Categories
podcast

How Pete Blair got 100% accuracy in a 2014 deception detection study 

In a 2014 deception detection study, Pete Blair made 33 judgments—and got all 33 right. That paper was titled “Expertise in deception detection involves actively prompting diagnostic information rather than passive behavioral observation.” I ask Pete about what the factors were in him getting 100% accuracy in that study, and we talk about the interplay between verbal and nonverbal behavior. Pete talks about the importance of context and domain knowledge in making high-confidence judgments, and he discusses how such insights will sometimes be based on aspects that we aren’t consciously aware of. 

Episode links:

Resources mentioned in or related to this talk:

TRANSCRIPT

(transcript is generated automatically and will contain errors)

Pete Blair: The research I did with Tim where we went really accurate on the… the cheating paradigm….   I think the results of that study where I, I went perfect, you know, I think it was 33 for 33, uh, in that particular study What exactly was I using?… I don’t know. It’s probably a constellation of things. There are certainly things that seemed to help, and we had a better question structure than we’d had previously. Um, but there were probably some of those moments of, you know, insight where I would be talking to somebody and whatever, I’d be like, “Oh, yep, you cheated….

Zach Elwood: That was a clip from my talk with Pete Blair, a former investigator and investigative trainer, who now is the executive director at the Advanced Law Enforcement Rapid Response Training Center at Texas State University, and also a professor of criminal justice at that school. 

The reason I wanted to talk to Pete was his involvement in a 2014 research paper titled “Expertise in Deception Detection Involves Actively Prompting Diagnostic Information Rather than Passive Behavioral Observation.” The first line of the abstract of that paper reads “In a proof-of-concept study, an expert obtained 100% deception-detection accuracy over 33 interviews.” Pete Blair was that expert. I wanted to know how he got 100% right; what approach did he take, and what strategies did he use.

I learned about that research in Tim Levine’s fantastic book Duped: Truth-Default Theory and the Social Science of Lying and Deception. I’ve had Levine on my podcast three times, to talk about what the research says about using nonverbal behavioral information to detect deception. 

While I’m on the topic, if you are interested in behavior and deception detection research, I think Tim’s book Duped is a must-read. It not only is a highly interesting tale of Tim’s journey to working through various areas of confusion and mystery in the deception detection space, but it’s also just a very informative book about how to approach intelligent research and how to avoid mis-steps and biases. 

Tim’s partnership with Blair was aimed at exploring ideas about what actually works in deception detection. We know that focusing on nonverbal behavior doesn’t work, so what does work? So their projects were focused on showcasing how much more effective deception detection comes about through focusing on what people say, and on prompting them to talk in ways that are likely to reveal information. Their goal, from my understanding, was not to say ‘the methods we used in this experiment are the best ways to detect deception’, but mainly to show that focusing on what people say, and methods for eliciting more useful information, is a hugely more effective approach than is focusing on trying to read nonverbal behavior. 

I’ll read a bit from Tim’s book Duped, where Tim talks about what he and Pete were trying to accomplish: 

Pete and I thought that accurate lie detection is often a high-context activity. Typical lie detection experiments deny judges the context to understand a sender’s statements. This makes content less useful. But, we thought, outside the lab much communication happens in situations where listeners have context. When we are talking to a person we know, for example, we have relational history with that person. In our professional lives, we have occupational experience. As members of a culture, we know the norms and the culture. A little bit of contextual understanding, we thought, might go a long way in improving the usefulness of communication content in lie detection.

Now let me skip ahead to where Tim leads up to talking about the study in question, the one where Pete got 100% correct. These were studies involving subjects who had been put into a trivia-test situation, with a cash prize at stake, in which their partner suggested they cheat when the test-giver was out of the room, and the subject would either cheat or not cheat, as was their decision. The subjects were then asked questions. 

Okay, the following is from Tim’s book:  

When Pete and I tried our hand at the fourth edition of the cheating interviews used in Truth Default Theory experiments forty-six through forty-eight, we each got 86% (twelve out of fourteen) correct.   We compared notes. We thought the questioning made ten out of the fourteen interviews pretty clear-cut. It turned out that we were both right on each of those ten interviews that we thought were pretty easy. There were four, however, that we thought were difficult. As it turned out, we both got only two of the four ambiguous ones correct. The thing was, we knew those four were the questionable ones before we scored ourselves. We knew we were guessing. Because they were taped interviews, we couldn’t go back and ask more questions and try to reduce the ambiguity. What if we’d had that option? Might we have done even better than 86%? 

In a related experience, I was at a conference, and in attendance was an expert who had been an interviewer in an expert deception detection experiment run at a different lab by different researchers. The expert told me that he thought he could have done much better than he did. He did not think the study he had participated in adequately assessed his ability. The research design, he said, thwarted him. He was given scripted questions and told to stay on script. A few times, he said, he went off script and got the sender to confess. In such cases the researchers had scolded him, said that he wasn’t following instructions, and that the data had to be tossed. He said he needed the freedom to resolve ambiguities and to persuade liars to be honest. Sticking to the script and excluding confessions meant that the results did not reflect his ability. 

Pete and I put our experiences together with the expert’s story of being thwarted by the researchers, and we decided to put our expert questioning ideas to the test. Truth Default Theory experiment fifty-two provided initial proof of concept. We thought that if we could produce some strong preliminary evidence, we could leverage that to get the funding we would need for a larger-scale experiment.

We ran thirty-three more subjects through the cheating experiment. This time Pete did the questioning. As a base script, Pete crafted the twelve questions listed in table 14.3, fifth set. However, he was free to improvise as needed. After each interview, he recorded his assessment as to who he thought cheated and who didn’t. Only four of the thirty-three participants (12%) cheated. All four honestly confessed. Even though Pete used a false-evidence ploy (he told them that their partner had said they cheated), there were no false confessions. When we tallied Pete’s judgments, he got all thirty-three correct. 100%! Wow! 

Both Pete and I were surprised by the results. We both thought he would do well. Based on Truth Default Theory experiments forty-six through fifty-one and our content-in-context experiments, we knew we could reliably get accuracy levels in the 70s. We were pretty sure we could beat that and hit accuracy in the 80s or maybe even 90%. But, thirty-three of thirty-three exceeded our expectations. Pete, in particular, was aware that he was assigning honesty at an unusually high rate. He thought that he was probably making too many honesty judgments. The cheating base-rates in the prior versions were always higher. He wondered if he was failing to catch some cheaters. He wasn’t. The base-rates were just unusual in that run, and yet he still performed to perfection. For Truth Default Theory experiment fifty-three, we showed Pete’s thirty-three interviews to 136 college students. The students got 79%. Students passively viewing Pete’s interviews did not do as well as Pete, but they did pretty darn well.

End quote. 

I’ll say that when I started talking to Pete for this interview, I was confused about how Pete had gotten 100% right in that study. I had misunderstood something Levine wrote, when he wrote that quote “pete used a false-evidence ploy” end quote. That made me think that that alone was how Pete made his guesses. I also hadn’t read the study itself, which I should have done. It turns out that there were two stages; in the first stage Pete just asked the subjects questions and listened to their responses. Then he went and made his guess. Then in the second stage, he came back and did the false-evidence ploy, which resulted in guilty people confessing. So his 100% guess rate had nothing to do with the false-evidence ploy. I just wanted to make that clear as that was something that confused me personally.

In this talk with Pete, we also talk about his views on piecing together logical deductions based on the things people say and do, and his view of the relative role of trying to interpret nonverbal behavior. We also get on the topic of how we may sometimes have flashes of insight, without really being consciously aware of all the reasons for that insight.  

If you want to learn more about Pete Blair and his work, check out his substack Tactical Science, which is at www.tacticalscience.substack.com

Here’s my talk with Pete Blair: 

Zach Elwood: Hi, Pete. Thanks for joining me.

Pete Blair: Thanks for having me on.

Zach Elwood: Uh, so maybe we could start with, uh, I, I was curious how you got into the, uh, more research side of, uh, law enforcement and interrogation related topics, if you’d care to share how that journey came about.

Pete Blair: Sure. In my, uh, misspent youth, I worked for a company called John Reid and Associates as a staff interviewer and trainer, and did a lot of work on interviewing and interrogation in the real world, and started to run into various research articles. Uh, I want– Probably ’cause of the master’s program that I was in at the time, and wanting to do some research related to my, my work, and I started to read the articles, and I just said, “You know, they, they don’t seem to be quite getting what’s really going on here, in my opinion.”

And that kinda launched me down that road.

Zach Elwood: Hmm. Yeah, I was actually just looking at the, uh, the Reid area. That, yeah, that would be a whole interesting topic for another day, but I’m, I’m curious, uh, yeah, what, what was your, uh, initial, uh, entry into the, into the, the research? Um, did you– were you researching a specific area at first?

Pete Blair: Yeah, so my, uh, master’s thesis was a basic, uh, test of the Reid technique training. So we had a bunch of videos of people who were actual suspects in criminal cases, and we did a basic experimental design of training some of the students to, uh, detect deception per the Reid technique, and the other students getting no training, and then, uh, having a pre/post-test design on accuracy about who was lying or telling the truth in the videos.

Zach Elwood: Hmm. Okay.

Pete Blair: That was the original study that I did.

Zach Elwood: Mm-hmm. Uh, when it comes to the, um, the work you’ve done with Tim Levine, um, obviously Tim is someone who was very skeptical of the idea that you can use non-verbal behavior to, uh, detect deception, and I assume that you would, uh, agree with him, but I’m curious if you’d like to give your overall thoughts on that area.

Pete Blair: Yeah, so I think if you, if you look at non-verbal the way it’s been researched, I think that that’s what the research shows. Uh, however, I think if you were to look at things in a… Uh, so I’ve, I’ve been leaning strongly toward an ecological approach to understanding not just interviewing interrogation, but skill acquisition and those sorts of things within law enforcement for a while.

And when you take that approach, it’s not clear that non-verbal is really distinguishable from verbal. They’re both part of how we communicate, and I don’t know that you can look at one without really looking at the other, uh, especially if the modality allows you to inter- you know, talk face to face.

Zach Elwood: Mm-hmm. Um, in that area, I think, um, some people that, you know, teach and talk about using non-verbal behavior, they will say something like, “Well, yeah, we know it’s not, you know, you c- there’s not specific behavioral cues you can use that are, you know, highly correlated with deception, but, you know, we can use behavior to, um, detect signs of anxiety which might be useful or detect other things.”

And but then the on- you know, the, the rejoinder to that, the, the criticism of that would be, well, in high-stakes situations like interrogation, people, even innocent people, have high anxiety, so that makes it very hard to use in a practical sense. But then on the– going back, you know, on the flip side, it’s like clearly every day we all have, have had, uh, situations where we’re talking to someone and they suddenly get an- anxious about a specific topic.

So it does seem possible that, you know, in interrogations that you sometimes would find yourself in a situation where you’re like, “Oh, this person seems especially, you know, uh, anxious about this topic, and maybe I’ll, I’ll follow it up with a few questions.” So that also seems quite plausible, but I’m curious what you think of that, that area there.

Pete Blair: Yeah, I’ve certainly had those moments of, uh, insight where I’m talking to somebody and then all of a sudden it’s like, “You’re lying.” You know, “You did this.” Uh, so certainly I can point to those experiences. I don’t know that I can put my finger on saying exactly what it was, uh, that made me trigger to that.

Uh, and I think one of the problems you have in trying to research nonverbal cues or even some of the verbal cues is that, uh, there’s a lot of individual variation, and then the context in which you’re doing this matters a lot, and it’s probably not the single indicators like fidgetiness or, you know, breaking eye contact or those sorts of things.

It’s probably, you know, people talk about clusters or constellations of behaviors that go together, but probably beyond what we’re able to code in regular interactions, ’cause a- as receivers of information, we’re taking in so much information all the time that’s being processed at a subconscious level.

And to be able to pick out and say, you know, “This is the thing that you’re keying on, and this is a general rule for things,” I, I don’t think that applies. And I, I think you start to see that as you look in the sports literature, that it’s, it’s kind of the same thing too, where you see that there, there’s just a lot more happening here than what we try to do in most of our analyses.

And so if you’re trying to do your analyses and break these things out, it doesn’t seem to be very successful.

Zach Elwood: Yeah, you said something that I think is a very good point about how hard it can be to disentangle the, the verbal from the nonverbal, and that’s, you know, I, I think that, I think that plays a huge role in why it’s so hard to talk about these [00:06:00] areas, because I think you’ll have, you know, for example, you might have somebody that does an interrogation and they think someone has acted in a suspicious way non-verbally, but it’s also like they could be mainly thinking about the actual things that person said, you know, like, because to me, and I think most people would agree, what, what’s being said is much, much more important than the nonverbal.

But I think what happens is because all that gets kind of clustered together, it’s like if somebody’s not answering questions in a, you know, forthright way and has patterns of deception by avoiding the question or, you know, these kinds of things, I think what happens is people can start thinking, and, and including people watching on interrogations on YouTube and such, they can be like, “Oh, I, I see these nonverbals of…”

But, but mainly they’re foc- mainly they’re getting the vibe from like how that person is actually answering the question and what they’re saying. But, but I think that gets into like why it can be muddy to kind of disentangle it because it is kind of like all jumbled together, you know?

Pete Blair: Yeah, I think that that makes sense to me that- There is that jumble that’s happening, but it, it’s even more than that.

So in some of the ecological psychology research that’s been done, uh, you know, I mentioned we’re picking up a lot of information we’re not even aware of. So there was a, a classic study done where they had three different shapes of metal. So say a, a rectangle, a circle, and a triangle. They blindfolded people so they didn’t know, uh, what the actual object was in front of them, and then somebody would hit it with a mallet so it made a gong sound, like boom, you could hear it hitting on there.

And then they had them indicate what is the shape of the object that was just struck? Okay? So they’re just hearing it. It’s a piece of metal being hit with a mallet, and they did better than chance at that. Not a lot better, but better than chance Now why is that? Well, it’s probably, assuming that it’s a real finding, and you can argue about those things obviously, but assuming that it’s a real finding, it’s probably that there’s information in the sound that’s different between those different shapes.

You’ve probably been exposed to that for your whole life, and now you’re hearing it and you’re focusing on it, and you can actually pick up that there’s something a little different about this shape, and it has some information about the shape in the sound, right? Mm-hmm. That’s something that no one would– none of us would have predicted you’d be able to do that anywhere above chance.

Mm-hmm. But there are, are several things that are like that, that are, you know, what are we picking up that we’re not really aware of and that’s not there? And a lot of these thoughts run together, and again, I haven’t been working in the deception world, uh, so much lately, but-

Zach Elwood: Mm-hmm …

Pete Blair: when we talk about coaching and proper technique and there’s one right way to do thing, do something, you know, we, we spend, spend so much time about this is the right way to do this, and then we make people rep that out, do this the right way.

Well, the right way for what, right? My body’s a little different than yours. We might be in a little different situation than we’re assuming here in this particular s- you know, if we’re doing golfing on a driving range and now I’m out actually playing, the, the field is not level, right? You know, the, the course is not level at that point, and there are all these little differences that are coming in about is there that optimal thing or that one right thing?

And what you’re trying to teach people to do in a lot of the sporting context, and now I’d argue in some of the police training that we’re doing, is to be sensitive to those variations in context and expose them to a lot of those variations in context, trusting that the machinery that you have is really, really good at doing this and that you’ll be able to figure it out.

Mm-hmm. And I, if we– I, I’m, I’m assuming at some point you’re gonna turn and talk about the, the research I did with Tim where we went really accurate on the- Mm-hmm … the cheating paradigm. And I think the thing that allows high accuracy there is that it’s a context that I had seen a whole lot, I was very familiar with.

We designed it to be similar to what my work experience was like too, so that it was closer to the ecology, the environment that I was– I had been trained in and I had worked in for several years. And then I got to see multiple runs of this study done in different ways with different things. And so I am picking up information this whole time about what truth tellers and what liars act like.

I’ve got very good feedback during those times because I, I know who’s lying and who’s telling the truth, right? And I’m able to, to develop that and, and then am I picking up information there? Well, I think the results of that study where I, I went perfect, you know, I think it was 33 for 33, uh, in that particular study What exactly was I using?

I don’t know. It’s probably a constellation of things. There are certainly things that seemed to help, and we had a better question structure than we’d had previously. Um, but there were probably some of those moments of, you know, insight where I would be talking to somebody and whatever, I’d be like, “Oh, yep, you cheated.”

You know? Um, so exactly what that is, I, I don’t know that I could put my finger on.

Zach Elwood: Now, I, I know it was a while ago, and I know there were several studies, um, so… But I was curious, I mean, the one study, correct me if I’m wrong, but I thought the one where you got them all right, I thought that was largely or entirely because, like, you– the, they confessed based on you saying that their partner, um, you know, saw the, uh, collaborated that they…

or said that they cheated. Am I getting that, that right? ‘Cause I think the specific study- So it’s, it’s… Yes … I’m thinking about was, it was only, like, four guilty out of 33, and you guessed-

Pete Blair: Right. Yeah … something like that. So, so what happened in that particular study is there was an, an interview portion, uh, upfront where I didn’t confront them.

I talked to them, and I asked them what had happened during the study and the questions that were on the, that question list that you sent me.

Zach Elwood: Right.

Pete Blair: Uh, went through those questions and I could ad-lib and go off of that. And then after that interview, there’s a break where I go out to talk to the conf- you know, their partner allegedly.

And during that break, I go out and I tell the person I think they did it or not And then I come back and I confront them regardless of whether I think they did it or not.

Zach Elwood: Oh, interesting. So- Okay. I didn’t realize that. Yeah. Yeah. I, I, I think maybe either I misread it or maybe Tim, no offense, Tim, maybe you didn’t do a great job explaining in the book.

But I, ’cause I thought… I, I hadn’t realized that you guessed after just doing the question answer.

Pete Blair: Yeah.

Zach Elwood: Okay. Um-

Pete Blair: Yeah, I guessed every time that I came out, and then I came back and I confronted everybody, uh, as if they had done it. Yeah.

Zach Elwood: So you, you got it right even before the confrontation is- Correct … is what you’re saying?

Okay. Wow, that’s, I mean, that’s even more, more impressive than I, than it much… Yeah, much more impressive than I, I had, uh… I mean, it was still impressive enough what I thought, but that’s… Yeah. And it, now it makes me wanna delve even deeper to it, deeper into it because I didn’t actually read the study, which I probably should have done.

So that, yeah, that’s really interesting. So yeah, it was really about the, the design of the questions and also probably your, your, the vibes that you’re used to picking up when people answer questions in those scenarios. Is, is that accurate to say? But those two things, yeah.

Pete Blair: Yeah, I think there’s a lot going on in that situation, for sure.

Zach Elwood: Right. And, and would you say… I, I, I know it’s hard, like we said, it’s hard to differentiate these things, but I would imagine, if I had to guess, I would imagine y- it would, it would mostly be, like, in the realm of, you know, 90%, like, the content of what people were saying versus, you know, other things.

Pete Blair: Again, how do you separate that when I’m face to face with you, right?

Zach Elwood: Right.

Pete Blair: So-

Zach Elwood: Right. Like pauses- So certainly- Like pauses. Yeah. Yeah. All the, all these kinds of things, uh, can, can be, can theoretically even be cons- seen as content of what they say. Right. Right.

Pete Blair: Yeah. And I’m, I’m definitely paying a lot of attention, was paying a lot of attention to what they were saying-

Zach Elwood: Mm-hmm

Pete Blair: and how they talked about what happened during the particular situation and that, and that definitely had value and was diagnostic, I think. But then there’s all this other stuff that’s bundled into it, uh, that, that’s also shaping that decision as I’m doing it. And, and how much of a, an influence did it have?

I’m not sure.

Zach Elwood: Mm-hmm. Mm-hmm. I’m curious, were there- I know it was a long time ago, but, um, maybe, and maybe a general question too, uh, for those other studies, similar studies, how, how often would it happen where you thought, you know, “Oh, this could go either way,” and it was a close decision? Or were you, were you pretty confident for, for that specific study or similar studies using those questions?

Pete Blair: I think I was confident on my decisions. I’m sure there were some edge cases where maybe there was more uncertainty that would be, be typical. Uh, I do remember that the base rate of cheating in that study was, like, half of what it normally was in a study. Mm-hmm. So I was worried as I was going through it that I was missing people.

Right. Uh, that there should be more cheaters here and I’m, I’m not seeing them. Um- Yeah, there’s

Zach Elwood: so few …

Pete Blair: you know?

Zach Elwood: Yeah.

Pete Blair: Yeah. And so I was really– I was worried about that, but at the end when I found out, I was like, “Oh, okay, great.” And as I recall, I think, I think Tim actually changed the, uh, the structure of that Uh, in the, a lot of the runs that we did, as I recall, there was a single answer that would be given to the person, um, by the confederate, by the, yeah, by the confederate, so they would get some payout for participating in the study.

And I think for the, the run he did with me, that he actually changed that number so that it was like they got two right. So it even throws that baseline about, you know, what’s normal or, or the floor of what people can get right. So-

Zach Elwood: Hmm …

Pete Blair: so some change there too as well.

Zach Elwood: Mm-hmm. Yeah, that made it, made it– That, that’s what made it especially, um, impressive because I think most people in your spot would’ve been like, “There’s so few, I need to start looking for more, uh, guilty people,” guessing more people guilty maybe, and having more of like a deception or y- bias towards, uh, detecting deception.

So that, that made it especially interesting. And, and I’m, I’m curious, uh, did you, um, did you have a specific strategy going in, like if they don’t meet these criteria to me, or is it more like a case-by-case basis, you know, on every- It’s- … every one?

Pete Blair: It’s a case-by-case basis. Yeah. And definitely there are certain things that are, are valuable, like knowing how many questions they got right is valuable, uh, because a lot of questions right becomes very suspicious, right?

Zach Elwood: Right.

Pete Blair: However, and it wasn’t in my run, but I think in one of the other runs we did of the cheating study, somebody got like seven or eight correct that did not cheat out of the 10. Hmm. And, uh, he was studying for Jeopardy!.

Zach Elwood: Oh.

Pete Blair: Okay, yeah. Outlier. And so, and so he knew- Yeah … all the answers to the questions and, and that, and so you might’ve said, “Oh, here’s this,” but, um, I think they correctly classified him in that, that particular study.

But-

Zach Elwood: Hmm …

Pete Blair: yeah, it, it’s, so there are things like that that are there that are helpful, that are, are leading you that way, uh, on certain things, but nothing is, is surefire, and there are definitely people who got a few questions correct that didn’t cheat, so.

Zach Elwood: Hmm. Mm-hmm. Uh, and do you remember, um, do you have a rough memory o-overall of like what the most, like what stood out as the, the biggest pattern, um, pre-confrontation just based on how- It’s-

they answered questions?

Pete Blair: It’s been so long.

Zach Elwood: Yeah, it’s been a while. I, and I could look- It’s hard to remember … I’m sure I could read the study and, and summarize- Yeah … some of the thing, the specifics in there for the intro too. Um- Yeah. Do you have any other, uh, any other observations that, that come to mind about that, about, about doing that specific study that might be interesting to people?

Pete Blair: Well, I think it was part of, uh, how I knew I had made it in the research world when people responded to it and called my research dangerous, so I, I thought that was pretty good.

Zach Elwood: Oh, and what, and what was the context for calling it dangerous?

Pete Blair: Uh, well, so, uh, the, uh, when you take a look at the article that has, you know, my run as kind of the, the pilot study and then the other, uh, professional investigators’ runs as the, the main study, uh, they weren’t perfect but they were also very accurate.

Uh, and then they also had the same situation where they got almost everybody who cheated to confess and nobody who didn’t cheat confessed. Uh, and particularly in my run, I was doing the things that in the literature had regularly caused false confessions in, uh, in lab settings at least. Uh, the false evidence ploys and, you know, accusations of, of guilt and then minimization, maximization stuff, all those things that were supposed to cause false confessions, and did that across my study and, and didn’t produce them.

And so we talked about that and I, and the response from the community that was involved in the, a lot of the false confessions work was, “Well, this is dangerous. You’re, you’re suggesting that people are very good at this and, and they do that.” And, and I, I think that, uh I- if you have that preconceived view that this is what’s happening and this is how it happens, um, you see contrary results as being something that, that’s dangerous to that finding.

But I think for us it was pointing at, look, we’ve done– we’ve seen deception research done for a couple decades, and it all kinda looks like this.

Zach Elwood: Mm-hmm.

Pete Blair: And here’s this outlier here-

Zach Elwood: Mm-hmm …

Pete Blair: that we designed to be more ecologically valid, more like the real world. Uh, and here are the– And we got real people who do this kind of thing in the real world.

And here are this, here’s this finding that’s way out there, and I think that was very threatening to a lot of people. So the, you know, there were counterarguments and straw man attacks that were made and that kind of thing. And, and, uh-

Zach Elwood: Yeah …

Pete Blair: you know, at the end– And that’s the, the limit. And I haven’t seen people really try to, you know, push on what Tim and I found back then, so.

Zach Elwood: Yeah. I d- I mean, I did see… I mean, Tim talks about in his book about various misinterpretations, and it does seem like they’re missing the point, that the point is just to show that domain expertise and experience can play a major role in improving how people do in such things. It’s– And it’s not making a statement about whether those things are right or wrong in specific situations.

It’s just saying, “Hey, this is a, th- this is a much more fruitful area for deception detection than the things that were usually being studied.”

Pete Blair: Yeah. Yeah. And I, I think it’s, uh, again, if you, if you go to the, the ecological kind of approach that we’ve been looking at for skill acquisition and, and just general training, the idea of perceptual cues being connected directly to action in such a way that it’s so deep, that connection, that if you change any of the perceptual cues, you by definition change the action.

Okay? So the idea that I can practice something in an isolated context that doesn’t have any of the features of the actual game or sport that I’m doing, and then I can get really good form doing that, and then that will allow me to perform well in the field, doesn’t work. Mm-hmm. Okay? At least doesn’t work particularly effectively or efficiently, okay?

Hmm. So that, that idea that the context is critically important, that the environment in which these things function is critically important, is really a key part of trying to research that. And, and you mentioned, uh, poker in your, your email with me, and I think there you have a really interesting controlled context, right?

That we’re in, this is the way it works, this is, these are the rules around how it plays. I have, uh, I know those general rules, but then also because you’re a player for a period of time, you encounter the same people on multiple occasions, plus you can watch video of them now, I’m sure, and get all that, that background information.

And there’s probably a ton of information there that is helpful for you making judgments at the table that, yeah, if I push you on it, you’ll, you’ll tell me, “This is what I saw,” or, “This is why.” But the reality’s probably just, “I don’t know. Something seemed like they were bluffing, and I called them on it.”

Zach Elwood: Well, yeah, to that, yeah, it’s a, it’s a, it’s actually a really interesting thing because I’ve barely– I’ve basically not seen anyone talk about it, but I- it’s something I wanna do more on. Like, I’m working on a, a book actually, and the, the difference between, like, game environments, you know, which would include sports, and the difference between those areas and, like, real world non-game scenarios is just so massively different, and that’s just something I haven’t seen people talk about.

You know, for example, like, yeah, in poker, there’s specific discrete granular things you’re trying to accomplish. There’s also, like, polarity of, like, you know, someone feeling very relaxed and happy because they have a good hand and they’re gonna beat their opponent, which doesn’t really play a role, doesn’t really have an analogy in, like, interrogation or interview settings, right?

Like somebody just being really like, “I’ve beat this person,” kind of thing, you know?

Pete Blair: Yeah, yeah. It’s gonna be different for sure because the context is different, but there are also a lot of things in context that can be very similar, you know, over time. Sure,

Zach Elwood: yeah. Sure.

Pete Blair: So if you’re, if you’re an investigator, you tend to do certain kinds of crimes that you draw.

Um, one of the things that’s really ignored in, in all the, uh, deception literature for criminal investigators is you don’t interview people at random, right? You’ve got a whole set of facts that are leading you to the people that you’re talking to, and in a fair number of cases, you actually know who did it and who didn’t do it when you sit down to start talking.

Right. That- that’s why you’re talking to this person to start with. You’ve got a

Zach Elwood: wealth of information usually.

Pete Blair: Yeah.

Zach Elwood: Yeah.

Pete Blair: Yeah. And so that’s how you ended up at that, that situation. And, and so this idea that I’m just doing random draws out of the population and then I’m trying to, to make sense out of what’s happening in there, I don’t think happens very often in criminal investigations.

Zach Elwood: Right.

Pete Blair: And I would actually argue in some of those worst cases that you see where there are just horrendous things that happen, there are probably cases where these investigators are being asked to do something that is not their normal thing that they do, and their expertise then doesn’t apply because it’s a different situation that they haven’t encountered before.

So I, I, I would go back to real clear case of, uh, JonBenét Ramsey and the whole situation there. Those investigators that ran that investigation screwed it up, uh, from the very beginning But why? They’d never dealt with a homicide in a house like– Or like, they just don’t deal with that. Right. And so they made tons of mistakes, and they made lots of errors, and their expertise as investigators didn’t carry through very well because it’s not the thing that they are expert at.

Zach Elwood: Mm-hmm. Mm-hmm. Yeah, the, uh, another thing, it might be a good segue to something that I, I’ve also barely seen discussed. I mean, there’s al- there’s often all of this talk about, like, how can we, you know, interpret behaviors to, uh, get clues about dots or, you know, deception, or how can we interpret statements?

Uh, but I think the underlying question that’s barely examined is like, well, all of these things are gonna be pretty far from 100%, you know, s- similar to the lie d- lie detector, the polygraph, for example. It’s like, I think the, the underlying question is basically like, in what situations is information that is, is far from 100% even useful, right?

Let’s, uh– A-and so I think– So for example, in, in games like poker, it’s very useful because you’re forced to make decisions where, from a game theory, uh, strategy perspective are very close, so a little bit information can go a long way. But like– A-and similarly, if you were in a situation in an investigation where you had like, say you were in this rare situation where you had a lot of suspects and you had to try to sort something out really quickly, you might need to rely on these things that were far from 100%, right?

Or if you- But you’re- Yeah.

Pete Blair: But you’re also very likely to not be expert at those things because you don’t encounter them often enough to be, to develop that expertise. And-

Zach Elwood: Right …

Pete Blair: and that’s really, uh, uh, to me, the, the thing when you start looking at ecological design and, and that, is you need to expose people to the thing that you’re asking them to do.

And you can guide them about, “Here are things to pay attention to. Here’s what’s important. Here are some approaches you might try.” But ultimately, you’ve gotta get in there and encounter it and experience it and figure out how it works, right? And- Mm-hmm … and trust that system that you have to sort things out and really make the connections that it needs to make.

Zach Elwood: Yeah, I guess the, uh… I mean, I, I keep– I tend to go towards, like, the practical question of, um, you know, how, how would these various reads and vibes that we get, say you’re an interrogator, you know, d-doing an inter- uh, interview interrogation, and you have various reads that you have. It’s like, when, when do they actually play a significant role?

And it seems to me like if we’re, if we’re talking about, like, little clues like, you know, non-verbal things or how someone has phrased something that’s separate from content, it seems to me the main practical thing, because these things are, are so, are gonna be so ambiguous, the main practical outcome is maybe like, “Oh, I’m gonna ask you a few more questions because I thought somebody was anxious.”

But like, le- if you don’t have any other confirmation in terms of, like, what somebody says or in terms of evidence, it seems like those, those kinds of things are really low on the, you know, on the practical use side other than, like, maybe I’ll follow up with a, a few questions. But would you agree with that?

Pete Blair: Yeah, I think when, uh, the idea that there’s something that piqued your interest, right? Like-

Zach Elwood: Mm-hmm …

Pete Blair: whatever, maybe they said something, maybe you saw some behavior, whatever it is, you’re gonna delve into that area more and try to explain what’s happening. It, it makes sense that that’s probably happening in our…

any-anything you’re doing, any conversation you’re having, you’re monitoring your partner and, and tracking what they’re saying and then trying to follow up on that and, and do those things. An investigation has a little bit different focus on that maybe, but it’s still that interaction that’s occurring.

And so, yeah, if you ask a question and you get an answer that for whatever reason doesn’t, doesn’t seem sufficient, you’re gonna follow up on that and try to get more information on it, for sure.

Zach Elwood: Mm-hmm. Mm-hmm. Um, yeah, I, I just want to ask you so many questions because I value your input. I, I’m curious to get your take on, um, statement analysis because I, you know, and, and, and similarly to the, the nonverbal, I think that stuff can go too far, sort of like the scan stuff goes too far to me because it makes these highly confident proclamations and says that you can, you know, very accurately determine guilt.

But I, I’ll say that I am a fan of statement analysis in the sense that I do think that there are patterns there, even if they’re, you know, far from 100%, that there’s, there are patterns that I find very interesting, like the ones that Mark McClish wrote about in his book, “I Know You Are Lying.” And hi- his book was actually the, uh…

inspired me to write my book, “Verbal Poker Tells,” where I studied all these, like, small turns of phrase that people use at the poker table, which, you know, I think definitely have meaning and a lot more meaning than nonverbal things do. But I’m curious to get your take on that area, like this, maybe the spectrum of, like, what people call statement analysis.

Pete Blair: Well, I think what we’ve, we’ve seen in the, the literature that’s out there about looking for cues and those sorts of things is that the verbal things seem to be better than the nonverbal things, at least the way that we code them now. And certainly there’s a lot of information encoded in words and how you talk about things that, that’s worth really thinking about, you know.

And just one of the, the key things I think for an investigator is to always ask yourself, “Did they even answer the question that I asked?”

Zach Elwood: Right.

Pete Blair: Or are we talking about something different? And if we’re talking about something different, why are we talking about something different? And what happens when I redirect back to that to try to get an answer to that question?

Zach Elwood: Yeah. Do they dodge again? Which is just a, which is just a great thing for interviewing politicians and stuff too. A lot of, a lot of interviewers- Sure … a lot of interviewers- Sure … fail on that basic thing of like, did they, did they actually answer the question? Yeah.

Pete Blair: Yeah.

Zach Elwood: Yeah.

Pete Blair: So I- I’ll keep, keep going

there are, there are a lot of things like that that I think are, are relevant. Uh, but I, again, trying to look for certainty, you’re gonna have to have that certainty be outside of the actual interaction itself, I think. You’re gonna have to- Mm-hmm … have some sort of evidence that confirms what’s there, uh, that you probably saw some of the coherence correspondence stuff that we, we wrote on so that, you know, does this thing they said correspond to what we know actually happened out in the real world from some other source of information we have?

Mm-hmm. If it does, then, then we’re much more likely to believe it than if it doesn’t. Um- Mm-hmm … coherence, are they consistent in what they say or are they shifting in what they say? Um-

Zach Elwood: Mm-hmm …

Pete Blair: that, those, those are things that matter to some extent. So that, that idea of having that out there, and if you’re talking about in a, a legal context, you’ve got to get to the level of proof.

And, you know- Right … the, one of the things I think that the false confession research people really added to the world is before that, I would say investigators would say, “I don’t I don’t believe that false confessions exist. They’re a unicorn. They’re, they’re not true. And I think after that, now I, I don’t know any investigators that I’ve talked to that say they don’t happen, right?

So-

Zach Elwood: Hmm …

Pete Blair: clearly I think false confessions happen. That line of research and, and exploring that has, has been very positive in making you acknowledge that. Uh, but then you also have the idea of, okay, if false confessions happen, how do we verify, uh, the false confes- the confession that I have, right? And there’s, there are those ideas of checking the information that they say in the confession against the case facts, and particularly if we have information that we held back, do they know that information that we held back about the crime that you assume only the criminal would actually know?

Hmm. Those are things that can strengthen your belief in that confession statement, but again, it’s an external referent that you’re looking at to, to confirm it as opposed to that. And then you need to make sure that if you’re, you’re assuming that this information didn’t get out, that it didn’t actually get out, ’cause we’ve seen in cases too where cops are either accidentally or intentionally leaking information to the suspect and that’s being integrated into a confession once the, the suspect defi- decides to confess.

Zach Elwood: Uh, I recently interviewed, uh, M- Maria Hartwig and, uh, Christian Corey together. Uh, it hasn’t come out yet, but, um, we– I know that they and other people talk about the, what they view as, uh, too much focus on getting confessions and how that can hurt, um, an interview. Uh, but I’m, uh, but on the flip side of that, it does seem like from a practical point of view, getting confessions is, is hugely important.

Um, and, and, and it seems like that can be hugely valuable as long as you’re avoiding doing things that would result in false confessions. So I’m a little bit… And, and maybe just, just be that I’m, you know, I’m, I’m pretty far from the, the discussion and pretty- Yeah … ignorant of all the discussion, but I’m curious, do you have any thoughts on that whole discussion about being too focused on confessions versus not?

Pete Blair: Yeah. So I think if you look at it from, uh, that ecological approach again, intention focuses attention determines the information that you pick up. So if my intention is to get a confession, that focuses my int- attention on the things that are gonna lead to that confession, and then the information that I pick up is all related to that confession thing.

It’s really like another, another way to think about that, have you seen the gorillas playing basketball thing that’s the perceptual blindness deal, right?

Zach Elwood: Yeah. People don’t notice- Yeah … when this gorilla walks in because- Yeah … they’re so focused on the players playing

Pete Blair: basketball. So it’s two… Yeah. It’s two teams of basketball players, one’s wearing white, one’s wearing black, and you’re told to pay attention to the pa- the players in white passing the ball back and forth.

And there are a couple balls that are going around between them, and you really have to focus on the players that are in white in order to track how many passes are happening amongst the white players, ’cause the black players are also passing the ball as well. In the middle of it, there’s a gorilla that walks out in the middle of the video and pounds his chest, and then walks off to the other side of the video.

And I, I’m always terrible about the exact percentage, but there’s a large percentage of people that when they watch that video, because their intention is to count passes, their attention is focused on the players that are in white. They’re only picking up information from the players that are in white, and they completely miss the gorilla that walks out.

And it’s, it’s, it’s a fun thing- Mm-hmm … to show in classes to students where they watch it and you just say, “Whatever happens, you’re just gonna watch this and, and do these counts.” And then the gorilla comes out, and then afterward you say, “How many passes?” And they’ll, they’ll start saying, “Oh, it was this many passes, that many passes.”

Then you’ll say, “I don’t care about the passes. What about the gorilla?”

Zach Elwood: Did you see anything

Pete Blair: else that was unusual? And they’ll go, “What?” You know? And, and that’s that idea of, of being focused on something, and really your intention focuses that attention and the world unfolds a- according to that kind of thing.

So I, I think it’s true that if you come in with this idea that I’m, I’m only looking for a confession, uh, then you’re much more likely to have a, a chance of a false confession happening. Uh, but I also would say that in at least the Reid training that, that I was engaged in, and that it’s not true for all interrogations ’cause policing’s very fragmented and, and there’s a lot of variation, but You start with a non-accusatory interview and you’re trying to make an assessment before you come in and you confront them.

And so after you confront them, their reaction to being confronted, I don’t know that there’s a lot of value in, in trying to pick up that ’cause you’ve just been confronted, right? And that’s not something you’ve experienced a lot probably, and so your reaction’s gonna be highly variable and, and, and that.

So I don’t know there’s much you pick up off of how they respond to being confronted as much as you do before you confront them and you’re trying to gather information and do that. So all the professional police training stuff that I’ve seen has always had that non-accusatory interview up front and then a switch to the confrontation afterward, uh, when it occurs.

And so I think that’s– I, I think it’s true. If you go in and your idea is just to get a confession, it’s there. And I, I think, uh, one of Tim’s insights into the runs that we did is that he thinks because we were interested in getting to the truth We didn’t produce false confessions, um, as one of his, his takeaways from it is that even though we were using these pressure tactics to do it, ultimately we were still concerned about did this person do it or not, as opposed to we just need them to say that they did it.

Zach Elwood: Yeah. And it, and it just seems like it was such a low stakes environment that it would be hard for me to imagine somebody, uh, doing a false confession to like a, an experimental, like cheating on a test kind of-

Pete Blair: Well- Um- … but yeah, but the… Yeah, it’s, it’s low stakes for it, but, and there’s also that problem you run into when you’re- ever you’re doing experimental work with college students, um, a lot of them just don’t give a crap about what you’re doing.

They just wanna get out of there. They don’t see it, uh, you know, just like, “Okay, what do I gotta do to get out of here and carry on with my life?” Oh, so

Zach Elwood: they, so they might, they might false confess ’cause they- ‘

Pete Blair: Cause it just- Yeah … doesn’t matter, right? Like I did my, my dissertation involved doing something that was a, a riff on the Alt key experiment that was the famous, uh, false confession one where you were t- you were typing on a keyboard, you were told not to press the Alt key because it would cause the computer to crash.

And then the computer would automatically crash no matter what you did, and then they would confront you and do different techniques when they confronted you, and that caused more, uh, false confessions, but-

Zach Elwood: Oh, interesting …

Pete Blair: in my, in my dissertation, my idea was, well, how often is it that you don’t know that you actually did the crime that you did?

And they definitely found people were sensitive to how fast you were typing. If you type faster, false confessions were more likely to occur ’cause there’s- Oh, interesting … more uncertainty about, about it. And so my idea was I’m gonna tell them to not press the Control, Alt, and Delete key at the same time ’cause that will cause a crash.

And then the task was totally mouse driven, so they weren’t even touching the keyboard. Yeah. And then it crashes, and then I come in and confront them. It’s about 25% of them confessed, uh, no matter what I said, whether I had false evidence or did any of these other ploys. They, they were just like, “Yeah, whatever,” and they signed the thing and got out of there.

Zach Elwood: So it’s kind of like they just know it’s an experiment, so they don’t really

Pete Blair: care basically. Yeah. It’s like, yeah. They, they just don’t… They’re not, yeah. They don’t think anything’s gonna happen to them, whereas in- … in the design that, the cheating design, it’s much more ambiguous ’cause there’s money involved.

It’s a federal study. It’s ri- you know, all this kind of stuff that it, it’s, it’s higher stakes than the typical, uh, lab study that you see, for sure.

Zach Elwood: A- and they might not have even known it was part of the study theoretically. Like, they thought they might have been being really pressured. That’s kind of how it was designed to feel that way, right?

Pete Blair: No. Yeah. So they were, th- they were told that the study was about a, uh, the impact of teamwork on being able to answer questions. That’s what they were told about.

Zach Elwood: Right. Yeah. So they’re, it’s an outside… Yeah. They’re, th- they’re being made to feel like, yeah, this is not part of the study, but some of them still might think this is still part of the study s- even with that.

Yeah. Yeah. Yeah.

Pete Blair: Yeah. Well, and Tim will point at, uh, some of the other interesting things that, that happened that, that we don’t talk about. So I would ask them how many questions they got right. I don’t think anybody lied to me about that But I didn’t have that information in front of me, so I don’t actually know what the ground truth was on that question.

Hmm. Um, they could’ve lied, but my impression was that none of them lied about how many they got correct.

Zach Elwood: Hmm.

Pete Blair: Um, nobody actually stole the money, and there were several times that the, uh, in different runs of that study that the cash that we were paying them off of was also left in the room when the experimenter left.

Nobody ever stole the money, right? And I don’t know that anybody ever pointed at the confederate and said they thought the confederate was trying to set them up.

Zach Elwood: Hmm.

Pete Blair: Uh, so that’s another thing, and Tim has– talks about that and some of the other famous, uh, social psych experiments about nobody pointed in, in a lot of those experiments at these other people as being involved in lying, uh, about it, that were involved with them in the study.

Hmm. So that’s part of his truth default theory, talking about that they’re just buying that this is the truth and it’s really happening as opposed to this is a, a setup-

Zach Elwood: Right … that’s happening. Right. Yeah. Yeah, I think, uh, yeah, even if you were, if you were unsure, I think most people would default to acting as if it was real, you know, ’cause that’s- Yep

you don’t wanna, yeah, you don’t wanna make a mistake on that. Yeah, makes sense. Yeah. Um, yeah, one second. Just looking at my notes here.

Do you have any, uh, anecdotes from your career, uh, where you made some really interesting deductions in the moment, like during a, an interview/interrogation where whether it was behavior- Definitely I’ve had- … or things people said, small clues, any anecdotes

Pete Blair: like that? Yeah, definitely I’ve had some of those moments of insight where it was just an aha moment.

Um, there, there’s one case that was an internal loss case I remember, and I don’t even remember what what it was, but there was this moment where I was like, “Oh, she took the money,” and she still has it. Um, so, so there are definitely some of those that I have. I don’t know that there are any of them I can point at and say, you know, here’s the, the actual constellation of, of what was going on at that time.

Mm-hmm. Um, there are a lot of, uh, other things. That’s the, you know, again, I’ve been out of the deception literature and, and, you know, when I talk to Tim, it makes my brain re-engage with it and, and, and think about those things and about how I’m thinking about these other things and how they connect with each other.

And I think there’s a lot of overlap in these different areas of research that, that really suggests this idea of it’s not the person that matters, it’s the person in the context that matters. That’s the unit of analysis. And if you try to remove the person from the context, you shouldn’t expect them to have any expertise.

And again, there’s a ton of research in the sports literature. If you’ve ever watched, uh, baseball players try to hit softball pitchers- … uh, it’s pretty interesting, right? You’d think, “Oh, it’s kind of the same thing,” and actually the softball’s moving slower and it’s bigger, so they should be good at this.

Zach Elwood: Mm-hmm.

Pete Blair: Uh, they’re not.

Zach Elwood: Yeah.

Pete Blair: They’re not, because it’s different. The whole pitch motion’s different and judging it is different. And so they can figure it out if they’re a good baseball player after a little bit, but it takes some calibration time to figure it out and, and make it actually, make it actually work.

And, but I just go back to that if you’re putting people in this very artificial environment, having them do something that’s not something that they have experience with and expertise in, it’s not surprising they’re not good at whatever it is that you’re asking them to do. That’s just- Mm-hmm … not how people work.

Um-

Zach Elwood: Right. The domain is, the domain knowledge is, is so important, yeah.

Pete Blair: It, it’s, it’s credibly important and, and this idea of it gets referred to as near transfer and far transfer. So near transfer is stuff that’s really close to what you’re already doing. That’s pretty good. Far transfer is not good. Uh, it’s, it doesn’t seem to work very well on very many situations at all.

And that idea of having intuition and, and judgment and those sorts of things is kind of like, uh, Gary Klein’s naturalistic decision-making stuff that he has about what are experts doing? Are they contr- are they comparing everything to the total body of what’s happening? Or have they been exposed to a lot of situations that are like this, and they grab one that this is similar to this and that’s how I’m gonna judge it, right?

And he has some, I think, pretty good research suggesting that that’s a lot of what happens in a lot of, of human decision-making on things. Mm-hmm. So I just, I really go dig into that and say that, you know, I wouldn’t expect somebody who was an expert, uh, police interrogator, for example, to be a good poker player.

It just- Sure, the idea that I, I, I should be able to tell when somebody’s lying or telling the truth-

Zach Elwood: Right …

Pete Blair: but it’s a different, totally different context and a totally different environment. And so those skills and that calibration you have for that other environment are not likely to work well in the, in the poker environment.

Zach Elwood: Yeah. The, um… I mean, I do think Related to what we’ve been talking about, I, I do think there, there’s a lot– there, there’s some people that have a lot more pessimism on the idea of reading people than I think they should have because, I mean, there’s, there’s valid reasons to be pessimistic about-

Pete Blair: You read-

for

Zach Elwood: example, reading non-verbal-

Pete Blair: So you- …

Zach Elwood: cues and relating that to deception. But like, I’ll, I’ll give an example. Like, I was watching this interrogation footage recently. I, I watch a good amount of it and, you know, the- there’s, there was this guy who came in, his mother had been killed, and just immediately, like, he’s not– he’s strangely unemotional, but he’s talking like as if everything’s normal.

And like, people might say like, you know, he could be in shock, but it’s like, I, I think that my, my point is police who see these kinds of things, you know, if you, if you’ve seen people act guilty and unusual in a lot of, lot of situations, you’re gonna have an opinion on that, and your opinion will– If you, if you’ve had that experience a few times, your opinion will be confident a lot of times.

And it also gets into like, that’s not– Some people might call that what he, what h- how he was acting unusual related to his non-verbal behavior because it was not related to his content. But then again, it’s like that’s also like a form of communication, like how we deliver the, you know, getting back to the all the things being jumbled.

It’s like he was-

Pete Blair: Yeah …

Zach Elwood: he was making some sort of communication in, in, in, in, in how he was communicating and his lack of emotion and hi- the fact- Yeah … that he was trying to act completely normal, you know? Um, so just to say they– I think people can be too pe-pessimistic because, you know, we know that the studies show specific non-verbal cues are not tied, correlated to deception, but it’s like that’s leaving out like the meta level thing of like how all those things bubble up to meta level, you know, uh, manifestations of behavior and demeanor and all those kinds of things.

Yeah.

Pete Blair: Yeah. So I think, I think there are two things going on there. One is also we gotta remember that in order to get better at something, you need feedback on whether you’re, you’re making good judgments or bad judgments, right? Uh, because just exposure without good feedback isn’t gonna help you. Um-

Zach Elwood: Right

Pete Blair: you know, the– If you go back to the, the baseball, softball example, you have immediate feedback. You hit the ball or you didn’t, right? And you know that immediately. And in poker, a lot of times you have… Although, sometimes you don’t get the feedback because of how, how the hand unfolds, right? And so- Right.

You don’t- You might- Yeah, you

Zach Elwood: often won’t. Yeah.

Pete Blair: You might, you might miss it, but the times that you do, you’re getting some feedback there, and it’s- Yeah … and it’s accurate when you get it, right? You know w- who really had what. Yeah. They were

Zach Elwood: bluffing or not. Yeah.

Pete Blair: Yeah. You know? Yeah. So you, you know that. So that feedback is important.

But the other thing I would say is that we are all reading each other all the time, every day. In this interview, we’re reading each other and, and paying attention to what’s happening, and I think if you get away from trying to Put it in the terms of detecting deception-

Zach Elwood: Right …

Pete Blair: to I’m just trying to interact with you about this, and what does that, where does that interaction take me and- Mm-hmm

and where are we going? Does it seem like I’m getting where I feel like I know more about this, uh, and what happened here? Or do I leave this interview with this person and I feel like I don’t know anything more than when I started talking to that person? Maybe even I think I know less uh- Mm-hmm … than, than that.

Those are those kind of monitoring questions that we’re doing in, in every conversation, and I would, I would steer people toward paying attention to those. And, and my, my take on it kind of is that you’re gonna pay attention to the verbal stuff because that’s the stuff you’re gonna be able to validate and confirm, right?

You’re gonna assume that these other things are coming along with that as you- as you’re in there, and they’re probably biasing how you interpret and, and attend to those pieces of information that are there. And that’s probably how things work successfully, at least in the legal context, um-

Zach Elwood: Mm-hmm Yeah, I would imagine, um, correct me if I’m wrong, but when you’ve had deductions, like you’ve mentioned that one where you suddenly were like, “Oh, I, I’m quite sure she’s took the money,” or whatever.

Uh, would you say that that was by and large based on, like, deductions based on what they were actually saying? W-

Pete Blair: I, I’m not even comfortable putting a finger on it- Hmm … right? To say that’s it. Yeah, definitely what she was saying was part of it. Was there other information that was going on? I’m not sure.

Zach Elwood: Hmm. Mm-hmm.

Pete Blair: Uh, it’s probably there. I, uh, there’s an example of a, of a Formula 1 race car driver, and he’s in the Monte Carlo, uh, Formula 1 race, and there’s this curve that when you come out of the tunnel, it’s kind of a blind curve that happens. And on this particular race, somebody crashed right around that curve, and it took out, like, half the participants in the race.

Like, it was a major crash, and everybody was just piling into it ’cause it’s blind going in there. And this particular guy, as he came out of that tunnel and was approaching that curve, all of a sudden he downshifted, slowed down, changed his, his lane position and posture, uh, and avoided the accident, was able to continue.

And afterward everybody said, “Well, how did you know? How did you know?” And he didn’t have an immediate answer, but he thought about it for a long period of time, and he came up with an answer and he said, “Well, you know, I think what happened is, as I came out of the cur- the tunnel, I’m used to there are all these spectator stands around, and people are usually looking at you when you come out of the tunnel.”

Oh. “So it’s, it’s brighter.”

Zach Elwood: Yeah. “

Pete Blair: And because the crash had happened, everybody was looking away, so it was darker.” Not that it makes much difference in my hair, but you know, “It was darker, uh, and that’s what I think I picked up as a clue-“

Zach Elwood: Something was

Pete Blair: weird immediately. Yeah … that it, that it, that it happened, and then that made me take caution ’cause I thought I was in a weird situation.

I don’t know that that e-explanation is true at all.

Zach Elwood: Hmm.

Pete Blair: It’s interesting. It makes a good story.

Zach Elwood: Mm-hmm.

Pete Blair: But there are hundreds of other pieces of information he might have been picking up on, like- Right. Some sort of

Zach Elwood: sounds-

Pete Blair: When you have a- … that he didn’t realize … yeah, the sound. Yeah. Yeah. Yeah. The, there’s, the race noise isn’t the same as the race noise was.

Yeah. The, you know- Yeah … all those things that could be going on. Mm-hmm. And we’re terrible as human beings about taking procedural knowledge and expressing it in propositions, turning it- Mm-hmm … into words and facts- Mm-hmm … about things. If you ask sports people how do you do these things that you do- A lot of times they do, “I just do it.”

If you make them think about it, they’ll, they’ll tell you a story.

Zach Elwood: Mm-hmm.

Pete Blair: But there’s a whole line of research in the sports literature that shows here’s this professional doing this, says, “This is what I do for this sport.” They explain it in great detail. They demonstrate it. They do all those kinds of things, and then we go watch them play, and that is not what they do at all.

Zach Elwood: Right.

Pete Blair: Yeah. And, and they’re not lying. They’re not lying. Right. It’s just we don’t have good- They’re bad at

Zach Elwood: explaining.

Pete Blair: Yeah. We don’t have good access to that for explaining. Yeah. Right? And so- Yeah … that’s a lot of what I think y- I, I caution people about is, you know, and even if you, you say, okay, in the non-verbal stuff, there’s not, um, a clear signal, we still have 5% above chance across a lot of studies for a long period of time, 4% or 5% above chance for a long period of time.

Right. There’s some kind of signal there. Um- Yeah.

Zach Elwood: Anx- anxious signs of anxiety and, and these kind of

Pete Blair: things. Yeah. Well, who knows, who knows what it is, but there’s, there’s some signal there because we’re seeing above– Otherwise, we’d be at 50% exactly, and we’re not at 50%. We’re slightly higher than that. And so- Mm-hmm

there’s some signal. We’ve tried to delve in to identify that, and it doesn’t seem there’s a clear signal from the way we’re coding it, but there’s some kind of signal there. Otherwise, you don’t get that, that finding. So I, I trust that my sensory apparatus are really good at this and really good at putting it together, and I’m just gonna allow it to work and, and delve into things and know also that I’m, I’m fallible.

I’ll make mistakes. Right.

Zach Elwood: Exactly. And I’m

Pete Blair: gonna

Zach Elwood: have- Yeah.

Pete Blair: Yeah.

Zach Elwood: As long as you’re… You know, the, I think that’s the thing. It’s like, I think some people think, you know, me or others who s- you know, talk about, uh, criticize some of the things people say about in the behavior space, which there’s a lot of just plain bullshit, but then there’s also some defensible stuff.

But it’s like, as long as you’re– It’s okay to have… You know, we’re all gonna have our reads and our vibes that we get from people through a number of ways, you know? A- a- and so as long as you’re, as long as you’re not doing things that are y- going out on a limb too much and gonna hurt people if you’re wrong, as, as long as you’re recognizing that you’re fallible, it seems like- Yeah

there’s no- there’s nothing wrong with having various reads and, and such. Yeah.

Pete Blair: Well, and then think about it in the, in the… It’s probably easier to think about in the sports context or the poker context is Do I know that you’re trying to pay attention to me and what my tells are? And so then do I start to broadcast false tells in order to lead you into a, you know, the wrong decision on what you’re doing?

Same thing happens in all kinds of sports, you know, feints and- Mm-hmm … and those sorts of things that you do in order to misdirect somebody in what they’re doing. There’s, there’s a huge amount of that that’s, that’s going on. Once I know that you’re paying attention to this and you think this is what I do, then I can change it up and throw you off, right?

Zach Elwood: Right.

Pete Blair: Yeah.

Zach Elwood: And then, then, yeah, and there’s even the level of, like, an unsophisticated opponent may think that they’re manipulating you, but their, their attempt at a false tell might actually become a tell because they’re so bad at-

Pete Blair: Sure …

Zach Elwood: uh, ma- you know, trying to, trying to manipulate you. It’s like to an experienced player or an experienced interviewer, it’s like, “Oh, I know exactly, we know what you’re trying- Sure

to do with this,” right? Yeah.

Pete Blair: Sure.

Zach Elwood: So, uh, um, well, this is, this is great. Do you wanna share, uh, briefly what you’re up to these days? I know you’ve gotten into other types of research.

Pete Blair: Yeah. So I, I, I really am in the, uh, active shooter world now, where we look at, uh, police response and EMS and fire response to these events.

That’s my primary line of research. That’s- Deeply tied in this idea of training and skill acquisition, ’cause the alert center that I work for is a training organization. That’s, that’s what we do. So I have my professor role, but I’m also running this big grant program full-time that does training. And so we’re always trying to look at how do we get the best training, the best bang for the buck out of the training that we do, which is what led me into the ecological dynamic stuff and all the sports literature, and that is looking at how do we develop skill and, and how do we do that.

Uh, I’ve got a substack called Tactical Science that’s, uh, you can go to and, and check out, uh, my publications up there. And sometimes it’s think pieces. Uh, a lot of times it’s me reviewing criminal justice related research that’s related to the topic areas that I, I think are interesting, and just kinda summarizing them in, in pretty easy to understand terms about this is what the research shows in that area.

Zach Elwood: Uh, very briefly, if you wanted to, uh, I, I saw– I happened to see one interesting thing you wrote about where something about throwing a chair to distract someone in a, in a, um, standoff kind of situation could, could save people’s lives. You wanna talk about what that was?

Pete Blair: Yeah. So that’s part of a, a long string of, uh, studies into reaction time.

And so it’s, it’s a truism in policing that action beats reaction. Uh, so if you, if you were to think back to the old Bruce Lee days where he would do that, that punch and see if you could block it, and he would catch everybody. Uh, if you understand action versus reaction, yeah, Bruce Lee was fast, but that’s not that impressive a display because the person who starts to act first has an advantage because they, they don’t have to decide what’s happening.

They don’t have to perceive what’s happening. They’re just acting, right? And so that’s a truism. We did some research back in the day, uh, on several different, you know, kind of environments that show that that plays out in multiple things, is that, uh, even a police officer with his gun extended, pointed at you, uh, if you were to have your gun down by your side and then you tried to bring the gun up and shoot him, you’re probably gonna be shot at the same time or faster than the actual police officer’s gonna shoot, ’cause the police officer has to perceive that you’re doing an action, the action’s hostile, decide they’re gonna fire, and then fire, right?

So that all has to go together. Um, to do that. So we, we had a bunch of research like that, and we know that we’re asking officers to go into rooms that could have hostile suspects in them. And if we know they’re at a reaction time disadvantage, that the bad guy’s waiting in the room for them to come in and they don’t know exactly where the bad guy is, and that, and that, is there anything we can do to help offset that disadvantage they have in reaction time?

And so the idea of… We actually did a chair and we did a coin. Turned out that the coin was too small. People didn’t see it and track it. Uh, but the chair was big enough that when you throw it, and we had a vision tracker on the bad guy standing in the corner waiting for the cop to come in, and he can see the bad guy as the chair comes in, like automatically track it and then come back to the, the police officers.

The o- idea was that that tracking slowed them down just enough to give the officer a little bit of time to get in the room and, and do that. That was the basic idea of the study.

Zach Elwood: Yeah. No, that’s great. Yeah. It seems like such a, you know, it seems almost common sense when you think about it in hindsight.

It’s like, what a great, what a great tactic. But yeah, there’s, uh, it feels like there’s probably many of these things that are little, little tips that help a lot that people just don’t think of and, yeah.

Pete Blair: So th- this goes to expertise, right? And, and what people do. You have problem sets you’re trying to deal with.

You come up with solutions to the problem sets you’re trying to deal with. So the, as a researcher, your most important job is to pay careful attention to what practitioners are doing, steal all their good ideas, and then publish them and say that, that you thought of it. I didn’t, I didn’t think of that.

It was somebody else who had thought of it, right? Nice. And all of our good ideas at, at, uh, at ALERT come from practitioners who are out in the field doing things, and then we just take them in an environment and validate those to see if they work, you know?

Zach Elwood: Mm-hmm. Mm-hmm.

Pete Blair: And I don’t know of any of them that we’ve tried to evaluate yet that don’t work because- Hmm

the practitioners have a pretty good feel for what’s happening.

Zach Elwood: Mm-hmm. Yeah, which gets into why Tim’s work was so great involving you and other people. He was really curious about what people were actually doing in, in the field and such. Yeah. Um, well, this, this has been great, Pete. Thanks a lot for, uh, for your time.

Pete Blair: Oh, thanks for having me on.