What is a Big Data Study and How Does it Help with Trial Strategy?

Most attorneys are familiar with focus groups and mock trials. Perhaps you’ve heard about a big data study but don’t know what it is. Perhaps you’ve run one but are still learning how to maximize the impact of the information you gleaned from it. In this post, I will go through what a Big Data study is, how it is different from a focus group or mock trial, and what it is best used for.

Until recently, the only method we had for testing a case was a traditional focus group or mock trial. These tools are still invaluable for testing the case, finding holes, case framing, and guiding you through discovery. But they always had limitations, namely that they could never predict the outcome of trial. Prediction was never their purpose: the purpose was stress testing and finding themes and worst-case scenarios. If any consultant ever told you that they could tell your chances of winning or losing, the value of the case, or a juror profile based off anything less than 6 panels of focus groups (and more like 20 panels for a juror profile) at a bare minimum, they were lying to you. The participant numbers in a focus group or mock trial are simply too small to have any way to determine statistical significance.

Enter, Big Data studies. Big Data fills the gaps that focus groups and mock trials could never fill and these tools all work synergistically together.

So, what is a Big Data study? While some attorneys call them online focus groups, this is somewhat of a misnomer. You won’t see these jurors or talk to them and they won’t talk to each other (this is their main limitation – you can’t get a good feel for WHY people feel a certain way or how they communicate with each other in a Big Data study). A Big Data study is really an online survey. Depending on who runs it (more on AI focus groups/data studies in the next post – be sure to read!), these should be sampled from real human beings and preferably in your venue or similar venues. The studies I run are through Campbell Law, so I will discuss how those are run as I personally believe they are the best quality Big Data studies.

What do jurors see and react to? 300+ jurors read scripts drafted by you (and edited/collaborated on by me if you run them through me) that lay out the plaintiff and defense arguments, including photos and video. Throughout, they are given attention checks to ensure they are paying attention. They answer 50+ juror intake questions, some of which I personally customize based on the case, and then answer leaning/attitudinal questions at the end along with a verdict form.

Importantly, jurors are broken into 3 groups of approximately 100 each. All jurors review the same information except that they are presented with different damages asks at the end. So, one group might be asked to award $10M, the second group asked for $25M, and the third group $40M. You might also add extra groups of 100 jurors to test different variables: what if defense admits liability? What if the video is in or out? What if a party settles? All of that information is analyzed and returned to us in a 60-70 page report.

What is in the report? The report contains a wealth of information, so let’s break down the important parts:

Win Rate Based on Damages Ask: The first thing you will see in the report is your win rate based on what amount jurors were asked to award and any other variables tested. Perhaps surprisingly, your chances of winning or losing can often depend on the number you throw out at jurors. Below, you can see three different damages asks as well as a test with one party removed (NoHouston). Since the high ask tested best overall, we then ran that same High ask with Houston out of the case so we can compare whether having Houston in or out affects the verdict. This chart shows that the best ask is the High ask at an 89% win rate and that if Houston is out of the case, that greatly hurts our chances of winning as the win rate drops by 13%.

Likely Verdict: The next thing you will see in the report is the amount of money you are likely to recover (which accounts for defense leaning jurors and the deliberative process) on your best day, worst day, and average/median. Here is the same case. You can see the average expected recovery is about the same from low to medium, though the medium drops slightly, and then climbs again in the high, but drops if Houston is out. Perhaps shocking is that the best case and worst case scenario is better in the low ask and high ask but worst in the medium ask. This shows how important the damages ask is. Selecting a number in the middle can tank your case. While the best case scenario is higher without Houston, it’s not worth having Houston out of the case as the win rate is lower and average recovery is also much lower:

Themes: The report will give you some information on framing. Jurors are asked to rank the two most important issues in making their decisions. Those are tallied and ranked to show you the most impactful themes, biggest danger points, and what potential issues rank low (and therefore should be eliminated from your main themes). While this is ranking is mostly accurate, there is a limitation in that humans are notoriously horrible self-reporters. Therefore, a juror may list something as the reason they came to their decision when in fact something else entirely was driving it. This is where I like to add in focus groups and mock trials to hear jurors talk. In watching their thought process, you can often determine if something hidden is driving them or if some other issue was more important but perhaps seemed so obvious that they skipped past it. Here are samples from a hospital sexual abuse case showing the top themes for plaintiff jurors and for defense jurors:

How Much Does It Matter?: The report will tell us what jurors think about the case issues (for example, “I think the defendant’s motive was pure but they were just being careless”) as well as how much those attitudes and beliefs matter in terms of your win rate or damages. The charts on these tend to be hard for attorneys to read and decipher, so I won’t share the chart, but in simple terms, it will tell us how much that sentiment mattered for win rate and damages. Someone who agrees the defendant was careless but had pure motives, for example, may be 60% worse for you on liability compared to the average juror in the study and may give 45% less money. Consider a statement “I think the plaintiff was likeable” garners 50% agreement and 50% disagreement. That means half the jurors dislike your plaintiff. But, consider if the data shows that you are still winning over the jurors who dislike your plaintiff. That may mean you don’t have to worry about your plaintiff being unlikeable as it’s not affecting win rate or damages.

Juror Profile: In focus groups and mock trials, we always warned against profiling jurors. It’s not always true that conservatives are bad for you or certain ethnicities and ages are good or bad; but, sometimes it is! The trouble with smaller sample research is that there is no way to know. If we run a four panel mock trial and 4 of the 5 African-Americans are bad for you, does that mean you should be wary of African-American jurors? Perhaps, but perhaps you simply flipped a coin too few times for the average to come out to 50%. If your first 4 flips are tails, it would be a gave mistake to assume a coin flip is 80% tails over time. We need over 200 jurors to find enough data points to compile a juror profile. Sometimes, the strangest things show up, such as people they admire (in one case, those who admired Angelina Jolie were 167% better on damages for us) or TV shows they regularly watch (NFL football watchers was a -44% on liability on another of my cases). But we also see traditional data points show up, such as tort reform attitudes or case-specific backgrounds and beliefs. This can help aid the jury selection process in addition to more holistic questioning. Here is a sample of a scoring sheet I compiled for a client based on their data. In this case, we had very little time for voir dire, so we ran through as many of these questions as we could in the time permitted:

Where Is Data Weak?

Data studies are powerful assets to trial preparation, but they do have some limitations. Personally, I need to hear people talk. It gives me a 3-D view of what the data is showing and allows me to better frame the case. As mentioned, you also have to be aware of self-reporting biases, which is again where mock trials and focus groups come in useful.

Data cannot capture strong personalities of counsel or defendants of whom you do not have video evidence. I had one case that I’m convinced we won due to the horrific personality of the defendant (along with a strong jury selection).

I also worry about really complex cases in data studies as it can be hard to determine when jurors are simply misunderstanding or not entirely following an argument. Mock trials and focus groups would reveal this, but it can hide in the data.

How Do I Use Data?

In very large cases (8-figures and above), I strongly prefer a mix of mock trials and data studies. I believe that captures the full picture and allows you the best all-encompassing view of framing as well as prediction. In 7-figure cases, we may run a smaller focus group and then data. In other cases, we may pick one or the other depending on what is most advantageous to the case.

I have a relationship with Campbell Law in which I can run data for/with you for the same cost as you would pay going direct to the Campbells. This is due to our synergistic relationship in that I can help prepare the data scripts with you, hand off a polished script to them to run, and then help you to understand the report at the end. They do less work, you get a consultant for free, and it gets me involved in case which may need further help such as jury selection or focus groups and mock trials. In fact, it often saves money in the end to have me run the data as it will decrease intake fees for me to get up to speed on the case if you need my assistance otherwise.

Stay tuned for the next post regarding AI testing of cases. If you have not yet subscribed, you may subscribe to this blog to get automatic updates in your email when I post.

If I can help with your case, please contact me at: Jessica@TrialDynamics.net

Leave a comment

Filed under Uncategorized

The Importance of Question Order in Jury Selection

Did you know that you could undercut your trial theme simply by the order in which you’re asking voir dire questions?

You have probably put thought into the content of your questions, but if you don’t consider the order, you’re missing a big part of the message and potentially sending the wrong message to your jurors.

Order tells jurors what you think matters most, what you’re worried about, and how confident you are in your own case. It gives them a sense of what the case is about before the evidence is put forth.

If you start your jury selection with your biggest issues, jurors may wonder if you have a good case at all.

Here’s a simple example. Say your client wasn’t wearing a seatbelt. Jurors may or may not know that from the mini opening or neutral statement. But either way, if you bring it up early in voir dire, you’re putting weight on that issue. You’re telling the room, “This is the thing I’m most concerned about.”

Consider instead putting questions about your themes first that will give jurors context.

Start with the product defect. Talk about what the company knew and when they knew it. Ask jurors how they feel about companies that cut corners when safety is on the line.

Then, once jurors have context for the case, work your way into the seatbelt issue. Now it’s one factor among many, not the headline. Jurors will think, “maybe the seatbelt wasn’t that important here.” Or, “maybe it didn’t matter if the car hadn’t failed.”

Same questions. Completely different message. If you put your worst facts first, you may eliminate plenty of jurors who would have been good for you if only they had some context.

If comparative fault is a secondary issue, it should feel secondary in the way you sequence it.

You’re not just gathering information in voir dire: you’re setting themes early on and building a framework for how jurors will process everything that comes after. Just like in opening and the rest of trial, order matters.

Leave a comment

Filed under Uncategorized

Mastering Jury Selection Strategies

Jury selection is about far more than finding and striking “bad” jurors. You should have several concerns when making strikes during jury selection.

Don’t be fooled by the defense strategies as they have an entirely different end game. They don’t need to win your whole jury; they just need one holdout. But plaintiffs don’t have that luxury. You need consensus.

That means jury selection isn’t just about removing bad jurors. It’s also about:

* Identifying leaders. A juror who says “I think companies should be held accountable” is helpful, but is that something you had to pull out of a quiet juror or is that coming from someone who is more outspoken and has a presence in the room? Do other jurors on the panel nod along? Which jurors are opening doors for others, helping to keep other jurors on schedule, coming into jury selection dressed up? Beware, however, that a loud but obnoxious juror is rarely a leader; they are a liability.

* Group dynamics. Find jurors who will work together as a group, particularly when there are a few strong personalities. If you have two leaders who will butt heads, you could end up with a hung jury and will lose cohesiveness. Look for jurors who are already forming a bond. When you’re doing side bars, who is talking with whom? Which jurors are keeping to themselves and refusing to become part of a group? Who do jurors talk with at breaks in the hallways?

* Identifying good jurors. While the main purpose (usually) is to find bad jurors, you also want to identify good jurors who will fight for you, rather than simply being neutral. While you can’t actively choose to keep someone, knowing who is good for you can play into your strike strategy. Sometimes, it’s better to leave a couple of somewhat bad jurors on the panel if you think you are able to save a couple of really good jurors. You would likely get a better outcome with two “A” jurors, a few “C” jurors, and one “D” juror than you would with all “C” jurors depending on leadership qualities. Do not leave bad leaders on the panel even if it means keeping a good juror.

If you are focusing only on bad jurors and failing to consider group dynamics and strike order/strategy, you’re missing half of jury selection.

Leave a comment

Filed under Uncategorized

Don’t Play in the Defense’s Sandbox: How to Properly Frame a Case

My client had 7-10 theories for the case. I told them to cut 80% of them.

More is not better.

More is the game of the defense.

If everything is important, then nothing is important.

In this case, the defendant had failed to check his tractor trailer before leaving, failed to fill up on fuel, failed to stop and refill, failed to find a truck stop as a safe place to stop, ignored low fuel lights, ignored his check engine light, and ended up stranded on the side of a narrow shoulder with his back end sticking out by a couple of feet into the slow lane. Our client (another tractor trailer) was left with a split second decision to avoid the truck and swerved into the wall, which killed him.

The defendant had used red fuel to avoid fees, failed to register his vehicle (and therefore was never inspected), and then claimed the truck died due to a sudden mechanical failure.

There were days of testimony to prove the cause of the crash (mechanical vs running out of fuel), whether the fuel gage was broken or not, whether it was an accurate read.

There were negligent hiring and trianing claims. Claims about whether warning triangles were required to be placed (or whether there was no time), whether his blinkers were on, whether he even had warning triangles in the truck.

The case became too complex. We didn’t have to fight causation if we framed it right.

The defendant ignored warnings. Period. He ignored low fuel lights and check engine lights. Whether the truck ran out of fuel or died from mechanical issues is irrelevant. Whether he had triangles and should have put them out doesn’t matter: had he pulled over when warning lights were buzzing and blinking, a life would have been saved.

If you focus on everything, jurors will lose sight of the end goal. The risk here was that jurors would decide the case based on the reason the truck stopped. By removing that as an issue, we refocused the jurors.

Those jurors walked in with their verdict wearing 9/11 shirts to honor our deceased who was a veteran.

Keep it simple.

Leave a comment

Filed under Uncategorized

How did we get a $10 million verdict for broken heel with $10k in meds?

A week before Thanksgiving, a trial team and I won a $10 million verdict in a case involving a broken talus bone in a conservative county (Indianapolis) with $10,000 in past medical bills (which we fought to keep out but which defense was allowed to bring in). This was a hard fought battle with a tough judge and some really bad rulings, but we prevailed with grit and foresight.

The case involved an admitted liability trial against Frito Lay. Plaintiff, a 47 year old female, shattered her talus bone in the wreck.

Her talus bone healed but post-traumatic arthritis became apparent later and progressed quickly from mild to severe over the course of two years to the point that she can no longer walk more than a block or two, cannot stand for longer periods of time to cook meals, cannot fall asleep until 4am due to the pain, and cannot sit to work longer than an hour or two without tears.

Her physicians all opined she needed a subtalar fusion which she had yet to get due to high blood pressure that predated the wreck but that she could not control post-wreck due to inability to exercise.

Defense tried to use blood pressure as its biggest red herring throughout the trial and said the fusion would fix it all. We argued that there was no magic pill for her and even if she got the fusion, she would forever have pain.

Paid medicals were $10,000, incurred medicals were $45,000. We wanted to waive these but defense was allowed to bring them in.

Defense IME opined that plaintiff had a new injury during the recovery period. Defense pulled the plug on him and didn’t call him to testify which likely backfired as he was their star witness.

One of the big claims for damages was the fact that the plaintiff was unable to complete her doctorate and work. The pain in her foot is so severe that she cannot sit for more than a couple of hours without tears and can’t walk without assistance. She had dreams of starting programs to help aged-out foster children which would help heal some of her wounds from her broken and abusive childhood. Lost earnings were $600k, but the loss was so much more than that. She lost her passion and purpose.

Defense economist refused to acknowledge that chronic pain and loss of sleep could make it more difficult to focus or sustain employment long term and that she could sit to work but she broke down in tears during testimony and needed a break. She was far from a complainer and was a stoic woman with a tough life. Jurors knew this was not theatrics.

The judge was so harsh on my attorney clients that they almost called off trial days before. “We can’t get a fair trial with this judge,” they said. After talking it through with them, I convinced them to push forward. Trials are never clean, judges often make bad rulings, and yet it all comes down to the story, the client and the jury. I knew we had a great client, a compelling story, and we had to get to trial to see the jury (who also turned out to be fantastic).

The judge denied demonstrative animations, including a surgical animation that even defense agreed was admissible. In opening, he cut us off on any part of undermining defense contentions stating that we could not preview the defense case at all. During jury selection, he did not seem to properly account for juror numbers and denied a very clear cause challenge (while granting a lesser one for defense).

He was good at engaging jurors, however, and did calm down when we got moving.

Final offer during trial was $200,000. The jury took 90 minutes to come forward with a $10,000,000 verdict.

I helped with case framing, witness prep of the client and her family members, review of client-conducted focus groups, aide in opening statement and closing argument preparation, and select the jury. Brandon Yosha of Yosha Law and his team took my advice and eagerly applied it. They weren’t afraid to put in the hours or change course and the results speak to that.

I’m very proud of this one: it could have been a case about a broken heel and speculative surgery but jurors knew it was about so much more.

A note about trying a case during holidays: trial timing is rarely convenient. But when your closing argument lands during the holidays, you can either ignore it or use it to anchor the jury in the real-world human consequences of their decision.

In closing, Brandon went through how long the verdict has to last. By showing a timeline of past events, he was able to get jurors to realize how long 40 years is. This gets jurors primed to think of the consequences of their verdict over time.

Since we closed a week before Thanksgiving, I told Brandon to add this:


“Thanksgiving and the holidays are upon us and we will all go back to our families and friends. In twenty years, when Thanksgiving rolls back around, maybe you will think of Karen and your verdict. I hope that you will be proud of the decisions you make here today and look back without any regret. Because somewhere, at that very same moment, at some other Thanksgiving table, Karen will be living out your verdict. What does that look like for her? That’s up to you.”

If you want to hear more about this trial, we will be speaking on Trial Lawyers University on December 18th at 10:30am PST and it will be recorded and available on demand thereafter.

If there is a case I can help you with, please reach out. I’d be honored to help.

Leave a comment

Filed under Uncategorized

How do AI jury models stack up?

AI is ever evolving, but for the time being, it is not proving to be accurate in predicting case outcomes. Credit to Nick Schweitzer at Trial by Data for running all of these tests on various AI models which show results are all over the place.

In this first image, you can see the win rate comparisons between AI and Big Data (with real people as jurors). I will add that I also ran a 4-panel mock trial on this case before we ran data and can attest to how strong the case was with jurors (though mocks are not statistically accurate):

Big Data win rate: 97-98%
AI win rates: 20-85%


In this next image, you can see the awards. This case went to trial so we can tell how accurate data was compared to AI:

Big Data predicted value: $22.8M
AI predicted value: $1M-$20M (I believe Nick asked the $1M AI to “think harder” and then it produced $100M!).
Trial result: $25M

We also have testing of some of the big named AI jury platforms which I won’t mention by name but which are not testing any better despite their claims of being accurate.


At the time being, AI is off by staggering amounts. It may not always be this way and I will stay on top of it but at this time, I cannot recommend AI testing of cases.

Leave a comment

Filed under Uncategorized

JuryBall Free Consulting

Are you going to JuryBall but don’t want to golf on Sept 30th? Would you like some free jury consulting advice?

I’ll be coming in on September 29th but I’m not a golfer, so to take advantage of the time, I’m offering to meet up with anyone else who will be there but not golfing all day.

Bring a case or general questions. Let’s talk shop! Message me to coordinate: Jessica@TrialDynamics.net

Leave a comment

Filed under Uncategorized

Big Data Just Got Better

Have you ever run a Big Data study (you can skip to the “win-win-win” section below)? Considered running a Big Data study? Have no idea what a Big Data Study is or how it can benefit you? I have exciting news. You can now run a data study through the Campbells AND have a consultant on board to help prepare the study (and thereby familiar with the case if you need additional help) for no additional cost.

What if you could predict your chances of winning a case, estimate the percentage of fault attributed to your client, or forecast the likely verdict? In today’s legal landscape, harnessing the power of big data is one of the most effective ways to gain a strategic advantage at trial or mediation. By combining vast datasets with advanced analytics, big data studies enable you to anticipate trial outcomes, understand juror profiles, and make data-driven decisions to optimize your case strategy.

What Big Data Offers

Big data provides unparalleled insights into factors that can influence a case. By analyzing large volumes of information, big data allows you to:

  • Identify potential biases based on demographic or attitudinal factors.
  • Predict case value and win rates with statistical accuracy.
  • Gauge prevailing attitudes toward specific legal issues.
  • Develop juror profiles for crafting tailored strategies in voir dire.

At Trial Dynamics, we are proud to partner with Campbell Law LLC, a leader in big data analysis, to deliver reliable data combined with actionable insights and implementation strategies tailored to your case.

What is Big Data, and Why is it Useful?

Big data refers to large-scale data collection and analysis designed to identify trends and predict outcomes. Corporations use it to anticipate consumer behavior, and now these methodologies are being applied to trial strategy.

Through big data studies, we test your case with hundreds of mock jurors to determine your best, worst, and average day in court. For example, we can simulate outcomes based on varying damages requests:

Damages AskWin RateAverage Verdict
$10 million75%$5.5 million
$40 million78%$30 million
$75 million60%$35 million

In this scenario, requesting $40 million yields the best result—a high win rate and substantial verdict—demonstrating how big data can inform optimal case strategies. The highest request of $75 million drops the win rate significantly while yielding only slightly higher damages. Without this information, you may go into trial asking for too little, thereby forfeiting tens of millions of dollars, or ask for too much and lose the trial.

Juror Profiling Through Data

Big data also provides detailed juror profiles by analyzing responses to extensive intake questionnaires. These insights go beyond surface-level observations to identify traits correlated with specific outcomes. For instance:

  • NFL fans might be 47% less likely to find in your favor.
  • Regular viewers of Jeopardy! may be 35% more inclined to award large damages.
  • Those who have been victims of a crime may be 117% more likely to find in your favor.
  • Those who have suffered similar injuries may be 79% lower on their damage awards.

Additionally, we can incorporate anticipated voir dire questions into the study to test which questions best predict liability or damages outcomes, giving you a roadmap for jury selection.

WIN-WIN-WIN

While you can run a data study straight through the Campbells, there is now no reason to do so. You can have a consultant on board to help with the data preparation for no additional cost (in the vast majority of cases).

While Campbell Law LLC provides technical expertise in data gathering and analysis, having a consultant invested in the study and implementing findings is crucial. At Trial Dynamics, we partner with Campbell Law to offer seamless support, from preparing scripts and exhibits for data studies to integrating findings into your case strategy.

Our process includes:

  • Adding tailored voir dire questions to test their predictive power for liability and damages outcomes.
  • Editing case presentations and exhibits before submission to Campbell Law.
  • Translating Campbell Law’s comprehensive data report into actionable strategies for opening statements, voir dire, and trial presentation.
  • Providing ongoing support, including jury selection, focus groups, mock trials, witness preparation, case framing, and opening statement development to ensure every insight is fully implemented.

While big data will provide juror profiling, deciding what weight to give a juror’s “score” compared to what they say in open court is an art. Data is just that – data – but the implementation of the data is much more nuanced. We understand and listen to the data, but also recognize the need to listen to jurors and make strategic decisions during jury selection that is not accounted for in the data results.

Further, if your case requires both mock trials and big data studies, involving us early allows for streamlined coordination and a consistent strategy from start to finish.

With big data studies powered by Campbell Law LLC and our tailored consulting, we offer a sophisticated, data-informed approach to jury consulting. By bridging raw data insights with practical trial strategies, we help you make confident, informed decisions that maximize your chances of success—whether at trial or mediation.

If you are considering running a Big Data study or have questions, contact me. There really is no downside.

Jessica@TrialDynamics.Net

303-653-2233

Leave a comment

Filed under Uncategorized

Openings: Less is More

Opening is NOT your opportunity to spew all your facts. Not even all your good facts. Jurors (humans) can only absorb so much at once. If you turn on the firehose, they will drown.

In your opening, aim to (1) capture their attention, (2) keep their attention, (3) tell them only what they need to know to decide the case in your favor (all other good facts that do not relate to your claims should be presented later or, gasp, not at all!), (4) tell a story in a sequence which allows jurors to know what they are looking for before you feed them important facts. Within the opening, you also need to counter defense points, but in a subtle manner that does not cause jurors to feel that you are playing defense or putting too much emphasis on the issues in your case. Otherwise, you may give jurors important facts and they will miss them because they had no groundwork to recognize they were importantly. Remember, you have lived the case. They know nothing.

The opening I crafted two nights ago was for a medical malpractice case and took no longer than 10-15 minutes to deliver. We cut over 45 minutes worth of information and I guarantee it will be multitudes more impactful.

If you need help with opening, framing, jury research, jury selection, or voir dire prep, reach out.

Leave a comment

Filed under Uncategorized

Get More Out of Your Data

Did you know that you can have me consult on your big data at no extra cost?

Really, there’s no catch.

In addition to my normal consulting practice, I also work regularly with the Campbells. They’re overwhelmed with work and send scripts to me to edit. I send them studies to run. We work as a team (and even better, we love working together – yes, you can fact check me by asking them).

Therefore, if you come through me to get to the Campbells, you can have a consultant on the case for no extra fee. By working as a team with the Campbells, I’m taking work off their plate, so our joint bill ends up the same as if you went to them directly, except now you have a jury consultant on the case.

This is a WIN-WIN-WIN (the best kind of business deal). A win for me as I make money, a win for Campbells who are inundated with work, and a win for you and the client. No one loses.

I will help to edit your scripts and think ahead to jury selection, adding intake questions to the preset questions the Campbells use. I can then help on the back end to strategize your case, run focus groups, or help select a jury using the data and my own intuition/knowledge. I’ll already be familiar with the case and be a part of your team having run the data with you.

And, if you want to do contingency, I also pair up with the Campbells on the right case.

If you plan to run a data study, reach out. There’s really no excuse not to. I look forward to helping you!

Leave a comment

Filed under Uncategorized