In the

    United States Court of Appeals
                 For the Seventh Circuit
                     ____________________
No. 18-3630
MOLLY JOLL,
                                                 Plaintiff-Appellant,
                                 v.

VALPARAISO COMMUNITY SCHOOLS,
                                                Defendant-Appellee.
                     ____________________

        Appeal from the United States District Court for the
          Northern District of Indiana, Hammond Division.
      No. 2:16-cv-00338-JEM — John E. Martin, Magistrate Judge.
                     ____________________

   ARGUED DECEMBER 4, 2019 — DECIDED MARCH 20, 2020
               ____________________

   Before FLAUM, RIPPLE, and HAMILTON, Circuit Judges.
    HAMILTON, Circuit Judge. Plaintiff Molly Joll is an accom-
plished runner and an experienced running coach. She ap-
plied for a job as the assistant coach of a high school girls’
cross-country team. The high school hired a younger man for
the job but invited Joll to apply for the same position on the
boys’ team. So she did—and the high school hired a younger
man again. She filed this suit for sex and age discrimination.
After discovery, the district court granted summary judgment
2                                                    No. 18-3630

for the school district, concluding that Joll had not offered
enough evidence of either form of discrimination to present
to a jury.
    We reverse the dismissal of Joll’s sex discrimination claim.
The district court appears to have erred by doing what we
have repeatedly said a court should not: “asking whether any
particular piece of evidence proves the case by itself,” rather
than aggregating the evidence “to find an overall likelihood
of discrimination.” Ortiz v. Werner Enterprises, Inc., 

, 763, 765 (7th Cir. 2016). Joll offered evidence that would
allow a reasonable jury to find that the school district used
hiring procedures tilted in favor of the male applicants, ap-
plied sex-role stereotypes during the interview process, and
manipulated the criteria for hiring in ways that were incon-
sistent except that they always favored the male applicants. A
reasonable jury might also find no sex discrimination, but on
this record, the decision belongs to a jury.
I. Facts for Summary Judgment
    We state the case in the light reasonably most favorable to
Joll, giving her the benefit of conflicts in the evidence and rea-
sonable inferences from the evidence, but without vouching
for the objective truth of any fact or expressing any opinion
on the weight of the evidence. Garofalo v. Village of Hazel Crest,

, 430 (7th Cir. 2014); Lewis v. City of Chicago, 

, 651 (7th Cir. 2007).
   Defendant Valparaiso Community Schools operates Val-
paraiso High School and Thomas Jefferson Middle School,
two public schools in northwestern Indiana. Joll has been a
teacher at the middle school for more than twenty-five years.
No. 18-3630                                                      3

For twelve years, from 2005 to 2013, she co-coached the mid-
dle school boys’ and girls’ cross-country team. In 2013 she re-
signed the middle school coaching position to better support
her two daughters in their high school and college running
careers. “It wasn’t fair” to the cross-country program, as Joll
thought and explained at the time, for her to divide her time
between the program and her daughters.
    By 2014 her older daughter’s running career had come to
a close, so Joll again had the time to devote to coaching. Her
old position at the middle school had been filled, but the high
school was hiring an assistant coach for the girls’ cross-coun-
try team. On June 17, Joll sent letters of interest in the position
to high school principal Reid Amones and high school athletic
director Herb Hofer. That was “the normal process” for ap-
plying, according to Hofer.
    At first Joll heard nothing back. Around the end of June,
she emailed Hofer and enlisted the help of her union presi-
dent and vice-president. Finally, on July 3, Joll received a text
message from the girls’ head coach, Adam Nellessen, asking
whether she would be available for an interview two days
later, on July 5. Joll said she would be. While athletic director
Hofer ordinarily would have conducted the interview with
coach Nellessen, knee surgery prevented Hofer’s attendance
on July 5. His place was taken by principal Amones.
    Normally, according to Hofer, interviewers would ask ap-
plicants “what their qualifications are,” “what experiences
they have, what drew them to the job.” Joll has substantial ex-
perience both as a coach and as a runner herself. She ran cross-
country and track all four years of high school and attended
Indiana State University, an NCAA Division I school, on a full
4                                                  No. 18-3630

athletic scholarship for both sports. After receiving her mas-
ter’s degree in education, Joll spent about five years as the
girls’ track coach at a junior high school in central Indiana,
with a one-year stint as a volunteer track coach at a school in
the United Kingdom as part of a Fulbright teachers’ exchange
program.
    Joll then began her more than twenty years with Val-
paraiso Community Schools and its athletics programs. From
1991 to 1995, Joll was the assistant girls’ track coach at Val-
paraiso High School; she was the co-coach of the girls’ track
team at Thomas Jefferson Middle School from 1995 to 2005.
For part of the same period, from 1994 to 2004, she was an
assistant women’s and men’s cross-country coach at Val-
paraiso University, also a Division I school. Finally, as men-
tioned above, Joll co-coached the middle school’s cross-coun-
try team from 2005 until 2013, when her daughters’ athletic
careers led her to resign that position. Throughout this period
Joll also coached a local youth running club as a volunteer.
   There is no evidence, however, that this wealth of experi-
ence was discussed much during Joll’s July 5, 2014 interview
with coach Nellessen and principal Amones for the position
with the high school girls’ team. Rather, Joll fielded questions
about resigning her middle school coaching position in 2013
and whether her parenting duties would permit her to devote
sufficient time to coaching at the high school. She was em-
phatic that they would.
   Given Joll’s 2013 resignation from the middle school
coaching position, such questions might have seemed unre-
markable. In this case, however, the only other applicant was
John Arredondo, a forty-year-old man who was eventually
hired for the position. Arredondo had also resigned a cross-
No. 18-3630                                                   5

country coaching position in 2013 for family reasons. During
his interview with Nellessen and Amones, however, Arre-
dondo was not asked about his family life. He was asked in-
stead about his “coaching experience, what my coaching phi-
losophy is, a lot of shop talk.” In other words, the three men
“talked shop.” Joll had to talk parenting.
    The differences in the high school’s hiring process did not
end there. According to athletic director Hofer, the school’s
ordinary course was to check an applicant’s references only
after the decision had been made to recommend his hiring to
the school board, the school district’s final decision-maker on
personnel matters. That is how the school proceeded with Ar-
redondo. Joll’s references, however, were contacted within
days of her interview on July 5.
    The high school heard from at least three of Joll’s seven
references. Two gave her unqualified recommendations, sub-
mitting lengthy narratives praising her as a “respected
leader” and a “great role model.” Aaron Crague, head coach
of the high school’s boys’ cross-country team, whom Joll had
coached when he was a student-athlete at Valparaiso Univer-
sity, pointed to Joll’s “authentic knowledge gleaned” as a run-
ner herself. He said “she would do a fine job collaborating
with Coach Nellessen.” It had been Crague’s experience that
Joll had “always done a great job supporting and compli-
menting [sic] the head coach.”
   A third reference was generally strong but less favorable
on the latter point. Jim Polite, the middle school’s principal,
opined in a brief, three-point email to Amones that Joll’s fit-
ness to serve in a subordinate role “would be my only con-
cern.” Joll, thought Polite, “has a dominate [sic] personality.”
(The parties assume he meant “dominant personality.”)
6                                                   No. 18-3630

Amones forwarded Polite’s email to Hofer with the one-sen-
tence comment, “The 3rd one I got,” seemingly referring to
the “dominate personality” comment, the third of Polite’s
comments.
    On July 18, Hofer called Joll to tell her the school had cho-
sen Arredondo because he had “more current experience
working with high school age athletes.” That was true as far
as it went, despite Joll’s much more extensive experience as a
coach and runner. Arredondo had coached the boys’ cross-
country team at a different high school outside the school dis-
trict from 2008 to 2013. Joll had last professionally coached
high schoolers in 1995. There is no evidence of how the differ-
ence between middle school and high school athletes or be-
tween high school and college athletes was material to the as-
sistant coach position, and Polite saw “no reason” why Joll’s
middle school experience “wouldn[’]t translate to the HS
level.”
    In the meantime, Joll had learned that there would also be
a not-yet-publicized opening for an assistant coach of the
boys’ high school cross-country team. She told athletic direc-
tor Hofer that she intended to apply for that position as well.
Around the end of July, Joll received a message from Crague,
the boys’ head coach, asking her to interview for the position
on the boys’ team.
    The interview was conducted by athletic director Hofer
and coach Crague. Again Joll was called on to convince her
interviewers that she was willing to devote the necessary time
to coaching, notwithstanding her family responsibilities.
Again there is no evidence she was asked to “talk shop” with
the two men interviewing her. Again Joll lost out to a younger
man who was the only other applicant, a twenty-eight-year-
No. 18-3630                                                     7

old named Ben Kerezman. Again Joll’s references were han-
dled differently than the successful applicant’s. Kerezman’s
references were not checked at all because, according to Ho-
fer, “he was a teacher within our system” (so was Joll) and
because Hofer knew him personally (as Crague knew Joll per-
sonally; he had been a reference for her for the girls’ team job).
   Unlike Arredondo, Kerezman did not have more recent
professional experience coaching high school cross-country
runners compared to Joll. But for this hiring decision, recency
of professional experience with high school athletes was no
longer the relevant criterion. Instead, Hofer told Joll at the
time, Kerezman had been preferred because he had “better
rapport with the boys.” Kerezman was a teacher at the high
school and so, says the school district, “was familiar with
many of the cross-country [team] members,” though there is
otherwise no evidence of what Kerezman’s “rapport with the
boys” was.
    There is also no evidence and no explanation why “rap-
port with the girls” did not assume similar importance in hir-
ing the assistant for the girls’ team. That would have been bad
for Arredondo and good for Joll: Arredondo was a teacher at
an out-of-district high school whose only opportunity to build
“rapport with the girls” had been a five-week stint as a volun-
teer coach on the girls’ team beginning June 2014, just before
he was hired. Joll, on the other hand, had coached many of the
high school runners, both boys and girls, as middle schoolers.
    As it turns out though, “rapport with the boys” may not
actually have been all that relevant to Crague and Hofer’s de-
cision about the boys’ team. Rather, Hofer later testified, the
“only factors” considered in hiring the boys’ assistant were
“being a teacher within the building, the HS building,” and
8                                                  No. 18-3630

“recent coaching experience in the high school ranks.” At least
with respect to professional coaching (both Joll and Kerezman
had recent volunteer experience with high school cross-coun-
try runners), the latter cannot be squared with the school dis-
trict’s own account of Kerezman’s resume, nor with Hofer’s
own statement that “coaching experience” was not consid-
ered in the hiring process “because we already decided from
working in the building as being [sic] a very valuable thing for
our program.”
    As for “being a teacher within the building,” as with “rap-
port,” there is no evidence or explanation why the putative
single most important qualification for the boys’ assistant ap-
parently played no role at all in hiring the girls’ assistant.
Again that would have been bad for Arredondo and good for
Joll. Though neither taught at the high school, Joll at least
taught within the school district and as noted had already
coached many of the high school runners as middle schoolers.
    In July 2016, Joll brought this suit against the school dis-
trict in the Northern District of Indiana alleging violations of
her rights under Title VII of the Civil Rights Act of 1964,
42 U.S.C. § 2000e-2, and the Age Discrimination in Employ-
ment Act (ADEA), 29 U.S.C. § 623. The district court had ju-
risdiction under 28 U.S.C. § 1331. Magistrate Judge Martin
heard and decided the case with the parties’ consent under
28 U.S.C. § 636(c)(1). After discovery, the court granted de-
fendant’s motion for summary judgment on both counts of
the complaint and entered final judgment against Joll. After
the court denied her motion to reconsider, Joll appealed. We
have jurisdiction under 28 U.S.C. § 1291.
No. 18-3630                                                      9

II. Analysis
    We review the district court’s grant of summary judgment
de novo. Garofalo v. Village of Hazel Crest, 

, 430 (7th
Cir. 2014). Summary judgment is appropriate if there are no
genuine disputes of material fact and the movant is entitled
to judgment as a matter of law. Fed. R. Civ. P. 56(a); 

. The maxims of summary judgment are famil-
iar—no weighing the evidence, no credibility determinations,
draw all non-speculative inferences in favor of the non-
movant—but they are more easily recited than applied. John-
son v. Advocate Health and Hosp. Corp., 

, 893–94 (7th
Cir. 2018). They define the outer limits of Joll’s right to “the
commonsense judgment of [her] community,” Taylor v. Loui-
siana, 

, 530 (1975), and to the generous latitude the
court accords it. We entrust a great deal to the jury’s human
experience: that “strong drink strips the mind of its pretences
and brings out into the open what is hidden in a man’s heart,”
for example, United States v. Bloch, 

, 643 (7th Cir.
2013); that four days shackled to the wall of an interrogation
room is unreasonable policing, Lopez v. City of Chicago, 

, 720 (7th Cir. 2006); that “a 16-year-old girl in the
company of three adult men” is “the least likely of the four to
be carrying … heavy handguns.” County Court v. Allen, 

, 163–64 (1979). We allow a jury to infer a great deal
without mathematical precision. Standard Oil Co. v. Van Etten,

, 334 (1882). “It is always possible for … logicians
to examine what appears to be a reasonable inference and to
show how that inference is actually the sum of several shorter
inferential steps,” but it “would ignore common sense” to
prohibit all inferences “except those which cannot be broken
down any further.” United States v. An Article of Device, 

, 1263 (7th Cir. 1984). In short, trials are stories, not
10                                                 No. 18-3630

syllogisms. Old Chief v. United States, 

, 187–89
(1997). Accordingly, the court must try to focus on the most
persuasive story possible on the non-movant’s behalf when
asking whether a verdict in her favor would be reasonable or
could result only from irrational speculation.
    Before answering that question here, however, we try to
clear up some confusion that has so far affected the arguments
in this case.
     A. Scope of Review
    When the school district moved for summary judgment, it
addressed essentially the same arguments to both of Joll’s
claims for sex and age discrimination, arguing that both
claims were subject to the same standard and analysis and
that both failed to satisfy the familiar standard, stemming
from McDonnell Douglas Corp. v. Green, 

(1973). Joll
might have opposed the motion in part by pointing out that
the McDonnell Douglas test is only one means, not an end in
itself; that Title VII and ADEA causation standards are not al-
ways the same because of the availability of mixed-motive
claims under Title VII but not the ADEA; and that determin-
ing whether one factor was causal demands a different factual
analysis than determining whether a different factor was
causal.
    That is not how Joll argued the case in the district court.
Instead, her opposition picked up where the school district
left off—without distinguishing between Title VII and the
ADEA or between sex and age. She argued that evidence of
sex stereotyping and pretextual justification presented triable
issues under McDonnell Douglas, with one fleeting reference
to “mixed-motives.” See Gross v. FBL Fin. Servs., Inc., 557 U.S.
No. 18-3630                                                     11

167, 173–76 (2009) (Title VII allows “mixed-motive” claims,
but with limited remedies, while ADEA requires plaintiff to
prove “but-for” causation).
    On appeal, Joll has spent a good bit of energy arguing that
the district court “improperly applied the outdated McDon-
nell Douglas framework to this mixed motive case of sexual
[sic] discrimination.” This assertion is triply curious given
Joll’s exclusive reliance on that same framework at summary
judgment, its plaintiff-friendly operation in the ordinary case,
and our recent reaffirmation of its continuing vitality. See

. In any event, Joll’s waivers “in the truest
sense” here and in the district court, G & S Holdings LLC v.
Cont’l Cas. Co., 

, 538 (7th Cir. 2012), preclude our
review of the ADEA claim at all and of the Title VII claim un-
der a mixed-motive standard. We are left with her argument
on appeal that a reasonable jury could find Joll would have
been hired as an assistant cross-country coach if she were
male “and everything else had remained the same.” 

. On that score, we agree with Joll.
   B. Title VII
    In general, Title VII prohibits intentional discrimination in
employment on the basis of statutorily proscribed factors, in-
cluding sex. 42 U.S.C. § 2000e-2(a)(1); Ripberger v. Corizon, Inc.,

, 877 (7th Cir. 2014); Atanus v. Perry, 

,
671–72 (7th Cir. 2008). In this case, as in many discrimination
cases, “the sole question that matters” is causation: whether a
statutorily proscribed factor caused a failure to hire. Ortiz v.
Werner Enterprises, Inc., 

, 764–65 (7th Cir. 2016).
Direct as well as circumstantial evidence may support an in-
ference of causation, and thus intent.

Troupe v. May
Dep’t Stores Co., 

, 736 (7th Cir. 1994). “To clarify
12                                                 No. 18-3630

and to simplify” her task, a plaintiff may choose to enlist the
burden-shifting framework of McDonnell Douglas. Coleman v.
Donahoe, 

, 863 (7th Cir. 2012) (Wood, J., concur-
ring); accord, 

.
    “In order to make out a case of sex discrimination without
resorting to McDonnell Douglas, a plaintiff must provide either
direct or circumstantial evidence that supports an inference of
intentional discrimination.” Coffman v. Indianapolis Fire Dep’t,

, 563 (7th Cir. 2009). We have identified three
broad types of circumstantial evidence that will support an
inference of intentional discrimination: ambiguous or sugges-
tive comments or conduct; better treatment of people simi-
larly situated but for the protected characteristic; and dishon-
est employer justifications for disparate treatment. 

. Joll has offered some of each type. Taken together,
her evidence would permit a reasonable jury to infer “an over-
all likelihood of discrimination” that merits a trial, not sum-
mary judgment. 

.
    A jury could reasonably infer that athletic director Hofer
and principal Amones simply did not want to hire a woman
for either assistant coach position, and that they did what they
could, short of admitting it, to ensure that the only woman
applicant would not be hired. (Employers long ago “taught
their supervisory employees not to put discriminatory beliefs
or attitudes into words oral or written.” 

.) Looking at the evidence as a whole, as we must, a jury
No. 18-3630                                                                13

could reasonably find that Hofer and Amones bent the rules
in Arredondo and Kerezman’s favor and to Joll’s detriment.1
    We can begin with differential treatment in the selection
process itself. Joll had trouble simply securing an initial inter-
view for the girls’ job, suggesting a baseline reluctance to en-
tertain her candidacy. After the interview, her references were
contacted sooner than was ordinary (with important conse-
quences for her), while Arredondo’s were contacted later, as
was usual, and Kerezman’s not at all—suggesting a higher
baseline of scrutiny for Joll.




    1  We agree with the dissenting opinion that discrete adverse employ-
ment actions give rise to discrete claims that must be evaluated individu-
ally. Post at 32. The jury accordingly may find the school district liable for
one, both, or neither of the two failures to hire alleged here. Nevertheless,
the closeness in time and the similarity of the positions permit, indeed in-
vite, a useful comparison between the two hiring processes in determining
what inferences the evidence reasonably permits.
     As for the identity of the decisionmaker, see post at 32 n.2, the record
permits the reasonable inference that Hofer and Amones made both hiring
decisions, with Hofer as the more active partner. Though there is a bit of
conflicting evidence on this point, some of which is pointed out below,
according to the school district the hiring decision for the girls’ position
“was made by” Nellessen, Hofer, and Amones; the hiring decision for the
boys’ position “was made by” Hofer and Crague. While Amones filled in
for Hofer at the interview for the girls’ position, it was usually Amones’s
more limited role to approve and forward Hofer’s hiring recommenda-
tions to the school board. With respect to the head coach’s usual role, Ho-
fer’s testimony that a head coach had “never” disagreed with his hiring
recommendation suggests the decision was effectively Hofer’s, subject
generally to Amones’s approval and, exceptionally in this case, his some-
what greater involvement in hiring the girls’ coach.
14                                                  No. 18-3630

    Evidence about the interviews does not require, but cer-
tainly permits, an inference that the decision-makers were in-
dulging sex-role stereotypes as they approached the deci-
sions. As Joll told the school district at the time, she resigned
the middle school position in 2013 precisely from a desire not
to give the program less than its due. Yet her 2014 declaration
of full commitment to coaching was met with skepticism and
discounted. Job interviewers are of course entitled to probe
applications in relevant ways, and they are not required to
take an applicant’s answers at face value. Those questions by
themselves would not necessarily show bias. But the telling
twist in this case is that, although family matters had recently
prompted his resignation from a similar position, Arre-
dondo’s commitment was not questioned. He “talked shop,”
not kids, with the interviewers.
    Basing an employment decision on an employer’s notions
of how women do or ought to behave—the employer’s sex-
role stereotypes—is discrimination “because of sex.” See
Hively v. Ivy Tech Comm. Coll. of Ind., 

, 345–47 (7th
Cir. 2017) (en banc), discussing among others Price Waterhouse
v. Hopkins, 

(1989). The notion that a woman is or
ought to be dedicated to family above work is one long-stand-
ing such stereotype. Nevada Dep’t of Human Res. v. Hibbs, 

, 731, 736 (2003) (holding FMLA family leave provi-
sion validly abrogated state sovereign immunity: “pervasive
sex-role stereotype that caring for family members is
women’s work,” in turn “foster[ing] employers’ stereotypical
views about women’s commitment to work and their value as
employees”).
No. 18-3630                                                    15

    A jury could find that the interviewers’ questions, at least
when they were asked only of Joll and not of a similarly situ-
ated male applicant, reflected such stereotyping. See Lust v.
Sealy, Inc., 

, 583 (7th Cir. 2004) (sustaining jury
verdict for plaintiff: “Most important, Penters admitted that
he didn’t consider recommending Lust for the Chicago posi-
tion because she had children and he didn’t think she’d want
to relocate her family, though she hadn’t told him that.”);
Bruno v. City of Crown Point, 

, 362 (7th Cir. 1991)
(reversing jury verdict on other grounds, but acknowledging:
“The fact that Pyle asked only Bruno family-oriented ques-
tions reveals that those questions were based on sex stereo-
types—namely, that females are the primary care providers
for children … .”);

(Easterbrook, J., dissenting) (“pa-
ternalistic questions” about husband’s views of job applica-
tion and plans to have additional children supported infer-
ence of sex discrimination so that verdict should have been
upheld).
    As noted, Joll’s references were contacted more quickly
than usual. A jury could find that this deviation from stand-
ard procedures also lends support to her claims. See, e.g.,
Baines v. Walgreen Co., 

, 664 (7th Cir. 2017) (revers-
ing summary judgment); 

(reversing
summary judgment; explaining that selective enforcement of
company policy can establish pretext); Gordon v. United Air-
lines, Inc., 

, 891–92 (7th Cir. 2001) (same); see also
United States ex rel. Hamrick v. GlaxoSmithKline LLC, 

, 22 (1st Cir. 2016) (“[D]eviations from standard procedures
can give rise to an inference of pretext.”) (quotations omitted);
Giacoletto v. Amax Zinc Co., 

, 427 (7th Cir. 1992)
(affirming jury verdict for age-discrimination plaintiff where
16                                                    No. 18-3630

employer fired him without following its standard proce-
dures, including helping employees overcome deficiencies),
citing among others Johnson v. Lehman, 

, 922 (D.C.
Cir. 1982) (“departure from internal hiring procedures is a fac-
tor that the trier of fact may deem probative”).
    With respect to the references, a jury could find further ev-
idence that the school district strained to reject Joll’s applica-
tions based on sex-role stereotypes. The decision-makers dis-
counted the two long and positive narratives from a runner
coached by Joll and from Crague, the head coach of the high
school boys’ team (whose opinion, recall, was that Joll had
“always done a great job supporting and complimenting [sic]
the head coach”), in favor of middle school principal Polite’s
isolated comment that Joll has a “dominate personality.”
    A jury could think that Polite’s phrase was an odd way to
say “insubordinate,” if that was what was understood. In-
deed, at the risk of indulging in some other stereotypes, it’s
easy to think that a dominant personality is often viewed as a
virtue of at least male athletic coaches. The point is not that the
school district hired “dominant” male coaches but not “dom-
inant” female coaches. The point is that a jury could draw on
its experience to conclude that the same behavior may be la-
beled “assertive” in a man and “aggressive” in a woman. In
other words, a reasonable jury could conclude that Joll was
being penalized for transgressing the age-old stereotype that
women are or ought to be submissive. See 

(plurality) (“[A]n employer who acts on the basis of a be-
lief that a woman cannot be aggressive, or that she must not
be, has acted on the basis of gender.”); 29 C.F.R. § 1604.2
No. 18-3630                                                                 17

(“Such stereotypes include, for example, … that women are
less capable of aggressive salesmanship.”).2
     A jury could infer further that the signal was received im-
mediately by principal Amones and athletic director Hofer
(“The 3rd one I got.”) and operated so powerfully that it over-
rode the contrary opinion of coach Crague (who would have
been the one to deal with any insubordination or “domina-
tion” by Joll, at least on the boys’ team) and even the more
complete and balanced opinion of Polite, which was that if Joll
“already has a good relationship with the head coach,” as she
did with Crague, “it would probably be fine.” Of course, we
agree with the dissenting opinion that a jury could reasonably
conclude that what we have characterized as evidence of sex-
stereotyping was in fact entirely innocent conduct. Some am-
biguity inheres in all employer remarks that fall short of ad-
mitting discrimination while still supporting an inference of
it. This ambiguity is the very reason the jury must be called.
Sheehan v. Donlen Corp., 

, 1044–45 (7th Cir. 1999)
(affirming plaintiff’s verdict for pregnancy discrimination;
jury was not required to accept proffered innocent interpreta-
tion of remarks that could also be understood as reflecting
sex-role stereotypes).
    The evidence reviewed so far is sufficient to warrant a
trial, but there is more support from comparing the hiring cri-
teria for the boys’ and girls’ team jobs. The defendant’s stated


    2  In the district court and on appeal, the parties have argued a good
deal about Joll’s proffered expert on implicit bias, but mutatis mutandis “[i]t
takes no special training to discern sex stereotyping in a description of an
aggressive female employee as requiring ‘a course at charm school.’” Hop-

(plurality).
18                                                  No. 18-3630

reasons for hiring for these nearly identical jobs were incon-
sistent with each other. What was consistent was that, in each
case, the stated reasons and criteria favored the male appli-
cant. From the shifting criteria that favored the male appli-
cants, a jury could infer that the different stated reasons were
not honest and were pretexts for sex discrimination.
    When the male applicant for the girls’ job (Arredondo)
had more recent experience than Joll professionally coaching
high school cross country (though much less experience over-
all), the school district said that was the controlling factor.
When the male applicant (Kerezman) was “in the building”
during the school day with the opportunity to build “rapport
with the boys,” that became the controlling factor instead.
Joll’s long presence in the middle school and in the school sys-
tem with opportunity to build “rapport with the girls” was
not considered in her favor vis-à-vis Arredondo. Her more ex-
tensive and apparently more recent professional experience
coaching high schoolers was not considered in her favor vis-
à-vis Kerezman. Further, Polite’s opinion, given such weight
as to Joll’s “dominate personality,” was apparently given no
weight at all as to the transferability of Joll’s middle school
experience to the high school setting. These inconsistent
choices of criteria for nearly identical jobs are not necessarily
decisive in Joll’s favor, but they could certainly support a
jury’s inference that the real reasons for the decisions were
based on sex.
   Employment discrimination law has long recognized that
an employer’s dishonest explanation of a decision can sup-
port an inference that its real reason was unlawful. If the jury
does “not believe the employer’s explanation for its decisions,
No. 18-3630                                                     19

it may infer that the employer is trying to cover up … discrim-
ination,” Gehring v. Case Corp., 

, 343 (7th Cir. 1994),
“particularly if disbelief is accompanied by a suspicion of
mendacity.” St. Mary’s Honor Ctr. v. Hicks, 

, 511
(1993). See generally McDonnell 

–05
(suggesting types of evidence that would support finding of
pretext). In other words, the jury may infer that “the real rea-
son was a forbidden one.” Testerman v. EDS Tech. Prods. Corp.,

, 303 (7th Cir. 1996); see also 

.
Again, however, we emphasize that Joll does not rest her case
on pretext alone. A plaintiff is required to show pretext when
her only other evidence of discrimination is the prima facie
McDonnell Douglas showing. Anderson v. Baxter Healthcare
Corp., 

, 1123–24 (7th Cir. 1994). Where the plaintiff
does not rely solely on McDonnell Douglas, she may survive
summary judgment even without evidence that the em-
ployer’s explanation is dishonest. Venters v. City of Delphi, 

, 974 (7th Cir. 1997).
     A further indicator of less than perfect forthrightness is
Hofer’s fluctuating accounts of his involvement in hiring the
girls’ coach. See, e.g., Stalter v. Wal-Mart Stores, Inc., 

, 291 (7th Cir. 1999) (reversing summary judgment for em-
ployer); E.E.O.C. v. C.G. Schmidt, Inc., 

, 869
(E.D. Wis. 2009) (denying summary judgment: “CGS has
shifted their story on who was involved in the decision to lay
off Jackson, providing some evidence, albeit not very strong
evidence on its own, that retaliation occurred.”). In his depo-
sition, Hofer testified that “Dr. Amones and the head coach
Adam Nellessen made that decision.” In his affidavit, he tes-
tified: “The decision as to which candidate to recommend for
the position was made by all three of us,” that is, Hofer,
Amones, and Nellessen.
20                                                    No. 18-3630

    Another indicator of pretext are the shifts in relevant qual-
ifications outlined above, with their consistent operation in
the men’s favor, both as between the girls’ and boys’ positions
and as between the event and the litigation. See, e.g., 

(“Stalter presents one more piece of persuasive
evidence, and that is that Wal-Mart changed its story between
the time of the state administrative proceeding and the federal
action.”), citing Emmel v. Coca-Cola Bottling Co. of Chi., 

, 634 (7th Cir. 1996), and Perfetti v. First Nat’l Bank of Chi.,

, 456 (7th Cir. 1991). We have said time and again
(in more than one hundred reported opinions, by our count)
that we are not a super-personnel department that will sub-
stitute our criteria for an employer’s for hiring, promoting, or
disciplining employees. See, e.g., Grant v. Trustees of Indiana
Univ., 

, 570 (7th Cir. 2017); Dale v. Chicago Tribune
Co., 

, 464 (7th Cir. 1986). That hesitation and the
often subjective character of many employment decisions can
make it difficult for a lone plaintiff to show that a single em-
ployment decision was made for unlawful reasons. But incon-
sistent explanations, especially from the same decision-mak-
ers about similar decisions near the same time, can support an
inference of unlawful intent.
    We conclude with a few observations on the district
court’s analysis. First, as a general matter it cannot be empha-
sized too strongly: “Evidence must be considered as a whole,
rather than asking whether any particular piece of evidence
proves the case by itself … . Evidence is evidence. Relevant
evidence must be considered and irrelevant evidence disre-
garded,” and the case must be assessed for an “overall likeli-
hood of discrimination.” 

, 763.
No. 18-3630                                                    21

    More specifically, the district court quoted our opinion in
Millbrook v. IBP, Inc., 

, 1180–81 (7th Cir. 2002):
“where an employer’s proffered non-discriminatory reason
for its employment decision is that it selected the most quali-
fied candidate, evidence of the applicants’ competing qualifi-
cations does not constitute evidence of pretext unless … no
reasonable person … could have chosen the candidate se-
lected over the plaintiff for the job in question” (internal quo-
tation marks omitted).
    We have tried to make clear that the Millbrook rule applies
only where “the plaintiff relies exclusively on evidence of the
applicants’ comparative qualifications.” David v. Caterpillar,
Inc., 

, 863 (7th Cir. 2003) (emphasis added); see
also, e.g., Robertson v. State of Wis. Dep’t of Health Servs., 

, 381 (7th Cir. 2020) (“the plaintiff’s qualifications
alone do not establish evidence of pretext”). This is not such
a case. Joll has presented much more evidence beyond the
comparison of qualifications. And again, Joll would be re-
quired to show pretext only if she were proceeding under
McDonnell Douglas and had no evidence of discrimination be-
yond her prima facie case. 

–24.
    Second, in response to Joll’s evidence of the shifting crite-
ria for the two hiring decisions, the district court cited two
non-precedential orders of this court for the proposition that
no inference of discrimination arises where the successful ap-
plicant turns out to have self-evidently better or broader qual-
ifications that merely happened not to be listed in the initial
job solicitation. Tai v. Shinseki, 325 F. App’x 444, 447 (7th Cir.
2009) (chosen applicant had “a more diverse range of experi-
ences and showed more enthusiasm”); Currie v. Paper Convert-
ing Mach. Co., 202 F. App’x 120, 122 (7th Cir. 2006) (“Ropson
22                                                   No. 18-3630

claimed in his application that he could operate both a jig bore
and a boring bar.”) These non-precedential orders are—well,
non-precedential. On the relevant point of comparative qual-
ifications, however, both orders quoted Mlynczak v. Bodman,

, 1059 (7th Cir. 2006), which in turn quoted
Millbrook. In Millbrook, as noted, the comparative qualifica-
tions were the plaintiff’s only evidence of unlawful motive in
addition to the prima facie case. Again, that does not describe
the facts of this case; evidence of Joll’s stronger qualifications
is only part of her evidence of sex discrimination. Neither of
the cited non-precedential orders involved inconsistent
changes in hiring criteria that had the effect of favoring one
sex or race over another across multiple decisions.
    Finally, the district court also erred to the extent it dis-
missed Joll’s evidence of sex-role stereotyping as mere “stray
remarks” that were not probative of unlawful intent in the hir-
ing decision. The “stray remarks” rationale can be applied
only to remarks that are not part of the decision-making pro-
cess itself. See, e.g., Overly v. KeyBank, N.A., 

, 865
(7th Cir. 2011) (affirming summary judgment where highly
offensive remark was made after plaintiff resigned); Davis v.
Time Warner Cable of Southeastern Wis., L.P., 

, 672–
73 (7th Cir. 2011); see also Hunt v. City of Markham, 

, 652–53 (7th Cir. 2000) (warning against over-reading
“stray-remarks” cases: “All that these cases hold—all that they
could hold and still make any sense—is that the fact that
someone who is not involved in the employment decision of
which the plaintiff complains expressed discriminatory feel-
ings is not evidence that the decision had a discriminatory mo-
tivation.”). But Joll’s evidence comes from the decision-mak-
ers during the decision process.
No. 18-3630                                                  23

    In considering a defendant’s attempt to neuter a sexist or
racist comment as a stray remark, a court deciding a summary
judgment motion must keep in mind its duty to consider the
evidence as a whole and to do so in the light reasonably most
favorable to the non-moving party. A remark or action by a
decision-maker reflecting unlawful animus may be evidence
of his or her attitudes more generally. See 

(“Emmel does not rely on these remarks alone to prove
anything. The remarks are evidence, which together with the
other evidence in this case could lead a jury to conclude …
that the company engaged in unlawful discrimination… .
[T]he statements in question are more than just stray com-
ments … . They are from the top policymakers in the company
… . They directly address the policy at issue … .”), discussing

(O’Connor, J., concurring in the judg-
ment); cf. Blasdel v. Northwestern Univ., 

, 820 (7th
Cir. 2012) (“same actor” inference permits but does not re-
quire inference that attitudes of person who hired plaintiff, for
example, would not have changed by time same person fired
plaintiff). Similarly, the district court strayed outside the
bounds of summary judgment practice by accepting that the
“dominate personality” comment spoke simply and inno-
cently to the employer’s nondiscriminatory interest in “per-
sonality fit” and could not support an inference of discrimi-
nation. See Hop

(plurality) (“[I]f an em-
ployee’s flawed ‘interpersonal skills’ can be corrected by a
soft-hued suit or a new shade of lipstick, perhaps it is the em-
ployee’s sex and not her interpersonal skills that has drawn
the criticism.”).
24                                                   No. 18-3630

                           Conclusion
    We recognize, of course, that there are ways to tell the
story of the school district’s hiring process and decision that
are entirely innocent, involving no unlawful discrimination.
The dissenting opinion illustrates several ways the defense
could argue this case to a jury. But because there is at least one
reasonable way to tell the story in favor of Joll’s claim of sex
discrimination, a jury rather than appellate judges must
choose among them.
    The grant of summary judgment for the defendant on
Joll’s age discrimination claim is AFFIRMED. The grant of
summary judgment on the sex discrimination claim is
REVERSED and the case is REMANDED for trial on that
claim.
18-3630                                                           25

    RIPPLE, Circuit Judge, dissenting. In response to a motion
for summary judgment, “the non-moving party is required
to marshal and present the court with the evidence she con-
tends will prove her case.” Goodman v. Nat’l Sec. Agency, Inc.,

, 654 (7th Cir. 2010). Here, the question is wheth-
er Ms. Joll came forth with sufficient evidence from which a
reasonable jury could conclude that VCS failed to hire her
for the girls’ assistant cross-country position or for the boys’
assistant cross-country position because she was a woman.
Ortiz v. Werner Enters., Inc., 

, 764 (7th Cir. 2016).
    On an appeal from the grant of summary judgment, we
view the record as a whole “and all justifiable inferences
drawn from it in the light most favorable to the party against
whom judgment was entered,” here Ms. Joll. Cheek v. W. & S.
Life Ins. Co., 

, 500 (7th Cir. 1994); see also Life Plans,
Inc. v. Sec. Life of Denver Ins. Co., 

, 349 (7th Cir.
2015). “[T]he court should give credence to the evidence fa-
voring the nonmovant as well as that evidence supporting
the moving party that is uncontradicted and unimpeached,
at least to the extent that the evidence comes from disinter-
ested witnesses.” Reeves v. Sanderson Plumbing Prods. Inc., 

, 151 (2000) (internal quotation marks omitted).
However, we have not “interpret[ed] the quoted language so
broadly as to require a court to ignore the uncontroverted
testimony of company employees or to conclude, where a
proffered reason is established through such testimony, that
it is necessarily pretextual.” Traylor v. Brown, 

,
791 (7th Cir. 2002).
   Based on a review of the entire record, I do not believe
Ms. Joll has come forward with evidence from which a rea-
sonable jury could conclude that, had she been male, she
26                                                  No. 18-3630

would have been awarded either of the positions for which
she applied. I therefore respectfully dissent.
                                I
   Ms. Joll has coached cross country and track for many
years and has a long tenure of coaching in Valparaiso. From
1991 to 1995, she was the assistant girls’ track coach at Val-
paraiso High School (“VHS”). Beginning in 1994 and contin-
uing through 2004, she was the assistant men’s and women’s
cross-country coach at Valparaiso University. During the
same relative time frame (1995 to 2005), she co-coached track
and field at Thomas Jefferson Middle School in Valparaiso;
and from 2005 to 2013, she co-coached cross country at
Thomas Jefferson Middle School. Ms. Joll described her
coaching responsibilities at Thomas Jefferson Middle School
accordingly: “The way the middle school program works, …
we were both equal coaches and it’s not girls and boys. It’s a
combined program at the middle school.” R.24-1 at 4 (Joll
Dep. 12).
    In 2013, Ms. Joll resigned her coaching position so that
she could travel with and watch both of her daughters com-
pete in cross country and track. At that point, one daughter
was running at Indiana State University, and one was run-
ning at VHS. Ms. Joll felt that the other coach at Thomas Jef-
ferson Middle School “needed somebody who was more
dedicated at that point in time to coaching.”

(Joll Dep.
16). In her resignation letter, she told the principal of Thom-
as Jefferson Middle School, Jim Polite, that she felt that
“couldn’t devote … [a] sufficient amount of time to … ade-
quately do the job.”
 

17).
No. 18-3630                                                  27

A. Girls’ Cross-Country Position
    In 2014, Ms. Joll’s older daughter stopped running at In-
diana State, and Ms. Joll wanted to coach again. On approx-
imately June 17 of that year, Ms. Joll applied for a position as
an assistant coach with the VHS girls’ cross-country team.
See
 id. at 7 
(Joll Dep. 22–23).
    In 2014, VHS did not have a “written hiring procedure
for coaches.” R.29-4 at 2 (Hofer Dep. 9). Herb Hofer, who
was Athletic Director of VHS at the time, explained he had a
“typical … process” that he “use[d] as an AD” for the hiring
of assistant coaches.
 

(Hofer Dep. 12). His “normal
process for hiring assistant coaches” began when the posi-
tion was posted on the school system’s website.
 Id. Resumes
would 
come either to him or to the VHS Principal, Dr. Reid
Amones. Mr. Hofer and the head coach then would review
the candidates and decide whom to interview. Mr. Hofer
and the head coach would conduct the interviews and come
to a consensus as to the best candidate for the position. See

id. Mr. Hofer 
contacted a candidate’s references after he and
the head coach had made a hiring decision, but prior to
bringing the matter before the school board.
 Id. at 10 
(Hofer
Dep. 32).
    It is undisputed that Mr. Hofer, the person who usually
oversaw the hiring process for assistant coaches, “was hav-
ing knee surgery” at the time the interviews and delibera-
tions for the girls’ assistant cross-country coaching position
were occurring.
 

(Hofer Dep. 12);
 id. at 14–15 
(Hofer
Dep. 36–37). Indeed, Ms. Joll emailed Mr. Hofer “towards
the end of June” to determine when interviews would be
held, and Mr. Hofer responded that “they were going to be
doing it that next week, but he had knee surgery and so he
28                                                No. 18-3630

was recovering from that.” R.24-1 at 7 (Joll Dep. 23). Instead,
Dr. Amones and the girls’ cross-country head coach, Adam
Nellessen, conducted the review of candidates, interviews,
and follow-up. The following week, Coach Nellessen con-
tacted Ms. Joll to ascertain if she would be available for an
interview on July 5th.
    Ms. Joll’s interview lasted “[m]ore than half [of an] hour,
but less than an hour.”
 

24). During her inter-
view, she was asked “[w]hy [she] was … applying when
[she] had resigned? What had changed?”
 

25).
Ms. Joll explained that one of her daughters was no longer
running at the collegiate level and so she “didn’t feel pulled
to go and see her compete.”
 Id. at 8 
(Joll Dep. 26). She also
was asked how being the parent of one of the athletes might
affect her ability to coach the rest of the team. See
 
id. (Joll
Dep. 
28–29).
    Following the interview, Ms. Joll was asked to provide
references, and Dr. Amones contacted those references.
Ms. Joll received favorable recommendations from several
individuals, including Jim Polite, the principal of Thomas
Jefferson Middle School, where Ms. Joll taught and where
she had coached for eighteen years. In emailing Polite,
Dr. Amones specifically inquired whether Ms. Joll had any
“past roles serving as an assistant, or other times she has not
been the one making the final decision.” R.29-6 at 5. Polite
responded that “[t]his area would be my only concern. She
has a dominate [sic] personality. If she … already has a good
relationship with the head coach, it would probably be fine.
If not, I could envision some issues arising.”
 Id.
    There was 
one other applicant interviewed for this posi-
tion: John Arredondo. From 2008 to 2014, Arredondo
No. 18-3630                                                                29

coached in the boys’ cross-country and girls’ track programs
at Portage High School.1 From 2008 to 2013, he also coached
boys’ cross country at Portage High School. He testified that
he had resigned his positions because he was going through
a divorce; the record does not reveal whether, in his resigna-
tion letter, Arredondo provided a reason for his departure.
     In the summer of 2014, Arredondo began volunteering
as a girls’ cross-country coach with VHS. When a paid posi-
tion came open, Coach Nellessen suggested to Arredondo
that he should apply. Arredondo testified that, during his
interview, he was asked about “[c]oaching experience, what
[his] coaching philosophy is, a lot of shop talk.” R.29-3 at 15
(Arredondo Dep. 37). Arredondo also was asked about
whether commuting from Portage to VHS would be a prob-
lem.
 
Id. at 16 
(Arredondo Dep. 38). There is no evidence in
the record—or at least none to which Ms. Joll has invited our
attention—that Arredondo gave Portage High School a spe-
cific reason, much less a specific family-related reason, for
resigning his coaching position. There also is no evidence in
the record that Arredondo made Coach Nellessen or
Dr. Amones aware of the reason he had resigned his position
at Portage. During Arredondo’s deposition, Ms. Joll’s coun-

1  It is unclear whether Arredondo had head coaching experience.
Ms. Joll states in her brief that he was the assistant in these positions, but
that he served one year as head coach of girls’ track in 2010. See Appel-
lant’s Br. 10. In support of this statement, Ms. Joll cites to Arredondo’s
deposition, see R.29-3 at 2 (Arredondo Dep. 5), and to his resume. How-
ever, Arredondo’s resume is not part of the record. In his deposition,
Arredondo simply states that he “coach[ed]” girls’ track and field and
boys’ cross country, but he does not specify (at least on the page refer-
enced) whether he served in a head coaching role.
 Id.
30                                                 No. 
18-3630

sel did not ask Arredondo whether he had informed
Coach Nellessen or Dr. Amones of the reason that he had
resigned his position. Moreover, Ms. Joll’s counsel did not
ask Arredondo whether during his interview he was ques-
tioned about familial obligations that might interfere coach-
ing responsibilities. Ms. Joll’s counsel simply asked Arre-
dondo to characterize the “type of questions” he was asked
during his interview. R.29-3 at 15 (Arredondo Dep. 37). Her
counsel’s only follow-up question regarding what Arredon-
do was asked during his interview was: “Did they ever ask
you if the commute would be an issue from Portage to Val-
paraiso?”
 
Id. at 16 
(Arredondo Dep. 38).
    The decision was made by Coach Nellessen, Dr. Amones,
and Mr. Hofer to hire Arredondo. In his call explaining the
hiring decision to Ms. Joll, Mr. Hofer stated that Arredondo
had “more current experience working with high school age
athletes.” R.29-6 at 7. In his deposition and affidavit, Mr. Ho-
fer testified that Polite’s statement also caused them pause as
to whether Ms. Joll could serve in the capacity of an assis-
tant, as opposed to head coach. See R.29-1 at 4 (Hofer Aff. 2);
R.24-2 at 7 (Hofer Dep. 31).
    After the decision was made to hire Arredondo,
Mr. Hofer checked Arredondo’s references in preparation
for proposing his appointment to the board. See R.29-4 at 14
(Hofer Dep. 36).
B. Boys’ Cross-Country Position
   After Ms. Joll was unsuccessful in obtaining the girls’
coaching position, she applied for a position as the assistant
boys’ cross-country coach at VHS. She was interviewed for
the position by the head coach of the boys’ cross-country
No. 18-3630                                                   31

team, Aaron Crague, as well as Mr. Hofer. Ms. Joll again was
asked about the reason she had resigned the position of
cross-country coach at Thomas Jefferson Middle School. See
R.24-1 at 13 (Joll Dep. 70). The other candidate for the posi-
tion was Ben Kerezman. Kerezman was a teacher at VHS,
had volunteered with the boys’ cross-country team and had
served as an assistant coach for the VHS boys’ track team.
Mr. Hofer and Coach Crague collectively decided to hire Ke-
rezman for the assistant coaching position.
    Mr. Hofer informed Ms. Joll that she did not receive the
job because Kerezman had better rapport with the boys on
the team. See
 
id. (Joll Dep. 
72). In his affidavit, Mr. Hofer ex-
plained that this rapport stemmed from Kerezman’s volun-
teering with the cross-country team and from his serving as
a high school teacher in the same building where the team
members went to school. R.29-1 at 5 (Hofer Aff. 3). Accord-
ing to Mr. Hofer, he and Coach Crague believed that Kerez-
man’s service on the faculty was “an advantage to building,
developing and maintaining the necessary relationship with
high school athletes.”
 Id.
    Unlike with 
Arredondo, Mr. Hofer did not check any ref-
erences for Kerezman. R.29-4 at 17 (Hofer Dep. 49).
Mr. Hofer explained that Kerezman’s references were not
checked because “he was a teacher at the high school,” he
had coached with the cross-country team on a voluntary ba-
sis, had been “a paid assistant for our track [program],” and
Mr. Hofer “personally kn[e]w him as a coach.”
 Id.
                               II
   Ms. 
Joll contends, and my colleagues conclude, that she
has come forward with sufficient evidence from which a jury
32                                                          No. 18-3630

reasonably could conclude that her gender was the reason
that she did not get either position at VHS. In doing so, they
treat her unsuccessful applications for the girls’
cross-country position and for the boys’ cross-country posi-
tion as a single determination. A “refusal to hire,” however,
is a “[d]iscrete act,” and “[e]ach incident of discrimination …
constitutes a separate actionable unlawful employment prac-
tice.” Nat’l R.R. Passenger Corp., 
536 U.S. 101
, 114 (2002) (in-
ternal quotation marks omitted). As such, we must evaluate
independently each of Ms. Joll’s failure-to-hire claims to de-
termine whether a statutory violation has occurred. Ford v.
Marion Cty. Sheriff’s Office, 
942 F.3d 839
, 858–59 (7th Cir.
2019) (evaluating each of the plaintiff’s four fail-
ure-to-promote claims, which involved positions open be-
tween March 2016 and February 2017, independently).2
A. Girls’ Cross-Country Position
    My colleagues conclude that, taken collectively, the evi-
dence of sex-stereotyping, deviation from established proce-
dure, changing rationales, and disparity in qualifications
raise an inference of discriminatory intent. I evaluate each
category of evidence individually and then consider them
collectively.




2 Even if it were not the established rule that we evaluate each failure to
hire independently, here the facts would warrant such an approach. Alt-
hough the positions at issue were similar, there were a different set of
decisionmakers for each position, there were a different set of candidates
for each position, and, given Mr. Hofer’s knee surgery, there also was a
different process that governed each decision.
No. 18-3630                                                   33

                               1.
    In Ms. Joll’s view, the strongest argument that her gender
played a role in the decision-making process for the girls’
cross-country position was VCS’s reliance on sex-
stereotypes. According to my colleagues, there is evidence of
sex-stereotyping both in the way the interviews were con-
ducted and in Polite’s use of the term “dominant” to de-
scribe Ms. Joll.
    Turning first to the interview process, my colleagues note
that the questions asked of Ms. Joll “would not necessarily
show bias.” Maj. at 14. The problem as they see it is that,
“although family matters had recently prompted his resigna-
tion from a similar position, Arredondo’s commitment was
not questioned. He ‘talked shop,’ not kids, with the inter-
viewers.”
 Id. However, there 
is no evidence in the record
that Coach Nellessen or Dr. Amones knew of the reason for
Arredondo’s resignation. Arredondo’s deposition does not
reveal what he said in his resignation letter, if that letter was
provided to Coach Nellessen or Dr. Amones, or if Arredon-
do informed Coach Nellessen or Dr. Amones about the rea-
son for his resignation. By contrast, Coach Nellessen and
Dr. Amones were aware of Ms. Joll’s specific reason for re-
signing: she wanted to be able to attend her daughters’
meets. Any focus on “kids” in the interview was prompted
by Ms. Joll’s resignation letter and does not raise an infer-
ence of discrimination.
    My colleagues also agree with Ms. Joll that Polite’s email,
describing Ms. Joll as having a dominant personality, consti-
tuted sex-stereotyping that infected the decision-making
process. There is no question Polite’s evaluation of Ms. Joll
was at least a factor in the decision-making process. Given
34                                                 No. 18-3630

that Polite had directly supervised Ms. Joll at Thomas Jeffer-
son Middle School (where Ms. Joll worked as a teacher and
had coached for eighteen years), it is not surprising that Po-
lite’s comments carried weight with Coach Nellessen and
Dr. Amones not matched by her other recommenders.
    Polite’s comments, however, do not fall into the category
of “sex-stereotyping,” at least as that term traditionally has
been understood. In Price Waterhouse v. Hopkins, 
490 U.S. 228
(1989), the Court held that the practice of gender stereotyp-
ing falls within Title VII’s prohibition against sex discrimina-
tion. In Hopkins, a female employee sued Price Waterhouse
for discrimination when she was not elected as a partner. In
assessing comments made by partners who evaluated Hop-
kins, the Court noted that there
       were clear signs … that some of the partners
       reacted negatively to Hopkins’ personality be-
       cause she was a woman. One partner described
       her as “macho”; another suggested that she
       “overcompensated for being a woman”; a third
       advised her to take “a course at charm school.”
       Several partners criticized her use of profanity;
       in response, one partner suggested that those
       partners objected to her swearing only “be-
       cause it’s a lady using foul language.” Another
       supporter explained that Hopkins “ha[d] ma-
       tured from a tough-talking somewhat mascu-
       line hard-nosed mgr to an authoritative, for-
       midable, but much more appealing lady ptr
       candidate.” But it was the man who … bore re-
       sponsibility for explaining to Hopkins the rea-
       sons for the Policy Board’s decision to place
No. 18-3630                                                 35

      her candidacy on hold who delivered the coup
      de grace: in order to improve her chances for
      partnership, Thomas Beyer advised, Hopkins
      should “walk more femininely, talk more fem-
      ininely, dress more femininely, wear make-up,
      have her hair styled, and wear jewelry.”

Id. at 235 
(internal citations omitted). The Court observed
that “[i]t [took] no special training to discern sex stereotyp-
ing in a description of an aggressive female employee as re-
quiring ‘a course at charm school,’” nor “to know that, if an
employee’s flawed ‘interpersonal skills’ can be corrected by
a soft-hued suit or a new shade of lipstick, perhaps it is the
employee’s sex and not her interpersonal skills that has
drawn the criticism.”
 Id. at 256.
    As Hopkins and our later cases make clear, it is an indi-
vidual’s failure to meet social preconceptions about what is
“male” or “female” that raises the inference of discrimina-
tion. There must be, therefore, some evidence of “preconcep-
tion.” In Hopkins, as already described, the Court relied on
overt references to male and female stereotypes. Gen-
der-neutral remarks also might reflect sex-stereotyping, but
there has to be some evidence that those gender-neutral
comments reflect preconceptions of how men and women
should conduct themselves. For instance, in Bellaver v.
Quanex, 
200 F.3d 485 
(7th Cir. 2000), a woman employee was
disciplined and eventually let go due to her lack of interper-
sonal skills. We observed that,
      [w]hile it is permissible to evaluate an employ-
      ee’s interpersonal skills when those skills are
      relevant to the job, evaluations may demon-
      strate discriminatory intent when employees
36                                                 No. 18-3630

       are evaluated on how their interpersonal skills
       match stereotyped, unequal ideas of how men
       and women should behave. …
           As in Price Waterhouse, the evidence sug-
       gests the employer here may have relied on
       impermissible stereotypes of how women
       should behave. Bellaver’s evaluations are
       marred only by the repeated references to her
       interpersonal skills, but these same types of de-
       ficiencies seemed to be tolerated in male em-
       ployees.

Id. at 492.
    Here, there is no evidence to suggest that Polite’s assess-
ment of Ms. Joll’s personality was based on a preconceived
notion of how women should behave. Polite was Ms. Joll’s
direct supervisor at Thomas Jefferson Middle School. He had
firsthand knowledge of how Ms. Joll interacted with others
and, specifically, with a supervisor. He provided his assess-
ment in response to a direct question of how Ms. Joll might
handle being in the role of an assistant, as opposed to being
in charge of a sports program. Thus, his assessment, based
on his experience supervising Ms. Joll, spoke directly to a
quality that would bear on her ability to perform the job for
which she was being considered. Polite’s use of a linguisti-
cally accurate word to describe a job-related trait of Ms. Joll
does not constitute sex-stereotyping. Indeed, it is difficult to
conceive of how Polite might have voiced his legitimate con-
cerns without running afoul of the majority’s conception of
sex-stereotyping. The majority, it would seem, would find it
equally troubling if Polite had described Ms. Joll as used to
No. 18-3630                                                   37

being in charge, accustomed to running a program her own
way, or not used to being in a subordinate position.
    There also is no evidence in the record that this quality—
a need or desire to be “in charge”—was tolerated in the male
candidate for the job. My colleagues explain that “it’s easy to
think that a dominant personality is often viewed as a virtue
of at least male athletic coaches.” Maj. at 16. But there is no
evidence in the record that Coach Nellessen, Dr. Amones, or
Mr. Hofer believed this to be the case—especially with re-
spect to professional relationships with other members of
the coaching staff. Inferences based on speculation and con-
jecture cannot be used to defeat a summary judgment mo-
tion. See, e.g., Lavite v. Dunstan, 
932 F.3d 1020
, 1029 (7th Cir.
2019). Additionally, the record does not support (or even
suggest) that Arredondo, the successful candidate, had a
“dominant” personality. Indeed, it appears that the wealth
of his experience was as an assistant coach.3 Because he had
not spent significant time in charge of his own program,
there was no need to question whether he easily could step
into an “assistant” position.
   At bottom, the comment from Polite is not the kind of
sex-stereotyping condemned in Price Waterhouse where an
employment decision was based on the plaintiff’s failure to
conform to the partners’ view of how a woman should act.
Instead, Polite’s reference spoke directly to whether Ms. Joll,
who had been a head coach for several years, could play a
successful supporting role in someone else’s program.



3   See supra note 1.
38                                                No. 18-3630

                              2.
    Ms. Joll also asserts, and my colleagues accept, that the
checking of Ms. Joll’s references prior to a hiring decision
being made was a “deviation from standard procedures”
which “lends support to her claims.” Maj. at 15. Mr. Hofer
testified that, at the time VHS was hiring an assistant girls’
cross-country coach, there were no written procedures gov-
erning the process of hiring assistant coaches. As a matter of
practice, however, he would check all candidates’ references
just prior to presenting their candidacies to the board. See
R.29-4 at 14 (Hofer Dep. 36). But it was not Mr. Hofer who
contacted Polite about a reference for Ms. Joll. Instead,
Dr. Amones and Coach Nellessen had taken charge of the
hiring process for the assistant girls’ cross-country position
while Mr. Hofer was out of the office due to knee surgery.
Although “[s]ignificant, unexplained or systematic devia-
tions from established policies or practices can no doubt be
relative and probative circumstantial evidence of discrimina-
tory intent,” Hanners v. Trent, 
674 F.3d 683
, 694 (7th Cir.
2012), there is nothing inherently suspect about
Dr. Amones’s proceeding in a different manner than Mr. Ho-
fer. Moreover, there is no reason why Dr. Amones and
Coach Nellessen should not have contacted candidates’ ref-
erences if they believed it would inform their hiring deci-
sion. Finally, there is no evidence in the record that
Dr. Amones and Coach Nellessen only contacted Ms. Joll’s
references. Mr. Hofer testified that both candidates’ refer-
ences were available to Dr. Amones and Coach Nellessen,
and Mr. Hofer did not know whether Principal Amones or
Coach Nellessen contacted Arredondo’s references. R.29-4 at
15 (Hofer Dep. 37). There is no evidence on this record that
No. 18-3630                                                   39

Dr. Amones and Coach Nellessen treated the candidates dif-
ferently with respect to checking references.
                               3.
     Ms. Joll also contends that VCS offered shifting criteria
for why it did not hire her, and that offering different expla-
nations supports her claim of discrimination. “One can rea-
sonably infer pretext from an employer’s shifting or incon-
sistent explanations for the challenged employment deci-
sion.” See, e.g., Appelbaum v. Milwaukee Metro. Sewerage Dist.,
340 F.3d 573
, 579 (7th Cir. 2003) (collecting cases). In Appel-
baum, for example, the employer had terminated the plain-
tiff’s employment and, initially, had told her that it was on
the basis of a breach of confidentiality and her poor perfor-
mance. Later, the employer “backed off” the performance
rationale, calling it just an “aggravating factor” in the deci-
sion to terminate her on the basis of the breach of confidenti-
ality.
 Id. At trial, 
however, the employer stated that her per-
formance played “zero role” in the decision to terminate her
employment. The employer’s shifting reason for the same
employment decision, the court held, raised an inference of
pretext.
 Id.
    Here, Ms. 
Joll does not claim that the defendant’s ra-
tionale for not hiring her for the assistant girls’ cross-country
team shifted over time. Instead, she claims, and my col-
leagues accept, that VCS’s hiring criteria shifted from one po-
sition to another:
       When Ms. Joll did not receive the Girls’
       Cross Country assistant coach position, the
       school stated it did not hire her because the
       male candidate who worked at Portage High
40                                                   No. 18-3630

       School, which was 9.5 miles away, had more
       recent high school coaching experience. How-
       ever, in subsequently failing to hire Ms. Joll for
       the Boys’ Cross Country assistant coach posi-
       tion at Valparaiso High School, the School stat-
       ed the male candidate hired worked in the
       building, and that proximity was a most im-
       portant factor in hiring its coaches.
Appellant’s Br. 35–36; see also Maj. at 17–18 (“The defend-
ant’s stated reasons for hiring for these nearly identical jobs
were inconsistent with each other.”).
    However, as noted previously, see supra at 32, we must
consider each failure to hire as a discrete act. This approach
is particularly sensible in a case such as this one where there
were different decisionmakers and candidates for each posi-
tion. Indeed, Ms. Joll has not come forward with any author-
ity in which we have suggested that different explanations for
different employment decisions were suspect.
    With respect to the girls’ cross-country position, the stat-
ed (and consistent) reason for hiring Arredondo was his
more recent experience coaching high school. In his affida-
vit, Mr. Hofer also explained that that there was concern
about Ms. Joll’s ability to step into the role of an assistant.
R.29-1 at 4 (Hofer Aff. 2). This supplemental rationale, how-
ever, does not raise the specter of pretext. See, e.g., Johnson v.
Nordstrom, Inc., 
260 F.3d 727
, 733 (7th Cir. 2001) (“Nordstrom
simply supplemented its explanations in the context of
EEOC charges and litigation; there has been no retraction of
No. 18-3630                                                         41

any of its reasons for failing to promote Johnson nor are any
of its reasons inconsistent or conflicting.”).4
                                  4.
    Finally, Ms. Joll maintains that her superior qualifications
raise an inference of discriminatory intent. We have ex-
plained that “evidence of the applicants’ competing qualifi-
cations does not constitute evidence of pretext ‘unless those
differences are so favorable to the plaintiff that there can be
no dispute among reasonable persons of impartial judgment
that the plaintiff was clearly better qualified for the position
at issue.” Millbrook v. IBP, Inc., 
280 F.3d 1169
, 1180 (7th Cir.
2002) (internal quotation marks omitted).
    Here, Ms. Joll has a wealth of experience coaching track
and cross country at multiple levels. Her most recent experi-
ence was as the head coach of Thomas Jefferson Middle
School, where she served from 2005 to 2013. In making the
determination that Arredondo was the better choice for the
girls’ cross-country assistant coaching position, Coach Nel-
lessen, Dr. Amones, and Mr. Hofer relied upon his recent
experience coaching cross country at the high school level


4 Even if it were appropriate to evaluate rationales across positions,
however, VCS’s rationales were not inconsistent. With respect to the
girls’ position, VCS had to choose between two individuals who did not
teach at VHS. As between those two individuals, the high school chose
the individual with the most recent high school coaching experience.
With respect to the boys’ position, VCS had to choose between an indi-
vidual who was a teacher at VHS, and therefore had daily contact with
the athletes he would coach, and an individual who was not a teacher at
VHS. As between those two individuals, VCS chose the individual in the
building.
42                                                   No. 18-3630

and the fact that, unlike Ms. Joll who had served as a head
coach, it would not be difficult for him to serve in an assis-
tant position. These two criteria are directly related to the
position for which Ms. Joll applied. With regard to the spe-
cific positions at issue, therefore, Ms. Joll was not clearly bet-
ter qualified such that the disparity in qualifications would
raise an inference of pretext or discrimination.
    The majority notes that “the Millbrook rule applies only
where ‘the plaintiff relies exclusively on evidence of the ap-
plicants’ comparative qualifications.” Maj. at 21 (quoting
David v. Caterpillar, Inc., 
324 F.3d 851
, 863 (7th Cir. 2003)).
The majority believes that Ms. “Joll has presented much
more evidence.”
 Id. However, as 
explained at length, I do
not believe there is evidence in this record to support the
other inferences of discrimination that Ms. Joll asks us to
draw.
   At bottom, based on the evidence presented by Ms. Joll, a
reasonable jury could not conclude that the decision to hire
Arredondo as the assistant girls’ cross-country coach was
based on her gender. I would therefore affirm the district
court’s grant of summary judgment to VCS on this claim.
B. Boys’ cross-country position
    Ms. Joll also contends that she suffered discrimination
when she was not hired for the assistant boys’ cross-country
position. She relies on many of the same arguments made
with respect to the girls’ cross-country position: that the de-
cision was affected by the use of sex-stereotypes, that differ-
ent procedures were followed with respect to checking ref-
erences, that there were shifting criteria for hiring, and that
No. 18-3630                                                            43

she was better qualified. Thus, I address only the evidence
unique to the boys’ cross-country position.
    Ms. Joll complains that different procedures were used to
check Kerezman’s references from those used to check hers.
Mr. Hofer explained that Kerezman’s references were not
checked because “he was a teacher at the high school,” he
had coached with the cross-country team on a voluntary ba-
sis and had been “a paid assistant for our track [program],”
and Mr. Hofer “personally kn[e]w him as a coach.” R.29-4 at
17 (Hofer Dep. 49). Kerezman was a known quantity to
Mr. Hofer. It indeed would be placing form over substance
to require Mr. Hofer to go through the motions of checking
the references of someone whom he knew personally and
supervised professionally.
    Ms. Joll also complains about shifting criteria for the
boys’ cross-country position. However, VCS consistently has
stated its primary reason for the hiring of Kerezman was
that he had better rapport with the boys on the team due to
his being a teacher in the building, having volunteered with
the program, and having been an assistant for the boys’ track
team. His presence in the building allowed him to build and
maintain this rapport in a way that a coach from outside the
building—whether male or female—could not.5


5 As noted previously, see supra at 32, Ms. Joll and the majority want to
evaluate whether the criteria shifts across positions and before different
decision-making teams. The majority therefore faults VHS for focusing
on recent high school experience with respect to Arredondo, but on be-
ing “‘in the building’ during the school day with the opportunity to
build ‘rapport with the boys’” with respect to Kerezman. Maj. at 18. Ac-
cording to the majority, these shifts worked to Ms. Joll’s disadvantage:
                                                           (continued … )
44                                                         No. 18-3630

    Finally, Ms. Joll compares her experience to Kerezman’s
and concludes that she clearly is the better qualified candi-
date. Again, however, the position for which Ms. Joll applied
was an assistant boys’ cross-country position at VHS. Alt-
hough she had a longer career coaching than Kerezman, she
was not a teacher at VHS; Kerezman was. Kerezman had
regular contact with the students and an opportunity to
build on this relationship in ways that Ms. Joll could not. It is
difficult to argue that this consideration is not a legitimate
one. Many, including Ms. Joll, might believe that Coach Cra-
gue and Mr. Hofer should have placed more emphasis on
experience than presence. However, “[a]n unwise employ-
ment decision does not automatically rise to the level of pre-

( … continued)
“Joll’s long presence in the middle school and in the school system with
the opportunity to build ‘rapport with the girls’ was not considered in
her favor vis-à-vis Arredondo. Her more extensive and apparently more
recent professional experience coaching high schoolers was not consid-
ered vis-à-vis Kerezman.”
 Id.
     As a 
preliminary matter, Ms. Joll did not have “more recent profes-
sional experience coaching high schoolers.” Kerezman recently had
served as an assistant coach for the VHS boys’ track team. See R.29-1 at 5
(Hofer Aff. at 3) (“Kerezman had also recently coached boys track as a
varsity assistant at Valparaiso which we felt was a plus.”); R.29-4 at 16
(Hofer Dep. 44) (explaining that Kerezman had been “a paid coach for
our track [program]”). However, even if this had not been the case, the
fact that Coach Nellessen, Dr. Amones, and Mr. Hofer applied different
criteria than Coach Crague and Mr. Hofer when faced with two different
sets of candidates does not suggest discriminatory intent. It is entirely
possible that, if presented with a qualified, in-building candidate,
Coach Nellessen, Dr. Amones, and Mr. Hofer would have chosen that
candidate over either Ms. Joll or Arredondo. This option, however, was
not available to them.
No. 18-3630                                                   45

text.” Teruggi v. CIT Grp./Capital Fin., Inc., 
709 F.3d 654
, 661
(7th Cir. 2013). Here, the record does not support an infer-
ence that Kerezman’s status as a teacher was not the true
reason that he was hired, nor does it support an inference
that Ms. Joll was not hired because she was a woman. Con-
sequently, I would affirm the district court’s summary
judgment for VHS on this claim as well.
   I close with a final observation concerning the effect of
our Ortiz decision on summary judgment decisions. We held
in Ortiz that, for purposes of summary judgment, we could
not cabin different types of evidence into categories and con-
sider those categories in isolation. We explained that the ap-
propriate legal standard
       is simply whether the evidence would permit a
       reasonable factfinder to conclude that the
       plaintiff’s race, ethnicity, sex, religion, or other
       proscribed factor caused the discharge or other
       adverse employment action. Evidence must be
       considered as a whole, rather than asking
       whether any particular piece of evidence
       proves the case by itself—or whether just the
       “direct” evidence does so, or the “indirect” ev-
       idence. Evidence is evidence.
Ortiz, 834 F.3d at 765
. However, we also were clear that
“[r]elevant evidence must be considered and irrelevant evi-
dence disregarded.”
 Id. Ortiz did 
not cast aside our well-
honed rules for determining when particular types of evi-
dence (the use of different procedures, shifting rationales,
etc.) are relevant and probative of discrimination. Today’s
decision demonstrates graphically how this principle can be
46                                                 No. 18-3630

disregarded when the Ortiz approach is employed to evalu-
ate a summary judgment motion.
    Here, the question is whether Ms. Joll came forth with
sufficient relevant and probative evidence from which a rea-
sonable jury could conclude that VCS failed to hire her for
the girls’ assistant cross-country position or for the boys’ as-
sistant cross-country position because she was a woman. For
the reasons set forth above, I do not believe Ms. Joll has met
her burden with respect to either position, and I therefore
would affirm the judgment of the district court.