HAMILTON, Circuit Judge.
Plaintiffs Immunosciences Lab, Inc. and its owner, Dr. Aristo Vojdani, were in the business of developing and selling medical tests and testing materials. In 2007, Pharmsan Labs, Inc. and NeuroScience, Inc. — sister companies offering medical testing to consumers — wanted to expand their offerings. Vojdani and Immunosci-ences (we refer to them collectively as “Vojdani”) and Pharmsan and NeuroScience (collectively “NeuroScience”) decided to collaborate, but the business relationship fell apart within two years.
These appeals concern two trials and two claims for breach of contract brought by Vojdani against NeuroScience. In the first trial, a jury decided the first claim— that NeuroScience did not pay Vojdani what it had contracted to pay for medical testing materials — in favor of NeuroScience. But the district court ordered a new trial on that claim, concluding that this verdict was undermined by flawed special verdict questions. The jury in the second trial found for Vojdani but awarded him much less money than he was seeking. NeuroScience contends on appeal that the court’s grant of a new trial was an abuse of discretion. Vojdani cross-appeals, arguing that the court abused its discretion by allowing NeuroScience to argue in the new trial that the parties had orally modified their written contract. The second claim is that NéuroScience breached a separate confidentiality agreement by continuing to use Vojdani’s testing methods after the parties ended their business relationship. The jury in the first trial awarded Vojdani nearly $1.2 million on this claim, but the district court granted judgment as a matter of law for NeuroScience, explaining that Vojdani had relied on an impermissible damages theory. As part of his cross-appeal, Vojdani seeks reinstatement of the original verdict on this claim.
We affirm the district court’s judgment in all respects. As we explain in Part I, the court acted within its discretion in
granting the new trial on the first claim and including NeuroScience’s contract modification theory within the scope of the second trial. We explain in Part II that the award for breach of the confidentiality agreement was not based on a permissible measure of damages that had actually been presented to the jury, so the district court correctly entered judgment as a matter of law for NeuroScience on that claim.
I.
First Claim
— Breach
of the Letter of Intent
A.
Factual and Procedural Background
Vojdani and NeuroScience signed a “letter of intent” in June 2007, and despite that provisional title, both sides agree it was a binding contract. The letter of intent provided in part that Vojdani’s company would ship medical testing plates and components to NeuroScience accompanied by “an invoice for the material at 50% of client price.” The parties would “absorb their own costs of operation,” and NeuroScience would pay the invoices “according to [NeuroScience’s] monthly sales.” The agreement in the letter was set to expire after 180 days, during which time a longer-term agreement was to be hammered out.
After the parties signed the letter of intent, Vojdani regularly shipped testing plates and components to NeuroScience and submitted monthly invoices for 50 percent of the price that NeuroScience would charge customers for each test. Neuroscience interpreted its obligation to pay Vojdani “50% of client price” according to its “monthly sales” as an obligation to pay only for those tests it actually sold. Vojdani accepted these payments without complaint and never included a past-due amount on any invoice. That pattern continued for the 180 days covered by the letter of intent and during an oral extension of the agreement that lasted for several months. No written agreement ever replaced the letter of intent, though several draft agreements were exchanged. In June 2009, NeuroScience notified Vojdani that it would no longer do business with him. All shipments and payments then ceased.
Vojdani responded to NeuroScience’s decision to sever ties with him by suing in federal court for breach of contract. Federal jurisdiction was based on diversity of citizenship. (Vojdani and Immunosciences are citizens of California; Pharmsan and NeuroScience are citizens of Wisconsin.) Of the numerous claims that went to trial, the first of the two involved in these appeals was that NeuroScience breached the letter of intent by not paying Vojdani’s invoices in full.
In the first trial, NeuroScience attempted to defeat this claim by arguing that the ambiguous terms of the letter of intent did not require full payment of the invoices. Vojdani’s acceptance of payments for only those tests that were actually sold, without complaining that he was being shortchanged, reflected the parties’ actual agreement according to NeuroScience. In the alternative, NeuroScience argued, even if the terms of the letter of intent did in fact require payment in full, the parties had modified the terms in line with what NeuroScience had actually paid.
On this claim, the first jury was instructed on contract modification, including the point that contracts can be modified orally or by “conduct or other means of expression” indicating that “strict performance was not insisted upon.” The problem here arose from the design of the special verdict form, which did not include any question about contract modification. Although NeuroScience had asked (somewhat vaguely) for questions that would allow the jury to consider “what happened” after the con
tract’s execution and had opined that the proposed questions were confusing, the court dismissed the company’s concerns. The jury was asked on this claim only: (1) whether Vojdani had proven by a preponderance of the evidence that NeuroScience “agreed in the June 21, 2007 letter of intent ... to pay plaintiff the invoiced amount for each [testing] plate sent to defendant NeuroScience whether the plate was sold to a client or not,” and (2) whether Vojdani had proven by a preponderance of the evidence that NeuroScience “did not pay the plaintiff the full amount of the invoicesThe jury answered the first question yes but answered the second no, thus reaching a verdict for NeuroScience on the claim.
Vojdani then filed a motion under Federal Rule of Civil Procedure 59(a) seeking a partial new trial on this claim. He argued that the manifest weight of the evidence was against the jury’s answer of no to the second verdict question. NeuroScience’s witnesses, he pointed out correctly, had conceded that the invoices were
not
paid in full. NeuroScience argued that the jury’s answer to the second question was its way of accounting for the contract’s modification.
The district court granted Vojdani’s motion because NeuroScience admittedly had not paid the full amount of the invoices. But the court settled on a broader scope for the new trial than Vojdani wanted. The new jury, unlike the first, would be asked expressly whether the parties had modified the written contract after its execution. The first jury’s yes answer to the first special verdict question — whether NeuroScience had agreed in the letter of intent to pay the invoices in full — would stand. Vojdani objected to any question about modification. He argued that NeuroScience had waived such questions by not requesting them in the first trial. The court responded that NeuroScience had sufficiently preserved the issue.
Just before the second trial, the case was reassigned from Judge Crabb to Chief Judge Conley for scheduling reasons. The second jury found in response to more specific verdict questions that the parties had- orally modified the contract six months after its execution so as to allow NeuroScience to pay for only those tests it actually sold. Based on stipulations about the amount invoiced before the modification and the amount NeuroScience actually paid, the district court entered judgment in favor of Vojdani for $187,000 on this claim.
B.
Analysis
NeuroScience argues on appeal that the district court’s grant of the partial new trial should be reversed. In its view, the jury’s finding that Vojdani had not “proven by a preponderance of the evidence that defendant NeuroScience did not pay the full amount of the invoices” could be understood as an acknowledgment that the parties had modified the contract. The court was obliged to read the verdict that way, NeuroScience contends, and thereby to reconcile the evidence with the verdict.
We review for abuse of discretion a district court’s decision to grant a new trial under Rule 59(a).
ABM Marking, Inc. v. Zanasi Fratelli, S.R.L.,
353 F.3d 541, 543 (7th Cir.2003);
Medcom Holding Co. v. Baxter Travenol Labs., Inc.,
106 F.3d 1388, 1397 (7th Cir.1997).
A jury’s
answers to flawed special verdict questions should stand if the answers can be reconciled with the evidence and with one another in any reasonable way,
Medcom Holding,
106 F.3d at 1401-02, but that standard allows for judgment and discretion.
The district court here used its superior vantage point to make a decision well within its discretion. The court reasonably concluded that the jury’s answer to the second question on the verdict form in the first trial was contrary not only to the manifest weight of the evidence but to the undisputed evidence. The jury’s answer to the second question&emdash;-that Vojdani had failed to prove that NeuroScience did not pay the invoices&emdash;was inconsistent with NeuroSeience’s admission that it did not pay the invoices in full. NeuroScience’s theory&emdash;that the jurors agreed with its modification argument and expressed their conclusion by answering no to the second question&emdash;might be true, but it remains purely speculative because of the way the special verdict questions were framed.
Confusing special verdict questions can require reversal on appeal. See
Burger v. Int’l Union of Elevator Constructors Local No. 2,
498 F.3d 750, 754 (7th Cir.2007) (new trial on damages needed because confusing verdict form may have caused jury to award damages on impermissible theory);
Mattson v. Schultz,
145 F.3d 937, 939 (7th Cir.1998) (“Ambiguous, biased, misleading or confusing [special verdict] questions may warrant reversal.”). We see no reason to reject the district court’s decision in this case to correct the problem on its own. Perhaps the district court also could have reasonably adopted NeuroScience’s theory, but borderline cases are precisely where the district court must be allowed to exercise its discretion. We find no abuse of that discretion in the grant of a partial new trial.
In his cross-appeal, Vojdani argues that the scope of the new trial was too broad. He insists that NeuroScience should not have been permitted to argue in the second trial that the agreement in the letter of intent was later modified. As he sees it, NeuroScience waived any right to special verdict questions on modification during the first trial. In support he cites Federal Rule of Civil Procedure 49(a)(3), which says that a “party waives the right to a jury trial on any issue of fact raised by the pleadings or evidence but not submitted to the jury” unless the party demands that the issue be submitted. He also relies on our statement in
In re Rhone-Poulenc Rorer, Inc.,
51 F.3d 1293, 1303 (7th Cir.1995), that the “right to a jury trial in federal civil cases, conferred by the Seventh Amendment, is a right to have juriable issues determined by the first jury impaneled to hear them (provided there are no errors warranting a new trial), and not reexamined by another finder of fact.”
Our deferential review of a district court’s decision to grant a partial new trial under Rule 59 naturally extends to the district court’s decisions about the scope of the new trial. See,
e.g., McClain v. Owens-Corning Fiberglas Corp.,
139 F.3d 1124, 1128 (7th Cir.1998) (affirming grant of new trial only on damages). The formulation of special verdict questions is also a matter committed to the district court’s sound discretion.
Mattson,
145 F.3d at 938;
U.S. Fire Ins. Co. v. Pressed Steel Tank Co.,
852 F.2d 313, 316 (7th Cir.1988).
The district court did not abuse its discretion by allowing the contract modification issue to be presented to the second jury. When flawed special verdict questions require a new trial, the district court itself can and should correct the problem. See
Fort Howard Paper Co. v. Standard Havens, Inc., 901
F.2d 1373, 1375-77 (7th Cir.1990) (upholding exclusion of special verdict questions from second trial that should not have been asked at first trial). Here the district court had assured NeuroScience before the first trial that it could argue its contract modification defense to the jury. At the same time the court appears to have mistakenly rebuffed NeuroScience’s concern that the special verdict questions were confusing and did not account adequately for the possibility of contract modification. Correcting that mistake was proper.
The authorities Vojdani cites do not undermine our conclusion. Rule 49, far from blocking answers to previously unsubmit-ted special verdict questions, allows the court to answer such questions on its own if it so chooses. Fed.R.Civ.P. 49(a)(3). Even if NeuroScience had waived under Rule 49 its right to special verdict questions on modification, a district court has broad discretion to relieve a party of waivers of issues, claims, or defenses so long as the other party is given sufficient notice. See
Pactiv Corp. v. Rupert,
724 F.3d 999, 1003 (7th Cir.2013) (district court had discretion to excuse plaintiffs waiver of his best theory for relief). And our conclusion in
Rhone-Poulenc
that issues decided by one jury should not be reconsidered by another “provided there are no errors warranting a new trial” does not bar asking the second jury in this case about NeuroScience’s contract modification theory. The district court reasonably concluded that NeuroScience had raised the modification issue, that the design of the first special verdict prevented a clear answer to the question, and that the modification issue simply was not resolved by the first jury. Judge Crabb acted within her discretion when she determined that fairness required the second jury to decide the modification issue. We affirm the award of $187,000 to Vojdani on this claim.
II.
Second Claim
— Breach
of the Confidentiality Agreement
Two months before Vojdani and NeuroScience executed the letter of intent in 2007, they entered into a confidentiality agreement. They agreed that while they explored a possible collaboration, they would exchange unspecified confidential information and that the exchanged information would be kept confidential and would not be used outside of the collaboration for five years, subject to certain exceptions. As part of the parties’ later letter of intent, the parties agreed that Vojdani would provide NeuroScience with certain testing methods he had developed for use with third-party testing kits and that the two would “split the revenue of such tests 50/50.” NeuroScience paid Vojdani to use his testing methods through the end of 2008. NeuroScience then stopped paying yet continued to use Vojdani’s methods. According to Vojdani’s testimony, he asked NeuroScience to “return our confidential and proprietary information,” but to no avail. (The damages verdict at issue on appeal concerns NeuroScience’s failure to pay only after June 5, 2009, when the business relationship was completely over.)
Vojdani claimed that the testing methods he supplied were confidential, making NeuroScience’s unauthorized use of them after the business relationship ended a violation of the confidentiality agreement. During the first trial, which was the only
trial on this claim, NeuroScience responded that all of the testing procedures Vojda-ni had provided were publicly available or had been significantly altered by NeuroScience, two conditions that would except them from the confidentiality agreement. NeuroScience also argued that Vojdani could not prove damages from any violation of the agreement. He was no longer selling the tests, so he could not show that NeuroScience’s use of his testing methods cost him any sales.
Vojdani countered that he had lost 50 percent of NeuroScience’s revenue from performing the tests, the amount that he had been receiving under the letter of intent before NeuroScience canceled that agreement. Vojdani made clear, though, that the breach he was alleging was of the confidentiality agreement alone. Under that theory, the payment term in the expired letter of intent could have been relevant only as a measure of damages. The confidentiality agreement contained no payment terms and specified that it was not a licensing agreement. To make matters even worse for Vojdani, he repeatedly disclaimed any theory that he was seeking a reasonable royalty as a measure of damages under the confidentiality agreement.
During the trial, NeuroScience moved for judgment as a matter of law on this claim under Federal Rule of Civil Procedure 50(a), arguing that Vojdani could not prove damages. The court denied the motion and the claim went to the jury. The jury’s special verdict on liability for breach of the confidentiality agreement found: (1) that Vojdani had “proven by a preponderance of the evidence that after June 5, 2009 defendant NeuroScience continued to use any testing methods provided” by Vojdani,” (2) that no exception to the confidentiality agreement permitted that use, and (3) that NeuroScience’s violation of the agreement harmed Vojdani.
During the damages phase of the trial, the jury was instructed as follows:
The fundamental basis for an award of damages for breach of contract is just compensation for losses necessarily flowing from the breach. A party whose contract has been breached is not entitled to be placed in a better position because of the breach than the party would have been had the contract been performed. The injured party is entitled to the benefit of its agreement, which is the net gain it would have realized from the contract but for the failure of the other party to perform.
The jury awarded Vojdani $1,165,230 in damages for NeuroScience’s breach of the confidentiality agreement. (The parties agree that some portion of this award was for prospective damages, though that is not a critical point on appeal.)
NeuroScience then renewed its motion for judgment as a matter of law on this claim. See Fed.R.Civ.P. 50(b). NeuroScience explained that Vojdani had presented no evidence of losses flowing from a breach of the confidentiality agreement. Although he was entitled to be put in the position he would have been in but for the breach, the confidentiality agreement merely prohibited NeuroScience from using the testing methods. NeuroScience’s compliance with the agreement would not have benefitted Vojdani financially unless they were in competition. And as NeuroScience reiterated, Vojdani had made no attempt to sell the tests himself during the alleged breach, so NeuroScience’s violation of the agreement could not have taken away sales he might have made.
Judge Crabb initially denied the Rule 50(b) motion, explaining that Vojdani was entitled to a reasonable royalty for the use of his testing methods and that the agreement in the expired letter of intent was good evidence of what a reasonable royalty
would have been. But on reconsideration, Judge Crabb granted the motion, explaining that Vojdani had never argued for a reasonable royalty and that the jury had not been instructed on a reasonable royalty or any other measure of damages that could justify the award.
Vojdani challenges the district court’s grant of NeuroScience’s renewed motion for judgment as a matter of law, which vacated the verdict of nearly $1.2 million for breach of the confidentiality agreement. Vojdani argues that the money would merely compensate him for the “actual loss” he suffered from the breach. NeuroScience responds that Vojdani did not and cannot prove any damages because he was not made worse off by the confidentiality agreement’s violation.
We review
de novo
the district court’s grant of judgment as a matter of law.
Khan v. Bland,
630 F.3d 519, 528 (7th Cir.2010). A damages award the district court has set aside will not be reinstated on appeal based on a theory that was never presented to the jury. See
Liu v. Price Waterhouse LLP,
302 F.3d 749, 756 (7th Cir.2002) (upholding remittitur because plaintiffs theory justifying the award was never presented to the jury).
The decisive issue, then, is whether the jury could have applied the instructions it received to the evidence in the record and concluded from it that Vojdani is entitled to $1.2 million for NeuroScience’s breach. The jury was instructed that the non-breaching party, Vo-jdani, should be placed in the position he would have occupied but for the breach, which is the canonical understanding of damages for breach of contract and also the standard in Wisconsin.
United Concrete & Const., Inc. v. Red-D-Mix Concrete, Inc.,
349 Wis.2d 587, 836 N.W.2d 807, 824 (2013). The non-breaching party may recover expectation damages and any other losses foreseeably flowing from the breach.
Id.
As the jury here was also instructed, though, the non-breaching party may not be “placed in a better position because of the breach than he would have been [in] had the contract been performed.”
Dehnart v. Waukesha Brewing Co.,
21 Wis.2d 583, 124 N.W.2d 664, 670-71 (1963).
The loss .Vojdani claims is 50 percent of NeuroScience’s revenues from tests it performed using his methods after their business relationship ended. This is the amount Vojdani would have received under the letter of intent, but that agreement had expired before the time in question. Vojdani has said repeatedly, both in this court and the district court, that his claim about the ongoing use of his testing methods is for breach of the confidentiality agreement alone. He has not argued that the payment agreement in the letter of intent was extended orally or otherwise beyond June 2009&emdash;in which case the confidentiality agreement would be irrelevant. Nor was the jury asked to decide whether the agreement was extended. (Perhaps Vojdani’s clearest statement on what he is arguing is found in his reply brief: “Contrary to Neuroscience’s misplaced characterizations, Vojdani’s damage theories do not assume that the relevant breach was failure to pay 50% of its gross revenues. Instead, the relevant breach was Neuroscience’s use of Vojdani’s confidential information after June 5, 2009, the day the parties’ business relationship terminated.”)
On this claim for breach of the confidentiality agreement, the expired letter of intent can be relevant only as a measure of damages. Yet NeuroScience is correct on a critical point. If it had complied with the confidentiality agreement’s requirement that the company not use Vojdani’s proprietary information outside of their
collaboration, Vojdani would have gained nothing. His methods simply would not have been used after the collaboration ended.
One possible answer to that argument might be a “diverted trade” theory, under which a breaching party’s profits may be a reasonable proxy for the non-breaching party’s losses and thus a suitable measure of damages. See
Blue Ribbon Feed Co. v. Farmers Union Cent. Exch., Inc.,
731 F.2d 415, 421 (7th Cir.1984) (trade name infringement under Wisconsin law). But the record in this case is bereft of any evidence that NeuroScience’s breach cost Vojdani a single sale. In fact, he seems to have made no attempt to compete with NeuroScience during the time it was breaching the confidentiality agreement. See
id.
(“Requisite to an award of the defendant’s profits under the theory of diverted trade ... is a finding that plaintiff and defendant were in competition.”); see also
U.S. Naval Inst. v. Charter Communications, Inc.,
936 F.2d 692, 696-97 (2d Cir.1991) (explaining that an award based on the breaching party’s profits is normally disallowed as punitive if it exceeds the nonbreaching party’s losses). In any event, Vojdani presented the jury with a calculation of NeuroScience’s gross revenues rather than its profits, so this theory is not available to sustain the verdict.
We do not hold that damages cannot be awarded for breach of a confidentiality agreement when the parties are not in direct competition. One obvious remedy for such a breach, when the confidential information is used by the defendant for its own commercial purposes, is a reasonable royalty. See
Celeritas Technologies, Ltd. v. Rockwell Int’l Corp.,
150 F.3d 1354, 1359 (Fed.Cir.1998) (affirming plaintiffs verdict in similar breach of contract case based on reasonable royalty theory). A reasonable royalty is the amount an unauthorized user of proprietary information would have agreed to pay if negotiating in good faith.
Forest Labs., Inc. v. Pillsbury Co.,
452 F.2d 621, 627 (7th Cir.1971). Judge Crabb relied upon this damages theory when she initially upheld this verdict, recognizing that the payment terms in the letter of intent provided unusually reliable evidence of the amount NeuroScience probably would have been willing to pay to continue using Vojdani’s testing methods.
Nevertheless, we agree with Judge Crabb’s ultimate conclusion that the award cannot be upheld as a reasonable royalty. Unlike the
Celeritas
plaintiff, who explicitly sought damages based on a reasonable royalty theory, Vojdani never pursued that theory at trial. He never argued it, never asked for instructions on it, and never presented evidence on it. NeuroScience thus had no reason or opportunity to argue against it. Even on appeal Vojdani argues that a reasonable royalty would not be the proper measure of his damages. He insists that he is seeking only standard expectation damages. Questions about a reasonable royalty, he says, are “nothing more than irrelevant distractions.”
We do not understand the reasons he has taken that position, but we cannot rescue the jury’s verdict based on a reasonable royalty theory that he has abjured repeatedly in the district court and in this court, so that NeuroScience never had an opportunity or a reason to respond to the theory. Cf.
Pactiv Corp. v. Rupert,
724 F.3d 999, 1001 (7th Cir.2013) (district judges “sometimes have the authority to relieve parties of their forfeitures ..., but if they do this they must notify the other side, so that it can meet the argument”);
Southern Illinois Riverboat Casino Cruises, Inc. v. Triangle Insulation and Sheet Metal Co.,
302 F.3d 667, 677-78 (7th Cir.
2002) (district judge may decide case based on issue the court has raised
sua sponte
but must first give parties notice and opportunity to respond)
In this case, the district court correctly found that Yojdani had simply failed to present to the jury the only theory that might have supported a damages award for the breach of the confidentiality agreement, so NeuroScience never had an opportunity to offer evidence or argument on the theory. Despite his awareness that NeuroScience continued to use his'testing methods, Vojdani also chose not to pursue an injunction. Instead, he sought to insert the payment terms of the expired letter of intent into the confidentiality agreement without justifying that synthesis under the law. The jury understandably may have thought that NeuroScience should not be allowed to benefit from its breach, but Vojdani simply did not supply the jury with the evidence, argument, or instructions to avoid that result.
The judgment of the district court is AFFIRMED in all respects. Each side shall bear its own costs.