The plaintiffs, residents of the City of Hattiesburg, Mississippi (the “City”), appeal the denial of their claims under § 2 of the Voting Rights Act of 1965 (“VBA”). We affirm.
*666I.
A.
This concerns the drawing of districts for municipal elections in the City, and the relevant facts of City government, demographics, and political history are easily summarized. The City employs a mayor-council form of government. See Miss. Code Ann. § 21-8-1. The mayor is elected at-large. The City Council has five members, each of whom is elected from one of the City’s five single-member wards. The wards are demarcated according to Mississippi law, which requires the City to place an equal number of citizens in each ward. See Miss. Code Ann. § 21-8-7.
The 2000 Census indicated a City population of 44,779, of whom 49.9% were white and 47.3% black. The Census also reported a voting-age population of 35,141, of whom 56.5% were white and 40.9% black. In 2001 the City held a general election. Johnny DuPree was elected the City’s first black mayor, receiving 53% of the vote and unseating the white incumbent, who received 46%. Three white City Council members were elected, from Wards 1, 3 and 4; two black members were elected, from Wards 2 and 5.
In 2003, the City annexed land, raising the population to 45,446, of whom 50.4% were white and 46.9% black. After the annexation, 56.9% of the voting-age population was white and 40.5% black.1 Based on the population following the 2003 annexation, the ideal population for each ward was 9,089. Id. The populations under the City’s current districting plan, implemented after pre-clearance in 2004 (the “2004 redistricting plan”), deviate from that ideal by 4.8%.
The black population now resides mostly in Wards 2 and 5, which have black voting-age populations of 76.2% and 73.1%, respectively. The white population lives mostly in Wards 1, 3, and 4, where there are white voting-age populations of 62.9%, 84.9%, and 77.4% respectively. In 2005 the City held another general election. DuPree, the black mayor, was re-elected with 60.9% of the vote, defeating a white challenger with 39.1%. All the incumbent City Council members were re-elected.
Ward 1, one of the City’s majority-white wards, is the site of the University of Southern Mississippi (“USM”). Ward 5, which is majority-black, is home to William Carey College (“WCU”). According to the district court and the 2000 Census, almost 3,500 students live in dormitories in the City. Most of those are white, and the vast majority — approximately 3,000 students — live in the USM dormitories in Ward 1.
There are over 5,000 USM students in Ward 1, comprising about 60% of the ward’s population. At least 570 students residing in the USM dormitories are registered to vote in the City. The district court, however, found that it was impossible, on the evidence presented, to determine the total number of college students registered to vote in the City or the total number of college students qualifying as City residents.
B.
In 2006, the plaintiffs sued to challenge the 2004 redistricting plan. In particular, they objected to the plan’s assignment of USM students to Ward 1 and alleged that the plan unnecessarily packed black voters into Wards 2 and 5. Those two elements of the 2004 redistricting plan, the plaintiffs claimed, left the City’s black voters unable to control Ward 1 and elect a third black City Council member. Plaintiffs alleged *667that, as a result, under the totality of the circumstances, the plan deprived black voters of the equal opportunity to participate in the political process and to elect representatives of their choice, thus violating § 2 of the VRA. The plaintiffs also asserted that the plan violates the “one person, one vote” principle of the Equal Protection Clause of the Fourteenth Amendment.
The district court issued judgment for the City. Plaintiffs appeal.
II.
Section 2 of the VRA forbids state and local voting procedures that “result[] in a denial or abridgement of the right of any citizen of the United States to vote on account of race[.]” 42 U.S.C. § 1973(a). A § 2 violation is shown if, “based on the totality of circumstances,” members of a racial group “have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice.” Id. § 1973(b). To establish a § 2 violation on a vote-dilution theory, courts apply the two-part framework in Thornburg v. Gingles, 478 U.S. 30, 106 S.Ct. 2752, 92 L.Ed.2d 25 (1986).
First, plaintiffs must satisfy, as a threshold matter, three preconditions. Specifically, the minority group must demonstrate that: (1) it is sufficiently large and geographically compact to constitute a majority in a[n additional] single-member district; (2) it is politically cohesive; and (3) the white majority votes sufficiently as a bloc to enable it — in the absence of special circumstances — usually to defeat the minority’s preferred candidates. Failure to establish all three of these elements defeats a [§ 2] claim. Second, if the preconditions are proved, plaintiffs must then prove that based on the totality of the circumstances, they have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice.
Sensley v. Albritton, 385 F.3d 591, 595 (5th Cir.2004) (alterations, citations, and quotation marks omitted).
This court reviews de novo the legal standards the district court applied to determine whether [§ 2] has been violated. However, because [§ 2] vote dilution disputes are determinations peculiarly dependent upon the facts of each case that require an intensely local appraisal of the design and impact of the contested electoral mechanisms, we review the district court’s findings on the Gingles threshold requirements and its ultimate findings on vote dilution for clear error.
Id. (citations and quotation marks omitted).
The district court found that the plaintiffs had not proven the first of the three Gingles preconditions and so had not made out a prima facie case for violation of the VRA.2 In the alternative, the court found that the plaintiffs had failed to prove deprivation of equal opportunity under the totality of the circumstances. The plaintiffs take issue with both findings.
A.
To prove the first Gingles precondition, the plaintiffs were required to show “the possibility of creating more than the existing number of reasonably compact districts with a sufficiently large minority population to elect candidates of its *668choice.” Johnson v. DeGrandy, 512 U.S. 997, 1008, 114 S.Ct. 2647, 129 L.Ed.2d 775 (1994). Such districts are called “majority-minority” districts.3
The plaintiffs speak of three redistricting plans they placed before the district court, each capable of creating a black majority in Ward 1: (1) the exclusion of dormitory college students from the City’s population base for districting purposes, (2) the equal division of USM dormitory students among all five wards instead of placing all of them in Ward 1, and (3) redrawing the City’s wards to create a third majority-minority district without excluding the students.4 The district court expressly rejected the first, which was the one the plaintiffs presented in an illustrative redistricting plan, but did not mention the other two. The plaintiffs, in their sole briefed argument countering the district court’s Gingles preconditions findings, take issue with that omission.
The district court in a VRA case is required to discuss all substantial evidence contrary to its opinion. “[FJailure to take note of substantial contrary evidence presented by” the losing party justifies “remand[ing] the case for further findings,” Velasquez v. City of Abilene, 725 F.2d 1017, 1020-21 (5th Cir.1984), because without discussion of such evidence “there is no way for this court to tell whether the court’s [findings of fact are] clearly erroneous,” id. at 1023.5 The question, then, is whether the evidence the plaintiffs want treated as redistricting plans constitutes substantial evidence that the court was bound to consider. The district court’s obligation to discuss the evidence before it, we note at the outset, was not unbounded. The court was certainly never required to discuss the entirety of the evidence presented or even the entirety of the plaintiffs’ own evidence,6 nor could the mere presence in the record of other evidence supporting the plaintiffs have been enough to trigger the court’s obligation to discuss contrary evidence. “We will not consider arguments ... that [were] not presented to the district court.”7
*669That general principle applies with Ml force to VRA claims; in fact, the need for a developed district court record is especially acute in VRA cases. Westwego, 872 F.2d at 1203 (citing Velasquez, 725 F.2d at 1021). Without sufficiently detailed evidence, a court is flatly unable to evaluate whether a possible redistricting scheme would establish legally adequate districts consistent with traditional dis-tricting principles such as compactness, contiguity, maintaining communities of interest, and respect for incumbency. Sensley, 385 F.3d at 597-98; Prejean v. Foster, 227 F.3d 504, 512 & n. 12 (5th Cir.2000). Requiring the district court to fish through the record for evidence that might conceivably support redistricting approaches that were never urged by the plaintiffs or presented as developed plans would be downright perverse.
For that reason, to establish the first Gingles precondition, plaintiffs typically have been required to propose hypothetical redistricting schemes and present them to the district court in the form of illustrative plans.8 After all, the plaintiffs bear the burden of proof in a VRA case, and any lack of record evidence on VRA violations is attributed to them, not the district court. Roscoe, 123 F.3d at 846.
In short, it was solely the plaintiffs’ responsibility to develop theories supporting their VRA claims and to provide the district court with the evidentiary basis for evaluating them. If VRA plaintiffs want the district court to consider plans for creation of additional majority-minority districts, or to rely on appeal on the district court’s obligation to consider evidence supporting those plans, they must press them by developing arguments and evidence in support.9 Unless the plaintiffs raised the two alternative approaches and supported them with evidence bearing on their legal adequacy, the district court was under no obligation to consider them or to go through the meaningless formality of reciting their insufficiency.
In finding that the plaintiffs had not established the first Gingles precondition, the district court examined the sole proposed redistricting plan that the plaintiffs brought as evidence. The plaintiffs’ expert as to that precondition, William Cooper, testified that he had been hired to determine “whether it would be possible ... [to create] a black majority district ... after removing the dorm students at” USM and WCU. R. 480.10 That was the question he studied and the only one he testified to on direct examination. Though Cooper, in his response to the plaintiffs’ last question on direct examination, suggested that there might be “other possibili*670ties” for redistricting that would create a third majority black ward, the plaintiffs did not follow up on that suggestion and, instead, immediately tendered Cooper for cross-examination. R. 520-21. Cooper then conceded on cross-examination that he “didn’t do an illustrative plan with the students” included and that he “would have to do another illustrative plan” to testify about such an approach to redistricting. R. 621.
Nor did plaintiffs’ other witnesses indicate any alternative redistricting plan. To the contrary, another of their witnesses, Deborah Delgado, testified that she thought it was “unfair” that inclusion of the USM students increased the City’s white voting age population. R. 354. Even the plaintiffs believed at trial that they were presenting only the student exclusion plan and no other: Plaintiffs’ counsel, in fact, referred to that plan as though it were the only one being propounded, stating to the court that “the existing plan is already in evidence” and that Cooper “ha[d] already covered it.” R. 653. Plaintiffs never presented the other two plans; they certainly did not “press” them. See Benefit Recovery, 521 F.3d at 329.
The plaintiffs base their supposed alternative plans primarily on colloquies during Cooper’s cross-examination. Cooper volunteered his suggestion that the dormitory students could be split among the five wards; the City did not elicit it, R. 545, and the plaintiffs never returned to it on cross-examination, R. 625. Cooper presented no evidentiary support for it; certainly neither he nor any other witness testified that it should actually be adopted.
Moreover, it is obviously questionable whether splitting members of geographically contiguous communities in interest, i.e., the colleges, into separate constituencies would have been legally satisfactory. Courts are expected, in evaluating redistricting plans, to take into account “traditional districting principles such as maintaining communities of interest and traditional boundaries,” Sensley, 385 F.3d at 596 (quoting Abrams v. Johnson, 521 U.S. 74, 92, 117 S.Ct. 1925, 138 L.Ed.2d 285 (1997)), but the plaintiffs never presented a map, for example, showing what such districts would look like, leaving the district court unable to determine whether those requirements would be met, id. (explaining that “the shape of a [proposed majority-minority] district certainly cannot be disregarded in a compactness inquiry”). Any plan that split USM into multiple wards, furthermore, would be highly suspect on its face. Id. at 598.
The plaintiffs have brought no legal argument, either here or in the district court, that justifies assigning dormitory students, for districting purposes, to wards in which they do not actually live.11 Cooper also conceded that, depending on how many of the college students are City residents, dividing the students up might have led to impermissible variations between the ward voting populations; Cooper acknowledges that he did not have the data necessary to predict whether it would. R. 585-88, 605-09.
The plaintiffs, then, failed to present evidence constituting a redistricting plan that would have shown either the factual or legal possibility of creating a third majority-minority district by dividing the dor*671mitory students among the wards. In no case have we remanded for consideration of evidence the plaintiffs themselves did not present, and we doubt that the district court’s failure to consider evidence that appeared only on cross-examination of the losing party’s witnesses could support remand under Velasquez. In the absence of an illustrative plan or testimony, the district court was not presented with substantial evidence it was obligated to discuss.12
As for the third “plan” — drawing three majority black districts without making any special arrangements for the students — the same observations and conclusions hold. In the sections of the trial record the plaintiffs cite as support, Cooper testified (again, for the first time on cross examination) that “maintaining] a third majority black ward” by that approach would be possible only if the black voting age population percentages in Wards 2 and 5 — the two majority-minority wards- — -were drawn down “to at least the high fifties,” in which case they would be “more competitive” between black and white candidates and “could well elect a white candidate.” R. 610-12. Cooper expressed doubt about whether Wards 2 or 5 would even count as majority-minority districts under such a plan;13 he did not urge the court to adopt that approach. R. 612. The third plan was therefore not before the district court sufficiently for the court to evaluate it.
In short, the only plan the plaintiffs developed in the district court was the one involving exclusion of dormitory students, including an unknown number of City residents, from the City’s population for redistricting decisions. The two alternative plans the plaintiffs press on appeal were never asserted in the district court. Without any possibility of performing the legally-required evaluation, the court could hardly have concluded, as the first Gingles precondition requires, that a third majority-minority district would have been feasible if the students were included.14 We therefore reject the plaintiffs’ claim that the district court was obligated to consider any approach to redistricting but the one they actually presented in their illustrative plan.
There is no error in the district court’s rejection of the plaintiffs’ plan involving student exclusion. Section 2 of the VRA “is not concerned with maximizing minority voting strength!.]” Strickland, 129 S.Ct. at 1248; see also De Grandy, 512 U.S. at 1017, 114 S.Ct. 2647. The plaintiffs’ theory, in their illustrative plan, seems to have been that removing the *672dormitory students, whether legal residents or not, from districting calculations would reduce the ideal number of officially-counted residents per district. Because the excluded students are mostly white, that would have made it easier to create a third majority black ward.
In other words, the plaintiffs wished the district court to order the City to redistrict based on fictional population data, namely the pretense that a number of City residents were not there. Although resident students would still have been actually present, and therefore able to vote, the plaintiffs evidently assumed that lower voter turnout among the students, R. 354, 365-66, would have protected the electoral power of the inflated black majorities. Nothing in Cooper’s testimony or report suggests how to avoid excluding resident students.
Though “aliens, transients, short-term or temporary residents, or persons denied the vote for conviction of crime” may be excluded for voter apportionment purposes, Burns v. Richardson, 384 U.S. 73, 92, 86 S.Ct. 1286, 16 L.Ed.2d 376 (1966); see also Boddie v. City of Cleveland, 297 F.Supp.2d 901, 905-06 (N.D.Miss.2004), bona fide City residents certainly may not be.15 No authority supports or permits the solution the plaintiffs proposed. Because they did not explain how the unconstitutional exclusion of residents in the course of implementing their desired redistricting plan could be avoided,16 the district court was correct in rejecting it.
B.
In the alternative, the district court concluded that even if the three Gingles preconditions were met, the plaintiffs still had not satisfied the totality-of-circumstances test. Plaintiffs argue that that was error. Having affirmed the district court’s conclusion with respect to the first Gingles precondition, we need not reach the totality analysis at all. Nonetheless, for the sake of completeness, we consider and reject the plaintiffs’ objections.
The Supreme Court has pointed to a number of factors derived from the VRA’s legislative history as essential for weighing the totality of circumstances:
[1] the history of voting-related discrimination in the State or political subdivision;
[2] the extent to which voting in the elections of the State or political subdivision is racially polarized;
[3] the extent to which the State or political subdivision has used voting practices or procedures that tend to enhance the opportunity for discrimination against the minority group[;]
[4] the extent to which minority group members bear the effects of past dis*673crimination in areas such as education, employment, and health, which hinder their ability to participate effectively in the political process;
[5] the use of overt or subtle racial appeals in political campaigns;
[6] the extent to which members of the minority group have been elected to public office in the jurisdiction[;]
[7] evidence demonstrating that elected officials are unresponsive to the particularized needs of the members of the minority group[;]
[8] [evidence] that the policy underlying the State’s or the political subdivision’s use of the contested practice or structure is tenuousQ]
[9] whether the number of districts in which the minority group forms an effective majority is roughly proportional to its share of the population in the relevant area[.]
LULAC v. Perry, 548 U.S. 399, 426, 126 S.Ct. 2594, 165 L.Ed.2d 609 (2006) (citing Gingles, 478 U.S. at 44-45, 106 S.Ct. 2752; De Grandy, 512 U.S. at 1000, 114 S.Ct. 2647).
The district court applied and interpreted these factors correctly as a matter of law — certainly the plaintiffs have not directed our attention to any legal errors — and so we review the district court’s determinations for clear error only. A clearly erroneous finding is one that leaves “the reviewing court on the entire evidence ... with the definite and firm conviction that a mistake has been committed.” Anderson v. City of Bessemer City, 470 U.S. 564, 573, 105 S.Ct. 1504, 84 L.Ed.2d 518 (1985) (quoting United States v. United States Gypsum Co., 333 U.S. 364, 395, 68 S.Ct. 525, 92 L.Ed. 746 (1948)).
If the district court’s account of the evidence is plausible in light of the record viewed in its entirety, the court of appeals may not reverse it even though convinced that had it been sitting as the trier of fact, it would have weighed the evidence differently. Where there are two permissible views of the evidence, the factfinder’s choice between them cannot be clearly erroneous.
Id. at 573-74, 105 S.Ct. 1504.
The district court relied on several pieces of evidence in reaching its conclusion: (1) “[0]fficial discrimination” in the political process “ceased long ago”; (2) “there is no indication that [the poverty of black City residents] hinders their ability to participate effectively in the political process”; (3) election turnout among black voters in the City is higher than among white voters; (4) there was insufficient evidence to establish “unresponsiveness” of the City to the needs of black residents; and (5) blacks are well-represented in City government, roughly in proportion to their population. Aggregated, these facts weigh directly against the plaintiffs on the first, third, fourth, sixth, seventh, and ninth of the totality-of-the-circumstances factors. The only factor found by the district court weighing in favor of the plaintiffs was the second, namely, the existence of polarized voting, a finding that neither side disagrees with.
Plaintiffs challenge the adequacy of the district court’s review on two bases.17 They argue first that the City’s reliance on Census data for purposes of *674districting was improper, but they bring only generalized suspicions about the Census’s accuracy instead of specific evidence. The Census is presumptively correct and typically must be rebutted with clear and convincing evidence. NAACP v. Fordice, 252 F.3d 361, 366 (5th Cir.2001). If generalized suspicions such as the plaintiffs’ were sufficient to cast doubt on the district court’s findings, then courts could hardly ever rely on the Census. In the absence of data contradicting the Census, district courts have no reasonable alternative but to rely on it, and we see no reason to question the court’s reliance in this case.
The plaintiffs also claim that the district court committed legal error with respect to the last factor in considering the proportionality of representation to voting age population (where the City’s white population has a majority), rather than to the registered voter population (which favors the City’s black residents). That argument is unsupported by the law. Voter registration figures are, at most, a factor within the district court’s case-by-case discretion to consider in the Gingles totality analysis, and the plaintiffs give no legal argument indicating that the court abused its discretion here. No authority requires the district court to give voter registration figures special attention. See De Grandy, 512 U.S. at 1017 n. 14, 114 S.Ct. 2647 (declining to decide which population figure is the best reference point for proportionality). In the end, “substantial proportionality” is what matters in the totality-of-circumstances analysis. Id. at 1015-16, 114 S.Ct. 2647.
Registration figures should be treated with skepticism: Because voter registration responds to incentives, including the results of the political process, registration figures are inherently unstable. As a result, if registration percentages were the predicate for a redistricting, the percentages might actually change in response to the redistricting, potentially leading to further § 2 claims. Moreover, because “[u]se of a registered voter or actual voter basis [for apportionment] depends not only upon criteria such as govern state citizenship, but also upon the extent of political activity of those eligible to register and vote,” those figures are “susceptible to improper influences” exerted by entrenched political powers. Richardson, 384 U.S. at 92, 86 S.Ct. 1286.
This court in Marshall v. Edwards, 582 F.2d 927, 937 (5th Cir.1978), followed the Richardson Court in urging against judicial reliance on voter registration data for similar reasons, stating that “reapportionment based on voter registration is allowable only if it produces a distribution of legislators not substantially different from that which would have resulted from the use of a permissible population basis.” Id. (quoting Richardson, 384 U.S. at 93, 86 S.Ct. 1286) (quotation marks omitted). Reliance on the more solid grounds of population figures therefore makes good sense in most cases, and the district court did not clearly err in its use of total voting age population.
III.
We affirm also the district court’s rejection of the plaintiffs’ claim that the 2004 redistricting plan violates the one person, one vote principle of the Equal Protection Clause of the Fourteenth Amendment. That clause guarantees the opportunity for equal participation by all voters in local government elections. See generally Avery v. Midland County, 390 U.S. 474, 88 S.Ct. 1114, 20 L.Ed.2d 45 (1968). Ideally, municipal apportionment schemes should distribute population equally across all voting districts or wards, but limited variances in population numbers per ward are permitted as long as *675they are unavoidable despite a municipality’s best efforts to achieve exact numerical equality. Sims, 377 U.S. at 578-79, 84 S.Ct. 1362. Some deviation may be permissible when it is in service of constitutionally permissible goals such as “main-tainting] the integrity of various political subdivisions” and “providing] for compact districts of contiguous territory.” Id.
If a population deviance exceeds 10%, it constitutes a prima facie case of invidious discrimination that requires the municipality to prove a legitimate reason for the discrepancy. See Brown v. Thomson, 462 U.S. 835, 842-43, 103 S.Ct. 2690, 77 L.Ed.2d 214 (1983). If a population deviance is less than 10%, it is considered minor and does not suffice, alone, to make out a prima facie case of discrimination. Although deviation below that threshold does not provide a safe harbor, a plaintiff in such a case “must prove that the redistricting process was tainted by arbitrariness or discrimination.” Moore v. Itawamba County, 431 F.3d 257, 259 (5th Cir.2005) (per curiam).
The City’s current districting plan, in which the dormitory students are included' in Ward 1, has a deviation of 4.8%. The district court held that because the inclusion of the dormitory students in the apportionment base was permissible, the plaintiffs had failed to make out a prima facie case, and because they had presented no evidence of arbitrariness or discrimination, their claim failed as a matter of law. Once again, we review conclusions of law de novo and findings of fact for clear error. Teague v. Attala County, 17 F.3d 796, 796-97 (5th Cir.1994) (per curiam).
The plaintiffs do not take issue with the district court’s calculation of population deviations, nor do they point to any record evidence of arbitrariness or discrimination in redistricting. Instead, they merely repeat their argument that the City’s dis-tricting should have been based on voter registration figures rather than the voting age population. Even if that were true— and once more we doubt it — that is not the evidence of arbitrariness or discrimination that they were required to bring. The plaintiffs needed to show some lack of honesty or good faith or the presence of discriminatory purpose in the City’s decision to draw its districts as it did, Moore, 431 F.3d at 259, but have not done so. The district court’s one person, one vote holding therefore stands.
In summary, the district court’s treatment of the plaintiffs’ claims is legally adequate, and there is no clear error. Its denial of those claims is therefore
AFFIRMED.
. But, 52% of registered voters in 2005 were black and 48% white.
. The district court found that the plaintiffs had proven the second and third Gingles preconditions.
. To the extent that the plaintiffs argue for creation of a ‘'swing” or "equal opportunity” ward in which the black population would form a larger minority and make black candidates more competitive than they are now, the argument is foreclosed by Bartlett v. Strickland, - U.S. -, 129 S.Ct. 1231, 1246, 173 L.Ed.2d 173 (2009) (plurality), which held that "a party asserting § 2 liability must show by a preponderance of the evidence that the minority population in the potential election district is greater than 50 percent,” at least where intentional discrimination is not at issue. Thus, the VRA does not permit this court to mandate creation of swing wards, but only majority-minority ones.
. The plaintiffs do not fully explain the difference between the second and third plans. Unless USM students living in Ward 1 dormitories were to be arbitrarily and counterfactually treated as residents of wards where they do not live (and there is some record evidence to suggest that is what the plaintiffs have in mind, R. 606), we see no way to divide those students among the five wards except by drawing ward lines through the campus. Nonetheless, we follow the plaintiffs' structure, noting that the vagueness of the two latter plans invites skepticism.
. The district court is also required to explain its reasoning in VRA cases in a way suitable for review, and if it does not, this court will remand for it to do so. Westwego Citizens for Better Government v. City of Westwego ("Westwego"), 872 F.2d 1201, 1203-04 (5th Cir.1989). Though the plaintiffs invoke Westwe-go, the district court's factual findings and legal analysis are thorough and detailed and so meet that standard easily.
. LULAC v. Roscoe Indep. Sch. Dist., 123 F.3d 843, 846 (5th Cir.1997) (citing Rollins v. Fort Bend Indep. Sch. Dist., 89 F.3d 1205, 1221 (5th Cir.1996)).
. See Benefit Recovery, Inc. v. Donelon, 521 F.3d 326, 329 (5th Cir.) ("[W]e require a *669party to do more than just raise an argument; the contention must be pressed so that the district court has had an opportunity to rule on it."), cert. denied, - U.S. -, 129 S.Ct. 201, 172 L.Ed.2d 142 (2008); see also Tel-Phonic Servs. Inc. v. TBS Int'l Inc., 975 F.2d 1134, 1137 (5th Cir.1992) (“A party will not be heard to appeal the propriety of an order to which it agreed.”). Thus, if the district court did not discuss contraiy evidence to which the plaintiffs did not draw the court’s attention, objections based on the court's Velasquez duty may be waived.
.See Magnolia Bar Ass’n, Inc. v. Lee, 994 F.2d 1143, 1151 n. 6 (5th Cir.1993) (observing that the first Gingles precondition "specifically contemplates the creation of hypothetical districts”).
. At the same time, VRA plaintiffs are entitled to make strategic decisions about which redistricting theories to press in court and which to omit.
. Cooper granted on direct examination that some City residents — at a minimum, those living in dormitories but not registered to vote — would be excluded in the process. R. 494. The district court noted and relied on that flaw.
. As Cooper testified on cross-examination,
[Y]ou can improve the plan in terms of the voting power of voters in Ward 1 because you're still going to get a turnout of several hundred registered voters who live in the donns in Ward 1 who have actually been apportioned to other wards in a sense. But it would — it would make the plan much more fair from that standpoint.
R. 606 (emphasis added).
. For similar reasons, we attach little evi-dentiary significance to plaintiffs’ counsel's cross-examining the City's expert about the physical and geographical possibility of redrawing the ward boundaries to include different portions of the USM campus. R. 778-80.
. That reticence is understandable, given that the ability of the cohesive black populations in the resulting districts to “elect candidates of [their] choice” would have been very uncertain. The plaintiffs may not even have wanted to present a plan that would create a third majority-minority ward at the expense of voting strength in the present ones.
.In Gingles, 478 U.S. at 50 n. 17, 106 S.Ct. 2752, the Court provided that it is sufficient that a plaintiff show that a workable plan for another minority-controlled voting district is possible; the plaintiff’s plan need not be an ultimate solution. Houston v. Lafayette County, 56 F.3d 606, 611 (5th Cir.1995). There is a world of difference, though, between well-developed, legally-adequate plans that can be adjusted during implementation and vague implications of hypothetical approaches to redistricting that might be inferred from the record on appeal by losing plaintiffs. Leaving details to the political process is one thing; leaving initial development of the plan to the district court is quite another.
. See, e.g., Reynolds v. Sims, 377 U.S. 533, 576, 84 S.Ct 1362, 12 L.Ed.2d 506 (1964) (discussing “[t]he right of a citizen to equal representation and to have his vote weighted equally with those of all other citizens”). In Fairley v. Patterson, 493 F.2d 598, 602 (5th Cir.1974), an earlier VRA case concerning the City, this court examined a redistricting plan that would have excluded all USM and WCU students living in dormitories or fraternity houses. The court never reached the merits of the plan, however, deciding the case instead on standing. Id. at 603-04.
. In their reply brief, plaintiffs note that they "could have undertaken a study” to determine which dormitory students were legal residents. That makes no difference; it was their burden actually to do so, and they bear the consequences of their failure to present evidence on the subject. Roscoe, 123 F.3d at 846. (We doubt, however, that the result would have changed.)
. The plaintiffs also mention pieces of evidence the district court allegedly failed to analyze sufficiently. Plaintiffs do not explain most of that evidence or show how it bears on the totality analysis. We decline to consider those evidentiary arguments, partly because we may not re-weigh the evidence as the plaintiffs ask, partly because they are waived for failure to brief, United States v. Beaumont, 972 F.2d 553, 563 (5th Cir.1992), and partly because we do not consider the evidence to be substantial under Velasquez.