Oracle Corporation et al v. SAP AG et al

Filing 885

Reply In Support re 769 MOTION No. 4: to Exclude Testimony of Defendants' Expert Donald Reifer filed by Oracle International Corporation, Oracle USA Inc., Siebel Systems, Inc.. (Alinder, Zachary) (Filed on 9/16/2010) Modified on 9/17/2010 (vlk, COURT STAFF).

Oracle Corporation et al v. SAP AG et al

Doc. 885

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

BINGHAM McCUTCHEN LLP DONN P. PICKETT (SBN 72257) GEOFFREY M. HOWARD (SBN 157468) HOLLY A. HOUSE (SBN 136045) ZACHARY J. ALINDER (SBN 209009) BREE HANN (SBN 215695) Three Embarcadero Center San Francisco, CA 94111-4067 Telephone: [phone] Facsimile: [phone] [e-mail] [e-mail] [e-mail] [e-mail] [e-mail] BOIES, SCHILLER & FLEXNER LLP DAVID BOIES (Admitted Pro Hac Vice) 333 Main Street Armonk, NY 10504 Telephone: [phone] Facsimile: [phone] [e-mail] STEVEN C. HOLTZMAN (SBN 144177) FRED NORTON (SBN 224725) 1999 Harrison St., Suite 900 Oakland, CA 94612 Telephone: [phone] Facsimile: [phone] [e-mail] [e-mail] DORIAN DALEY (SBN 129049) JENNIFER GLOSS (SBN 154227) 500 Oracle Parkway, M/S 5op7 Redwood City, CA 94070 Telephone: [phone] Facsimile: [phone] [e-mail] [e-mail] Attorneys for Plaintiffs Oracle USA, Inc., et al. UNITED STATES DISTRICT COURT NORTHERN DISTRICT OF CALIFORNIA OAKLAND DIVISION ORACLE USA, INC., et al., v. Plaintiffs, No. 07-CV-01658 PJH (EDL) REPLY MEMORANDUM IN SUPPORT OF MOTION NO. 4 TO EXCLUDE TESTIMONY OF DEFENDANTS' EXPERT DONALD REIFER Date: Time: Place: Judge: September 30, 2010 2:30 p.m. Courtroom 3. 3rd Floor Hon. Phyllis J. Hamilton
Case No. 07-CV-01658 PJH (EDL)

SAP AG, et al, Defendants.

REPLY MEMO RE MOT. TO EXCLUDE TESTIMONY OF DEFENDANTS' EXPERT DONALD REIFER
Dockets.Justia.com

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 III. I. II.

TABLE OF CONTENTS Page INTRODUCTION ............................................................................................................. 1 ARGUMENT ..................................................................................................................... 2 A. Oracle's Motion to Exclude Testimony Rebutting Pinto Is Not Moot .................. 2 B. Reifer's Opinions About Pinto's Code Counters Are Not Reliable, Because Reifer Never Looked at Them ................................................................. 3 1. Reifer's Own Delays and Choices Prevented Him From Doing A Proper Expert Analysis .............................................................................. 4 2. Reifer's Analysis of Tan's Counters Is Not a "Reliable" Analysis of Pinto's Code Counters ........................................................................... 6 3. Reifer's Data Is Not "Flawed;" It Is Just The Wrong Data ....................... 8 4. Reifer's Analysis Lacks Any Factual Basis ............................................... 8 5. Reifer Is Not Qualified To Rebut Pinto's Code Counts Based On An Analysis Of Tan's Code Counters........................................................ 8 6. Reifer's Opinions About Pinto's Code Counts Are Irrelevant, Because They Are Based On Tan's Code Counters, Not Pinto's .............. 9 CONCLUSION ................................................................................................................ 10

i

Case No. 07-CV-01658 PJH (EDL)

REPLY MEMO RE MOT. TO EXCLUDE TESTIMONY OF DEFENDANTS' EXPERT DONALD REIFER

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 CASES

TABLE OF AUTHORITIES Page

Andrews v. E.I. Du Pont De Nemours and Co., 447 F.3d 510 (7th Cir. 2006)................................................................................................ 8, 9 Fresenius Medical Care Holdings, Inc., v. Baxter Intern. Inc., No. C 03-01431, 2006 WL 1646113 (N.D. Cal. June 12, 2006) ............................................. 3 Hanson v. Alpine Valley Ski Area, Inc., 718 F.2d 1075 (Fed. Cir. 1983)................................................................................................ 3 Playtex Products, Inc. v. Georgia-Pacific Corp., 390 F.3d 158 (2d Cir. 2004)..................................................................................................... 8 Tyger Const. Co. Inc. v. Pensacola Const. Co., 29 F.3d 137 (4th Cir. 1994)...................................................................................................... 9 United States v. Chang, 207 F.3d 1169 (9th Cir. 2000).................................................................................................. 9 RULES Fed. R. Evid. 702 ....................................................................................................................... 2, 8

ii

Case No. 07-CV-01658 PJH (EDL)

REPLY MEMO RE MOT. TO EXCLUDE TESTIMONY OF DEFENDANTS' EXPERT DONALD REIFER

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

I.

INTRODUCTION SAP's expert, Donald Reifer, purports to rebut only one opinion of Oracle's expert, Paul

Pinto ­ his software development cost estimate. A central element of Reifer's attack on Pinto is Reifer's claim that Pinto over-estimated the size of the relevant software programs. The Court should not permit Reifer to offer this limited opinion, because rather than rebut Pinto, Reifer criticizes only the irrelevant and unreliable code counting programs that Reifer's own graduate student assistant developed. Pinto calculated what it would have cost SAP to develop, on its own, the copyrighted Oracle software that SAP simply took and used without authorization. In his report and deposition, Pinto presented his opinions regarding additional costs, risks and delays inherent in software development, and the impact those would have on the fair market value of any negotiated software license, including between Oracle and SAP. Reifer attempts to rebut only Pinto's cost estimate, and that rebuttal is premised on a fatal mis-step. Reifer tries to argue that Pinto overestimated the project's size and cost by using (allegedly) flawed computer programs, called "code counters," to establish the size of the Oracle code. Yet Reifer concedes that he never used, tested, or analyzed Pinto's code counters. Instead, Reifer ­ who has never built a code counter himself ­ asked a graduate student assistant, Tom Tan, to try to "replicate" Pinto's code counters by building new ones. Tan failed in that task. The code counters that he built are not the same as Pinto's, as Pinto demonstrated at his deposition. Reifer nonetheless confined his analysis to Tan's counters, and, finding them flawed, argued that since Tan's code counters are flawed, Pinto's must be as well. SAP does not dispute any of the above facts. Instead, they make the following six arguments: First, SAP contends that Pinto's opinions are moot. They are not. Pinto's opinions about the costs and risks associated with developing independent software, as well as the delays in obtaining a non-infringing alternative to an Oracle license, remain relevant and go to the heart of the Georgia Pacific analysis that the jury must consider. Second, SAP argues that even though Reifer and Tan did not use Pinto's actual counters, 1
Case No. 07-CV-01658 PJH (EDL)

REPLY MEMO RE MOT. TO EXCLUDE TESTIMONY OF DEFENDANTS' EXPERT DONALD REIFER

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

the counters Tan built on his own are "accurate replicas." In reality, as Pinto demonstrated at his deposition, Tan's counters yield much larger code counts than Pinto's actual counters. They are not the same. Moreover, there is no adequate excuse for Reifer's failure to analyze or use the counters that Pinto actually used. Third, SAP argues that "Reifer is qualified to develop replica counters." Dkt. 831 (Def's. Opp. to Mot. to Exclude Reifer) ("Opp.") at 6-7. He is not, and he did not. SAP concedes that Reifer has never personally built a code counter, arguing only that he "led teams" that did. In this case, Reifer did not build the counter, Tan did. The only evidence of Tan's ability to build counters is Reifer's claim that Tan has "helped" do so before, in some undefined capacity as a graduate student. Such cursory evidence of such scant experience does not make Tan an expert under Rule 702. Even if Tan did qualify, he was not disclosed by SAP. Fourth, SAP claims that "flaws in Reifer's data do not render his opinion inadmissible." This is misdirection masquerading as argument. Oracle does not argue that Reifer's data was "flawed"; Oracle established Reifer looked at the entirely wrong data. Fifth, SAP argues that the cases cited by Oracle, barring expert testimony by witnesses like Reifer who examined the wrong thing, are "inapposite." But SAP concedes that expert opinion must be based in fact, and Reifer's opinions concerning Pinto's code counts are not based in fact. They are based on a mistake of fact. Sixth, SAP argues that "Reifer's opinions are highly relevant to the jury's analysis." Reifer's only opinions relevant to this motion ­ opinions about Pinto's code counters ­ are irrelevant, because Reifer did not analyze Pinto's code counters and Tan's replicas do not match Pinto's counters. II. ARGUMENT A. Oracle's Motion to Exclude Testimony Rebutting Pinto Is Not Moot

SAP has moved separately to exclude Pinto's testimony concerning the amounts that SAP would have spent to independently develop the software it infringed, and its mootness argument in defense of Reifer is simply a shorthand repetition of that Daubert motion. See Dkt. 774 (Def's Mot. to Exclude Pinto) at 3:3-20. As Oracle explained in its opposition to SAP's motion to 2
Case No. 07-CV-01658 PJH (EDL)

REPLY MEMO RE MOT. TO EXCLUDE TESTIMONY OF DEFENDANTS' EXPERT DONALD REIFER

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

exclude Pinto, Pinto estimated the amount SAP would have spent to develop non-infringing alternative software products to be between $1.134 and $3.477 billion, depending on the labor source and associated costs. See Dkt. 843 (Oracle's Opp. to Mot. to Exclude Pinto) at 5:7-9; see also Alinder Decl., Ex. C (Pinto Report) at 43-44. Pinto further opined that such a development effort would be large, aggressive, risky and "exceedingly difficult" to complete within the two year window for a time sensitive market opportunity such as this one. See Dkt. 843 (Oracle's Opp. to Mot. to Exclude Pinto) at 5:9-12; see also Alinder Decl., Ex. C (Pinto Report) at 7. Indeed, the experts that SAP retained agree with Pinto that the delay and risk of developing software equivalent to Oracle's would be so great that no business would even attempt it. See id., Ex. D (Garmus Report) at 1, 26; Ex. A (Reifer Report) at 5, 56, 86-87. All of these opinions are relevant to the analyses and testimony of the principal damages experts, Paul Meyer and Stephen Clarke, concerning the outcome of a hypothetical license negotiation between Oracle and SAP. And the summary judgment order did not preclude opinions, like these, that go to the heart of the Georgia Pacific analysis. See Dkt. 843 (Oracle's Opp. to Mot. to Exclude Pinto) at 9:12-11:9; see also Fresenius Medical Care Holdings, Inc., v. Baxter Intern. Inc., 2006 WL 1646113, at *1 (N.D. Cal.) ("[A] key part of the reasonable royalty determination under Georgia Pacific is whether the accused infringer had acceptable non-infringing alternatives available to it at the time of the hypothetical negotiation."); Hanson v. Alpine Valley Ski Area, Inc., 718 F.2d 1075, 1080-81 (Fed. Cir. 1983) ("Reliance upon [infringer's] estimated cost savings from use of the infringing product is a well settled method of determining a reasonable royalty."). Pinto's testimony is not moot. B. Reifer's Opinions About Pinto's Code Counters Are Not Reliable, Because Reifer Never Looked at Them

Although SAP's brief repeatedly ­ and misleadingly ­ refers to Reifer's supposed analysis of "Pinto's counters,"1 it is undisputed that Reifer never analyzed, used, tested, or ran

Opp. at 4:22-23 (asserting that "Pinto's counters fail to comply with" the standards and conventions cited by Pinto); id. at 4:27:5:1 ("Reifer concluded that it was necessary to analyze the accuracy of the Pinto counters themselves"); id. at 5:2 ("By analyzing the counters that Pinto (Footnote Continued on Next Page.) 3
Case No. 07-CV-01658 PJH (EDL)

1

REPLY MEMO RE MOT. TO EXCLUDE TESTIMONY OF DEFENDANTS' EXPERT DONALD REIFER

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

Pinto's counters. Everything that Reifer did and said with respect to line counts is based on purported "replica" counters developed by his graduate-student assistant, Tom Tan, not Pinto's actual code counters. 1. Reifer's Own Delays and Choices Prevented Him From Doing A Proper Expert Analysis

Recognizing that it cannot ultimately avoid the fact that Reifer never used Pinto's counters, SAP tries to blame Pinto for Reifer's failure. Opp. at 5. SAP asserts that Reifer was not provided with "executables" of Pinto's counters until February 24, 2010, that Reifer could not get the executable code to work, and that Reifer's "requests for further information were not effectively addressed." Id. at 5:9-15. Even if Reifer had a good excuse for using the wrong code counters, it would not make his mistaken analyses reliable or relevant under Daubert. But Reifer has no excuse, and SAP's self-serving account omits the undisputed facts that show why: Reifer waited until the last minute to start work, obtained executable copies of Pinto's counters as soon as SAP asked Oracle for them, and never asked Oracle for any further information. Pinto timely provided his detailed report on November 16, 2009, along with the documents he relied upon. See Alinder Decl., Ex. C (Pinto Report) at 45, Appendix. B. Pinto's report makes clear that as the first step in sizing the code, he extracted the source code from the same underlying Oracle computer programs that had been produced to SAP on hard drives; for PeopleCode, he provided SAP with the actual extracted code. Id. at 12, 14 & n.9. Pinto's report further makes clear that he used specially-designed code counting utilities to count the lines of code. See id. at 15-17. Upon receiving Pinto's report, SAP's expert did nothing for months. Reifer says he (Footnote Continued from Previous Page.) used . . . ."); id. at 5:4-5 ("Reifer was able to use Pinto's counters . . . ."); id. at 6:1-2 ("prompted Reifer to continue his evaluation of Pinto's counters"); id. at 6:8-9 ("Because there appeared to be a significant anomaly in Pinto's counters . . . ."); id. at 7:15-17 (referring to use of irrelevant software program to test "the accuracy of Pinto's counters"); id. at 7:17 ("an initial experiment to compare Pinto's counters with the industry counters"). 4

Case No. 07-CV-01658 PJH (EDL)

REPLY MEMO RE MOT. TO EXCLUDE TESTIMONY OF DEFENDANTS' EXPERT DONALD REIFER

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

assumed his task would be a "simple job," and that he could perform all of the necessary analyses to understand and confirm Pinto's conclusions in "a week to two weeks." Id., Ex. B (Reifer Depo.) at 174:3-175:18. Only in February 2010, just weeks before his report was due, did Reifer begin to try to extract source code from the underlying programs in order to count it. Id. at 88:11-15. Only in mid-February did Reifer ask Tan to assist him. Id. at 137:6-12. Although SAP argues that "Reifer needed software files that he could execute," neither Reifer nor SAP made any request for executable versions of Pinto's counters until Sunday, February 21st. Id., Ex. E (Feb. 21, 2010 email from Jeff Butler to Geoff Howard). Oracle quickly responded with the requested executable files on the morning of Wednesday, February 24th, and included an exact description of the machine environment in which they ran. Id., Ex. F (Feb. 24, 2010 email from Amy Donnelly to Jeff Butler). SAP claims that Reifer could not get the files to work and that his "requests for further information were not effectively addressed." Opp. at 5. SAP's vague, passive-voice formulation conceals the fact that neither Reifer nor SAP ever asked Oracle or its counsel for "further information." Alinder Decl., ¶ 8. Whatever requests Reifer may have made, he made them to SAP's counsel, who chose to remain silent rather than assist Reifer in obtaining what he now says he needed. Id., Ex. B (Reifer Depo.) at 158:2-23.2 After Reifer chose to wait until the last few weeks to start work, and after SAP's counsel chose not to seek assistance with Pinto's code counters, Reifer "ran out of time" to do the job he had been designated to do. Id., Ex. B (Reifer Depo.) at 124:20-125:13 (Reifer testimony that he "ran out of time" without ever trying to analyze the PeopleCode that Pinto produced in November); id. at 136:16-137:12 (Reifer testimony that Tan started in mid-February and did not have time to build replicas of all of Pinto's counters); id. at 137:23-138:10 ("we just had a

Pinto appeared for a full day of deposition on May 20, 2010, and brought with him detailed notes explaining the operation of his counters compared to Tan's "replicas." See generally, Dkt. 770 (Alinder Decl.) Ex. E (Reifer Rebuttal Notes). In the entire day of deposition, SAP's counsel never asked any questions about how to make Pinto's counters operate. Alinder Decl., ¶ 8. 5

2

Case No. 07-CV-01658 PJH (EDL)

REPLY MEMO RE MOT. TO EXCLUDE TESTIMONY OF DEFENDANTS' EXPERT DONALD REIFER

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

month, didn't have enough time"); id. at 168:1-11 (Reifer testimony that he sampled no Java code because "[w]e ran out of time."). In short, if Reifer was unable to use Pinto's counters to analyze the source code, it was because Reifer chose not to start doing the work of a responsible expert until it was too late. When Reifer could not run the counters, defense counsel made a choice not to help him, but instead to blame Pinto for Reifer's failures. In any event, SAP's attempt at blame-shifting is misdirection. The issue before the Court now is whether Reifer's opinions about Pinto's code counts are relevant, reliable, and based in fact. They are not, and are therefore inadmissible. 2. Reifer's Analysis of Tan's Counters Is Not a "Reliable" Analysis of Pinto's Code Counters

SAP next argues that Reifer's analysis of Pinto's line counts is reliable. But Reifer did not provide any analysis of Pinto's code counters, and neither Reifer nor Tan would be qualified to offer one. First, SAP says that Reifer's graduate student assistant, Tan, correctly replicated Pinto's descriptions of his code counters. Opp. at 6:13-24. SAP offers no evidence to establish that Tan has the necessary expertise to develop such counters at all, much less in the specific programming languages relevant here. Alinder Decl., Ex. B (Reifer Depo.) at 138:11-139:1 (describing Tan as merely a Ph.D candidate who "is one of the people who help write the code counters"); id. at 323:19-324:9 (Reifer testimony that Tan has an undergraduate degree in an unknown field and no master's degree). It is clear that Tan's counters fail to replicate Pinto's, counters because Pinto ­ unlike Reifer ­ analyzed the data using both sets of code counters and found that in every case, Tan's counters consistently yielded significantly higher code counts than Pinto's. Dkt. 770 (Alinder Decl.) Ex. E (Reifer Rebuttal Notes) at 2-3. Tan's counters simply are not Pinto's counters.3

3

SAP never addressed Pinto's comparison of the two counters. Instead, they repeat Reifer's illogical argument that "the only way" that Tan's replicas could be defective would be if Pinto did not build his own counters as he described. Opp. at 8:15-18. But Pinto's comparison demonstrates that Tan did not accurately implement Pinto's parsing rules. See Dkt. 770 (Footnote Continued on Next Page.) 6
Case No. 07-CV-01658 PJH (EDL)

REPLY MEMO RE MOT. TO EXCLUDE TESTIMONY OF DEFENDANTS' EXPERT DONALD REIFER

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

Next, SAP argues that Reifer is qualified to develop replica counters. Opp. at 6:25-7:11; see also id. at 9:17-10:8. But Reifer did not develop the counters ­ Tan did. Alinder Decl., Ex. B (Reifer Depo.) at 130:18-24. As noted above, SAP has not offered any evidence that would qualify Reifer or Tan as an expert in building code counters.4 Indeed, the only experience that Reifer himself claims with respect to code counters is that he has "led teams that have developed them, but I have not personally written the code for those." Id. SAP offers nothing else in support of the reliability of Tan's counters or Reifer's purported analysis of Pinto's counters. Instead, they digress into a defense of Reifer's use of an open source flight simulator program to compare Tan's counters to other counters developed by USC. Opp. at 7:12-22. But Reifer himself refused to opine that that comparison had any relevance to this case. Alinder Decl., Ex. B (Reifer Depo.) at 150:13-24 ("So the experiment was not germane to anything in my report."); see also Dkt. 769 (Oracle's Mot. to Exclude Reifer) at 6, n.1. SAP apparently intends to "rebut" Pinto by offering an opinion about how Reifer's graduate student assistant's code counters performed on irrelevant code written for a different purpose in a different programming language in an experiment that Reifer himself said was "not germane." Id. That position alone demonstrates how far SAP has drifted from the principles of Daubert, and even simple relevance. None of these arguments establish that Tan's counters are the same as Pinto's, that they can be used to test the accuracy of Pinto's counters, or that they can be used to derive any reliable, relevant, or useful conclusions about the quality of Pinto's work. 3. Reifer's Data Is Not "Flawed;" It Is Just The Wrong Data

SAP next falls back to an argument that even if Tan's counters are "flawed," alleged flaws in the data affect only the weight of Reifer's opinion, not its admissibility. Opp. at 7:24-

(Footnote Continued from Previous Page.) (Alinder Decl.) Ex. E (Reifer Rebuttal Notes) at 2-3. Reifer apparently contends that it is a logical impossibility that he himself or Tan made a mistake. That contention demonstrates arrogance, not expertise. 4 Nor was Tan even disclosed as an expert at all. 7

Case No. 07-CV-01658 PJH (EDL)

REPLY MEMO RE MOT. TO EXCLUDE TESTIMONY OF DEFENDANTS' EXPERT DONALD REIFER

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

8:21. SAP misses the point. Oracle does not claim that Reifer made a methodological error in his analysis of Pinto's counters. Oracle argues that Reifer cannot offer opinions about Pinto's counters without ever analyzing them. Reifer does not have "flawed" data, he has the wrong data. See, e.g., Andrews v. E.I. Du Pont De Nemours and Co., 447 F.3d 510, 513 (7th Cir. 2006) (expert testimony excluded under Daubert because expert based his calculations on data from the wrong highway ramp); Playtex Products, Inc. v. Georgia-Pacific Corp., 390 F.3d 158, 167-68 (2d Cir. 2004) (in trademark infringement case, disregarding expert testimony where expert analyzed wrong mark), superseded by statute on other grounds. As a result, Reifer's analysis does not suffer from "weakness," Opp. at 8:6, it suffers from irrelevance. 4. Reifer's Analysis Lacks Any Factual Basis

In it opening brief, Oracle cited cases establishing that an expert's analysis should be excluded under Daubert and Rule 702 when the proffered opinions have no factual basis. Dkt. 769 (Oracle's Mot. to Exclude Reifer) at 8:24-9:6 (citing cases). SAP does not dispute this principle. Opp. at 9:1-5 (conceding that when expert relies on facts that are "directly contradicted" or bases an opinion on "clear factual error," opinions based on those errors are properly excluded). Reifer's factual error is no different. His entire critique of Pinto's code counts is based on an erroneous factual assumption that Tan's code counters are identical to Pinto's. Pinto's deposition established that the counters are not the same. See Dkt. 770 (Alinder Decl.) Ex. E (Reifer Rebuttal Notes), at 1-3. Reifer's convenient fiction is not a proper basis for expert testimony in federal court, and his testimony must be excluded. 5. Reifer Is Not Qualified To Rebut Pinto's Code Counts Based On An Analysis Of Tan's Code Counters

Oracle's Daubert motion seeks to exclude Reifer's opinions that Pinto overestimated the size of the software development project by using counters that overstated the number of lines of code. Dkt. 769 (Oracle's Mot. to Exclude Reifer) at 7:23-9:12. At pages 9-10 of their Opposition, SAP argues that "Reifer is qualified to rebut Pinto's opinion" concerning code counters. Opp. at 9-10. In support, they cite his years "in the field of software," his development experience, and his experience with COCOMO. Id. But the only experience that 8
Case No. 07-CV-01658 PJH (EDL)

REPLY MEMO RE MOT. TO EXCLUDE TESTIMONY OF DEFENDANTS' EXPERT DONALD REIFER

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

they cite related to code counters is Reifer's statement at deposition that he has "led teams that have developed them, but I have not personally written the code for those." Id. at 9:23; see also Alinder Decl., Ex. B (Reifer Depo.) at 130:18-24. SAP has the burden of establishing Reifer's relevant expertise. Generalized experience in the software industry, or with software estimation models, does not, by itself, establish expertise in building or testing code counting utilities. See United States v. Chang, 207 F.3d 1169, 1172-73 (9th Cir. 2000) (expert "qualified" in the history and purpose of certain certificates was excluded from testifying on the identification of counterfeit certificates). 6. Reifer's Opinions About Pinto's Code Counts Are Irrelevant, Because They Are Based On Tan's Code Counters, Not Pinto's

Lastly, SAP argues that Reifer's opinion is "highly relevant" because it attacks an important component of Pinto's analysis. Opp. at 10:9-25. This argument merely repeats the error that permeates Reifer's analysis and SAP's opposition brief. Tan's counters and Pinto's counters do not match. Attacking Tan's counters does not tell the jury how accurate Pinto's counters are at all. See Andrews v. E.I. Du Pont De Nemours and Co., 447 F.3d 510, 513 (7th Cir. 2006) (expert testimony excluded under Daubert because expert based his calculations on data from the wrong highway ramp); Tyger Const. Co. Inc. v. Pensacola Const. Co., 29 F.3d 137, 144 (4th Cir. 1994) ("When the assumptions made by an expert are not based on fact, the expert's testimony is likely to mislead a jury, and should be excluded by the district court.") It makes no difference that SAP claims that Tan faithfully applied Pinto's parsing rules for the code counters. Opp. at 10:13-16. Even if SAP were correct in asserting that Pinto's description of his parsing rules had errors in them ­ and SAP has failed to offer any competent evidence of such a claim ­ Pinto did not use parsing rules to count code. He used code counters. See Alinder Decl., Ex. C (Pinto Report) at 15-17. While SAP is free to try to cross-examine Pinto in an effort to show that his parsing rules do not comply with the standards and conventions he cites, they cannot offer an expert opinion that Pinto's code counters are flawed based on anything but Pinto's actual code counters. Because Reifer never analyzed, used or tested Pinto's actual code counters, he cannot offer such an opinion. 9
Case No. 07-CV-01658 PJH (EDL)

REPLY MEMO RE MOT. TO EXCLUDE TESTIMONY OF DEFENDANTS' EXPERT DONALD REIFER

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

III.

CONCLUSION For the reasons stated above, Plaintiffs ask the Court to order that neither Reifer, nor any

witness relying on Reifer, be permitted to offer any testimony or evidence that Pinto's counts of source lines of code for Oracle software applications are in any way incorrect.

DATED: September 16, 2010 Bingham McCutchen LLP

By:

/s/ Zachary J. Alinder Zachary J. Alinder Attorneys for Plaintiffs Oracle USA, Inc., et al.

10

Case No. 07-CV-01658 PJH (EDL)

REPLY MEMO RE MOT. TO EXCLUDE TESTIMONY OF DEFENDANTS' EXPERT DONALD REIFER