State v. Craig S. Lillge

COURT OF APPEALS
DECISION NOTICE
DATED AND FILED This opinion is subject to further editing. If
published, the official version will appear in
the bound volume of the Official Reports.
January 6, 2023
A party may file with the Supreme Court a
Sheila T. Reiff petition to review an adverse decision by the
Clerk of Court of Appeals Court of Appeals. See WIS. STAT. § 808.10 and
RULE 809.62.

Appeal No. 2022AP192-CR Cir. Ct. No. 2017CF365

STATE OF WISCONSIN IN COURT OF APPEALS
DISTRICT IV

STATE OF WISCONSIN,

PLAINTIFF-RESPONDENT,

V.

CRAIG S. LILLGE,

DEFENDANT-APPELLANT.

APPEAL from a judgment and an order of the circuit court for
Jefferson County: WILLIAM F. HUE and WILLIAM V. GRUBER, Judges.
Affirmed.

Before Blanchard, P.J., Kloppenburg, and Graham, JJ.

Per curiam opinions may not be cited in any court of this state as precedent

or authority, except for the limited purposes specified in WIS. STAT. RULE 809.23(3).
No. 2022AP192-CR

¶1 PER CURIAM. Craig Lillge appeals a judgment of conviction and a
circuit court order denying his postconviction motion for a new trial.1 At a bench
trial the circuit court found Lillge guilty of one count of burglary of a dwelling in
violation of WIS. STAT. § 943.10(1m)(a) (2019-20) and one count of arson of a
building without the owner’s consent in violation of WIS. STAT. § 943.02(1)(a).2
Both counts were based on the allegation that Lillge entered the house of his on-
and-off romantic partner without her permission and intentionally caused a fire.
Lillge argues that the circuit court erred in denying his postconviction motion, which
was based on the contention that his trial counsel provided constitutionally
ineffective assistance by failing to identify and rely on particular expert testimony
regarding the cause of the fire that would be exculpatory. We conclude that the
court properly denied Lillge’s motion because he fails to show a substantial
probability of a different outcome if his trial counsel had relied on the particular
expert testimony that he identifies.

BACKGROUND

¶2 In August 2017, the State filed a criminal complaint that alleged, as
pertinent to this appeal, one count each of burglary of a dwelling and arson of a
building without consent, the latter as an act of domestic abuse. Each count relates

1
The Honorable William F. Hue presided over trial and entered the judgment of
conviction, and we refer to him as the trial court. The Honorable William V. Gruber entered the
order denying Lillge’s postconviction motion, and we refer to him as the postconviction court.
2
All references to the Wisconsin Statutes are to the 2019-20 version unless otherwise
noted.

2
No. 2022AP192-CR

to alleged conduct involving the same victim, who we will refer to as A.B. 3 The
complaint alleges in pertinent part that, on the night of December 20, 2014, and into
the next day, Lillge intentionally and without A.B.’s consent entered her rented
house in Watertown and caused an explosion and fire. The complaint alleged that
Lillge opened a valve on an uncapped line carrying natural gas into the basement
and shut the door to the basement, which allowed gas to concentrate in the basement
until it was explosively ignited by the operation of an appliance or some other source
in the basement, resulting in a fire.4

¶3 The circuit court held a three-day bench trial. The prosecution called
17 witnesses, including A.B., her daughter, and several law enforcement
investigators. We provide additional details from the trial evidence in the discussion
below, but for background purposes we note the following.

¶4 There was no dispute at trial or now on appeal regarding the
following: (1) there was an explosion at A.B.’s house sometime between 7:00 and
8:00 a.m. on the morning of December 21, and (2) this explosion resulted in a
significant fire at the house. Moreover, as discussed further below, there has never
been any dispute that this explosion in the house was the result of a natural gas
build-up in the basement. The disputed issue raised in the postconviction motion

3
The complaint also charged Lillge with one count of stalking A.B., resulting in bodily
harm, a count on which Lillge was acquitted at trial. Given the acquittal, the parties do not raise
any issues related to the stalking charge, although some evidence related to the stalking charge was
relevant to the burglary and arson charges.
4
A brief word on this reference to an “uncapped line.” Testimony given at trial and in the
hearing on Lillge’s postconviction motion established the following. Appliances powered by
natural gas lines that run into houses are of course sometimes removed and when that happens, a
best-practice step for safety is to “cap and cover” the end of the line to prevent gas from escaping
into the air in the house. In this case there was undisputed evidence that the line did not have a cap
securing its terminal point in the basement of the house, but instead it had a valve that could be
opened or shut by hand.

3
No. 2022AP192-CR

and now on appeal involves evidence bearing on the cause or causes of the gas build-
up.

¶5 Prosecution witnesses included Special Agent Joshua Pudlowski of
the Wisconsin Department of Justice’s Division of Criminal Investigation.
Pudlowski offered expert testimony regarding his investigation of the explosion and
fire. In addition to the undisputed opinion which we have just noted—that the origin
of the fire was a natural gas explosion in the basement of the house—Pudlowski
further testified that the fire is properly categorized as “incendiary”; that is, it was
caused by the intentional act of a person. That intentional act, according to
Pudlowski, was someone fully opening a valve on an uncapped natural gas line that
had a terminal point in the basement. Previewing discussion below, what Lillge
contested in his postconviction motion and now renews on appeal is Pudlowski’s
incendiary classification.

¶6 At trial, A.B. testified in part as follows. Beginning in around 2011,
A.B. and Lillge had a romantic relationship with many break ups and
reconciliations. Lillge frequently accused A.B. of cheating on him. He would
become angry when A.B. spoke with other men and sometimes he would act
violently toward her. A.B. was living with Lillge during some of these incidents,
along with A.B.’s two children, but A.B. and her children were living in a rental
house at the time of the fire. Days before the fire, on December 15, 2014, A.B.
broke up with Lillge. On the night of December 20-21, neither A.B. nor her children
were staying at the house A.B. was renting. A.B. did not consent to Lillge entering
her house on December 20 or 21, or consent to him causing a fire in the house.

¶7 Richard Hall testified as a partial alibi witness for the defense. Hall
testified that he spent much of the night of December 20 and early morning hours

4
No. 2022AP192-CR

of December 21 with Lillge and that both men drank a lot of alcohol. But Hall
provided only a partial alibi because he did not purport to account for Lillge’s
whereabouts the entire night and morning. Lillge did not testify. The defense did
not present expert testimony, including on the topic of what caused the explosion
and fire.

¶8 The prosecution argued that the evidence established that Lillge,
motivated by anger over the December 2014 breakup, entered A.B.’s house twice
on the night of December 20-21: first breaking into the house and taking some of
A.B.’s clothes and throwing them onto the street; and then later returning in the
early morning hours and fully opening the valve of the uncapped gas line with intent
to cause an explosion or fire, which resulted in an explosion followed by a fire. The
prosecution emphasized Pudlowski’s testimony that the fire should be characterized
as “incendiary,” together with separate evidence supporting the finding that Lillge
was the person who opened the valve on the uncapped line.

¶9 The defense did not directly challenge the prosecution theory,
supported by Pudlowski’s testimony, that the fire was intentionally caused by
someone. Instead, the defense centered on the lack of direct evidence showing that
Lillge visited A.B.’s house on December 20-21 together with partial alibi witness
Hall, which the defense argued created reasonable doubt that he would have been
the person who opened the valve on the uncapped line. As part of this strategy, the
defense emphasized that, in order to visit A.B.’s house that night, Lillge would have
had to travel two miles from his house and back, on a night when he had been
drinking heavily, during the limited time periods when his whereabouts were
unaccounted for according to Hall.

¶10 The circuit court found Lillge guilty of the burglary and arson charges.

5
No. 2022AP192-CR

¶11 Lillge filed a postconviction motion for a new trial on the ground that
his trial counsel provided ineffective assistance. More specifically he argued that it
was ineffective for his trial counsel to “fail[] to hire an expert witness to highlight
deficiencies and failures in the state’s investigation as to the cause [of] the fire” and
to “fail[] to cross-examine the State’s expert as to deficiencies in his investigation
of the fire.”

¶12 At a Machner hearing, Lillge’s trial counsel testified that he did not
consult any experts in preparation for trial.5

¶13 Lillge also presented the report and testimony of an expert “fire scene
examiner,” John Agosti, based on Agosti’s review of the discovery materials
provided to the defense by the prosecutor before trial and Pudlowski’s trial
testimony. Agosti testified to the opinion that the fire had an “undetermined” cause,
not an incendiary one. Agosti criticized Pudlowski’s investigation. Agosti opined
that Pudlowski failed to gather sufficient evidence to establish that an opened gas
valve in the basement caused the explosion and failed to properly rule out potential
accidental causes.

¶14 The prosecution called Pudlowski, who disputed each of Agosti’s
opinions. Pudlowski testified that Agosti failed to undermine Pudlowski’s view that
the fire had an incendiary origin and that Agosti made incorrect factual assumptions.

¶15 The postconviction court found that both Agosti and Pudlowski were
credible. However, the court further reasoned that it would not have made a

5
See State v. Machner, 101 Wis. 2d 79, 303 N.W.2d 633 (1981).

6
No. 2022AP192-CR

difference at trial if the experts had “dueled” at trial in the manner that they did in
connection with the postconviction motion.

DISCUSSION

¶16 Lillge argues that his trial counsel provided constitutionally
ineffective assistance by failing “to consult with or hire an expert witness who could
have provided exculpatory evidence as to the cause of the explosion.” “To establish
ineffective assistance of counsel, a defendant must prove both that counsel’s
performance was deficient and that the deficient performance prejudiced the
defense.” State v. Chu, 2002 WI App 98, ¶47, 253 Wis. 2d 666, 643 N.W.2d 878
(citing Strickland v. Washington, 466 U.S. 668, 687 (1984)). We assume without
deciding that Lillge’s trial counsel performed deficiently. With that assumption, we
conclude that Lillge fails to show through Agosti’s testimony that Lillge was
prejudiced. See State v. Breitzman, 2017 WI 100, ¶37, 378 Wis. 2d 431, 904
N.W.2d 93 (citing Strickland, 466 U.S. at 697) (“If the defendant fails to satisfy
either prong [of an ineffective assistance of counsel claim], we need not consider
the other.”).

¶17 Regarding the prejudice prong, Lillge “must show that ‘there is a
reasonable probability that, but for counsel’s unprofessional errors, the result of the
proceeding would have been different. A reasonable probability is a probability
sufficient to undermine confidence in the outcome.’” See State v. Love, 2005 WI
116, ¶30, 284 Wis. 2d 111, 700 N.W.2d 62 (quoting authority that quotes
Strickland, 466 U.S. at 694). The likelihood of a different result must be substantial,
not merely conceivable. See State v. Cooper, 2019 WI 73, ¶29, 387 Wis. 2d 439,
929 N.W.2d 192 (citing Cullen v. Pinholster, 563 U.S. 170, 189 (2011)). The
prejudice prong is examined under “the totality of the circumstances to determine

7
No. 2022AP192-CR

whether counsel’s errors, in the context of the entire case, deprived the defendant
[of] a fair trial.” State v. Domke, 2011 WI 95, ¶54, 337 Wis. 2d 268, 805 N.W.2d
364.

¶18 “This court will uphold the circuit court’s findings of fact unless they
are clearly erroneous.” State v. Thiel, 2003 WI 111, ¶21, 264 Wis. 2d 571, 665
N.W.2d 305. “Whether any deficient performance was prejudicial is … a question
of law we review de novo.” Breitzman, 378 Wis. 2d 431, ¶39.

¶19 Lillge argues that trial counsel should have consulted Agosti and
provided evidence at trial that the cause of the fire could be categorized in only one
way, as “undetermined,” and not as incendiary, because Agosti would have proved
that the cause might have been accidental.6 More specifically, Lillge’s position is
that Agosti would agree with Pudlowski that a gas-fueled explosion in the basement
of the house occurred and caused the fire, but Agosti would convince a factfinder
that this could have been the result of one or more accidental gas leaks instead of
the intentional opening of the gas valve, as under Pudlowski’s theory.

¶20 The State argues that Agosti’s conclusion “that the cause of the fire
was undetermined” had little or no weight as an “exculpatory” opinion because
Agosti did not purport to disprove that the valve was intentionally opened by
someone, and therefore the strong evidence that Lillge intentionally opened it that
night was not seriously rebutted. Regarding the first part of this argument, the State
contends that Pudlowski “refuted Agosti’s criticisms and demonstrated [that
Pudlowski] properly inspected and tested the gas lines [including the one with] the

6
The State does not argue that trial counsel could not have located and consulted with
Agosti himself, or with another expert who has similar opinions, in advance of trial and made use
of such expertise with admissible evidence at trial. For ease of reference we speak in terms of
consulting Agosti in particular, without further referencing the concept of a different expert.

8
No. 2022AP192-CR

open valve” in A.B.’s basement. Regarding the second part, the State acknowledges
that it had no direct incriminating evidence to offer at trial, such as a witness seeing
Lillge in the house at a key time or a confession, but contends that there is
nonetheless a vivid pattern of incriminating evidence.

¶21 We conclude that Agosti’s testimony could have increased to a small
degree the probability of a different outcome at trial. However, we also conclude
that the increased probability of a different outcome would not have been a
substantial probability. This is due to two factors: the limited scope of Agosti’s
testimony and the totality of the incriminating evidence against Lillge.7 Given that
our analysis of the prejudice prong depends in part on the totality of evidence at
trial, we summarize some of that evidence in additional detail. Then we provide
additional background regarding the evidence presented at the postconviction
hearing. Finally, we explain further our conclusion and our rejection of Lillge’s
arguments to the contrary.

I. Additional Background

Evidence At Trial

¶22 To repeat, A.B. testified that over the years of their on-again, off-again
relationship Lillge sometimes expressed anger, to the point of acting violently
toward her, about her purported interactions with other men. The following are
additional details from A.B.’s testimony. In February 2014, Lillge accused A.B. of

7
In reaching this conclusion, we do not place weight on an argument by the State that we
should rely on findings of fact made by the trial court regarding Lillge’s motive and opportunity to
intentionally break into A.B.’s house and cause a fire, as opposed to relying on our independent
view of the totality of the evidence adduced at trial. That is, we assume in Lillge’s favor without
deciding any related legal issues that, in themselves, the findings of the trial court acting as the
factfinder are not significant to the prejudice prong analysis of Lillge’s postconviction claim.

9
No. 2022AP192-CR

flirting with one of his friends, knocked drinks off a table at a bar, and later, at
Lillge’s house, knocked a tray over and threw a remote into a wall while yelling at
A.B. about her supposed flirting. In another incident also involving a jealous rage,
in March or April 2014, Lillge grabbed A.B. by the neck, held her up against a wall
and threatened to “‘f-ing kill [her],’” said that “‘no one would ever find the body,’”
and punched a hole in the wall next to A.B.’s head. During the fall of 2014, Lillge
again said to A.B. words to the effect that he “can kill” her “and no one would ever
find the body.” On multiple occasions, Lillge let air out of the tires of A.B.’s car.8

¶23 A.B. and her daughter testified that, when they lived with Lillge, he
would, in anger, frequently throw out onto the front yard A.B.’s possessions,
including some of her clothes, as well as clothes belonging to her daughter.

¶24 A.B. further testified to the following. In either the summer or fall of
2014, Lillge told A.B. that when he and his friends were younger they found an
abandoned house that they “blew … up and watched … burn.” In the summer of
2014, A.B. and her children moved into the rental house in Watertown where the
fire would occur in December 2014. In the fall of 2014, during a period when A.B.
and Lillge had resumed their romantic relationship, A.B. gave him a key to the
house. In November 2014, A.B. asked Lillge to return the key. Lillge gave A.B. a
key that was different in color from the one she had given him, and when she
questioned this he said words to the effect of, “‘I don’t need a key to get into your
house anyway, so it doesn’t really matter if I possess a key.’” A.B. recalled that
there were at least two occasions later in the fall of 2014 when lights were

8
A.B. also testified that, before Lillge was charged in this case, the two reconciled again,
but also that in early 2016 Lillge assaulted her while calling her a “slut” and a “whore.” This
testimony was offered by the State as part of evidence of Lillge’s post-fire activities relating to the
stalking charge, but the evidence could also bear on the burglary and arson charges.

10
No. 2022AP192-CR

unexpectedly on in the house and certain items seemed to be missing or out of place.
When A.B. texted Lillge about this, he responded, “‘I already told you I don’t need
a key to get into the house.’”

¶25 After the fire, investigators found in Lillge’s bedroom four “bump
keys,” i.e., keys “typically used by a locksmith to gain access to a pin-and-tumbler-
style lock when [the locksmith lacks] the key … cut” for the lock.

¶26 A.B. testified that approximately six months before the fire, both A.B.
and Lillge smelled natural gas in the basement of the house. After reporting this to
the landlord, A.B. scheduled a time for a utility company representative to visit, but
then cancelled the appointment because she no longer detected the smell. At no
other time did A.B. smell gas in the house.

¶27 We turn now to testimony of a chronology of events more directly
related to the fire itself. A.B.’s daughter testified that, during the daytime on
December 20, 2014, she received a phone call from Lillge during which he asked
her where she, her brother, and A.B. would be that day. The daughter told him that
the children would be staying at a family friend’s house and A.B. “was going out
with friends, and … would be home later that night.”9 It was not unusual for Lillge
to ask A.B.’s children what they were up to generally, but it was unusual for him to
ask the daughter about A.B.’s whereabouts instead of asking A.B. directly.

¶28 Still on December 20th, between 4:00 and 5:00 p.m., a teacher of
A.B.’s daughter picked up both of A.B.’s children from their house to go Christmas
shopping and to buy groceries. After these errands, between 7:00 and 8:00 p.m., the

9
A.B. testified that she ultimately spent the entire night away from the house and, as
discussed below, did not return until after the explosion and fire had occurred the next morning.

11
No. 2022AP192-CR

teacher dropped them back off at their house, at which time the children took the
teacher around the house, including through the basement, to look at toys and art
supplies. While in the basement for about two or three minutes, the teacher, who is
generally familiar with the smell of natural gas, did not smell it. A.B.’s daughter
did not smell gas at this time either. At about 8:30 p.m., A.B.’s daughter made sure
that the lights in the house were off and that the doors were locked, and the teacher
took them to the family friend’s house for the night.

¶29 Lillge’s friend, Hall, testified that on December 20, at about 6:00 p.m.,
he went to Lillge’s house and they started drinking. Hall further testified that at
approximately 9:30 or 10:00 p.m., Lillge and Hall went to pick up another person
and then visited some bars in Watertown. An officer later testified that this was
inconsistent with what Hall told investigators: Hall said in the interview that he left
Lillge’s house between 7:30 and 9:30 p.m. to pick up the other person, but that
Lillge remained at his house.

¶30 Between 10:30 and 11:00 p.m. on December 20, someone who lived
in A.B.’s neighborhood was driving past A.B.’s house and came upon items of
clothing lying on the street. The neighbor testified that she drove around the clothes.

¶31 Hall testified to the following. At “bar time,” or about 2:15 or
2:30 a.m., he and Lillge separated, leaving Lillge’s truck parked by the bars.
Approximately 60 to 90 minutes later, Hall met Lillge at a mutual friend’s house,
and they walked to Lillge’s house, where they continued drinking. Hall last saw
Lillge at about 5:00 a.m., when Lillge was going upstairs, and Hall then passed out.

¶32 At about 3:54 a.m., Lillge sent two texts to A.B.: “Lmao still being a
hoe I hear. Goodbye sluts”; and “Great mother! Lmao not!” About two hours later
Lillge sent more texts to A.B., including “I’ll never understand the things that you

12
No. 2022AP192-CR

do …” and including a frowning, crying emoji symbol. Five minutes after this,
Lillge sent A.B. one more text that was lengthier:

I hope he realizes what he has and treats you better than I
have. You were my world and you deserve to be happy.
[I’]m deleting your number and will never contact you again.
Tonight, you with him… I can’t even breath[e] … you just
hurt me more than ever and I know you don’t care. I[’]m so
sorry for everything I just didn’t know how to handle losing
you but you just made it clear that this is who you really are
and I deserve someone with my morals, someone that values
their body and doesn’t just let anyone sleep with them. Good
luck to you and the kids. I love you 3 but cannot have you
in my life any longer. You hurt me to[o] much. God I hate
that you are doing this ….

(First ellipsis in original.)

¶33 A next door neighbor to A.B. (different than the neighbor who
testified about driving around clothing lying on the street that night), told an officer
responding to the fire that there were lights on in A.B.’s house at about 5:30 a.m.,
before the explosion occurred.

¶34 Shortly before 7:45 a.m., police and other first responders received a
call of an explosion and fire at A.B.’s house. Police arrived to find the house in
flames, with the back half “essentially demolished.” Responding officers noticed
and documented clothing lying on the street a short distance from the front of A.B.’s
house. A.B. later identified some of the articles of clothing on the street as
belonging to her or her children and said that these articles had been taken from the
basement of the house.

¶35 After firefighters and police responded to the scene, multiple
witnesses saw Lillge and interacted with him at the scene. Hall testified that
between 7:00 and 8:00 a.m., Lillge “came downstairs” and woke Hall up. Lillge
asked Hall for a ride to A.B.’s house after A.B.’s daughter and a firefighter asked

13
No. 2022AP192-CR

Lillge to come over. Lillge asked one police officer, who he happened to know, if
the officer knew how the house had blown up. Lillge later called the same officer
back over and suggested that investigators “check out the gas lines in the basement”
because A.B. and her aunt had smelled gas earlier in the summer. A.B. arrived at
the scene while Lillge was still there but the two did not interact. A.B. and her
daughter testified that they each observed Lillge “smirking” while looking at the
house and at times pointing at a hole in the roof.

¶36 Turning to the expert testimony presented at trial, the following is an
additional summary of Special Agent Pudlowski’s trial testimony. Pudlowski began
his examination of the scene on the day of the explosion and fire and finished it the
next day.

¶37 Pudlowski testified that the goal of a fire investigation is to use “a
scientific method in the collection of data[] and … conduct an analysis of the” scene
to determine the “origin and cause” of the fire. This involves assessing damage at
the scene by moving from the outside of the structure to the inside and from points
of the least damage to areas of heavier damage, narrowing down potential points of
fire origin. Using this process, Pudlowski concluded that the fire originated in the
basement with an explosion; he eliminated the other floors of the house and its
exterior (including the gas service meter) as potential points of fire origin.

¶38 Pudlowski eliminated several areas of the basement as potential points
of fire origin based on observable damage. He inspected the appliances in the
basement, which included two natural gas-powered appliances (a water heater and
boiler), and he determined that no appliance had been a source of leaking gas.

¶39 Pudlowski identified one natural gas line that had no cap, which had
previously apparently been used to power a dryer but was no longer hooked up to

14
No. 2022AP192-CR

any appliance. The uncapped line terminated in the basement and had a valve at its
end. Pudlowski’s discussion of the valve made clear that he determined that it could
be opened and closed by hand. Investigators found the valve in the fully open
position. Pudlowski did not determine the rate at which gas would move through
the uncapped line, specifically under the following pertinent conditions: the valve
was in the fully open position and the gas in the uncapped line was under pressure.
Pudlowski did not know how long the valve had been left in the fully open position.

¶40 Pudlowski removed the open valve and capped the uncapped line.
The gas service system for the house was then repressurized to determine if there
were leaks in the system, besides the valve when open. Pudlowski determined that
there was a leak in the piping extending from the gas service meter into the house
due to a “slight decrease in pressure in the system.” Pudlowski then conducted a
test using a soapy liquid to check the fittings in the gas piping inside the house,
which revealed a slight leak in another valve located on the branch of gas piping
that led to the water heater. This second valve was “an older style” that uses grease
to “complete the seal” on the valve when closed. It was possible that the fire burned
off the grease in the valve, but Pudlowski could not determine whether the leak
predated the explosion and fire. However, in any case, Pudlowski considered the
leak “[e]xtremely slow” and, even assuming that it predated the explosion and fire,
it could not have caused the build-up of enough gas in the basement to have caused
combustion given “the ambient flow of air in and out of that area.”

¶41 Pudlowski determined that the door leading to the basement was
closed and locked before the explosion. The closing of the door would have
prevented the gas—which is lighter than air and therefore tends to rise—from
escaping the basement. This would have led to the gas “pool[ing]” sufficiently in
the basement so that it exploded when ignited by one of a number of potential

15
No. 2022AP192-CR

sources, such as a pilot light. In sum, Pudlowski concluded that the origin of the
fire was gas flowing through the fully open valve on the uncapped line into the
basement.

¶42 Pudlowski further testified that the next step in a fire investigation is
to classify the cause of the “fire event” as either “natural, accidental, incendiary, or
undetermined.” Pudlowski classified the fire as “incendiary,” meaning it was
intentionally caused through a “human act,” specifically someone fully opening the
valve on the uncapped line. Pudlowski testified that there was no evidence of a
natural cause, notably a lightning strike, and he did not find any evidence that the
valve in the basement was turned to the fully open position only by accident.

¶43 Pudlowski further opined that the clothing observed lying on the street
did not land there as a result of the explosion. This opinion was based on the
observation that “there was very little debris outside the backyard” and his inability
to understand how, as a result of the explosion, “clothing would make, essentially,
a 90-degree turn outside of that debris field” and end up landing down the street that
adjoined the front yard of the house.

Evidence At Postconviction Hearing

¶44 Because we resolve this appeal based on the prejudice prong of an
ineffective assistance of counsel claim, we do not further address the postconviction
testimony of Lillge’s trial counsel.

¶45 Agosti testified in pertinent part as follows. To repeat, Agosti
concluded that the fire should have been ruled “undetermined” and not
“incendiary.” He explained that the undetermined classification applies when more
than one of the other potential classifications cannot be ruled out based on the

16
No. 2022AP192-CR

evidence. Agosti opined that several deficiencies with Pudlowski’s fire
investigation resulted in Agosti’s inability to rule out several potential accidental
fire causes.

¶46 Consistent with his written report, Agosti testified to purported
deficiencies in Pudlowski’s investigation. Pudlowski, in his testimony responded
to each purported deficiency, followed by further testimony from Agosti. We boil
down the back and forth to the following.

 Agosti: Pudlowski failed to personally conduct, or become familiar with
others conducting, bar-hole testing of the ground around the gas piping
from the street to the gas service meter on the house.10 But Agosti
acknowledged receiving for the first time at the hearing what purported
to be a utility incident report that states the utility conducted bar-hole
testing at the house following the fire. Agosti then clarified that his
opinion was limited to the criticism that Pudlowski “had not followed the
scientific method” because Pudlowski apparently formulated his opinion
regarding the explosion’s cause before he knew that the bar-hole testing
was apparently done.

Pudlowski response: Utility companies routinely conduct such testing
when responding to a gas explosion. Here, if such testing had raised any
concerns, Pudlowski would have expected the utility company employee
to bring them to Pudlowski’s attention.11

 Agosti: Pudlowski did not conduct a “flow test” to determine if the fully
open valve on the uncapped line actually permitted the flow of gas,
raising the possibility that there were disconnects or shutoff valves
“upstream” in the piping that prevented gas from flowing through the
uncapped line and open valve. However, Agosti acknowledged that if
Pudlowski’s report could be taken as true in representing that there were

10
There is no dispute that bar-hole testing involves physical tests to rule out the possibility
that gas leaking from piping is seeping into the ground and potentially moving from the ground
into an adjoining structure.
11
Pudlowski testified that he observed flags and paint on the ground at A.B.’s house,
which in his experience are used by utility company employees to mark the location of their utility
lines and which “would have been done prior to the bar-hole testing.” However, Pudlowski testified
that he did not actually observe anyone conduct bar-hole testing.

17
No. 2022AP192-CR

no such shutoffs that would undermine Agosti’s concern that the flow of
gas could have been interrupted.

Pudlowski response: Pudlowski noted the representation in his report
indicating that his inspection revealed no disconnects or shut-off valves
on the uncapped line between the gas service meter and the open valve.

 Agosti: Pudlowski worked with a technician employed by the gas utility
company instead of with an independent engineer, which created a
conflict of interest given that the utility company could have liability for
an accidentally caused fire.

Pudlowski responses: Forensic fire investigation standards do not require
consulting an independent engineer, doing so is not always feasible given
limited resources to conduct an investigation, and in any case the failure
to work with an engineer here did not cause Pudlowski’s investigation to
be biased or incomplete.

 Agosti: Pudlowski did not properly test the piping leading to the gas-
powered appliances or piping and components of the appliances, nor did
he adequately test the gas-powered appliances due to a safety mechanism
that would have prevented the pressure test of the overall system from
revealing leaks “inside of the appliances.” Agosti could not determine
from Pudlowski’s report whether Pudlowski properly inspected
components of the appliances.

Pudlowski responses: The gas-powered appliances were tested as part of
the pressure-test of the overall system and no leaks were discovered.
Further, Pudlowski inspected “the components” of each appliance for
“damage or obstruction.”

 Agosti: Pudlowski failed to consider that a repair to the gas-powered
boiler two years before the explosion could have created a leak that
warranted further testing of the appliances.

Pudlowski response: The maintenance involved a repair of the
thermostat, which Pudlowski determined could not have been related to
the explosion.

 Agosti: Pudlowski failed to sufficiently consider the report of the smell
of gas in the basement six months before the fire.

Pudlowski response: Pudlowski did consider the report but determined
that it was not significant given testimony that there was no smell before

18
No. 2022AP192-CR

or after that incident, notably including on the day and night leading up
to the explosion.

 Agosti: Pudlowski did not remove the gas-powered appliances and gas
piping for preservation or provide more thorough documentation than a
sketch that Pudlowski made, which Agosti characterized as “not
discernible,” and failing to completely account for all gas piping in the
house.

Pudlowski response: The results of Pudlowski’s pressure testing and
inspection of the appliances made it unnecessary to remove the
appliances or additional piping for further testing. Pudlowski’s field
notes and photographs of the gas piping system accounted for all piping
in the house.

 Agosti: Pudlowski did not generate a “debris field diagram” detailing
how far and in what direction explosion debris traveled from the house or
additional information about how far the clothing found in the street in
front of the house was from the part of the house that was most damaged.

Pudlowski response: Pudlowski’s opinion regarding the clothes was not
based on the distance traveled by the clothes. Instead, it was based on:
the direction the clothes would have had to travel relative to the directions
in which all the other explosion debris went; the particular side of the
house that was damaged in the explosion; and the estimated strength of
the explosion. Based on those factors, it was unlikely that the clothes
would have landed on the street located on the opposite side of the house
from the explosion damage, particularly given that there was no explosion
debris in the front yard, adjacent to the street where the clothing was
found.

¶47 On cross-examination, Agosti clarified that his opinion that the fire
should have been classified “undetermined” was not based on the conclusion that
the evidence could not support the reasonable inference that the fire was
intentionally caused, such as by someone intentionally opening the valve on the
uncapped line and closing the door to the basement.

19
No. 2022AP192-CR

II. Prejudice Prong Analysis

¶48 We conclude that, if the defense presented at a new trial the additional
expert testimony that Agosti provided at the postconviction hearing, it could make
a conceivable difference, but not a substantial difference. We reach this conclusion
in light of two factors, taken together: the limited nature of Agosti’s testimony and
the strength of the evidence pointing to Lillge having entered A.B.’s house without
her consent, opened the valve on the uncapped line, and closed the door to the
basement.

¶49 We begin with the narrow and limited scope of Agosti’s relevant
conclusions. To repeat, he contended that there may have been alternative,
accidental causes for the pooling of gas in the basement, leading to the explosion
and fire, which were not properly ruled out by Pudlowski’s investigation.
Specifically, Agosti testified that the open valve on the uncapped line may not have
been the source of the gas that caused the explosion. But even in this regard,
Agosti’s testimony ultimately proved narrow and limited.

¶50 Regarding the presence of potentially hidden shutoff valves or
appliance leaks, Agosti at most purported to raise issues of credibility regarding
Pudlowski’s testimony based on alleged incompleteness in Pudlowski’s
investigation, not actual errors, erroneous assumptions, or a faulty understanding of
any scientific or engineering principles. Agosti’s criticisms regarding these points
largely focused on an alleged lack of detail in Pudlowski’s documentation, but did
not come to grips with the substance of Pudlowski’s testimony regarding what
Pudlowski represented were his personal observations. For example, Agosti’s
testimony regarding the flow test focused on the possibility of additional shutoffs
that Pudlowski testified were not present based on his personal investigation, and

20
No. 2022AP192-CR

Lillge does not point to any testimony by Agosti or other evidence suggesting that
some other type of obstruction could have prevented gas from flowing to and then
through the open valve on the uncapped line.

¶51 Lillge emphasizes Agosti’s point that Pudlowski did not conduct a test
to determine if gas could and did flow through the open valve on the uncapped line.
But Agosti’s postconviction testimony on this point was limited to the notion that
there may have been other shutoff valves further upstream in the gas piping, a point
we have just addressed. More generally, Agosti conceded explicitly that it was not
his opinion that the fire could not have been intentionally set, and he conceded
implicitly that a factfinder could infer from the open position of the valve coinciding
with the occurrence of the gas explosion that the valve had been intentionally left
open.

¶52 Regarding bar-hole testing, we assume without deciding that the State
did not establish that bar-hole testing was performed that confirmed there were no
ground leaks. Still, it remains that Agosti’s testimony did not raise a significant
issue about a possible utility line leak from the ground into the basement.

¶53 Echoing a point from Agosti’s testimony, Lillge argues that it
prejudiced him that trial counsel did not cross-examine Pudlowski regarding
Pudlowski’s failure to “engage the services of a qualified forensic engineer” to assist
in the investigation because of the utility company’s inherent conflicts of interest.
But this is mere speculation. Lillge does not identify evidence that any meaningful
act or omission of the utility worker who Pudlowski testified assisted him caused
Pudlowski’s investigation to be incomplete or biased.

¶54 Agosti acknowledged that his review was limited to studying the fire
investigation and expert testimony presented by the State at trial. In doing so, Agosti

21
No. 2022AP192-CR

further acknowledged that he did not take into account much of the incriminating
evidence from sources other than what Pudlowski presented at trial. For example,
in challenging Pudlowski’s testimony regarding the clothes lying on the street that
night, Agosti did not weigh the testimony of the neighbor who observed clothing on
the street hours before the explosion occurred.

¶55 Turning to the incriminating evidence presented at trial, as
summarized above, that evidence included the following, which if credited by the
circuit court was strong evidence for the prosecution: testimony from A.B. that
Lillge was strongly motivated by what he described as feelings of jealousy and was
capable of becoming violent; Lillge’s statements that he did not need keys to enter
A.B.’s house, signs that he entered the house without consent, and his possession of
bump keys; Lillge’s text messages over the night of the fire showing that his
jealousy motivations were driving his actions during the pertinent timeframe; the
clothing seemingly thrown onto the street before the explosion in a way that was
consistent with Lillge’s behavior during prior conflicts with A.B.; even assuming
that Hall was a credible defense witness, the inability of Hall to account for Lillge’s
whereabouts at multiple points during the evening that would have given Lillge time
to walk or drive to A.B.’s house; and Lillge’s conduct upon arriving at the fire scene.
All this was combined with evidence that more generally appeared to establish the
intentional use of the uncapped line to concentrate gas in the basement to eventually
cause an explosion: the fully open position of the gas valve, with the door to the
basement being closed before the explosion, and multiple witnesses reporting that
they did not smell gas in the basement earlier in the day or evening of the 20th.

¶56 To be clear, we do not understand either party to challenge the
postconviction court’s findings that both Agosti and Pudlowski were generally
credible and reliable witnesses, nor do we have an independent reason to question

22
No. 2022AP192-CR

those findings. Rather, our analysis is based, in part, both on the ultimately limited
nature of Agosti’s differences with Pudlowski and on Agosti’s indisputably limited
ability to contest testimony based on personal observations of other witnesses,
including Pudlowski.12

CONCLUSION

¶57 For all of these reasons we affirm the judgment of conviction and the
circuit court’s order denying Lillge’s postconviction motion for a new trial.

By the Court.—Judgment and order affirmed.

This opinion will not be published. See WIS. STAT.
RULE 809.23(1)(b)5.

12
Lillge also argues that this court should order a new trial in the interest of justice under
WIS. STAT. § 752.35 because the issue of whether the explosion and fire were intentionally caused
was not fully tried. “The power to grant a new trial when it appears the real controversy has not
been fully tried ‘is formidable, and should be exercised sparingly and with great caution.’” State
v. Sugden, 2010 WI App 166, ¶37, 330 Wis. 2d 628, 795 N.W.2d 456 (quoted source omitted).
We exercise our power to grant a discretionary reversal only in exceptional cases. Id. For the
reasons stated in the text, we conclude that this is not an appropriate case to exercise our power of
discretionary reversal.

23