C
(Project Armstrong projection).5 Dr. Cockburn cannot reasonably base any upward adjustment
on the Project Armstrong projection, for the reasons discussed above.
D. Dr. Cockburn’s opinion fails to separate out patent damages on a claim by claim
basis or copyright damages among the different categories of allegedly infringed
Oracle concedes that Dr. Cockburn failed to separate out his patent damages calculation
on a claim-by-claim basis, Oracle Opp’n at 16, as this Court’s July 22, 2011 Order directed him
to do. July 22, 2011 Order [Dkt. No. 230] at 7. Oracle just argues that this oversight doesn’t
matter. Oracle asserts in its brief that it has limited its trial case to certain asserted claims that
“are basically representative of the claimed invention, and vary only as to their type.” Of course,
it cites nothing from Dr. Cockburn, any of its technical experts, or even any Oracle employees, to
support that statement. Oracle has never explained through its experts exactly how the claims it
plans to try fully represent the patents-in-suit or cover the valuable features disclosed by those
Dr. Cockburn begins his calculation of the copyright license, as he did his hypothetical patent
license, with a $100 million starting point. But after that his approach to copyright is different.
In his patent calculation he had first attributed 30% of that total to the patents-in-suit before
adjusting upward based on the Project Armstrong projection. In his copyright calculation, he
does the reverse—first adjusting upward based on the Project Armstrong projection, then
attributing 15% of the adjusted total to the copyrights. Purcell Decl. Ex. A (Cockburn Dep.) at
246:22-247:17. He never explains why he calculated his copyright license using the same inputs
as, but a different order of operation from, the patent license.
patents. Dr. Cockburn has never considered the possibility that some of the asserted claims may
be less valuable, or easier to design around, than other claims contained within that same patent.
Indeed, neither he nor Oracle appears to have considered whether this might make a difference to
the value of the hypothetical patent license.
Accordingly, because Cockburn has refused to apportion patent value among the asserted
claims in his report, the Court should also preclude him from doing so at trial. Google renews its
request that the Court instruct the jury that, because Oracle has no evidence of the value of
individual patent claims, if the jury finds any asserted claim of any patent invalid or not
infringed, it may assume that the rejected claim represented the full value of that patent, and thus
that Oracle is entitled to no damages for infringement of other claims of that patent.
Similarly, the allegedly copyrighted material at issue here includes not only the “structure
and arrangement” of 37 API packages (to the extent that merits copyright protection), but also a
handful of lines of Android code that Oracle contends were literally copied. Oracle asserts that
the APIs are critically important to Android, and Dr. Cockburn’s copyright damages discussion
focuses almost entirely on the APIs. For example, in his reply to Google’s copyright damages
expert, Dr. Cockburn mentions the literally copied code only in one paragraph, Purcell Decl. Ex.
E (Cockburn Reply to Cox) ¶ 21, compared to dozens of mentions of the APIs. Dr. Cockburn
asserts that this doesn’t matter, because “damages must be assessed based on the copyright
infringement as a whole.” Id. But it is at least possible, if not probable, that portions of Oracle’s
copyright claim will be eliminated before trial, or that the jury will find that Google infringed
some copyrights but not others. If the Court concludes that the “structure and arrangement” of
the API packages is not copyrightable, or the jury concludes Google engaged in fair use of those
packages, Oracle’s copyright claim will be limited to a dozen code files. None of that code was
either inventive or original, most of it was never implemented in Android, and all of it either has
been removed from Android or will be in the next release of the Android software. If the alleged
literal copying were the sole basis of copyright liability, there would be nothing in Dr.
Cockburn’s opinion to assist the jury in calculating copyright damages. The Court’s July 22,
2011 order directed Dr. Cockburn to be specific about how much damages flow from the various
elements of Oracle’s case, but Dr. Cockburn has never made any attempt to draw any distinction
for damages purposes among the different categories of copyrighted material at issue here.
E. Cockburn should not be permitted to prejudice the jury by presenting data about
licenses for noncomparable technologies or settlements of noncomparable litigation.
Finally, Oracle fails to address Google’s argument explaining why Dr. Cockburn should
be barred from referring to various licenses involving unrelated technology and parties—most
obviously, a semiconductor license negotiated between Nokia and Qualcomm. That license
involves two companies who are not parties and an entirely different technology than the one at
issue here. Oracle does not dispute that the Federal Circuit generally condemns expert testimony
about “licenses to technologies other than the patent in suit” as part of a reasonable royalty
analysis. ResQNet.com, Inc. v. Lansa, Inc., 594 F.3d 860, 869 (Fed. Cir. 2010) (emphasis in
original); see also Lucent Techs., Inc. v. Gateway, Inc., 580 F.3d 1301, 1329 (Fed. Cir. 2009).
Oracle doesn’t engage with this argument at all. The Court should bar Dr. Cockburn from
testifying about licenses for technologies other than those allegedly embodied by the patents-insuit,
or licenses involving neither Sun nor Google.
Instead, Oracle devotes most of its section on other licenses to a screed about the evils of
so-called fragmentation and the Sun-Microsoft settlement. This argument is premised on the
same error the Court condemned in its Daubert order: it asserts that both this action and the
Microsoft case involved “the same technology (Java).” But this case is not about “Java.” It is
about specific, identified functionality in Google’s Dalvik virtual machine that allegedly
infringes Sun patents, and specific API packages and lines of code in Android that allegedly
infringe Sun copyrights. In any event, Oracle doesn’t dispute that the Sun-Microsoft case was
about more than “fragmentation,” or deny that Cockburn has failed to separate out the values of
the various components of Sun’s settlement with Microsoft. Again, this blasé and broad-brush
approach shows that Dr. Cockburn and Oracle are less interested in presenting relevant evidence
to the jury than in causing prejudice to Google by flashing a nine-figure number. The Court
should exclude any reference at trial to the Sun-Microsoft settlement.
This Court should exclude Dr. Cockburn’s testimony for all the reasons presented above.
Dated: October 20, 2011
KEKER & VAN NEST LLP
By: s/ Robert A. Van Nest
ROBERT A. VAN NEST
Attorneys for Defendant