00:00:00.017 --> 00:00:05.517
Have you ever found yourself trying to select one thing, but there were too

00:00:05.517 --> 00:00:07.577
many options and so it took a long time?

00:00:07.897 --> 00:00:12.757
Think of a restaurant menu or the toothbrush aisle at a store.

00:00:13.057 --> 00:00:17.697
A time where you saw something right away that you liked, but it still took

00:00:17.697 --> 00:00:21.717
a long time to make a choice with all those options in front of you.

00:00:22.477 --> 00:00:27.497
Now, even though the reason might feel obvious, more choices equals more time,

00:00:27.817 --> 00:00:33.837
what if there's a more nuanced relationship at play, maybe a mathematical one,

00:00:34.277 --> 00:00:36.337
between how many options are

00:00:36.337 --> 00:00:41.117
sitting in front of you and how long it takes for you to make a choice?

00:00:41.317 --> 00:00:46.937
And how exactly does that calculation change when the options are presented differently?

00:00:46.937 --> 00:00:52.097
What are the other ways available to us where we can bring ease and clarity

00:00:52.097 --> 00:00:57.957
into the process of making a choice, even when we're faced with too many options?

00:00:58.437 --> 00:01:04.177
This episode is about Hick's Law, a key principle of human decision-making.

00:01:04.297 --> 00:01:10.257
We'll explore how it works. We'll even pit a robot's logic against a human's

00:01:10.257 --> 00:01:13.157
logic and see which one wins on efficiency.

00:01:13.157 --> 00:01:19.777
And then we'll dive into how we as design psychologists can create smoother

00:01:19.777 --> 00:01:25.017
user experiences even when the complexity of choices is unavoidable.

00:01:25.597 --> 00:01:31.737
Okay, to start off this episode, imagine that you're participating in an experiment.

00:01:32.037 --> 00:01:37.257
You're standing in a room in front of a table and on the table there are four

00:01:37.257 --> 00:01:39.577
options of fruit that you can eat.

00:01:39.757 --> 00:01:43.677
There's an apple, an orange, a peach and a plum.

00:01:44.137 --> 00:01:48.957
You quickly pick one and the experimenter clicks their stopwatch and the amount

00:01:48.957 --> 00:01:51.017
of time it took you has been recorded.

00:01:51.497 --> 00:01:55.457
Now imagine that instead you have eight options in front of you.

00:01:55.637 --> 00:02:00.857
There are the first four fruits plus berries, bananas, kiwis, and mangoes.

00:02:00.997 --> 00:02:05.197
The decision won't be as quick and if you're considering them all,

00:02:05.397 --> 00:02:08.617
it will take longer by a certain amount of time.

00:02:09.419 --> 00:02:15.399
But why is that? What is it about simply having more choices that slows us down?

00:02:15.659 --> 00:02:21.899
How does the mere number of options contribute to decision-making time?

00:02:22.199 --> 00:02:25.919
Let's break this down with a simple graph that I'll ask you to imagine.

00:02:26.419 --> 00:02:28.279
Picture a graph with two axes.

00:02:28.859 --> 00:02:33.459
On the x-axis, running from left to right, we have the number of choices,

00:02:33.759 --> 00:02:39.099
starting from zero on the left and increasing in the number of choices as you move further right.

00:02:39.299 --> 00:02:44.219
On the y-axis, running from bottom to top, we have decision time,

00:02:44.539 --> 00:02:50.019
faster at the bottom, small amount of time, and slower as you move to the top, more time.

00:02:50.419 --> 00:02:55.739
Now to really wrap our minds around this concept, let's look at three different

00:02:55.739 --> 00:03:01.799
potential relationships we might have between the number of options and decision time.

00:03:02.059 --> 00:03:07.419
Let's look at these potential relationships as three different slopes on our

00:03:07.419 --> 00:03:13.799
graph that show how decision-making time could be related to the number of options.

00:03:14.179 --> 00:03:19.639
We'll start with a simple one, a linear slope, a straight slanted line that

00:03:19.639 --> 00:03:22.879
climbs from the bottom left to the top right.

00:03:23.079 --> 00:03:29.579
This represents a system where each additional option adds one additional unit

00:03:29.579 --> 00:03:34.499
of decision time. To be clear, the straight line is not what Hicks law gives

00:03:34.499 --> 00:03:36.499
us. We'll get to Hicks law a little later.

00:03:37.272 --> 00:03:43.452
So let's apply this function to our four fruits, the apple, orange, peach, and plum.

00:03:43.692 --> 00:03:48.732
We're going to compare algorithms here. So instead of us selecting the fruit

00:03:48.732 --> 00:03:51.232
this time, we're going to use a small robot.

00:03:51.652 --> 00:03:57.532
It's been placed on the table in front of you, and it's been programmed to pick for us.

00:03:58.032 --> 00:04:03.412
So to test our robot, we're going to try to get it to pick the plum.

00:04:03.412 --> 00:04:08.512
We're going to tell the robot, pick the fruit that is smooth and purple.

00:04:09.032 --> 00:04:14.512
So first it rolls up to the apple and asks itself, is this fruit smooth and

00:04:14.512 --> 00:04:16.812
purple? For the apple, that's a no.

00:04:17.172 --> 00:04:21.792
So then the robot scoots back and rolls over to the orange and asks itself,

00:04:22.012 --> 00:04:26.032
is this fruit smooth and purple? And for the orange, that's another no.

00:04:26.632 --> 00:04:30.352
Then it tries the peach, and the answer is no again. And finally,

00:04:30.352 --> 00:04:35.252
it tries the plum, and that's a yes. the robot has found the plum.

00:04:35.732 --> 00:04:40.552
In this example, the robot's approach is the straight line on our graph.

00:04:40.772 --> 00:04:43.752
Each choice adds a constant amount of time.

00:04:44.032 --> 00:04:51.072
So, let's say for easy math that it took one second for the robot to evaluate each fruit.

00:04:51.392 --> 00:04:54.572
For the four fruits together, then it took four seconds.

00:04:54.912 --> 00:04:58.592
But now let's add four more fruits, a strawberry, a blueberry,

00:04:58.952 --> 00:05:03.472
a blackberry, and a raspberry. If the plum is at the end, forcing the robot

00:05:03.472 --> 00:05:06.712
to go through all the options, it's going to take 8 seconds.

00:05:07.072 --> 00:05:12.172
So there's our straight line on the graph. Each fruit adds a constant amount of time.

00:05:13.209 --> 00:05:16.069
So now we're going to do something cool. We're going to pick up the robot,

00:05:16.349 --> 00:05:20.909
press a few buttons, and now the robot is using a different algorithm.

00:05:21.269 --> 00:05:25.389
This new algorithm is for maximizing thoroughness.

00:05:25.669 --> 00:05:31.409
This time, it's going to ask more questions to make absolutely sure of its choice.

00:05:31.929 --> 00:05:35.629
So it starts with the apple. Is it purple? No.

00:05:36.129 --> 00:05:39.949
Is it red? Yes, but it doesn't stop there. It keeps ongoing.

00:05:40.249 --> 00:05:44.909
Is this fruit fuzzy, checking for a peach? Is this fruit lumpy,

00:05:45.069 --> 00:05:46.549
checking for a berry, and so on?

00:05:46.889 --> 00:05:51.969
With each new fruit, the robot adds more questions. In this example,

00:05:52.349 --> 00:05:55.789
each additional fruit adds more factors to the algorithm.

00:05:56.089 --> 00:06:01.949
Since this robot has to consider more factors, decision making time rapidly

00:06:01.949 --> 00:06:04.329
increases with this thorough algorithm.

00:06:04.669 --> 00:06:10.569
When plotted on our graph now, this slope is not a simple straight line. It curves upward.

00:06:10.829 --> 00:06:14.289
With each additional option, the slope gets steeper and steeper.

00:06:14.369 --> 00:06:20.829
And with this algorithm, the act of adding choices sharply punishes decision-making time.

00:06:21.089 --> 00:06:23.289
So what about the human brain?

00:06:23.569 --> 00:06:26.789
What algorithm does the human brain run on?

00:06:27.349 --> 00:06:33.189
Hicks law warns us that more options means longer, slower decision-making times,

00:06:33.209 --> 00:06:38.649
which can imply that humans cannot handle lots of choices. But is that true?

00:06:38.949 --> 00:06:44.929
Are we more or less efficient than the robot, either in the first or second example?

00:06:45.109 --> 00:06:51.689
And what exactly is the relationship between the number of options and our decision-making time?

00:06:52.109 --> 00:06:57.029
So to think about this more holistically, I want to add one more dimension to

00:06:57.029 --> 00:07:00.269
this discussion, and that's information theory.

00:07:00.469 --> 00:07:05.069
So we have bits, which are units to measure information.

00:07:05.869 --> 00:07:10.229
So just as we have grams to measure weight or centimeters to measure length,

00:07:10.389 --> 00:07:13.549
we have bits to measure information.

00:07:14.416 --> 00:07:20.696
So in the examples we're using with our robot, each question that is asked is a bit.

00:07:20.896 --> 00:07:24.716
It's a no or yes, zero or one outcome.

00:07:25.236 --> 00:07:30.256
So when it asked the first four fruits questions, that was four bits.

00:07:30.536 --> 00:07:37.016
And when we added the four more fruits, the berries, that was eight bits in total.

00:07:37.456 --> 00:07:42.776
But when we had the upward curving curve, that was much worse.

00:07:42.776 --> 00:07:48.556
The system got punished, the algorithm got worse, and it was asking more questions

00:07:48.556 --> 00:07:51.416
every time more fruits were added to the set.

00:07:51.736 --> 00:07:54.056
So what is the human algorithm?

00:07:54.416 --> 00:08:00.496
Looking through the lens of information theory, how many bits does it take for

00:08:00.496 --> 00:08:02.256
the human to solve this problem?

00:08:02.456 --> 00:08:05.616
So just as a reminder of an important caveat,

00:08:05.896 --> 00:08:11.376
for simplicity, we are ignoring the actual complexity of this decision-making

00:08:11.376 --> 00:08:17.956
process, and we're only looking at what does the number of options add to decision-making time.

00:08:18.316 --> 00:08:22.836
So we already know that according to Hicks law, the decision-making process

00:08:22.836 --> 00:08:30.916
is punished in terms of time by adding extra choices, but by how much and in what way.

00:08:31.516 --> 00:08:37.376
Okay, so to test this out, we've downloaded into our robot a human-like algorithm.

00:08:37.616 --> 00:08:40.016
Okay, so let's watch it and see what it does.

00:08:40.416 --> 00:08:46.316
The robot rolls up and it takes one look at the fruits and it asks, do I want a berry?

00:08:47.056 --> 00:08:51.976
No. And it cuts the set in half. And then now with the remaining set,

00:08:52.116 --> 00:08:57.196
it asks, do I want to have an ordinary fruit that I see all the time,

00:08:57.396 --> 00:09:00.236
like an apple or an orange? And the answer is no.

00:09:00.516 --> 00:09:05.036
So now the set has been cut in half again. And now it looks at the final two

00:09:05.036 --> 00:09:07.376
and it says, well, I want the juicier fruit.

00:09:07.496 --> 00:09:10.196
So that rules out the peach and it grabs the plum.

00:09:10.856 --> 00:09:16.796
So that was three decisions and it only used three bits for this set of eight options.

00:09:17.616 --> 00:09:22.976
So to be clear, the way the robots used its bits is that when starting with

00:09:22.976 --> 00:09:26.196
eight options, it chopped it immediately in half with one decision.

00:09:26.696 --> 00:09:31.576
And with four options left, it chopped it in half again. And with only two options

00:09:31.576 --> 00:09:33.916
left, it chose a final option.

00:09:34.959 --> 00:09:39.819
So when we plot this on our original graph, we get a logarithmic curve.

00:09:39.819 --> 00:09:46.439
It starts out angled up, but then it flattens out underneath the linear straight line.

00:09:47.079 --> 00:09:50.559
It gets less and less steep as you add more options.

00:09:51.499 --> 00:09:55.539
So in other words, as the penalty increases because you're adding more options,

00:09:55.779 --> 00:09:59.419
it does so by less and less as you're adding those options.

00:09:59.939 --> 00:10:03.739
So it's efficient, unlike what it could have been.

00:10:03.739 --> 00:10:10.759
It could have been that with our very intelligent human minds that we're considering

00:10:10.759 --> 00:10:15.959
so many options and that every time you add options, it goes up and up at a steep slope.

00:10:16.179 --> 00:10:20.199
Or it could have been that it's constant and every time you add an option,

00:10:20.399 --> 00:10:25.179
we're forced to consider that with an equal weight as all of the other options.

00:10:25.759 --> 00:10:30.839
But instead, we are efficient with how we deal with increased options.

00:10:30.839 --> 00:10:35.419
That's the mathematical relationship illustrated by Hicks law.

00:10:35.779 --> 00:10:38.659
So that's interesting, but the

00:10:38.659 --> 00:10:44.339
applied and practical reality is still that more options adds more time.

00:10:44.539 --> 00:10:48.379
But I think it's important that we take some time to understand what Hicks law

00:10:48.379 --> 00:10:52.839
is actually saying and what it implies by our decision-making process.

00:10:53.579 --> 00:11:00.019
So the easy answer for this problem is to remove options whenever you can.

00:11:00.839 --> 00:11:05.079
But what if there's just a lot of options? And what if you can't just remove options?

00:11:05.239 --> 00:11:10.119
You might be in a design scenario where that's not a thing that you can do.

00:11:10.779 --> 00:11:15.859
Plus, removing options should not be the only tool in our toolbox as design psychologists.

00:11:16.159 --> 00:11:21.999
We want to achieve simplicity through good design and not just merely relying

00:11:21.999 --> 00:11:24.359
on the strategy of minimalism.

00:11:24.759 --> 00:11:28.839
In addition to that, sometimes removing options creates other problems,

00:11:28.859 --> 00:11:34.299
like maybe there's really important options that become invisible or unavailable.

00:11:34.839 --> 00:11:38.839
Or maybe the user can get to an option, but it's unclear how to.

00:11:39.479 --> 00:11:43.399
So solutions may vary because design is not a silver bullet,

00:11:43.419 --> 00:11:47.679
but we're going to walk through some specific examples to see what can we do

00:11:47.679 --> 00:11:51.199
when we're faced with a lot of options in an interface.

00:11:51.199 --> 00:11:58.039
Okay, so for these examples, we want to design things in such a way that we

00:11:58.039 --> 00:12:02.879
are somehow reducing the burden of lots of options.

00:12:03.819 --> 00:12:07.479
And so for this discussion, we'll use the example of a restaurant menu.

00:12:07.939 --> 00:12:12.219
Okay, so our first strategy will be chunking and grouping.

00:12:12.219 --> 00:12:18.539
So, on our restaurant menu, instead of just printing out a long list of dishes,

00:12:18.979 --> 00:12:23.119
we're going to group the dishes into categories like appetizers,

00:12:23.459 --> 00:12:26.379
rice dishes, noodle dishes, desserts, and drinks.

00:12:26.659 --> 00:12:31.479
So, this creates a different decision landscape where the user doesn't have

00:12:31.479 --> 00:12:33.659
to face everything at once.

00:12:34.591 --> 00:12:40.271
An earlier decision at the category level can help with deciding what to pay

00:12:40.271 --> 00:12:41.331
attention to and whatnot.

00:12:41.871 --> 00:12:45.331
Or how about the strategy of progressive disclosure?

00:12:45.871 --> 00:12:51.451
This time you're using an interactive digital menu, and you only see the next

00:12:51.451 --> 00:12:53.811
option after making a previous choice.

00:12:54.011 --> 00:13:00.531
So you choose a burger, and because you chose a burger, it asks you,

00:13:00.631 --> 00:13:02.071
what type of patty do you want?

00:13:02.191 --> 00:13:08.571
And then what type of bun do you want? and these options become available only when they're relevant.

00:13:09.531 --> 00:13:13.371
In this system, you'll probably never even be exposed to all the options.

00:13:13.951 --> 00:13:19.091
Another strategy would be pre-filling the interface with good defaults.

00:13:19.191 --> 00:13:23.011
So let's say we've got our interactive menu at the restaurant,

00:13:23.011 --> 00:13:27.831
and the dishes have popular choices that most people choose.

00:13:27.831 --> 00:13:33.451
So when you choose the most popular dish, it might be pre-filled with the most

00:13:33.451 --> 00:13:38.111
popular choices. So let's say you picked pad thai, and then it pre-fills with

00:13:38.111 --> 00:13:40.151
medium spicy and with chicken.

00:13:40.351 --> 00:13:45.031
Or better yet, it fills with your own choices from a previous time.

00:13:45.631 --> 00:13:48.751
Since that's the best guess of how you're going to choose a dish now,

00:13:48.911 --> 00:13:50.131
it might be a good default.

00:13:51.191 --> 00:13:56.511
So pre-filling options reduces the number of choices you have to make.

00:13:57.031 --> 00:14:01.691
And when a person sees each choice, they simply verify that the choice is correct.

00:14:02.411 --> 00:14:07.131
And once the person realizes that the system is filling in with their previous

00:14:07.131 --> 00:14:11.891
options, they might even trust the system enough to not even look at the rest of the options.

00:14:12.631 --> 00:14:16.811
Another strategy would be providing recommendations. What if you're in a new

00:14:16.811 --> 00:14:22.131
restaurant and you're not feeling particularly decisive, and all you know is

00:14:22.131 --> 00:14:25.691
that you're hungry and you see the chef's recommendation at the top,

00:14:25.811 --> 00:14:29.511
and the chef has a great reputation here, so you trust their judgment,

00:14:29.751 --> 00:14:30.751
and you just go with that.

00:14:30.751 --> 00:14:36.451
This is another example where you are shielded from all of the options that

00:14:36.451 --> 00:14:39.091
you could go through and you can bypass them altogether.

00:14:40.244 --> 00:14:46.324
Yet another strategy is choosing good visual hierarchy and creating a visual

00:14:46.324 --> 00:14:50.244
pattern that helps guide the user through looking at the options.

00:14:50.684 --> 00:14:57.824
So imagine that the restaurant menu is laid out in a visually well-organized

00:14:57.824 --> 00:15:00.124
manner, and the prices are on the right,

00:15:00.324 --> 00:15:03.324
and the customer rating is in the same place for every dish,

00:15:03.464 --> 00:15:08.044
and that there are icons that help point out different features of the dish,

00:15:08.084 --> 00:15:13.784
and you find that this helps you focus your attention on the aspects of the

00:15:13.784 --> 00:15:15.504
choice that you actually care about,

00:15:15.944 --> 00:15:17.744
reducing the cognitive load.

00:15:18.144 --> 00:15:23.564
So this is similar to chunking and grouping, except that the layout allows you

00:15:23.564 --> 00:15:28.484
to direct your attention to where it needs to be directed, which is features

00:15:28.484 --> 00:15:32.004
of each dish that is relevant to your decision.

00:15:32.004 --> 00:15:36.304
The last example that I'll give here is a guided flow.

00:15:36.704 --> 00:15:40.944
So imagine that you're being guided through a set of steps.

00:15:41.124 --> 00:15:45.604
And so you're still making choices, but they're split out into steps.

00:15:45.904 --> 00:15:50.564
So let's say you're at a barbecue spot and the menu starts with asking,

00:15:50.784 --> 00:15:53.764
which regional flavor profile do you want?

00:15:54.584 --> 00:15:58.304
Carolina, Memphis, Kansas City, Texas.

00:15:58.724 --> 00:16:03.844
And then you choose your type of protein, and then you choose sauce versus dry

00:16:03.844 --> 00:16:05.484
rub, and then you choose sides.

00:16:06.044 --> 00:16:12.124
So the entire complexity of the task is now rendered down into easy steps that

00:16:12.124 --> 00:16:14.284
on their own are easier to think about.

00:16:15.038 --> 00:16:21.018
So let's recap what we've covered. Hicks law tells us that decision time increases

00:16:21.018 --> 00:16:23.278
as the number of options grows.

00:16:23.678 --> 00:16:29.298
More choices equals more time and by extension effort for a decision.

00:16:29.618 --> 00:16:35.318
So Hicks law reminds us that the number of choices punishes decision making time.

00:16:35.558 --> 00:16:40.478
This implies that too many options and too much complexity hurts us.

00:16:40.698 --> 00:16:45.898
This is true and that's why we want to reduce complexity whenever possible.

00:16:46.418 --> 00:16:49.718
But Hicks law also reveals something unexpected.

00:16:50.218 --> 00:16:53.498
Humans are surprisingly efficient decision makers.

00:16:53.838 --> 00:16:59.518
We don't evaluate every option equally. Instead, we narrow the choices quickly,

00:16:59.818 --> 00:17:04.238
cutting through the complexity faster than a simple algorithm would.

00:17:04.638 --> 00:17:09.518
Even so, too many choices slows us down. So what can we do about it?

00:17:09.938 --> 00:17:15.298
Yes, removing options helps, but good design is more than just removing choices.

00:17:15.298 --> 00:17:19.298
It's about making those choices easier to navigate.

00:17:19.658 --> 00:17:24.878
So when we can't reduce options, we can reduce the effort it takes to decide.

00:17:25.538 --> 00:17:30.858
Here's a quick summary of the strategies designers can use to reduce the time

00:17:30.858 --> 00:17:32.678
it takes people to make decisions.

00:17:33.338 --> 00:17:37.878
One strategy is chunking, breaking choices down into categories.

00:17:38.138 --> 00:17:44.278
An example of this is grouping dishes in a menu. This helps the user focus and decide faster.

00:17:44.538 --> 00:17:51.298
Another strategy is progressive disclosure, showing only what's needed step-by-step

00:17:51.298 --> 00:17:53.458
prevents decision overload.

00:17:53.758 --> 00:18:00.078
It's also useful to have good defaults. In other words, you can pre-fill forms

00:18:00.078 --> 00:18:05.978
with common choices or the user's own previous choices. Another strategy to

00:18:05.978 --> 00:18:07.638
use is recommendations.

00:18:08.498 --> 00:18:14.858
Highlight popular or expert picks. You can also leverage visual hierarchy,

00:18:15.318 --> 00:18:19.478
organize options with clear patterns and guide users' attention,

00:18:19.778 --> 00:18:21.598
and simplify comparisons.

00:18:21.598 --> 00:18:26.538
The last strategy we'll mention here is to use guided flows,

00:18:26.818 --> 00:18:29.898
break decisions down into step-by-step choices.

00:18:30.498 --> 00:18:34.678
In the end, Hicks Law isn't just a caution, it's an insight.

00:18:34.998 --> 00:18:40.718
Decision-making isn't about avoiding complexity, but rather moving through it with clarity.

00:18:41.198 --> 00:18:46.898
The best designs don't overwhelm, they don't oversimplify, they guide and they empower.

00:18:46.898 --> 00:18:52.558
As design psychologists, we hold the ability and even the responsibility to

00:18:52.558 --> 00:18:58.218
shape experiences that honor how people think, not just how systems work.

00:18:58.478 --> 00:19:03.018
Every choice we design is an opportunity to reduce friction,

00:19:03.338 --> 00:19:08.698
build confidence, and create moments of clarity in an otherwise noisy world.

00:19:08.918 --> 00:19:14.198
Because when we get it right, we don't just speed up decisions, we make them better.