Annie: I'm excited to welcome my guest today, Daryl Morey. Daryl Morey is President of Basketball Operations for the Philadelphia 76ers. Daryl was profiled in Michael Lewis's book, The Undoing Project, and his analytical approach to basketball is often labeled “Morey Ball” in reference to Lewis’s earlier bestseller Moneyball. Daryl studied computer science at Northwestern University and has an MBA from MIT Sloan School of Management.
His career began with three seasons as Senior Vice President of Operations for the Boston Celtics in 2002, and was then followed by 13 seasons as general manager for the Houston Rockets from 2007 to 2012. The Rockets’ eight straight playoff appearances from 2013 to 2020 represent the longest streak in the NBA at that time.
In 2006, Daryl co-founded the MIT Sloan Sports Analytics Conference, an annual event that brings industry professionals and students together to discuss the increasing role of analytics in the sports industry. In 2018, Daryl was named NBA Executive of the Year, among executives with at least a thousand games of experience, Morey’s record is fifth best in league history.
Hi, Daryl. It's so nice of you to come on this podcast. Thank you so much.
Daryl: Thanks for having me, Annie. I appreciate it.
Annie: So we got introduced through Michael Mauboussin, who I know you're a huge fan of, and I am as well. So thank you, Michael! Maybe we'll get into some of the things that he's written about. And I think we may have a couple of other people in common that you know, Cade Massey, Richard Thaler...
Daryl: Of course! Dick [Richard Thaler] and Cade wrote the best sports paper ever: the one on the first pick in the NFL being a bad pick [Overconfidence vs. Market Efficiency in the National Football League]
Annie: Right. Didn't Thaler just recently present at the MIT Sports Analytics Conference on the question of why the NFL has been so slow to adopt analytics?
Daryl: Yeah, it was more about the NBA and how the NBA is pretty far behind. I would say for the audience that might be familiar with blackjack, it would be like if blackjack was not doubling down on 11, “Maybe we should double down on 11” and then everyone's like, “Oh, he's so smart, they're doubling down on 11!”
Annie: Wouldn’t it be more like if nobody doubled down on 11 for years, and then some mathematicians, or data scientists, came along and said, “You really shouldn't be doubling down on 11!” and then the whole league said “nah!”
Daryl: I think you have it more right. They said “nah” for a long time, and then a few teams started to do it, and then everyone copied. That's right.
Yeah. They had amazing shooters, like Larry Bird who they'd actively tell not shoot 3-pointers. So yeah, it was even worse.
Annie: So Michael Mauboussin actually just wrote about this, thinking about that in relation to investing. He was looking at the history of the 3-point shot.
So why don't we just start there. If we take something like the 3-point shot… people can understand the concept that three is more than two.
Daryl: You’d be surprised, Annie!
Annie: I think it's such a good example. I've talked to a few people from the NBA and asked them the same question. You're considered one of the key individuals who's brought sports analytics to the game, and you've got a competitive advantage from the fact that you understand this stuff and you're willing to be an early adopter of it.
When you think about something like the 3-point shot and you have somebody like Larry Bird who clearly could shoot from outside that line, why do you think it is that even though the math was very clear, people weren't doing it? Maybe you could just walk a little bit through the math. I mean, “three is more than two” is a little bit of a simplification because it depends on the probability for the shot. So can you walk us through why we know that you should be shooting more 3-pointers and how that got adopted within the NBA?
Daryl: Sure. Yeah. I’ll keep the math simple, and then I'll go into why, because the “why” is more interesting. If you look at any point in NBA history, and you just take all the shots behind the 3-point line, they make it somewhere between 35 to 40%. And then you take all the shots away from the rim, but inside the 3-point arc. And you see that, league-wide, teams are making about 40% or a little bit over 40% of those shots. And the math isn't that hard. I wish it was harder!
It's really 3 x 35-40% compared to 2 x low-40s, and one's bigger. Part of what people miss, that adds to the 3-point shot being worth a lot, is that for the same number of points generated, if you go 2 for 6 from the 3-point line, it gets 6 points. If you go 3 for 6 from the 2-point line, you also get 6 points. But there's one more miss for the offense to try to offensive rebound. So it actually boosts it even more. And the reason people didn’t do it was a lot of psychology, more than anything. To stick with a blackjack example, imagine… blackjack has been played for a long time. People who are very smart have written books on how to do it, and they've got strategies and everything. And then all of a sudden, next year a big casino says, “We're going to have to change it up a little bit. If you hit on 16 or higher and win, we'll pay you triple.” People would say that you probably still shouldn't, you’re going to lose. But if you make that benefit big enough, people should start hitting on that number, even though you're going to lose at a higher rate.
That's pretty similar to what happened to the 3-point shot. There were coaching philosophies, books, clinics, everything written about how to play without the 3-point shot. All the philosophies were based on how to get the best shot as close to the basket as possible. That makes sense. If every shot is worth two, getting the most open, close shot to the hoop is the right thing to do. So every coaching clinic and every system was set up to do that. And then all of a sudden, they say, “We're going to make this shot from way out here worth three [points].” All those people who are invested in that culture and how it's done, they said “It's just a gimmick. Of course you'll make a few, but you can't systematically do it.” And they were somewhat right. [But] what they missed is that the benefit was so big, that you should still be doing it. And then it took a long time for coaching philosophies to see any change, or player skills to change, to take advantage of this shot.
Right out of the gate, the [3-point shot] was worth so much more than two points. 50% more, it's not that hard to figure out! All the reasons why people weren't using it were trumped by how big the reward was.
“Right out of the gate, the [3-point shot] was worth so much more than two points.
50% more, it's not that hard to figure out! All the reasons why people weren't using it
were trumped by how big the reward was.” — Daryl Morey
Annie: So I have two different questions that I want to ask about this. And the first is: how much do you think asymmetric loss aversion plays into this?
For the listeners, loss aversion is: we don't like to make choices that carry some great chance of loss. So when you're talking about a 3-point shot, you're going to miss it about 65% of the time. So [due to] loss aversion, you would tend to want to take the option that carried a lower risk of loss rather than [the one with] a higher risk of loss, even if the expected value—the number of points that you would make in the long run—might be greater. But then there's also asymmetric loss aversion, which is that when we choose a status quo choice, loss aversion isn't triggered as much. In other words, if we usually take a 2-point shot and you miss it, people say “Ah, what could you do? You missed…” but if instead you do something new, like this new 3-point shot, it's much more in view when that basket is missed, and you're much more worried about the miss, so you lose sight of the expected value.
Daryl: Everything you said makes sense. I can tell you're a decision science person! You're a hundred percent right that if you're the gazelle in the herd, with the lions [chasing you] and you're in the middle of the herd, you're much less likely to be scrutinized than if you're taking a correct but radical strategy—at least perceived to be radical at the time—but no one else is taking it, then you have a decent chance of getting tackled by the lion and eaten, even if it was the right approach. In this case, getting tackled and eaten by the lion is getting fired, getting kicked out of this sport you love, when you’re someone who makes their livelihood off of it.
So all those factors came into play. The other factor that came into play is that for the first three quarters of a basketball game, expected value optimization is the right approach. But when you get to the end of the game, you can actually get to win probability maximizing choices that aren't taking the three. If you're ahead, for example, then you're trying to collapse the variance and have less standard deviation around your points scoring, even if your expected value is slightly lower. People who don't like the 3-point shot use those one-off anecdotal times to argue that [logic] into the rest of the game. Because there are these times when a 2-pointer is better than a 3-pointer, [they argue that] that ergo all the time the 2 is better than the 3. This, along with loss aversion, was a factor for sure.
Annie: So people are using arguments that in one situation are totally valid, and in order to rationalize the bias, they're bringing it into other situations that are not valid.
Daryl: Correct. I bet a lot of people who are listening are basketball fans, I hope. Another big one is they'll say, “well, when you were giving those stats earlier about the 3-point percentage from three versus two, on average: that's just an average. What about if it's the very open shot from 18 feet versus the very guarded shot from 3 [feet]? That's actually a really good question. It's one of the first ones you should look at. Averages can be super misleading, as you know. You might have a bimodal distribution, you might have all these different distributions that make it a really dumb thing to look at the average. It’s like looking at the average of a basketball team, when you have a 5'5" guy and a 7'3" guy. That average is not very useful!
But it turns out that if you say open shot versus [a] guarded [shot within] three feet, it is true that it tightens the gap of the advantage, right? The problem is that the benefit of 50% that they applied to the 3-point shot, trumped all that. So they were correct in the direction of their arguments, but they missed that there is such a massive benefit to the three that even when you adjusted for openness and things like that, it was still optimal to generally take a lot more threes.
Annie: So the other thing that I'm thinking about is the career risk side, which you mentioned. The gazelle gets eaten by the lion, which in the NBA means that you get fired. And I think about some of the cases of these radical adoptions of analytics that are occurring by people who aren't carrying the same kind of career risk. [For example] the coach in Texas who just didn't have a punter. I think he's now a college coach...
Daryl: Yes it's Kevin Kelley I think his name is. Yeah.
Annie: Here's somebody who's just coaching high school. Right? He's obviously read some interesting math and comes up with the idea that usually punting is not the best choice, you ought to be going forward a lot more in 4th down. So he doesn't even have a punter. But he's coaching in high school. This isn't, for example, on the NFL stage, right? I think about the difference in Belichick's early career, in terms of his hewing to the analytics, versus his mid-career. So early in his career, he actually wasn't going forward on 4th down in a lot of situations where you ought to, and he was much more risk averse. And then you see this big change, once he wins a Super Bowl. And you see him start to really stick to those analytics, but of course he's not carrying career risk. So [I’d like to get your] thoughts on that. And then from your perspective… you are a computer scientist, you went to the Sloan Business School at MIT, the Sloan School of Management, so you didn’t rise up through the ranks of basketball in the same way. So do you feel like that was also an advantage for you in the way? You came into the league and you were thinking about bringing analytics, that allowed you to be absolutely one of the pioneers...
Daryl: Yeah. I absolutely benefited, first with Wyc Grousbeck and Steve Pagliuca, the owners of the Celtics. I always came in with a lot of senior level support and, you know, anyone listening to this knows how important that is. So I first came in with a lot of support at the Celtics, but my thoughts, and the direction I wanted to go in, were seen as radical. So I was with a governor, Danny Ainge, who is a great general manager, but very open. He would take the good and throw out what he didn't want, what he thought was too aggressive. And that was the comfortable thing.
And then again, I had huge senior level support when I was hired at the Rockets. The owner at the time, Leslie Alexander, basically said, “Look, I'm sick of all this ‘normal’ that's going on. I want to be completely radical. You go!” Without that [support], change is extremely difficult. And change is hard even when you do have that [support], as seen by how long these adoptions have [taken to] come, even with very senior level people and owners saying, “We want to move on this.” And huge incentives, by the way.
“[The NBA is] the kind of industry that should move the quickest to new ideas, because the results of our changes on the floor are seen nightly. If we take a strategy and put it on the floor and win the game partially because of that strategy, that's instant feedback that we're moving in the right direction. Whereas in most businesses, when they're trying some new organizational process, it might be 18 months before their supply chain that they're working on delivers some new advantage. We see results instantly… so our changes should be even quicker. But even with all that, the change was slower. I think there are bottom-up changes that can happen in industries, but it's generally senior level support and resources, along with someone who can execute, that makes it happen. And it's sad that it often doesn't happen without that.” — Daryl Morey
[The NBA is] the kind of industry that should move the quickest to new ideas, because the results of our changes on the floor are seen nightly. If we take a strategy and put it on the floor and win the game partially because of that strategy, that's instant feedback that we're moving in the right direction. Whereas in most businesses, when they're trying some new organizational process, it might be 18 months before their supply chain that they're working on delivers some new advantage. We see results instantly. We see benefits quicker, so our chains should be even quicker. But even with all that, the change was slower. I think there are bottom-up changes that can happen in industries, but it's generally senior level support and resources, along with someone who can execute, that makes it happen. And it's sad that it often doesn't happen without that.
Annie: Yeah. I think there's such a good lesson in that. Across your decision-making life, it is so incredibly important to have a really good decision-making environment and support for you to be able to do the right thing, from expected value or maximizing win probability.
How do you deal with [the fact] that there's the support within the organization itself, but then you have this problem of the demands of the fans. So I think about this in terms of team building, right? So sometimes you’ve got to strip it down in order to build it back up, and then people can be pretty intolerant. Of course, just coming to Philly, you know, that “trust the process” used to be spit out as an insult by fans.
Daryl: Or positive. I think there was a dichotomy there.
Annie: Initially it was a compliment. And then, you know, it takes a while. And there's some intolerance... and then “trust the process” was an insult.. and then it went back to being a compliment again. But I think there are different time horizons. This is something that Michael Mauboussin has talked about as well. There are different time horizons for what the fans want. There are different time horizons for what the owner wants. There are different time horizons for what the GM wants. There are different time horizons for what the coach wants. And that's creating a lot of friction!
As someone who's managing a team and making these kinds of decisions, how do you allow yourself to then still make the right decisions? How are you thinking about this balance? Are you even thinking about that balance of the different time horizons and how do you deal with the fallout when the outcome goes in a way that doesn't look great for you?
Daryl: When you're making decisions to be a little bit out of the norm and be aggressive... And I won't say specifically, we just had one in Philadelphia where we took a step back and said, “Hey, let's be more aggressive in area X.” I spend a lot of time saying “Okay, everyone in this room, we're all together. We're going to take this more aggressive approach. It also means that we're probably going to fail at a much higher rate than before. And we can't be freaking out the first one, two or three times if things don't work out.” And that's an important thing for the senior leadership, like myself, to set that tone. [I tell] the coach and also ownership, “We're taking this [decision] as a collective and when it goes bad, you're going to be able to blame Daryl. And if it goes well, it's going to be a collective decision.” We need to take this bet because our alternative—and it's always important to look at your alternative—is a lower odds to win the title. We might have a higher expected value [with a safer option], but we have to be the best of 30. So if you have to be the best out of 30, by definition, you have to inject more variance into your system. You have to take bigger bets. If it's the Ricky Bobby of the NBA with “if you ain't first you're last,” you have to inject more variance into your system all the time. In fact, no team even injects enough. If you take the actual math of the NBA, teams should be doing radical things constantly because they're just trying to be the best of the 30. You don't actually see that behavior, however.
“When you're making decisions to be a little bit out of the norm and be aggressive… We might have a higher expected value [with a safer option], but we have to be the best of 30 … So you have to inject more variance into your system. You have to take bigger bets … In fact, no team even injects enough.
If you take the actual math of the NBA, teams should be doing radical things constantly because they're just trying to be the best of the 30.” — Daryl Morey
Annie: Do you know what's so funny is that I hadn't really considered that in relation to the NBA, but that's a total poker concept. So there are two different types of tournaments you can play in. One where there are graduated payouts. So you know, first place might get like 40% of the prize pool, and second place gets 20% and third place gets 10%. And then there are dribbles all the way down so that the top 10% of people who finish in the tournament will get some money. And if you're right on the bubble and you just cash, you might get your money back. So obviously you're still motivated to come in first place, but it's not like you get nothing for the other places, and that actually changes how much variance you're willing to take on at different points in the tournament. But for a winner-take-all tournament... now you can kind of think about it this way, that in a graduated tournament, there are certain points where you might take like a 51-49 favorite for yourself because the variance is too high. It's too high volatility. And if you're out, you're out, right? So you don't want to do that in certain situations, but in a winner-take-all, you would do that every single time. You have to take on that volatility because there's no point in getting the second most chips. It does not help you.
Daryl: A hundred percent. And actually I'm bad at poker, right?
Annie: Well I'm very bad at anything to do with basketball!
Daryl: I'm fairly intuitive with probabilities, but I ended up in a Poker After Dark… I think maybe you did the first one ever, or something like that...
Annie: I might have. It was a long time ago!
Daryl: So I was in it with Scott Blumstein, former World Series Poker Champion, Maria Ho, and Nate Silver, who I think is on his way to doing really well at the World Series of Poker right now. And so I consciously knew I was worse than them, but it was winner-take-all. I think this episode is coming out... you'll see me going all in, all the time. It's probably even overly risky, but I figured my only chance was to get a huge stack against the one player there who was worse than me, according to Scott and those guys who really knew every player at the table. And so I was actively trying to take all his chips, figuring that I had no shot against anyone else. And I was going all in, all the time. I rationalized it in my head that this was my only shot at beating these amazing pros that I'm absolutely at a disadvantage to!
Annie: Well, so I have a famous story of that strategy. Just as a side note. I was in the NBC Heads-Up National Championship and I was in the finals. So it's just me and one other player. And that other player was Erik Seidel, who for people who don't know is a nine-time champion at the World Series of Poker. And like he's won 40-something million dollars playing poker. He's amazing. He also happened to be like my mentor. So you play three heads-up matches to determine. So it's the best of three. So we played the first one and I actually beat him pretty soundly because he didn't quite have my pattern down yet. But we came back for the second match and it was like he had spent the time unlocking my style and just wiped the table with me. So I got into the third match and I did what you did. I said, “Okay, I have a problem because as fast as I can adjust to what he's doing, he's going to adjust faster to me. Because he's just a much better player than I am.” So I said, “I'm just going to look for an opportunity to get all my chips in. And I would rather get all my chips in at 40% to win, than try to make hundreds of small decisions where he's maybe 55-45 or 52-48, where over those hundreds of decisions I'm guaranteed to lose. I'd rather just take my chances one time. So I did that, and I won!
Daryl: Yes! That's awesome! And is it similar to when you get short-stacked and one of those things you just go all in, even if you know, you're going to ground out?
Annie: It doesn't really help you to win yet. You need to be able to play a handout. Like you can't get to real decision-making if you can't actually play a handout. So you need to actually try as much as you can to say to the person that you're playing against, “I could damage you,” so it still gives you a little bit of power. Anyway, that's so fun.
So what I'd love to understand is, as you're thinking about these trades, what is your process? Because there is obviously variance, so what does the likelihood that a person actually does become a star have to be? How does that come into the equation?
Daryl: Sure. It depends on where your team's at. So if you're a team that already has some foundational pieces... it's a different optimization once you have like your first foundational piece, then getting 5% edges can make a real difference, right? It can actually move you from being like last year where we were the one of the final teams standing, but not quite good enough to win it. If we could take our team last year and add 5-10%, that might be enough to win the title. But if we were a different team, a team out of the playoffs, doing these incremental moves doesn't do a whole lot.
Annie: Oh yeah. Would you relate that in terms of the NFL to… in terms of the quarterback, it's very high risk, there's a lot of stuff that can go wrong, but high reward. If you can get that quarterback and you're going to have a lot of misses because it is high volatility, but if you can get them in place, then you can start to make a lot more low volatility team-building around that person. Would you say that that's a similar way?
Daryl: Yeah I'm not an expert in the NFL, but that seems very similar. I do know there's only been a very small number of Super Bowls won by what people would consider an average or below-average quarterback. It shows how critical that position is. What's really tough for NFL team-building is, at least as far as I know, generally the prediction rate is very low about which quarterbacks are going to be good. So you have this thing that you have to get, to have a chance to win, but at the same time, by definition, there is high variance coming out of the draft. There have been [studies], this was in Dick [Richard] Thaler and Cade Massey's paper, that I think the fail rate on succeeding and picking quarterbacks might be the lowest of all the rates. On some level it makes sense, because you're a crucible into one, whereas if you're a wide receiver, you have more chances to get on the field. But some of it's just that they're very hard to predict.
Annie: Yeah. That's really, that's so interesting. I hadn't really thought about it that way. So you think about it as I want to get that anchor player and then build around them. And then I assume when you're building around them, sometimes you get lucky.
Daryl: It doesn't happen too often. In fact, it's happened to opponents of mine, and it's really frustrating when it does happen.. when Golden State famously won 73 games and then added the best free agents in the off season because there was a windfall of money that came into the league. And so that was super frustrating.
Annie: So I grew up in New Hampshire, so I grew up watching the Celtics. I just think about the Celtics when I was growing up, obviously Larry Bird was amazing, but it's not like Kevin McHale was a slouch. Or Robert Parrish.
Daryl: Or DJ. Danny Ainge, Bill Walton...
Annie: Then you had John Havlicek as the sixth man.
Daryl: Well, Havlicek was before.
Annie: A little before, but they were building into that, right?
Daryl: So you're right. He was a sixth man on another great team, you know, with Jo Jo White. I mean, you can keep going … I got this incredibly lucky chance to work with Red Auerbach. Very limited, because he was very old when I worked with them at the Celtics. But the one thing I learned in my very brief interactions with the Celtics was that it was no accident the Celtics were good consistently. As far as I could tell, it was because Red Auerbach was an innovator for years, and ahead of his time… picking Larry Bird by drafting a year early before they changed the rules. Frankly, in these sports and places where the playing field is pretty level, taking advantage of the rules that aren't in place yet, or are in place, but you take advantage of it more than others, is actually a more robust strategy than quite a few other strategies because it's very tough to get an advantage when 30 pretty intelligent organizations are all competing with each other.
Annie: I would imagine that's a place where really being deep in the analytics can be helpful. Because if there was a rule change, I imagine understanding what the consequences of that rule would be, unintended or otherwise, and how that might change the structure of the team in general... Being able to see what that impact might be, and being able to forecast better than other teams, is then going to open up an advantage for you.
“Being deep in the analytics can be helpful. Because if there was a rule change,
I imagine understanding what the consequences of that rule would be, unintended or otherwise,
and how that might change the structure of the team in general...and being able to forecast
better than other teams, is then going to open up an advantage for you.” — Annie Duke
Daryl: It's a huge advantage. The NBA has gone through it right now. It's early, we're only five or so games into the season for the league, but the interpretation from a lot of the league has changed, pretty radically from my eyes. The data's too early to really say whether the change is fundamental or not? But from my eyes, baseball went through it. Their ball will change radically from year to year. And I don't even think it's on purpose. I think just slight changes in the manufacturing, even unintentional, can really change Major League Baseball. So we had a change in the NBA where all of a sudden in 2007, you could no longer touch a player on the perimeter without a being a foul, and that radically changed the league. So you're right. Being able to rapidly adapt to these rule changes becomes a big factor.
Annie: So we're talking about the idea of how might a rule change, or the change in the ball, impact things in a way where you can see it in a way that other people don't see. I think this is such an important concept for gaining an advantage in decision-making.
So when you think about decisions in basketball, is there any that you can point to that are much more important or impactful than people realize? Maybe executives within the NBA [realize], but fans who are watching the game, they don't understand how important those decisions are.
Daryl: Yeah. A big one that I think fans see, but they don't know how important it is, is the pace of play. So because two teams are playing each other … and let’s say it's a Tuesday, and it’s a good team playing a bad team where the good team is at 85% odds to win.
There's one [team that has] played back-to-back, one’s played a bunch of games, and they're playing each other. If you're just looking at the score, it'll actually look like a very similar score to a game that you might see in a playoffs, like 101-98. But the immediate thing that we can see when we watch the game is the energy level, the effort, and how hard it is for someone to beat someone on their first move. All these things pop out immediately. And you can tell that both teams are sort of mailing it in tonight. Whereas for the casual fan, if both teams are mailing it and it sort of cancels out, then they can't really tell. They're like, “Oh, he got to the basket!” but on one night, it could be an amazing move, he beat a good defender. On another night, a guy getting the basket could just be that both guys are tired and the defender was just lazy. And I'll be able to pick those up, whereas someone else won't.
Annie: So when one team is really good, they're kind of taking the night off because they want to save their energy for when they're playing someone who's closer in skill. And when the other team's really bad they're also kind of taking the night off because they know they can't win anyway!
Daryl: Yeah, it's not necessarily conscious. It's a lot of just human psychology. And we know that humans are very good when there's a rush of adrenaline. If you're playing a team that you expect to beat, you don't come in with the same level of “This possession is the one I'm going to win!” That happens a lot in every sport. I think in the NBA, it's even more pronounced, you know, given the number of games we play, the amount of minutes that our players are expected to play. You definitely get some dragged out, back-to-back games between a great team and a worse team that isn't as high quality of a game, for someone who watches it for a living.
Annie: So it's funny because when you said “pacing,” I thought of something that's been on my mind, that I've always wondered if it’s kind of a nutty thought. So when we were talking about the match that I played against Erik Seidel or what you were doing in the Poker After Dark, which was a correct thing for you to do, you are essentially trying to reduce the number of decisions that you have to make against people who are going to be better at decision-making than you are.
So this is a very simple concept. If you force me to flip an unfair coin and we're betting $10 on it, and I get to choose how many times, but it has to be at least once, I should want to do it only once. But if I bet on the good side of that flip, I want to do it a hundred times.
So I've wondered about this in sports, whether it's NFL or NBA. If you know that you have an advantage over the team that you're playing, do you speed up play? I know at the end of the game, there's things that happen that are going to distort this, but do you consciously try to speed up play so you're getting as many touches of the ball as you possibly can, and vice versa? That's always been a nutty thought I've had in my head.
Daryl: Teams don't think as intentionally about that as they should. I would say the reason it hasn't come up in the NBA is that playing at a higher pace confers an advantage to the team that plays at it. If you play four on four or three on three, the offense has an advantage more than in five on five, or if somehow we were to play six on six, the offense would get hurt more than the defense. And so, if you're the better team you want to play at a higher pace to get more back and forth, a hundred percent. But you also want to play at a higher pace because it's actually just better.
And it's also true the other direction. It’s interesting. Before they had a shot clock, that's why you saw four corners offense, things like that. In theory, before there was a shot clock—and soccer teams should try this—if you're the worst team you should just play “keep away” if you can, and then take one shot, and it would be the worst thing ever to watch!
“If you're the better team you want to play at a higher pace to get more back and forth... And it's also true the other direction. It’s interesting. Before they had a shot clock, that's why you saw four corners offense, things like that. In theory, before there was a shot clock, if you're the worst team you should just play ‘keep away’ if you can, and then take one shot, and it would be the worst thing ever to watch!” — Daryl Morey
Annie: One flip of the coin. Just see what happens.
Daryl: I'm actually shocked that high school basketball isn’t like this because most high schools can't afford a shot clock, or it's not implemented in their leagues. High school basketball should be absolutely awful if they were just trying to optimize winning. Now, we know that there's more things to high school basketball than just winning games.
Annie: Getting scouted...
Daryl: ... entertainment, all these things. But the reality is the worst team should always just hold the ball the whole quarter if they can. Often you can press and trap and things like that. But in theory, if you could just take one shot per quarter. That would be the ideal, obviously.
Annie: Yeah. I love the way this conceptually crosses so many different things: the idea of “how good am I?” and then the second question “How much variance do I want to bring into the situation? Do I want one coin flip or do I want a hundred?” This is generally a decision-making framework. It's a good mental model to apply to decision-making and of course it takes some humility because you have to recognize, “Erik Seidel is way better than me. He's just going to crush me.”
“You are essentially trying to reduce the number of decisions that you have to make against people who are going to be better at decision-making than you are... [Ask yourself] “how good am I?”
and then the second question “How much variance do I want to bring into the situation? Do I want one coin flip or do I want a hundred?” This is generally a decision-making framework. It's a good mental model to apply to decision-making and of course it takes some humility” — Annie Duke
Daryl: Yeah, I find that there are a lot of poker players who have been in professional sports. I find that they just intuitively understand risk, reward, variance, payouts, bet sizing, all these kinds of things. So poker is—and I think you've written about it—probably one of the best games as a framework for life decisions and things like that
Annie: Hopefully! I wrote a whole book on that premise ...
Daryl: Yes, exactly. So I'm preaching to the choir. Buy Annie's book!
Annie: Well, thank you. So I cannot get away without asking you about Ben Simmons because my brother relentlessly pings my husband, since my husband has been a Sixers fan since 1982. So he always says this: “He's in, he's out, he's in, he's out!” So I'm just curious how's that going for you? How are you thinking about his future with the 76ers? And actually, not just in terms of trying to get some inside info here, but this is a decision that's receiving tremendous attention as you're making it. And I'm just wondering how does that also affect your decision-making?
Daryl: Yeah. I think if I was earlier in my career, the last part you asked about would have been more challenging.
I think in this case, the approach we need to take is so obvious to me that you're going to be comfortable with the criticism. I think, just like in poker when you got to the final table with Erik and you made the +EV [positive expected value] bet against him even adjusting for how he plays and how you play, and you made it. I would hope that you're going to go back and sleep that night and be like, “Hey, I took the right play, it just didn't work out.”
And so I think we're at that place where we know Ben Simmons is an extremely good player. We know he's someone who committed to the 76ers for five years and has four more years. We know that he brings elite defense and he brings pace to the game and actually early returns on the season. Those are the two areas where we're struggling the most relative to the past. No surprise, given how elite Ben is at those two things. And so we know that our job is to win the title, and we know that Ben Simmons makes us a much better team to win it. So when you get into a situation with a player who wants to leave, we have to focus on the 76ers. Now we do care about Ben Simmons, but my first job is to the Philadelphia 76ers and their fans.
Annie: I'm going to let my husband know that answer. He's very, very interested in this!
Daryl: Well, I think there's more than just him who's interested, so.
Annie: I know. He's a representative of all of Philadelphia though.
Daryl: Yes. I love it. Actually I grew up in Cleveland and so the fans here remind me so much of Cleveland. Hopefully I'm not insulting Philly fans, but I didn't grow up here, so I can only compare to Cleveland Browns fans who are very passionate, let you know when they don't agree ...
Annie: Well as you know, Philly is known to boo its own teams, so ...
Daryl: They do, and that's unique, but I think it comes from a place of love though, where they're trying to push the team to a higher level of play. So I appreciate it!
Annie: Philly fans are passionate and they stick with the teams. They are not fair weather fans. So I agree with you, I think when they're booing, it's like, “Come on, we want to win, because we're never going to leave you!” I think that that's true. I suffered through like Buddy Ryan at the Eagles a long time ago.
This has been so amazing. I want to just wrap up with a couple of things. So obviously you have such an incredible, varied background that you've taken. With you, I'm thinking of something like David Epstein's Range. You take all these different perspectives and backgrounds and you bring them into something new and it helps you to succeed. For people who don't know, Range is [a book about] why generalists thrive in a specialist's world. Amazing book, people should check it out. But let me ask you: what book would you recommend for listeners who really want to improve their decision-making?
Daryl: That's such a good question. I'm going to send you a list of books because I've done that before, so you can put it on your site. If I only get one book... I do think Daniel Kahneman's Thinking, Fast and Slow is probably the best of that.
The greatest thing ever just happened with Daniel Kahneman, which was that he basically said that one of the chapters of his book isn't right anymore… which you see so little of. It shows how great he is that he's updating his beliefs as new information comes in. As anyone who's written a book [knows], which I have, and you've done multiple, you know you're like, “Oh man, I wish I could rewrite that chapter because we've got all this new research.” And so the fact that Daniel Kahneman isn't a zealot and says that the rough frameworks that were the foundations of Behavioral Economics and Decision Science are right, but we've been wrong here, and this is where we're updating. That just shows why he's so important. So I would read the whole book and I would also read his recent, you know, “Hey, I think we were a little [wrong with one part]”. Was it actually loss aversion?
Daryl: But loss aversion has had a ton of stuff too. And in fact, again, poker players or blackjack players would know where loss aversion was wrong. So a lot of the early studies in loss aversion were not taking risk of ruin into account. So then we're taking Kelly criterion, and it turns out humans often are a little more rational than people think. We screw things up a lot too, but essentially humans were using a crude or bad heuristic version of Kelly criterion in a lot of their loss aversion. And when you adjust for that, it's not quite as bad as some of the original research.
Annie: So, what, what Kahneman talks about now, I think, which is super interesting related to loss aversion. There's what he's termed Sure Loss Aversion, which deals with that. You lose this problem: You owe me a hundred dollars. Do you want to flip a coin, double or nothing, or I owe you a hundred dollars, double or nothing. Do you want to flip a coin, double or nothing? Now those are identical propositions. If you flip a coin, double or nothing for a hundred dollar gain, the expected value is a hundred dollars. Vice versa, if you do it on the loss side, the expected value is a loss of a hundred dollars. What happens is that on the gain side, people say, no, they want to bring the volatility to zero and take the gain. But on the loss side, they say yes to that and they want to gamble for it. Now what's interesting because they don't want to turn a paper loss into a realized loss. So that's why it's called Sure Loss Aversion. I don't want to turn it into a sure loss.
Daryl: Makes sense.
Annie: What's super interesting is that if I say to you, “I owe you a hundred dollars. Do you want to flip a coin $220 or 0?” I know Daryl Morey is going to say yes, but most people will say no. So they're giving up $10. Right. But if I say “You owe me a hundred dollars. Do you want to flip a coin, negative $220 or $0?” Daryl Morey is certainly going to say no, because you're not going to pay me $10 for the opportunity, but most people will say yes, that's how much that's how strong Sure Loss Aversion is. So this gets rid of this Kelly criterion problem.
Daryl: Oh, interesting. You're one step past me. I'm learning something, this is great.
[The concept of] hot streaks have also been updated in that they had the math wrong, and then they had another thing wrong. We knew hot streaks existed, but people still think the effect is way bigger than it is. We knew they existed just from the second free throw which always had a slightly higher percentage. There’s a whole muscle memory aspect to it. And mostly why hot streaks didn't exist was because: A) they had the math wrong and B) people are taking much harder shots, they're like, “Oh, I'm hot so I'll just take really hard stuff.”
It'd be like [in blackjack if] you had an 11? He got a 10, 21… “Oh, I'm hot, I'm going to hit on 15 now!” Of course in blackjack, there are no hot streaks, let's be clear. But in the sports where there is a physical element especially whether there's muscle memory involved, it has been pretty much proven that there's these slight hot streak effects.
Annie: Yeah, I've thought about that also, in terms of poker. And I think about that with football and basketball, how it's an interaction of other people with you as well. In poker it's called a rush, as opposed to a hot streak. If you're perceived to be hot, then people attribute to you a stronger hand that you actually have, so it makes it easier to bluff. And the easier it is to bluff, the more likely you're going to win, because bluffing is a very key component of the game. So it kind of becomes self-generating because once they perceive that you're lucky...
Daryl: Can I ask is that even going to work on pros or do pros not have that effect?
Annie: You know, I think there is some subconscious [thought of] “I don't want to get in that person's way, they're winning every hand” that goes on. I hope I'm not susceptible to it, but I bet you I was. I would put a bet on it, because that's the way the mind works.
One last question. So I agree with you about Thinking, Fast and Slow. If I had to choose one book, it would be the one that I would choose. I always throw in Michael [Mauboussin]'s book too.
Daryl: Which one? He has so many good ones!
Annie: The Success Equation, on the influence of luck and skill. Because I just think it's so interesting to understand how luck and skill influence your life, and about how you should think about those two different things. And then I've just got to give a plug for Phil Tetlock's Superforecasting.
Daryl: I love Superforecasting! I think Phil lives in Brooklyn and he lives close by too.
Annie: Phil's a Philly guy!
Daryl: Does he live in Philly?
Annie: He's at University of Pennsylvania. You know why? I work in his lab!
Daryl: Oh wow. I actually think Superforecasting was a big breakthrough. And the one I read cover to cover most intensely when it came out in the last 10 years. Did it come out five years ago now? Maybe six?
Annie: Yeah it was probably six or seven. It was a while ago. Yeah. It's such a good book.
Last question: what decision-making tool, idea, or strategy would you want to pass down to the next generation of decision makers?
Daryl: I would say probably the main thing, especially if you get into a leadership role, is to make sure you're hanging back on your opinions before you ask people their opinions. That's probably the number one. Even people who are experienced and brave decision-makers, once they hear the direction the lead person is going, they can't help but line up behind that.
“If you get into a leadership role, make sure you're hanging back on your opinions before you ask people their opinions... Even people who are experienced and brave decision-makers, once they hear the direction the lead person is going, they can't help but line up behind that.” — Daryl Morey
Annie: You are a man after my own heart. I have a line in How to Decide. Which is, “If you want to know what someone thinks, don't tell them what you think first.”
Daryl: A hundred percent, you can’t. You have to catch yourself because often you feel like you know the answer. On small decisions that don't matter sometimes for speed, I'll just be like, “Hey, if anyone disagrees [let me know].” But for big ones, it's very important that you create a space to allow people to disagree.
Annie: One trick that I use is that I think about the feedback that I'm trying to get from the group, and actually send [the question] out to them in advance of the meeting and have them answer it to me asynchronously.
Daryl: That's great, yeah.
Annie: Because I can withhold my opinion, but as soon as Person A speaks, they're already influencing everybody else. It also helps my thinking because I get to think about what the feedback is that I need, and then send [the question] out. And then I bring [the answers to the meeting] together and everybody gets to examine it before any discussion. It really reduces cross-contamination. So I'll pass that on to you.
If listeners want to go online and learn more about your work, or follow you on social media, where should they start?
Daryl: Probably just Twitter, I'm @dmorey. I'll tweet out a lot of nonsense, but hopefully some good stuff as well. That’s probably the main place to go.
Annie: Well, I follow you. You're a good follow, I like your nonsense!
Oh for people who don't know, Daryl and I are one day apart on our birthday. So [we had a] mind meld before we even started talking!
I am so appreciative. I know how incredibly busy you are. You're obviously in the middle of everything and I'm so appreciative of you joining and having this really cool conversation with me. Thank you.
Daryl: Thanks for having me on, I appreciate it, Annie. Thanks.
Published December 8, 2021