concurring in part and dissenting in part.
In this summary judgment case, I view the applicable law and the record differently from the majority and must, therefore, respectfully dissent. I agree with the majority that this is a circumstantial case, and I agree with the majority’s disposition of all claims except the gender discrimination claims of Ms. Sanders and Ms. Coffey, and I conclude Ms. Sanders has established sufficient circumstantial evidence of age discrimination in addition to her direct evidence to survive summary judgment.
We are required to consider the totality of the circumstantial evidence when considering claims of pretext. Beaird v. Sea-gate Technology, Inc., 145 F.3d 1159, 1174 (10th Cir.1998). If the proffered evidence, when viewed in the aggregate, is “sufficient to raise a genuine doubt about a [defendant's motivation,” we may conclude a plaintiff “has met the requirements necessary to a showing of pretext.” Garrett v. Hewlett-Packard Co., 305 F.3d 1210, *11121220 (10th Cir.2002) (internal quotation marks omitted). “A showing that the employer’s justifications for its behavior are pretextual permits a finding of intentional discrimination, but does not compel it.” Exum v. U.S. Olympic Comm., 389 F.3d 1130, 1135 (10th Cir.2004).
Generally, “[a] plaintiff demonstrates pretext by producing evidence of such weaknesses, implausibilities, inconsistencies, incoherencies, or contradictions in the employer’s proffered legitimate reasons for its action that a reasonable factfinder could rationally find them unworthy of credence and hence infer that the employer did not act for the asserted non-diseriminatory reasons.” Jaramillo v. Colo. Judicial Dep’t, 427 F.3d 1303, 1308 (10th Cir.2005) (internal quotation marks omitted); Garrett, 305 F.3d at 1217. “[EJvidence of pretext may include, but is not limited to ... prior treatment of plaintiff; ... disturbing procedural irregularities (e.g., falsifying or manipulating ... criteria); and the use of subjective criteria.” Garrett, 305 F.3d at 1217 (final alteration in original) (internal quotation marks omitted).
Specifically, in Beaird v. Seagate Technology, Inc., we delineated three principal ways a plaintiff can demonstrate pretext in a RIF case. 145 F.3d at 1168. In my view, two of them apply in this instance. “First, [a plaintiff] can argue that her own termination does not accord with the RIF criteria supposedly employed,”1 even when an employer gives a business judgment reason for changing its RIF criteria midstream. Id. Although an employer “may choose to conduct its RIF according to its preferred criteria” rather than criteria previously documented by the employer, the employer’s application of or deviation from the documented RIF criteria is not immune to judicial review and may be “so idiosyncratic or questionable that a factfin-der could reasonably find that it is a pretext for illegal discrimination.” Id. at 1169. Selective application of documented or undocumented RIF criteria may indicate pretext. See id. at 1173-74.
Here, as in Beaird, Defendant claims to have followed its written RIF criteria. Specifically, Defendant’s four main RIF criteria are listed in the MSG as “performance,” “skills,” “experience,” and “training.” (R. at 370, Management Staffing Guidelines at 9.) The MSG expressly allows Defendant to add RIF criteria, and Mr. McNeely did so in a ranking considerations instruction sheet he sent to area managers. The ranking considerations sheet listed specific subcategories of the four general RIF criteria listed in the MSG for area managers to consider when ranking Band C managers. After litigation ensued, Defendant provided affidavits from several managers listing RIF subcategories previously undocumented in the MSG or on the ranking considerations sheet as their reasons for ranking certain employees above others within Band C. Although these reasons were certainly related to the four main RIF criteria, in my view, they were not consistent with the written subcategories found in Mr. McNeely’s ranking instructions. I refer to these criteria as the “undocumented RIF criteria” because they were not specifically documented in any of the contemporaneous documents relied on by Defendant when ranking employees.
The second method a plaintiff can employ to demonstrate pretext in a RIF case is to present evidence that her “evaluation under [a] defendant’s RIF criteria was deliberately falsified or manipulated” to affect her employment status negatively. Id. at 1168. “One method of demon*1113strating manipulation or falsification of evaluation is to produce evidence that a supervisor responsible for assessing her performance displayed ageist [or gender] animus.” Id.
In addition to these principal methods for demonstrating pretext, our cases explain three other relevant indications of pretext. First, “[w]hen an employer’s RIF criteria include job categorization, an employer must explain the basis for that categorization or risk a finding of pretext.” Id. at 1170. Second, the extent of apparent procedural irregularities with respect to the employee’s selection for the RIF may indicate pretext. Whittington v. Nordam Group Inc., 429 F.3d 986, 994 (10th Cir.2005); see also Plotke v. White, 405 F.3d 1092, 1104 (10th Cir.2005); Garrett, 305 F.3d at 1220. Third, inconsistencies in an employer’s explanation of the reasons for its negative employment action may also indicate pretext. Id.
Thus, in analyzing Plaintiffs’ claims, the court should focus its pretext analysis on the aforementioned five categories: (1) selective application of RIF criteria, (2) manipulation of employee evaluations under RIF criteria, (3) inadequate employer explanation for job categorizations, (4) RIF procedural irregularities, and (5) inconsistent employer reasons for surplussing.
Circumstantial Evidence as to Ms. Sanders
Ms. Sanders produced circumstantial evidence related to age and gender discrimination in four of the five pretext categories the court should consider. I conclude Ms. Sanders has produced evidence supporting a finding of pretext in three of those four categories.
Selective Application of RIF Criteria
To survive summary judgment in this category, Ms. Sanders needed to show Defendant’s application of or deviation from its RIF criteria was so “idiosyncratic or questionable that a factfinder could reasonably find that it is pretext for illegal discrimination.” Beaird, 145 F.3d at 1169. Deviation from documented RIF criteria alone does not support an inference of pretext. However, selective or non-uniform application of either documented or undocumented criteria may be sufficient to support such an inference.
When an employee provides sufficient circumstantial evidence to support a finding of pretext due to an employer’s selective application of its chosen RIF criteria, the employer may then attempt to meet the Reeves burden of providing abundant and uncontroverted evidence that no discrimination occurred to rebut the employee’s pretext claim and to preserve the employer’s claim of being entitled to summary judgment. I would hold that an employer has provided abundant and uncontroverted evidence of its nondiscriminatory RIF criteria application if the employer can document an observable pattern of RIF criteria application that shows the employer applied the same criteria to a large majority of its at-risk employees. When, as I see in this case, an employer selectively applies any RIF criteria favoring a minority of employees, who then finish with a higher rank than those surplussed, I would hold an employer did not meet its burden of presenting abundant and uncontroverted evidence of no discrimination to preserve its claim for summary judgment.
In this case, Defendant’s MSG states repeatedly that managers should use criteria such as performance, skills, experience and training as benchmark RIF criteria.2 Early in litigation, Defendant explained managers ranked Band C employees “from *1114high to low based on their performance, skills, experience, and training.” (R. at 181, Def. SWBT’s Mot. for Summ. J. and Br. in Supp. at 10.) Mr. McNeely claims he followed the MSG process for the RIF and instructed area managers to rank Band C employees using performance, skills, experience, and training criteria. Two of Defendant’s area managers, Mr. Alfred Saenz and Mr. Mike Harris, also claimed they used the four primary documented RIF ranking criteria.
Mr. McNeely also sent managers a ranking considerations sheet with a list of subset categories for the four main criteria to use when ranking their Band C managers.3 The ranking considerations sheet explained that, because Defendant had already used the performance criterion to calculate the band ratings, managers should use only the remaining three RIF criteria and their subcategories for ranking purposes. The ranking considerations sheet expressly instructed area managers to “[u]se the Skills, Experience and Training sections below in order to help in determining the ranking of the C band.” (R. at 233, Ranking Considerations at 1.)
Defendant claims it used its RIF criteria to rank Ms. Sanders and the other employees in her group. However, Defendant’s specific explanation, proffered after litigation ensued, for why certain employees were ranked above Ms. Sanders included several factors not listed in either the MSG or on the ranking considerations sheet provided by Mr. McNeely. Specifically, Defendant proffered twelve categories of undocumented criteria and one category of documented criteria to explain why twelve of Ms. Sanders’ peers, four of whom were younger and male, six of whom were older and male, and two of whom were younger and female, were ranked ahead of her. Although this fact alone does not support an inference of pretext, the record strongly supports an inference that Defendant did not apply these criteria uniformly to each employee when ranking twelve employees ahead of Ms. Sanders. In my view, using criteria not listed on the ranking considerations sheet and not considering all three content areas as instructed constituted a deviation from the RIF criteria “supposedly employed.” Beaird, 145 F.3d at 1168. I conclude Defendant then applied its chosen RIF criteria selectively.
The highest level of selectivity Defendant employed was when it applied six of the twelve undocumented criteria only once to six different employees ranked above Ms. Sanders and did not apply any of those six undocumented categories to any of the other eleven employees ranked ahead of Ms. Sanders.4 Defendant applied *1115three of the undocumented criteria to only two other employees.5 Defendant applied one undocumented criterion to three other employees who outranked Ms. Sanders.6 At best, the largest number of employees out of the twelve ranked ahead of Ms. Sanders that Defendant applied undocumented criteria to was five employees.7
Moreover, I am of the view that this inference of a selective, non-uniform approach is uncontroverted by the record. Indeed, there is little to no evidence from the area managers’ meeting indicating they discussed the undocumented criteria at all. The sketchy notes from that meeting support an inference that they did not uniformly apply their chosen criteria to a large majority of at-risk employees.
Specific examples of employees who outranked Ms. Sanders according to these undocumented criteria are as follows. First, Defendant ranked Mr. Michael Heatley higher than Ms. Sanders using the undocumented criteria of having “greater” skills. Even though Mr. Heatley did not have an extensive engineering background, he was considered to have greater skills because he learned quickly. (R. at 304, Harris Aff. 4 ¶ 11.) Yet the record indicates Defendant used Ms. Sanders to train new engineers rather than Mr. Heatley, whom Defendant indicated had greater skills.
Second, Defendant listed outside construction background as the undocumented criteria for ranking Mr. Alton Miller ahead of Ms. Sanders. Ms. Sanders trained Mr. Miller in 2002 when he was on a PIP for misconduct in his prior supervisory position. Mr. Miller had less time in his engineering position than Ms. Sanders. He had three months’ experience in 1983 and eight months’ in 2002 contrasted with Ms. Sanders’ ten years as an engineer. Length of time as an engineer was one of the undocumented criteria area managers used to place Ms. Sanders below some of her other peers, yet they did not uniformly apply that same criteria when comparing her to Mr. Miller. In addition, Ms. Sanders’ productivity was nearly three times higher than Mr. Miller’s. Productivity was another undocumented RIF criterion Defendant used in its considerations when ranking employees because it appeared in handwritten notes from the ranking meeting next to Ms. Sanders’ name.
Third, like Ms. Sanders, Mr. Woody Harjo had no outside plant technical experience, but Defendant retained him. Defendant ranked Mr. Harjo higher than Ms. Sanders and Ms. Nancy Bounds, who was also surplussed. According to the undocumented criteria, Ms. Bounds, like Ms. Sanders, was ranked lower because of having no outside plant experience. Even though neither Ms. Sanders, Ms. Bounds, nor Mr. Harjo had outside plant technical experience, Mr. Harjo was retained for his drafting and CAD background which Defendant claimed made him more versatile than either Ms. Sanders or Ms. Bounds. (R. at 305-06, Harris Aff. 5-6 ¶ 11.) Ms. Bounds had a drafting background, and as noted above, Ms. Sanders supervised drafting clerks. The record does not indicate if Ms. Sanders’ or Ms. Bounds’ drafting experience included CAD work.
*1116Thus, the non-uniform application of undocumented RIF criteria when cross-comparing employees during ranking, little to no evidence from the area managers’ ranking meetings indicating that they discussed the undocumented criteria at all and, finally, the existence of only sketchy notes regarding the criteria they did discuss when ranking employees all would combine to provide a reasonable juror with sufficient circumstantial evidence to support a finding of pretext in this area. In my view, Ms. Sanders’ circumstantial evidence of this selective application strongly supports an inference of pretext.
Finally, because Defendant applied each of its chosen RIF criterion to less than a large majority of ranked employees, I conclude it did not provide abundant and un-controverted independent evidence of nondiscrimination and, therefore, cannot preserve its claim for summary judgment in this area.
RIF Procedural Irregularities
I am of the view that Ms. Sanders provided circumstantial evidence of RIF procedural irregularities in two areas. First, Ms. Sanders points to Mr. Griffith’s participation in the ranking meeting in which Ms. Sanders was ranked thirteenth in the Band C pool of employees. The area managers knew before the ranking meeting Defendant would retain a specified number of first-level managers in Bands A-C. In fact, Defendant planned to retain all Band A and B managers but only thirteen Band C managers. The evidence supports an inference that Mr. Griffith knew he was likely to be surplussed from his position as an area manager because he was the least senior area manager. After participating in the ranking meeting, Mr. Griffith requested, pursuant to a policy adopted after the ranking meeting, a voluntary demotion from an area manager to a first-level manager in Ms. Sanders’ group to avoid being surplussed as an area manager during the RIF. Even though he was demoted, Mr. Griffith experienced no change in salary. I agree with the majority that Mr. Griffith’s demotion does not, by itself, provide support for a finding of pretext under irregular RIF procedures because Defendant is entitled to use its business judgment to create policies regulating transfers and to transfer employees according to those policies for legitimate business purposes. Moreover, the record provides no evidence that Mr. Griffith’s demotion was not conducted pursuant to a legitimate business decision.
Nevertheless, in my view, an inference of pretext arises from Mr. Griffith’s participation in the area managers’ ranking meeting at a time he knew his job was vulnerable and knew he might be transferring into the group of employees he helped to rank. Mr. Griffith was a decision maker at the meeting in which Defendant ranked Ms. Sanders thirteenth, after twelve other individuals who were either male or younger than Ms. Sanders, or both. I find Mr. Griffith’s involvement in the decision to rank Ms. Sanders thirteenth to be irregular when he knew she would be the next person surplussed if the total size of her manager group increased. That knowledge and Mr. Griffith’s subsequent request for a transfer to Ms. Sanders’ group combine to constitute a procedural irregularity that provides some support for a finding of pretext under the totality of the circumstances standard.
Second, Ms. Sanders provided circumstantial evidence of a procedural irregularity with respect to one topic of possible discussion at the area managers’ ranking meeting. That topic was age. Mr. Wooten claimed he did not know the ages of the engineers who reported to him and that age was not a factor in the rankings, yet his notes from the meeting indicated *1117six pension-eligible employees with a star. Although pension eligibility is not necessarily connected to age, it is sufficiently related under the circumstances of this case, to allow a reasonable juror to infer the area managers discussed age when discussing pension eligibility during the meeting. While this evidence, standing alone, would be insufficient to support a finding of pretext, under the totality of the circumstances standard, a reasonable juror could examine it along with the other evidence and infer from it that the area managers discussed age at the meeting.
Inconsistent Employer Reasons for Surplussing
If an employer offers inconsistent reasons for surplussing, we consider “(1) the timing of the change in position and (2) the evidentiary basis for the new rationale.” Jaramillo, 427 F.3d at 1311. A change in explanation occurring after significant legal proceedings have occurred supports an inference of pretext. Id.
Defendant gave numerous reasons at different times for surplussing Ms. Sanders. First, Mr. Wooten allegedly told Ms. Sanders directly that she was surplussed because of her age. However, in his affidavit and on his ranking notes, Mr. Wooten indicated Ms. Sanders’ productivity was the problem. Defendant elsewhere stated that Ms. Sanders’ lack of outside plant technical experience, work experience, and technical skills were the factors that led to her being surplussed. For instance, Mr. Harris stated that “Ms. Sanders’ ranking was primarily attributable to her non-technical background. A technical background is extremely valuable to an engineer’s working knowledge of outside plant facilities. Ms. Sanders was at a disadvantage since she did not have outside plant experience.” (R. at 304, Harris Aff. 4 ¶ 11.)
Defendant provided its age and productivity reasons during the RIF process both in Mr. Wooten’s conversation with Ms. Sanders and in the notes from the ranking meeting. However, Defendant’s later, inconsistent explanation was provided after Ms. Sanders filed suit. I find this change in explanation supports a finding of pretext because Defendant’s rationale changed after significant legal proceedings had occurred.
Conclusion as to Ms. Sanders
Viewing the evidence in the light most favorable to Ms. Sanders and under the totality of the circumstances standard, I conclude Ms. Sanders has produced a strong case of circumstantial evidence to support a finding of pretext. I am of the view that Defendant has not produced abundant and uncontroverted evidence showing discrimination did not occur and cannot preserve its claim of being entitled to summary judgment as a matter of law. Therefore, I would reverse and remand on both Ms. Sanders’ age and gender discrimination claims.
Circumstantial Evidence as to Ms. Coffey
Ms. Coffey produced circumstantial evidence of discrimination in four of the five categories the court should analyze. However, because Ms. Coffey was not ranked against any younger employees, I agree with the majority that her age discrimination claim cannot survive summary judgment. I therefore analyze the circumstantial evidence only with regard to her gender discrimination claim.
Selective Application of RIF Criteria
In Ms. Coffey’s case, as in Ms. Sanders’, I find Defendant’s explanation for its rankings is based on specific criteria undocumented in the MSG or on the ranking considerations sheet. Like Ms. Sanders, Ms. Coffey claims Defendant selectively applied its undocumented RIF criteria to preserve jobs for male employees.
*1118The five undocumented criteria Defendant proffered for ranking five men ahead of Ms. Coffey were having engineering experience, managing the job well, being more versatile, getting along better with a boss, and having outside technical experience. Defendant also considered supervisory experience, one of the documented criteria found on the ranking considerations sheet. Like Ms. Sanders, Ms. Coffey provides sufficient circumstantial evidence to support a finding of pretext in Defendant’s selective application of its chosen RIF criteria.
Specifically, Defendant applied three of the five undocumented criteria only once to three different employees ranked ahead of Ms. Coffey and did not apply those same criteria to the other two employees ranked ahead of Ms. Coffey.8 Defendant applied the other two criteria to only two employees ranked ahead of Ms. Coffey.9 Defendant applied none of these undocumented criteria to a large majority of at-risk employees.
Specific examples of the selective applications are as follows. Defendant ranked Mr. Roger Cox ahead of Ms. Coffey using one undocumented criterion. It stated Mr. Cox had a diverse working background and was thus more versatile than Ms. Coffey. Defendant cites to this criterion only one other time, when describing Mr. Steven Parrot. There is no mention of this undocumented criterion for any of the other three candidates ranked ahead of Ms. Coffey. Thus, the court has no way of knowing whether any of the other candidates ranked ahead of Ms. Coffey also had a more diverse work background. In my view, therefore, it was not uniformly applied.
Defendant also ranked Mr. Rolland King higher than Ms. Coffey because he managed his job well and because his boss was not having problems with him like Mr. Harris claimed he was having with Ms. Coffey. These two undocumented RIF criteria do not appear in the record in regard to any other person ranked higher than Ms. Coffey. Thus, they were not uniformly applied. In addition, Mr. Harris’ statement that he was having problems with Ms. Coffey is possibly in conflict with Mr. Harris’ evaluation of Ms. Coffey just nine months before when he wrote, “Ms. Coffey continues to learn and grow in her job. She works well with others and is a team player.” (R. at 311, Exhibit 7-B at 2.)
Finally, Defendant ranked Mr. John Starwalt higher than Ms. Coffey because he had more engineering experience and an outside technical background. The record indicates Defendant applied the engineering experience criterion only to Mr. Starwalt, and it applied the outside technical background criterion only to one other person ranked higher than • Ms. Coffey. Again, I conclude Defendant selectively applied undocumented RIF criteria.
Accordingly, I am of the view that Ms. Coffey has provided sufficient circumstantial evidence to support a finding of pretext in Defendant’s selective application of undocumented RIF criteria. Because I see no observable pattern of RIF criteria application to a large majority of Defendant’s at-risk employees, I conclude Defendant has not produced abundant and uncontroverted evidence of non-discrimination and thus cannot preserve its claim for summary judgment in this area.
*1119 Manipulation of Employee Evaluation
A reasonable juror can draw an inference that an employer has manipulated its evaluation of an employee under its RIF criteria if the record indicates evidence of gender animus on the part of a supervisor with power to assess an employee’s job performance. See Beaird, 145 F.3d at 1168. Inherent gender animus can warp a supervisor’s evaluation of an employee. In this case, Mr. Harris was Ms. Coffey’s immediate supervisor, and Ms. Kaylan Collins stated that she personally overheard Mr. Harris make “sexist remarks about female employees/managers.” (R. at 1003, Collins Aff. 2 ¶ 10.) Ms. Collins’ statement alone is sufficient to support a finding of pretext in this area.
Moreover, Ms. Coffey presented other evidence that, viewed in the light most favorable to her and under the totality of the circumstances, provides some support for an inference of pretext in this category. For instance, Ms. Coffey stated in her affidavit that Mr. Harris scrutinized female employees more closely than male employees. She claimed a male employee noticed the same behavior in Mr. Harris and mentioned it to her. Ms. Coffey also claimed that Mr. Harris rarely encouraged her in her work, and she believed he did not want to deal with women. Finally, Ms. Coffey stated in her deposition that another female employee had been told by Mr. Harris that she was a nice looking woman, and that if she dressed up and went for an “after hours’ interview” with the hiring manager, he thought she could probably get the job she was seeking. (R. at 759, Coffey Dep. 109:11-15.)
Inadequate Employer Explanation for Job Categorization
Defendant grouped employees in Ms. Coffey’s job category by job title, Manager — Construction. That job title included two subsets: construction managers, who supervised Defendant’s internal cable technician employees, and contract coordinators, who oversaw outside independent contractors in fulfilling Defendant’s cable work orders. Ms. Coffey was a contract coordinator, as were three of the five employees against whom she was ranked. The record indicates that a fourth employee was a construction manager, but there is a dispute regarding whether the fifth employee was a construction manager or a contract coordinator.
The Defendant’s MSG required it to group employees according to their “affected work group,” which the MSG defined as a portion of the organization at the same level, job title, similar job function, geography, and line of organization. (R. at 471, P’s. Combined Resp. to D’s. Mot. for Summ. J. at 15; R. at 363, MSG at 2 (Note).) Defendant argues it permissibly grouped construction managers and contract coordinators under the same job title because they had the same job function. Ms. Coffey asserts her day-to-day functions as a contract coordinator differed significantly from that of a construction manager. For example, construction managers worked with Defendant’s internal crews of cable technicians and provided supervision for them. Whereas Ms. Coffey, as a contract coordinator, had no supervisory authority and worked with several independent contractors at various job sites on any given day to ensure their compliance with Defendant’s contracted work orders.
This contrast is important because Defendant’s explanation for ranking two employees above Ms. Coffey — the two employees who were or who were alleged to be construction managers — included the criterion of prior supervisory experience, which Ms. Coffey allegedly could not obtain as a contract coordinator. Based on Defendant’s use of this criterion to rank construction managers over Ms. Coffey, in *1120my view, Defendant’s grouping of the two job subsets under one title could support a finding of pretext.
RIF Procedural Irregularities
Ms. Coffey alleges Defendant applied a permissible RIF procedure in an irregular manner. Specifically, Defendant allowed its area managers to reevaluate all employees prior to the RIF if the managers thought a reevaluation was necessary. In so doing, they had the latitude to raise or to lower an employee’s rating.10 This was likely to affect the employee’s risk of being surplussed, given that only employees rated as “average” and therefore placed in Band C were vulnerable during this RIF. The reevaluating procedure itself was not irregular. However, Ms. Coffey argues, and I agree, that she has provided sufficient evidence to support an inference that her supervisor used this procedure in an irregular manner.
Mr. Harris reevaluated one of Ms. Coffey’s younger, male peers who had been rated as “average” before the RIF, and raised him to an “above average,” Band B rating, thus saving him from possible surplussing. Yet, Mr. Harris did not lower the ranking of another male employee, Mr. Dewayne Hendricks, who had demonstrably poor performance during the RIF period. The action or inaction standing alone would not suffice to show irregularity. However, a jury could find it to be irregular that Mr. Harris exercised his power to improve his evaluation of one employee, based on his subjective determination that the employee’s performance merited reappraisal, yet did not exercise that same power to reevaluate another employee once he learned about that employee’s mismanagement during the RIF period.11 In my view, Ms. Coffey has provided sufficient evidence to support a finding of pretext based upon this procedural irregularity.
Conclusion as to Ms. Coffey
Viewing this evidence in the light most favorable to Ms. Coffey and under the totality of the circumstances standard, I conclude Ms. Coffey has produced sufficient circumstantial evidence in all four categories for which she produced evidence to support a reasonable juror’s finding of pretext as to gender discrimination. In addition, I find Defendant has not produced abundant and uncontroverted evidence showing discrimination did not occur and, therefore, cannot preserve its claim of being entitled to summary judgment as a matter of law.
For the reasons stated, I would hold that both Ms. Sanders and Ms. Coffey are entitled to trial on the gender claim and that Ms. Sanders has presented additional circumstantial support for her age claim.
. In Beaird, the RIF criteria supposedly employed were the procedures described in the employers handbook. See id. at 1162. Beaird was analyzing a termination not in accord with that policy. See id. at 1168.
. Organizations with a surplus condition must identify those employees who will be surplussed.
*1114Criteria such as performance, skills, experience and training, remain key to the selection of managers for staffing the organization. ...
In order to evaluate employees, the following steps must be followed and documented as indicated:
1. Those managers in an affected work group potentially subject to being declared surplus should be selected based on criteria such as performance, skills, experience and training. Initially, employees should be assigned to one of ... four bands: ...
Managers should be listed in order from high to low based upon criteria such as performance, skills, experience and training. ...
(R. at 370, Management Staffing Guidelines at 9.)
. The subcategories for skills consisted of management skills, technical skills, supervisory skills, effectiveness with others, ability to handle a broader scope of responsibility, and promotional possibility. Experience subsets were years of service and equivalent work experience. Training subcategories were formal company training, formal technical or trade school, and any degrees earned in school.
. These criteria were as follows: having a civic engineering background, learning the job quickly, having a technical splicing back*1115ground, being experienced in cable repair, having experience in the control management center, having outside plant technical experience, and having drafting and CAD experience.
.The three undocumented criteria used only twice were as follows: having longer time in an engineering position, installation and repair experience, and construction experience.
. The one undocumented criterion Defendant applied to three employees was having outside plant technical experience.
. These two criteria were having greater skill level and being more versatile.
. The three undocumented criteria Defendant used only once were having engineering experience, managing the job well, and getting along better with a boss.
. These categories were being more versatile and having outside technical experience.
. Absent evidence that the area manager’s subjective power to raise or lower any employee’s evaluation lapses once a RIF is declared, we must assume that power continues.
. I must note that Defendant disputes Mr. Harris' knowledge of the employee's misbehavior during the RIF period. However, we must view the facts in the light most favorable to Plaintiffs in this appeal from summary judgment. According to Ms. Brooks’ deposition testimony, Mr. Hendricks reported Mr. Harris told him during the RIF process that “if his production did not improve, that [Mr. Harris] could easily put his name on the list and take [Ms. Coffey] off.” (R. at 912, Brooks Aff. 137:1-8.) This testimony creates a triable issue of fact on this point.