In the
 United States Court of Appeals
                For the Seventh Circuit
                          ____________

No. 05-3300
ANDREW BOURNE, et al.,
                                            Plaintiffs-Appellants,
                                 v.

MARTY GILMAN, INCORPORATED,
doing business as Gilman Gear,
                                              Defendant-Appellee.
                          ____________
            Appeal from the United States District Court
     for the Southern District of Indiana, Indianapolis Division.
             No. 03 C 1375—David F. Hamilton, Judge.
                          ____________
       ARGUED MAY 9, 2006—DECIDED JUNE 20, 2006
                     ____________


  Before CUDAHY, KANNE, and WOOD, Circuit Judges.
  KANNE, Circuit Judge. When Ball State student Andrew
Bourne rushed onto a football field with a crowd that tore
down a goalpost, the post fell on his back and rendered him
paraplegic. He and his parents sued Gilman Gear, manufac-
turer of the post, in diversity under Indiana law arguing
that the post was defective and unreasonably dangerous
because (1) it was foreseeable that fans will tear down
goalposts, (2) the average fan would not understand the
extent of the risk, and (3) there are alternative designs that
would reduce that risk. The district court granted summary
judgment for Gilman Gear because the risk was obvious.
We affirm.
2                                                No. 05-3300

                       I. HISTORY
   We have taken the facts of this sad but straightforward
case from the parties’ summary judgment papers, beginning
with Bourne’s testimony that, in October 2001 when he was
21-years old, he attended his first-ever tailgating party
outside the game. Near the end of the fourth quarter, he
joined a crowd to storm the field in celebration of an
imminent Ball State victory. Bourne himself did not
rip down the post. He jumped and tried to grab it, missed,
and walked away. With his back to the post, he heard a
snap, and the post fell on his back, causing his injuries.
Although he knew that the post would collapse, he expected
it to do so gradually.
  As both parties agree, Ball State itself encouraged the
crowd to pull down goalposts with a flashing sign on the
scoreboard that read, “The goalpost looks lonely.” Indeed,
the school had earlier resolved that controlling the crowd
might prove even more dangerous than letting it tear down
the goalposts. (Ball State is not a party now because it
settled for a paltry $300,000, a limit imposed by state
tort reform in the 1970s.)
   Neil Gilman, the president of Gilman Gear, testified that
his company has known all along that fans sometimes tear
down posts; he also described his company’s posts. The
posts, he explained, are about 40-feet tall and weigh 470
pounds. They are aluminum rather than steel because steel
is heavier, harder to install, and tends to rust. And they are
the so-called “slingshot” style with one vertical support
holding up the structure. This slingshot style was intro-
duced in 1969 so as to minimize the danger posed to players
in the end zone by the old H-shaped goalposts with two
vertical supports. Notably, Gilman Gear did not design the
posts itself; instead, it bought the design in 1985. To
facilitate “rolling” of the metal in its newly assumed
manufacturing process, Gilman Gear switched to a differ-
No. 05-3300                                                3

ent, less-brittle type of aluminum alloy than was used by
the prior maker. When asked if his company had “consid-
ered engineering controls” to address hazards created by
pulling down posts, Gilman said no.
  To avert summary judgment, the Bournes submitted
the affidavit of their expert, Vaughn Adams, a Ph.D. in
Safety Engineering, who testified that reasonable manufac-
turers should foresee that goalposts will be torn down by
fans. Adams compiled non-exhaustive numbers of football
games in which students tore down posts: 16 in 2000, 10 in
2001, 17 in 2002, 12 in 2003, and 3 by October 2004. Adams
also noted Gilman’s testimony that he knew about some or
all of those tear-downs (though not all were Gilman Gear
posts). Additionally, Adams cited two newspaper articles
reporting incidents of injury other than Bourne’s, though he
did not attempt to compile statistics.
  In short, Adams’s—and the Bournes’—theory is that,
when fans try to pull them down, Gilman Gear’s aluminum
posts will at first bend but then suddenly “snap,” abruptly
falling on unwary fans whose lay knowledge of metallurgy
lulls them into believing that goalposts fall gradually
enough to permit a safe retreat. Adams, however, did not
testify to any science on which he based his opinion. For
example, he offered only speculation to support his premise
that social and cultural pressure misleads the average fan
into believing that goalposts collapse slowly enough that
ripping them down is safe. Moreover, although he hinted
that Gilman Gear’s change in aluminum alloy in 1985
rendered the posts more dangerous, he cited no evidence
comparing the posts before and after the change. Instead,
his conclusions apparently rested on availability of alterna-
tive designs. The first of these alternative designs is the
“double-offset gooseneck,” which reinforces the single
vertical support with another support right next to it.
Second is a “hinged” goalpost, first introduced by
the University of Iowa in the 1990s, which permits the
4                                                 No. 05-3300

athletic facility to lower the posts immediately after a game.
(Gilman Gear itself began making and selling these posts
after Bourne’s injury; at least one other company makes
them, too.) Third, there is the “fan-resistant” or “indestructi-
ble” goalpost made by Merchants Environmental Industries,
Inc. This third kind is made out of steel, less likely to break
than aluminum. But just as Adams did not conduct tests on
any posts manufactured by Gilman Gear, he did not test
any other company’s posts or cite to any scientific data.
Instead, he presented just a few marketing materials
distributed by makers of these alternative designs. While
posts like the one that injured Bourne cost $4,700 per pair,
the hinged posts cost $6,500 and the “indestructible” posts
between $23,000 and $32,000. The cost of the double-
gooseneck rigs is not in the record. Adams assumed that a
cost-benefit analysis shows the pricier alternatives to be
preferable in light of their greater safety and lower rate of
replacement. He also opined that Gilman Gear was negli-
gent for failing to test its posts to determine when they
would break.
  In granting summary judgment for Gilman Gear, the
district court held that Indiana law barred recovery for the
Bournes because it was obvious to a reasonable person that
a collapsing goalpost poses a risk of serious injury. The
court reasoned that Andrew Bourne’s subjective failure to
appreciate the magnitude of the risk that a collapsing post
might strike his back and take away the use of his legs did
not alter the fact that the risk of injury was obvious as a
matter of law and, consequently, that the post was not
unreasonably dangerous. In so holding the district court
acknowledged that in Indiana the so-called “open and
obvious” rule is no longer an absolute bar to a claim under
the Products Liability Act against a manufacturer, but the
court reasoned that the principle remains relevant and, in
this case, was decisive.
No. 05-3300                                                 5

                      II. ANALYSIS
   On appeal the Bournes maintain that the “open and
obvious” rule cannot bar a claim for defective design under
the Indiana Products Liability Act. Relying on Mesman v.
Crane Pro Servs., 

, 849-52 (7th Cir. 2005), they
insist that they can win despite the obviousness of the risk
if they can nonetheless prove through the application of the
classic formulation of negligence that Gilman Gear should
have adopted a reasonable alternative design.
  The relevant law is codified in the Indiana Products
Liability Act. Ind. Code §§ 34-20-1-1 to 34-20-9-1. Although
the Act originally applied only to strict liability (for manu-
facturing defects and failure to warn), it was amended in
1995 to apply to claims of defective design, which tradition-
ally sound in negligence. 

. Com-
pare Ind. Code § 33-1-1.5-1 (1995) with 

law
was re-codified in 1998, but without relevant change.
Compare Ind. Code §§ 34-20-1-1 to 34-20-9-1 (effective July
1, 1998) with Ind. Code §§ 33-1-1.5-1 to 33-1-1.5-10 (1997).
   A plaintiff bringing an action under the Act must estab-
lish that (1) he or she was harmed by a product; (2) the
product was sold “in a defective condition unreason-
ably dangerous to any user or consumer”; (3) the plaintiff
was a foreseeable user or consumer; (4) the defendant was
in the business of selling the product; and (5) the product
reached the consumer or user in the condition it was sold.
See Ind. Code § 34-20-2-1; see also Moss v. Crosman Corp.,

, 1171 (7th Cir. 1998).
  At the outset, we note that Indiana is a comparative-fault
state and contributory negligence is not a complete bar
unless the plaintiff bears more than 50% of the blame for
his own injury. Ind. Code. §§ 34-20-8-1; 34-51-2-7, -8; see
also Smith v. Baxter, 

, 244-45 (Ind. 2003).
What is more, misuse is not a bar unless the misuse was
6                                                No. 05-3300

“not reasonably expected by the seller.” Ind. Code
§ 34-20-6-4; see also Morgen v. Ford Motor Co., 

, 1149-50 (Ind. 2003). Likewise, the statute protects
“any bystander injured by the product who would reason-
ably be expected to be in the vicinity of the product during
its reasonably expected use.” Ind. Code § 34-6-2-29; see also
Stegemoller v. ACandS, Inc., 

, 975 (Ind.
2002). Mindful of these rules and Neil Gilman’s testimony
that his company actually foresaw the fans’ vandalism,
Gilman Gear does not argue that the claim should be barred
on the basis of misuse or Bourne’s fault. Consequently, we
need not pass on whether this is a case in which no reason-
able jury could find that the plaintiff was less responsible
for his own injury than others were, see, e.g., Barnard v.
Saturn Corp., a Div. of Gen. Motors Corp., 

,
1031 (Ind. Ct. App. 2003).
  The only question presented by the parties is whether the
goalpost was “in a defective condition unreasonably danger-
ous to any user or consumer.” Actually, this is two questions
because Indiana law requires the plaintiff to show that a
product is both “in a defective condition” and that it is
“unreasonably dangerous.” McMahon v. Bunn-O-Matic
Corp., 

, 657 (7th Cir. 1998) (citing Koske v.
Townsend Eng’g Co., 

, 440-41 (Ind. 1990));

, 1174; see also Baker v.
Heye-America, 

, 1140 (Ind. Ct. App. 2003)
(applying the statute after the 1998 recodification).
   The district court started and finished its inquiry with the
first prong, whether the post was “unreasonably danger-
ous.” “Unreasonably dangerous” means “any situation in
which the use of a product exposes the user or consumer to
a risk of physical harm to an extent beyond that contem-
plated by the ordinary consumer who purchases the product
with the ordinary knowledge about the product’s character-
istics common to the community of consumers.” 

146. Applying that rule in this case, the district court
No. 05-3300                                                    7

decided that any reasonable person on the field should have
known the general danger posed by a falling goalpost.
Consequently, the court concluded, recovery was barred
under precedent holding that a user’s knowledge of a
general risk precludes recovery even if he did not know the
extent or specific degree of that risk.1 For example, a family
whose child was killed by a BB gun could not prevail on the
theory that, although they knew when they bought the gun
that it could seriously injure him, they did not know it could
kill. 

. Nor could an electrician’s
estate succeed on a claim against the manufacturer of a
metal crane on the theory that he knew he could be shocked
if it touched power wires, but not that he could be killed.
Anderson v. P.A. Radocy & Sons, Inc., 

, 625-26
(7th Cir. 1995). Whether or not Andrew knew the post could
suddenly “snap” and paralyze him, he should have known
that it could fall and seriously injure him, and the dis-
trict court considered that the end of the matter.
  The Bournes’ principal objection to this ruling is that the
district court explained that their recovery was barred
because the danger was “obvious” as a matter of law. They
rely on our recent opinion in Mesman explaining that, after
the Indiana legislature in 1995 expanded its code of prod-
ucts liability to cover all theories of liability includ-
ing defective design, Indiana law no longer permits a manu-
facturer to avoid liability in a design defect case simply
because a defect is “open and obvious.” See 

. Compare Ind. Code § 33-1-1.5-1 (1995) with

all, a product may be designed with a



1
  There are very few cases about goalposts being torn down, e.g.,
Cimino v. Yale Univ., 

(D. Conn. 1986); Univ. of
Tex. at El Paso v. Moreno, 

(Tex. App. 2005);
Pallazola v. Town of Foxborough, 

(Mass. 1994),
and none is informative here.
8                                                No. 05-3300

feature that, although obvious, is nonetheless unreasonably
prone to cause accidents. For example, a machine may have
an exposed moving blade or other part such that the user,
though he knows of it, may nonetheless slip and fall and cut
off his hand. 

Since that injury is easily foresee-
able and cheaply preventable by attaching a guard, the
manufacturer ought not get off the hook. 

interpretation makes sense; the accident magnet is just as
obvious to the designer as the user, and the rule should not
work just one way.
  Rather than the open-and-obvious defense, the statute
creates the so-called “incurred risk defense,” which requires
the defendant to establish that the user actually knew of
the product’s danger. Ind. Code § 34-20-6-3; 

. But like the defendant in Mesman, Gilman
Gear did not plead this defense and does not argue its
application now.
  Despite the use of some imprecise language here (the
court should have said that the goalpost was not unrea-
sonably dangerous as a matter of law, rather than declaring
that the danger posed by the goalpost was obvious as a
matter of law), the gist of the district court’s ruling is
sound. Indeed, the district court, like the Mesman court and
the Indiana Supreme Court, expressly recognized that the
“open and obvious” rule has been abrogated. The district
court was correct, furthermore, that obviousness remains a
relevant inquiry because, as noted above, the question of
what is unreasonably dangerous depends upon the reason-
able expectations of consumers and expected uses. See Ind.
Code §§ 34-20-4-1, 34-6-2-146; 

;
FMC Corp. v. Brown, 

, 446 (Ind. 1990). In
some cases, the obviousness of the risk will obviate the need
for any further protective measures, or obviousness may
prove that an injured user knew about a risk but nonethe-
less chose to incur it. 

; FMC

. Although obviousness typically
No. 05-3300                                                 9

factors in the equation for the jury (it is evidence but “not
conclusive evidence,” 

), there are
some cases where the case is so one-sided that there is no
possibility of the plaintiff’s recovery. See 

. And the bottom line is that Indiana law does not
permit someone to engage in an inherently dangerous
activity and then blame the manufacturer. See 

Bournes nevertheless maintain that,
because the goalpost can be made safe (unlike a BB gun), a
window remains open for them to show defective design
because the goalpost exposed Andrew to a greater risk than
he should have expected. In other words, the product
exposed him “to a risk of physical harm to an extent beyond
that contemplated by the ordinary consumer who purchases
the product with the ordinary knowledge about the prod-
uct’s characteristics common to the community
of consumers.” Ind. Code § 34-6-2-146. Even indulging
that argument, cf. 

, the Bournes
must lose because they cannot show a defect with the
evidence that they have adduced.
  A defective product is one sold in a condition “(1) not
contemplated by reasonable persons among those consid-
ered expected users or consumers of the product; and (2)
that will be unreasonably dangerous to the expected user or
consumer when used in reasonably expectable ways of
handling or consumption.” Ind. Code § 34-20-4-1. That
definition is decidedly unhelpful. But fortunately the
statute more clearly explains that a plaintiff alleging a
design defect cannot prevail without showing that the
manufacturer was negligent. See 

(citing cases). That requires applying the
classic formulation of negligence: BMesman, 409 F.3d
at 849 
(citing United States v. Carroll Towing Co., 
159 F.2d
169
, 173 (2d Cir. 1947)). A caveat (hinted at above) is that
there is no duty for a manufacturer to redesign a product
that cannot be made safe, like a BB gun. See Ind. Code § 34-
10                                               No. 05-3300

20-4-4 (“A product is not defective under this article if the
product is incapable of being made safe for its reasonably
expectable use, when manufactured, sold, handled, and
packaged properly.”); cf. 
Moss, 136 F.3d at 1173
.
  The Bournes are not the first to make this type of argu-
ment. In McMahon, a woman injured by hot coffee that
spilled into her lap sued the manufacturer of the coffeepot
on the theory that its design was defective insofar as it
made the coffee hotter than necessary, and hotter than she,
as a reasonable consumer, expected. 
McMahon, 150 F.3d at
657
-59. A better design, she argued, would produce a
slightly cooler cup of coffee. 
Id. at 657-58. 
To this end, she
submitted the testimony of an expert who opined that the
coffeepot could easily, and cost effectively, be made to
produce a cooler, yet tasty, cup of coffee. 
Id. But just
because the safer pot could be made did not mean that the
manufacturer’s pot was defective. 
Id. at 657-59. 
Instead, it
was her burden to show that the cost-benefit formula
demanded adopting the alternative design. 
Id. Yet her
expert did not explain the basis for his conclusion regarding
the risks, benefits, and costs of reducing the temperature of
the coffee. 
Id. at 658. 
Nor was the case one in which there
was no possible benefit from hotter coffee such that res ipsa
loquitur might apply. 
Id. at 658-59. 
For that reason, and
because an expert’s conclusory assertions are of no eviden-
tiary value, summary judgment was affirmed. 
Id. at 657-59.
  The Bournes’ case shares the same fatal flaw. Their
expert’s affidavit is their only evidence that the design is
defective. But just like the expert in McMahon, Adams’s
testimony is comprised of mere conclusions. For the premise
that fans are unaware of the risks, he offers only specula-
tion that social pressure and publicity falsely assure them
that pulling down posts is safe. (Perhaps seeing the weak-
ness, the Bournes contend simply that people would not rip
down posts if they knew the risks.) As mentioned above,
Adams’s suggestion that Gilman Gear’s change in alumi-
No. 05-3300                                                 11

num alloy in 1985 made the product less safe is nothing but
innuendo. Moreover, Adams does not provide a basis on
which a finder of fact could evaluate the frequency of
injuries caused by goalposts, or calculate the extent to
which risk would actually be reduced by the alternative
designs, or justify the cost of those alternatives relative to
the benefits of aluminum posts. Although Gilman Gear
points out such flaws, explaining that Adams’s affidavit
actually proves the infrequency of injury relative to the
number of games, the Bournes retort simply that Adams’s
testimony was not meant to provide those statistics. As if
unaware of their burden, they say neither statistics nor
testing is required because the competitors actually sell
safer (according to Adams) posts (although they are 38% to
700% more expensive). But that will not do: mere existence
of a safer product is not sufficient to establish liability. See
McMahon, 150 F.3d at 657
-59; Whitted v. Gen. Motors
Corp., 
58 F.3d 1200
, 1206 (7th Cir. 1995); Pries v. Honda
Motor Co., Ltd., 
31 F.3d 543
, 545 (7th Cir. 1994). Otherwise,
the bare fact of a Volvo would render every KIA defective.
  Finally, Adams does not even consider the possibility of
unintended increases in risk to intended users, like the
students or staff who would have to hurriedly lower the
hinged post to police the crowd at the end of a game. But
the costs of those incidental effects must be weighed in the
balance. See 
Pries, 31 F.3d at 545 
(criticizing an expert’s
testimony for failing to consider whether a proposed
alternative design to protect the victim of a particular auto
accident might not increase risks to other users). After all,
Indiana neither requires manufacturers to be insurers
nor to guard against all risks by altering the qualities
sought by intended users. See 
McMahon, 150 F.3d at 659
.


                    III. CONCLUSION
  Because the district court’s conclusion that Indiana law
does not require manufacturers to protect consumers and
12                                           No. 05-3300

users from themselves is fundamentally correct, and
because any jury’s application of the B