ROGERS, Circuit Judge:

The Federal Aviation Administration ("FAA") awarded a contract to Multimax, Inc. ("Multimax"), but reversed its decision after an unsuccessful bidder, Informatica of America, Inc. ("Informatica"), protested the award to FAA's Office of Dispute Resolution for Acquisition ("ODRA"). ODRA found the contract award to Multimax to be "without a rational basis and contrary to the [required procurement] evaluation criteria." The FAA Administrator issued a final order adopting ODRA's findings and recommendations to terminate Multimax's contract and hire Informatica to complete the contract. On appeal, Multimax contends that ODRA applied the wrong standard of review, that its key findings of fact are unsupported by substantial evidence, and that its disregard of evidence was arbitrary and capricious, resulting in an abuse of discretion and a decision contrary to law. We deny the petition for review.

I.

On May 11, 1999, the FAA announced a solicitation for bids on a contract to provide information technology support services at the William J. Hughes Technical Center (the "Center"), a New Jersey facility for research, development, and testing of aviation programs. The solicitation informed prospective bidders that their proposals would be subjected to a "best value" evaluation — a type of bidding process in which "combined technical criteria are more important than price." Thus, the solicitation stated that "the successful offeror may not necessarily be the [one providing the] lowest priced offer," but also advised that "[p]rice may become more important as the difference between competing technical scores decrease[s]."

At the Center, a five-member Technical Evaluation Team (the "TET"), chaired by Dennis Steelman, evaluated the contract proposals. After the TET eliminated noncompetitive proposals, four offerors remained, including Informatica and Multimax. The TET considered and ranked the proposals according to five factors: (1) Management Plan/Technical Approach; (2) Key Personnel; (3) Oral Presentation; (4) Staffing Plan; and (5) Past Performance. The offerors were also required to submit a price proposal, separately from their technical proposal, about which the TET was not supposed to learn or take into account until its technical report was completed. By late August 1999, the TET had reached a consensus that Informatica had provided the "technically superior offer," and recommended that Informatica be awarded the contract. Informatica had scored about three and a half points higher than its closest competitor on TET's combined evaluation, and its proposal was the only one ranked "excellent" (the other three fell into the "good" range).

The TET reported to the Integrated Products Team (the "IPT"), which consisted of Dennis Steelman and Anne Marie Ternay. The IPT was responsible for completing a Technical Evaluation Report and an Award Recommendation and Determination. On Friday, August 27, 1999, Steelman submitted an initial version of the Technical Evaluation Report to the appropriate FAA contracting official, Michael Ward. This report recommended awarding the contract to Informatica, based on its clear technical superiority. The same day, Steelman learned price information for the four finalists. The following Tuesday, August 31, 1999, Ternay notified Steelman that the Factor 5 scoring had been completed incorrectly, because Informatica had been judged on the past performance of its subcontractor.

The next day, September 1, 1999, the TET was reconvened, and its members agreed that a Factor 5 scoring error had been made. Informatica was given the opportunity to submit additional past performance references. However, the TET found these references lacking, and Informatica's score on Factor Five was reduced considerably, from average to poor. After the Factor 5 rescoring, on September 3, 1999, the TET characterized Informatica as the "marginally technically superior offeror." Informatica's edge over its closest competitor had dropped from three and a half to two points, although its proposal was still the only one that had won an "excellent" rating.

However, the revised Technical Evaluation Report completed by the IPT after the Factor 5 rescoring did not reflect the TET's consensus that Informatica still possessed a measurable and meaningful technical edge over its competitors. Instead, this report stated that "the overall technical ratings of each proposal are too close to identify the superior offeror, that each is capable of providing the services called for . . . and for the purposes of this technical evaluation each is the equal of the other." The final Award Recommendation and Determination, prepared by Steelman, Ternay, and Ward, awarded the contract to Multimax. It stated that "there is insufficient demonstrated technical superiority on the part of Informatica to justify the additional expense of awarding the IT contract to that offeror. Over the life of the contract, Informatica would cost almost a million dollars more than Multimax. [$23,926,570.40 for [Informatica]; $22,960,852.00 for Multimax. . . .]" A contract with Multimax was executed on September 15, 1999.

Informatica filed a protest with ODRA on September 28, 1999, charging that the Center had departed from the evaluation criteria set forth in the Screening Information Request, and in subsequent materials stating that its proposals would have "looked very different" if it had understood price would be given a central role in the selection decision. Multimax intervened, and after receiving the parties' comments and responses, ODRA issued its findings and recommendations. ODRA's report suggested that the IPT allowed concerns about price to corrupt the evaluation process. ODRA determined that "the Center improperly `departed' from the Solicitation's evaluation criteria, abandoning the `best value' scheme and effectively making an award to the lowest price/technically acceptable offeror." Additionally, the ODRA report stated,

without a rational basis, [Steelman] reworded the [TET] report, systematically eradicating the previous conclusion that [Informatica's] proposal was technically superior to that of Multimax. The ODRA concludes that Mr. Steelman's actions, which were taken `with the result [of awarding the contract to Multimax] in mind,' leveled the technical field completely (at least on paper), allowing low price to determine the award decision.

ODRA bolstered its conclusions by referring to an email message that was sent from Ternay to Steelman on September 8, 1999, after the Factor 5 rescoring that narrowed Informatica's edge over its competitors. This message included a draft award recommendation that stated, in part, "given the solicitation's emphasis on technical superiority over price, the qualitative differences between the two companies fully justifies [sic] paying a 4% premium for Informatica over Multimax." Based on this email message, ODRA drew the inference that on September 8 Ternay believed that even though Informatica's margin of technical superiority over its competitors had shrunk, Informatica still provided the best value. However, draft documents that Steelman transmitted to Ternay, Ward, and other officials at the Center on September 8 characterized the offerors as equally capable and recommended awarding the contract to Multimax. ODRA interpreted the events of September 8 as evidence of a dramatic difference in viewpoint between Steelman and Ternay on that date. Thus, ODRA considered Steelman's drafts expressing a preference for Multimax to be a "total about-face" from the judgment of Ternay that was unduly influenced by price information.

In view of its findings, ODRA recommended that the contract with Multimax be terminated, and remaining parts of the contract be awarded to Informatica. ODRA's report stated that "it is clear that [Informatica] should have been awarded the contract, had the Solicitation's evaluation criteria been followed, and there is nothing that would render impracticable an [Informatica] takeover at this stage." The FAA Administrator adopted ODRA's findings and recommendations, and Multimax challenges the Administrator's order, requesting that it be vacated with instructions to terminate the contract with Informatica and to execute a new contract with Multimax.

II.

On appeal, Multimax challenges the FAA's Order in three respects. First, Multimax contends that ODRA impermissibly substituted its judgment for that of the Center's personnel by relying on inferences from an incomplete review of the evolving evaluation and source selection process, rather than considering the reasonableness of the Center's final evaluation and award decisions. Second, Multimax contends that ODRA erred as a matter of law in concluding that the Center's award decision was contrary to the "best value" criteria based on an inference drawn from the events of one day, September 8, 1999. Third, Multimax contends that ODRA's key findings are unsupported by substantial evidence. Multimax asserts that ODRA ignored Steelman's sworn testimony and documentary evidence available in discovery materials provided to ODRA to which that testimony referred, namely that the so-called critical "Price/Technical Trade-off" section in the September 8 version of the award recommendation document was created before the Factor 5 rescoring.

Our review is confined to determining whether the FAA's order adopting the ODRA's findings and recommendations is arbitrary or capricious or contrary to law. See 5 U.S.C. § 706(2)(A). Thus, as the court recently stated in J.A. Jones Management Services v. FAA, 225 F.3d 761 (D.C. Cir. 2000), "[u]nder this standard, [the court] `may reverse only if the agency's decision is not supported by substantial evidence, or the agency has made a clear error in judgment.'" Jones, 225 F.3d at 764 (quoting Kisser v. Cisneros, 14 F.3d 615, 619 (D.C. Cir. 1994)); see also 49 U.S.C.A. § 46110(c) (West 1997). The court's scrutiny is highly deferential because the court is reviewing an agency procurement decision that implicates the agency's technical expertise. "Where a procurement decision requires an agency to assess an offeror's qualifications to perform a contract, our review is `especially deferential.' See Iceland S.S. Co., Ltd. — Eimskip v. United States Dept. of the Army, 201 F.3d 451, 461 (D.C. Cir. 2000). [The court is] particularly reluctant to second-guess agency decisionmaking on these "`delicate questions.'" Elcon Enters., Inc. v. Washington Metro. Area Transit Auth., 977 F.2d 1472, 1479 (D.C. Cir. 1992) (quoting Delta Data Sys. Corp. v. Webster, 744 F.2d 197, 203 (D.C. Cir. 1984))." Jones, 225 F.3d at 765.

Thus, Multimax's contention that ODRA applied the wrong standard of review is misplaced. To the extent that ODRA was required to apply an arbitrary and capricious standard of review, ODRA met this requirement. Because ODRA determined that the IPT had impermissibly deviated from a "best value" evaluation scheme, it was not unreasonable for ODRA to recommend that Informatica's protest be sustained. Given our highly deferential scrutiny, we conclude, in the absence of evidence that ODRA's decision process "involved a clear and prejudicial violation of applicable statutes or regulations," that ODRA's findings and recommendations are consistent with arbitrary and capricious review. Elcon Enters., 977 F.2d at 1478 (quoting Kentron Hawaii, Ltd. v. Warner, 480 F.2d 1166, 1169 (D.C. Cir. 1973) (internal quotation marks omitted)); see also Kisser, 14 F.3d at 618.

Similarly, Multimax fails to demonstrate that ODRA erred by substituting its technical judgment for the Center's assessment of price and technical tradeoffs, instead of simply checking to see that the Center's final procurement decision was rational. ODRA's report focused on the integrity of the procurement process, and correctly scrutinized the actions of the Center's officials to determine whether they adhered to a "best value" evaluation. Furthermore, the regulations that govern ODRA afford it broad discretion to choose an appropriate remedy for a successful protest. See 14 C.F.R. § 17.21 (2000).

Multimax's contention that ODRA's findings are not supported by substantial evidence also fails. Substantial evidence is defined as "more than a `scintilla,' but less than a preponderance of the evidence." Evans Fin. Corp. v. Director, Office of Workers Compensation Programs, 161 F.3d 30, 34 (D.C. Cir. 1998) (citations and internal quotation marks omitted). Again, Jones provides useful guidance. "[T]he question [the court] face[s] is `not whether [petitioner's] view of the facts supports its version of what happened, but rather whether the [agency's] interpretation of the facts is reasonably defensible.' Harter Tomato Prods. Co. v. NLRB, 133 F.3d 934, 938 (D.C. Cir. 1998) (internal quotation marks omitted)." Jones, 225 F.3d at 765.

Multimax relies heavily on a document that is not in the administrative record to dispute ODRA's conclusions about the email messages that Ternay and Steelman sent on September 8, 1999. Multimax attempts to call our attention to this new evidence to prop up its contention that ODRA mistakenly inferred that on this date disagreement about the merits of the offerors' proposals existed between Ternay and Steelman, and mistakenly concluded that Steelman's September 8 drafts should be characterized as a dramatic "about-face." It would, however, be fundamentally inappropriate for the court to consider this new evidence. Although this evidence was made available to Multimax during discovery, it was not made part of the administrative record, and hence ODRA had no opportunity to rule on it. Multimax was a party to the ODRA adjudication, received during discovery the document that it now uses to dispute ODRA's findings, and is unable to provide any excuse for its failure to bring the document to ODRA's attention. "The burden of uncovering and pointing to the facts relevant to the case before the agency [belongs to] the parties most concerned in the matter . . . Our role is to review the agency's handling of the objections put before it, not to provide a forum for new arguments based upon different facts that the petitioner could have but did not bring out below." Sprint Communications Co., L.P. v. FCC, 76 F.3d 1221, 1227-28 (D.C. Cir. 1996).

In view of the record before it, ODRA's inference that the language in Ternay's September 8 email message supported the conclusion that the Technical Evaluation Report departed from the principles of "best value" proposal evaluation was reasonable. Furthermore, even if that particular inference about the events of September 8 could be considered dubious, ODRA's decision remains supported by substantial evidence. Even if there was no disagreement between Ternay and Steelman on September 8, there is still evidence to support the conclusion that the IPT deviated from the principles of "best value" assessment. Contrary to Multimax's contention, ODRA's findings are not solely based on its interpretation of the email messages sent by the IPT on September 8. Rather, the bulk of ODRA's findings and the core elements of its recommendations are supported by at least three other considerations.

First, ODRA noted the dichotomy between the consensus the TET reached, that Informatica's proposal was measurably superior to the other offerors' proposals, and the conclusion the IPT presented in the final Technical Evaluation Report, that each offeror was equally "capable." The use of the term "capable" is strongly indicative of a "lowest priced/technically acceptable" evaluation rather than a "best value" analysis. Second, ODRA pointed to Steelman's admission that he had drafted the final TET report, which was supposed to present an objective perspective on the technical merits of the proposals, with the particular result of awarding the contract to Multimax in mind. Third, ODRA was influenced by the way that Steelman revised the Technical Evaluation Report and the Award Recommendation and Determination after he learned about the prices of the proposals, making Informatica's strengths seem less significant and making Multimax's weaknesses appear less problematic. These three factors serve as substantial evidence to support ODRA's conclusion that the IPT abandoned the "best value" analysis to which it was required to adhere. Thus, any mistaken inference that ODRA may have made does not require reversal. See Jones, 225 F.3d at 764, citing 5 U.S.C. § 706.

Accordingly, we deny the petition for review.