MEMORANDUM AND ORDER
GREGG COSTA, Circuit Judge.*After an eight day trial in this patent infringement case, the jury returned a verdict finding that Plaintiff RLIS’s patents were both invalid and not infringed. RLIS now moves for a new trial (Docket Entry No. 242). RLIS raises five issues, *965many of which the Court addressed in oral rulings at trial, and none of which persuade the Court that a new trial should be granted.
I. Background
This patent infringement case relates to software used to generate electronic medical records. The landscape of the case changed considerably during the course of the trial. RLIS initially accused Cerner of willfully infringing nineteen claims in U.S. Patents 5,823,948 and 7,076,436 with its PowerNote product, and Cerner has consistently contested RLIS’s infringement contentions and also asserted various invalidity theories. The day after the jury was impaneled and informed generally about the case by the Court, RLIS dropped all claims from the '948 Patent and many of the claims from the '436 Patent, leaving induced infringement of claims 1, 2, 3, 6, 7, and 8, and direct infringement of claims 24, 25, 27, 28, 29, 30, and 32. Telephone Conference (01/06/15) at 6. On the third day of trial, after Dr. Ross testified and discussed an email to which Cerner had vigorously objected and the Court had admitted because of its potential relevance to willfulness, RLIS dropped its claim for willfulness. Trial Transcript Vol. 2 (01/08/15) at 4.
At the close of RLIS’s case-in-chief on the fifth day of trial, Cerner moved under Rule 50 for judgment as a matter of law on the inducement claims. Trial Transcript Vol. 4 (01/13/15) at 75-79. The Court granted that motion. Trial Transcript Vol. 6 (01/15/15) at 164-66. And just before closing arguments on the final day of trial, RLIS agreed to abandon Claim 29. Trial Transcript Vol. 7 (01/16/15) at 9. The remaining direct infringement claims were submitted to the jury, which found all claims both invalid and not infringed. Docket Entry No. 223 at 12, 24.
RLIS now moves for relief under Federal Rule of Civil Procedure 59, which permits a new trial to be granted “for any reason for which a new trial has heretofore been granted.” Fed.R.Civ.P. 59(a)(1)(A); see also Riverwood Int’l Corp. v. R.A. Jones & Co., Inc., 324 F.3d 1346, 1352 (Fed.Cir.2003) (holding that regional circuit law applies to the procedural aspects of a motion for new trial). A new trial may be granted if, for example, the “trial was unfair,” “prejudicial error was committed in its course,” or “the verdict is against the great weight of the evidence.” Smith v. Transworld Drilling Co., 773 F.2d 610, 613 (5th Cir.1985); see also Pryor v. Trane Co., 138 F.3d 1024, 1026 (5th Cir.1998) (explaining that courts affirm the verdict unless the evidence “points so strongly and overwhelmingly in favor of one party that the court believes that reasonable men could not arrive at a contrary conclusion.” (internal citations and quotation marks omitted)). Significant deference is owed to the jury’s verdict; “a new trial should not be granted unless the verdict is against the great weight of the evidence, not merely the preponderance of the evidence.” Dresser-Rand Co. v. Virtual Automation, Inc., 361 F.3d 831, 838-39 (5th Cir.2004) (internal citation omitted). RLIS challenges both the verdicts of invalidity and noninfringement; either one is sufficient to uphold the take-nothing judgment.
II. Invalidity
Cerner asserted several invalidity theories at trial, including obviousness, the on-sale bar, the public disclosure bar, and anticipation by disclosure in the prior art. RLIS argues a new trial is warranted on the invalidity defenses for three reasons: (1) the Court erroneously asked a general jury question on invalidity rather than asking special questions about each theory of invalidity; (2) the jury’s invalidity and non-*966inMngement verdicts are inconsistent and against the great weight of the evidence; and (3) the Court erroneously excluded evidence related to the public disclosure bar in RLIS’s rebuttal case.
A. Verdict Form
Recognizing the practical difficulty of challenging the sufficiency of the evidence on every theory of invalidity, RLIS largely hangs its hat on the argument that the Court should not have submitted a general question to the jury. RLIS argued that a general verdict was inferior to special questions on this issue, and submitted a jury charge of its own to that effect.
A general question on invalidity was best suited for this case. Submitting a question on each theory would have required the jury to make 24 specific findings on invalidity alone.1 This burdensome task could have required the jury to unnecessarily answer dozens of difficult questions even after they came to a consensus about certain invalidity defenses that alone would have been sufficient to invalidate the patent. And a general verdict question on invalidity is not uncommon in patent cases. See, e.g., Cordance Corp. v. Amazon.com, Inc., 658 F.3d 1330, 1337 (Fed.Cir.2011) (reviewing a verdict that asked the jury a single invalidity question even though the defendant presented two theories of invalidity); i4i Ltd. P’ship v. Microsoft Corp., 598 F.3d 831, 845 (Fed.Cir.2010) (same); Wi-LAN Inc. v. Apple, Inc., No. 2:11-cv-00068-JRG, Docket Entry No. 627 (E.D.Tex. Oct. 23, 2013) (submitting a single invalidity question to the jury); TQP Dev., LLC v. 1-800-FLOWERS.COM, Inc., No. 2:11-cv-00248-JRG, Docket Entry No. 407 (E.D.Tex. Nov. 25, 2013) (same); see also Finjan, Inc. v. Symantec Corp., 2013 WL 5302560, at *47 (D.Del. Sept. 19, 2013) (observing that “the jury [is] not required to identify the basis for invalidity”).
But whatever the best course may have been as a matter of court administration, RLIS cites no authority finding legal error resulting from a general instruction on validity. Indeed, RLIS conceded as much at the charge conference, acknowledging that this is a choice within the discretion of the trial court. Trial Transcript Vol. 4 (01/13/15) at 284 (“[Absolutely we agree it’s in your discretion, you can ask the simple is it invalid or not question.”); see Hoechst Celanese Corp. v. BP Chems. Ltd., 78 F.3d 1575, 1581 (Fed.Cir.1996) (“The specificity of the verdict is within the discretion of the trial judge.”); Structural Rubber Prods. Co. v. Park Rubber Co., 749 F.2d 707, 720 (Fed.Cir.1984) (“[I]t must be left to the sound discretion of the trial court what form of verdict to request of a jury.”); Weinar v. Rollform Inc., 744 F.2d 797, 808-09 (Fed.Cir.1984) (rejecting the argument that special interrogatories must be submitted on different theories of invalidity); see also Warner-Jenkinson Co., Inc. v. Hilton Davis Chem. Co., 520 U.S. 17, 39 n. 8, 117 S.Ct. 1040, 137 L.Ed.2d 146 (1997) (“With regard to the concern over unreviewability due to black-box jury verdicts, we offer only guidance, not a specific mandate.”).
B. Invalidity Theories
Assuming the general invalidity instruction was not error, a “general jury verdict of invalidity should be upheld if there was sufficient evidence to support any of the alternative theories of invalidity.” See Cordance, 658 F.3d at 1339. *967RLIS’s motion for new trial does not convince the Court that there are evidentiary issues with any of the invalidity theories, let alone with all of them.
For starters, RLIS has not raised anything more than a cursory challenge to obviousness, and more than ample evidence existed on that theory alone.
With respect to anticipation, RLIS does challenge the existence of claimed features in the prior art, arguing that the jury’s verdict on invalidity contradicts its verdict on infringement. RLIS does not argue the inconsistency of the verdict itself justifies a new trial, as it waived that argument by failing to raise it in a timely fashion. See Giddy Up, LLC v. Prism Graphics, Inc., 2008 WL 656504, at *2 (N.D.Tex. Mar. 12, 2008) (collecting cases that explain that the failure to object to inconsistency prior to the jury’s discharge waives a challenge to the inconsistency of the verdict). Instead, RLIS argues that the inconsistency demonstrates the lack of evidence supporting either verdict. But even if there was an inconsistency, a new trial is not warranted because there are invalidity theories other than those based on prior art which do not contradict the infringement verdict. And at any rate, no inconsistency existed. The jury could have easily concluded, based on the extensive expert testimony regarding the software, that Cerner Classic met the limitations described in the claims but PowerNote did not.2 RLIS’s argument depends on an isolated view of the testimony related to input text limitations, and other testimony on the same subject supported the jury finding for Cerner on both this theory of invalidity and infringement.3 A jury finding of invalidity based on prior art thus was not against the great weight of the evidence.
The same is true of the public disclosure and on-sale bars, which were the most hotly contested issues at trial and defenses on which Cemer focused heavily in its opening statement and closing argument. Cerner presented compelling evidence that TeleMed embodied the invention and that the invention was ready for patenting long before the critical date, including an expert review of the TeleMed source code at that time, a copyright registration and brochure indicating TeleMed was complete, printouts virtually identical to those in the patent specification, and numerous attempts to sell the completed product.4 Cerner then pointed to numerous documents showing that the invention was both on sale and publicly demonstrated. For example, RLIS indisputably demonstrated the claimed invention at trade shows and elsewhere, and Cemer walked through a litany of marketing materials, business plans, and other documents suggesting that RLIS hoped to publicize the invention rather than protect its confidentiality.5 In terms of the on-sale bar, Cemer presented everything from profit analysis to bidding, price, and installation discussions in an effort to show that RLIS offered to sell TeleMed to various prospective buyers.6
*968Every step of the way on these issues, RLIS tried to impeach Cerner’s evidence or present testimony to the contrary. Dr. Ross and Mr. Lynch maintained that RLIS only demonstrated the product to those willing to sign Nondisclosure Agreements (NDAs), and that those NDAs had been lost when the company folded.7 They vehemently denied making any offers for sale, trying to convince the jury that Tel-eMed was not ready for such significant steps.8 And they denied that, at the time /of the alleged demonstrations and sales, TeleMed embodied the invention or that it was ready for patenting.9
These sort of credibility determinations involving memories of events occurring twenty years before trial-a dispute divorced from any significant technical knowledge and thus more typical of a drug trafficking case than a modern-day patent one-are issues for which deference to the jury verdict should be at its zenith. See Miller v. Borden, Inc., 664 F.2d 543, 546 (5th Cir.1981) (“[The jury] weighs the contradictory evidence and inferences, judges the credibility of witnesses, receives expert instructions, and draws the ultimate conclusion as to the facts. The very essence of its function is to select from among conflicting inferences and conclusions that which it considers most reasonable.” (quoting Tennant v. Peoria & P. U. Ry. Co., 321 U.S. 29, 35, 64 S.Ct. 409, 88 L.Ed. 520 (1944))). Depending on which evidence they found most believable and trustworthy, the jury could have come out either way on these issues. Its decision to make these classic credibility calls in favor of Cerner was hardly against the great weight of evidence.
C. Exclusion of Evidence
Finally, RLIS challenges an evidentiary ruling related to the public disclosure bar. RLIS claimed Cerner’s evidence regarding the public nature of the demonstrations came as a “surprise,”10 and, at the start of the second week of trial (day 5), asked for the Court’s permission to introduce testimony in rebuttal from previously undisclosed witnesses who would say that the demonstrations took place pursuant to since-lost confidentiality ■ agreements. Docket Entry No. 206. Cerner moved to exclude this testimony, and the Court granted Cerner’s motion. Docket Entry No. 204; Trial Transcript Vol. 4 (01/13/15) at 267-70. RLIS argues that the exclusion of this evidence was harmful error that requires a new trial.
Rule 26(a) requires each party to provide the name of each witness it may call at trial. Fed.R.Civ.P. Rule 26(a)(3)(A)(i). “If a party fails to ... identify a witness as required by Rule 26(a) ... the party is not allowed to use that ... witness to supply evidence ... at a trial, unless the failure was substantially justified or is harmless.” Fed.R.Civ.P. 37(c)(1). Four factors guide the “substantially justified or harmless” inquiry: “(1) the importance of the evidence; (2) the prejudice to the opposing party of including the evidence; (3) the possibility of curing such prejudice by granting a continuance; and (4) the explanation for the party’s failure to disclose.” Texas A & M *969Research Found. v. Magna Transp., Inc., 338 F.3d 394, 402 (5th Cir.2003); see also Tokai Corp. v. Easton Enters., Inc., 632 F.3d 1358, 1364 (Fed.Cir.2011) (holding that regional circuit law governs the decision to exclude evidence).
A balancing of these factors did not warrant allowing RLIS to present testimony from undisclosed witnesses. First, although the evidence was on an important issue related to one of the many invalidity theories, it was cumulative of testimony from several witnesses (including Dr. Ross, Mr. Lynch, and Mr. Gerson) who stated that confidentiality agreements were signed prior to the demonstrations. Compare Garton v. United States, 203 F.3d 828, at *3 (5th Cir.1999) (finding the evidence important because, without it, the plaintiff had no witnesses to testify on a vital issue).
Second, permitting designation in the middle of trial would have caused Cerner substantial prejudice. Testimony would have been allowed from witnesses who were not deposed and whom Cerner had no time to investigate and discover any impeachment material. Though RLIS disclosed three of these doctors in response to one of Cerner’s interrogatories, RLIS never designated any as witnesses or gave any other indication that it might call them as witnesses at trial. Much of the case law rejects “last-minute efforts to designate” new witnesses on the eve of trial. See Bradley v. United States, 866 F.2d 120, 125 (5th Cir.1989) (“The government’s [late designation] nonetheless put the Bradleys at a distinct disadvantage throughout the trial. Because they could not depose Brekken and Bernell until approximately 2-3 days before trial, the Bradleys had very little time in which to prepare their cross-examination.”). RLIS’s request was far more prejudicial because RLIS’s disclosure was not made until the second week of trial. Perhaps even more troubling, allowing witnesses who were disclosed for the first time during the second week of trial could have given RLIS a substantial tactical advantage by potentially undermining a defense on which Cerner had placed much of its focus. That was an especially salient consideration given the concerns about sandbagging that the Court had already identified based on the protean case RLIS presented.11 See Innogenetics, N.V. v. Abbott Labs., 512 F.3d 1363, 1376 n. 4 (Fed.Cir.2008) (admonishing “eleventh hour disclosures, and attempts to proffer ... testimony without compliance with Rule 26” and noting that “[exclusion and forfeiture are appropriate consequences to avoid repeated occurrences of such manipulation of the litigation process”).
Next, a continuance was not practical given that RLIS did not inform the Court of its intention to introduce this testimony until the second week of trial. See Martinez v. Union Carbide Corp., 62 F.3d 392, at *2 (5th Cir.1995) (holding that a continuance would not cure the significant prejudice from “last-minute efforts to designate” new witnesses).
*970Finally, RLIS’s explanation that it was “surprised” by Cerner’s arguments is belied by the record, including Dr. Ross’s testimony indicating that RLIS had tried to find people who signed the NDAs prior to trial, but “didn’t know how to get in contact with anybody”;12 the fact that RLIS presented witnesses in its case-in-chief to testify that these demonstrations took place under conditions of confidentiality;13 and pretrial depositions, interrogatories, and motions that showed RLIS was on notice of the importance of this issue prior to trial.14 See 1488, Inc. v. Philsec Inv. Corp., 939 F.2d 1281, 1288 (5th Cir. 1991) (holding that the defendants failed to provide an adequate explanation for late disclosure of a witness when “[they] were aware” they needed to do so). The Court therefore reaffirms its conclusion that RLIS should not have been able to present testimony from persons whom they did not identify as potential witnesses until the second week of trial.
For these reasons, the Court will not grant a new trial based on any issues with the jury’s verdict invalidating the claims.
III. Induced Infringement
RLIS challenges the Court’s granting judgment as a matter of law on the claims for induced infringement. The claims for induced infringement related to Claims 1, 2, 3, 6, 7, and 8 of the patent. Liability for induced infringement arises if “the inducer ... persuade[s] another to engage in conduct that the inducer knows is infringement.” Global-Tech Appliances, Inc. v. SEB S.A., 563 U.S. 754, 131 S.Ct. 2060, 2065, 2068, 179 L.Ed.2d 1167 (2011).
Cerner began selling PowerNote around 1996-the conduct that is the basis for the allegation that it induced its customers’ infringement of the '436 Patent— long before it could have had any awareness of the '436 Patent which was filed in 1998 and issued in 2006. RLIS relies on one meeting in 1997 and two later prior art references in the patents of a Cerner affiliate. The meeting took place nine years before the '436 Patent issued, referred at most to the '948 Patent,15 and thus provides no evidence that Cerner knew the '436 Patent would issue or the scope of its potential claims. See State Indus., Inc. v. A.O. Smith Corp., 751 F.2d 1226, 1236 (Fed.Cir.1985) (“Filing an application is no guarantee any patent will issue and a very substantial percentage of applications never result in patents. What the scope of the claims in patents that do issue will be is something totally unforeseeable.”). And given how far removed they were from the meeting, the prior art references do not create an inference that Cerner monitored the scope of the '436 Patent claims, let alone suggest that Cerner believed its products infringed. The sale of Power-Note was thus set in motion long before the '436 Patent existed. Even after RLIS filed the lawsuit and Cerner became aware of the patent, RLIS presented no evidence that Cerner’s conduct or intent changed after the suit so as to support a claim for inducement. Accordingly, the JMOL was properly granted.
The Court’s dismissal of the induced infringement claims was not based on a good faith invalidity defense, which was later rejected in Commil USA, LLC v. *971Cisco Sys., Inc., — U.S. -, 135 S.Ct. 1920, 1927-28, 191 L.Ed.2d 883 (2015). In fact, the Court avoided ruling in a way that could have been impacted by the then-pending Commil case. See Trial Transcript Vol. 6 (01/15/15) at 164-66, 254-58. Commil did not disrupt the principle on which the Court relied, which is that “Global-Tech requires ... proof the defendant knew the acts were infringing.” 135 S.Ct. at 1928. Indeed, RLIS has not submitted any supplemental briefing since Commil was decided to argue that it impacts the ruling in this case.
IV. Direct Infringement
Finally, RLIS challenges an oral instruction to the jury that the infringement theory for Claim 24 related to HotSpot Dictation.16 Of course, this matters only if there is a problem with the verdict of invalidity for Claim 24, but the Court nonetheless will address it.
On the morning of closing arguments, Cerner sought to exclude RLIS from making an argument about an infringement theory for Claim 24 unrelated to HotSpot dictation. Docket Entry No. 215. RLIS responded that it should be allowed to make such an argument, Docket Entry No. 216, but the Court granted Cer-ner’s motion and stated it would issue an' oral instruction to that effect. Trial Transcript Vol. 7 (01/23/15) at 14-32. The issue is whether the instruction — read in the context of the trial and charge as a whole — misled the jury. See SynQor, Inc. v. Artesyn Technologies, Inc., 709 F.3d 1365, 1379 (Fed.Cir.2013) (analyzing whether “the instructions as a whole in the context of the trial [correctly] informed the jury”).
It did not. The instruction was both accurate and necessary to avoid jury confusion arising from RLIS’s ever-shifting theories of infringement. RLIS’s counsel conceded at trial that RLIS’s infringement theory for Claim 24 related to HotSpot dictation. Trial Transcript Vol. 7 (01/16/15) at 27. RLIS’s infringement expert, Dr. Rhyne, opined that the feature infringing Claim 24 was the HotSpot dictation feature, and RLIS presented no other evidence of an infringing feature during its case-in-chief. Trial Transcript Vol. 7 (01/16/15) at 29; see also Trial Transcript Vol. 2 (01/08/15) at 133. The instruction thus conveyed to the jury that the infringement theory for Claim 24 depended on the HotSpot dictation feature. The instruction plainly did not mean that RLIS had to prove HotSpot by itself satisfied every limitation of Claim 24. The only thing RLIS points to is a jury note about Claims 25-32, but that shows the jury had questions about how dependent claims functioned, not confusion regarding Claim 24. RLIS did an excellent job framing the issues, and had every opportunity to discuss all evidence in the record. RLIS simply has not shown that the instruction distorted its position or altered the outcome of the verdict on infringement.
RLIS’s Motion for a New Trial (Docket Entry No. 242) is DENIED.
Sitting by designation.
. At the time the Court considered whether to submit questions on each invalidity theory, all of the inducement claims remained in the case, meaning that the jury would have had to make about 70 different invalidity findings. Trial Transcript Vol. 4 (01/13/15) at 283-86. Granting the JMOL and the withdrawal of Claim 29 cut that number down to 24.
. E.g., Trial Transcript Vol. 4 (01/13/14) at 115-16, 119, 123-66.
. E.g., Trial Transcript Vol. 4 (01/13/14) at 128-29, 146; Trial Transcript Vol. 5 (01/15/14) at 37, 48, 91, 131, 137, 195.
. E.g., Trial Transcript Vol. 1 (01/07/15) at 139-40, 158-60; Trial Transcript Vol. 5 (01/14/15) at 131, 137-38
. E.g., Trial Transcript Vol. 1 (01/07/15) at 149-53.
. E.g., Trial Transcript Vol. 1 (01/07/15) at 148, 158, 161-62, 190-94.
. E.g., Trial Transcript Vol. 1 (01/07/15) at 96, 199.
. E.g., Trial Transcript Vol. 1 (01/07/15) at 108.
. E.g., Trial Transcript Vol. 1 (01/07/15) at 141.
.Docket Entry No. 232 at 260, 262 ("[RLIS’s Counsel]: I’ll repeat what I said, which is, I was surprised that this [public demonstration evidence] did come in because it was not disclosed properly under their invalidity contentions.”).
. To give some context to the strategic motive Cerner attributed to RLIS’s tactics, the parties hotly disputed the admissibility of a piece of evidence — the "McCallie email”— that allegedly showed Cerner copied RLIS's technology after a meeting between inventors at RLIS and Cerner. E.g., Docket Entry Nos. 113 & 124. The Court viewed the email's admissibility as a close call, but admitted the email as relevant to willfulness. But the day after Mr. Lynch testified about the email, RLIS dropped the willfulness claim, leading Cerner to allege that RLIS used a weak willfulness claim to ensure the Court would admit the email, and then jettisoned that legal theory once it served this evidentiary purpose. See Trial Transcript Vol. 1 (01/13/15) at 71, 81-82 (providing an overview of these allegations).
. See Trial Transcript Vol. 2 (01/08/15) at 223-24.
. See, e.g., Trial Transcript Vol. 1 (01/07/15) at 96.
. See Trial Transcript Vol. 4 (01/13/15) at 270 (Court’s findings on this issue); Docket Entry No. 155-7 (RLlS's motions in limine).
. The meeting took place on September 30, 1997. The application for the '948 Patent had been filed, but it had not yet issued.
. "I want to instruct you that RLIS’s position on Claim 24 has — is that only the HotSpot Dictation feature is an infringing feature in the PowerNote product of Claim 24. So Claim 24 relates to the HotSpot dictation feature.” Trial Transcript Vol. 7 (01/16/15) at 47.