---
title: 'The Anti Trampoline Effect'
source: 'https://youtube.com/watch?v=EP1mYq8hLIY'
video_id: 'EP1mYq8hLIY'
date: 2026-06-28
duration_sec: 1004
---

# The Anti Trampoline Effect

> Source: [The Anti Trampoline Effect](https://youtube.com/watch?v=EP1mYq8hLIY)

## Summary

The video explores a counterintuitive phenomenon where combining two highly bouncy objects results in a poor bounce. Through experiments and simulations, the creator investigates the physics of bouncing, focusing on the interplay between the stiffness and mass of a ball and a surface.

### Key Points

- **Bouncy Paradox** [0:00] — A very bouncy ball on a very bouncy surface can result in a dead stop, contrary to intuition.
- **Problem with Similar Flexibility** [1:30] — A poor bounce occurs when the ball and surface have similar flexibility. A good bounce requires a mismatch in flexibility (e.g., hard ball/flexible surface or vice versa).
- **Simplifying Model** [2:26] — A research paper models the bounce using four key factors: mass and stiffness of the ball, and mass and stiffness of the club/surface.
- **Optimal Bounce Conditions** [3:03] — The quality of the bounce oscillates in a counterintuitive pattern as mass and stiffness are varied, showing multiple points of optimal and terrible bounces.
- **Perfect Bounce Explanation** [5:07] — A perfect bounce occurs when the club's spring returns to its rest length and the club mass is at rest at the exact moment of ball separation, transferring all energy to the ball.
- **Simple Rule for Good/Bad Bounces** [7:07] — A good bounce requires the club to go through a whole number of oscillations plus a half. A bad bounce results from a whole number of oscillations during the ball's half-oscillation.
- **Fractal Structure** [10:14] — Heat maps of bounce quality show complex, fractal-like patterns when multiple parameters are varied.
- **Spiky Sensitivity** [12:17] — The quality of the bounce is very sensitive (spiky) to small changes in parameters, which matches the real-world data collected by the creator.
- **Real-World Implication** [13:06] — For combinations like a squishy ball on concrete or a stiff ball on a squishy surface, the period-matching effect is negligible, and dissipation is the primary factor.
- **Analogy to Razor Blades** [14:26] — The problem of bouncy combinations is compared to cartridge razors, where adding complexity (more blades) stems from an inability to fix the fundamental issue (holding a single blade firmly).

### Conclusion

To maximize bounce, you cannot simply combine two high-bounce materials; you must carefully tune their vibrational frequencies. This principle is also a metaphor for solving problems at their root cause rather than piling on complexity.

## Transcript

What happens when you combine a really
bouncy ball with a really bouncy
surface?
It stops dead. But look, if I tweak it
slightly,
now it's a really good bounce. This is
when I realized I don't really
understand bouncing. Because it turns
out if you want to maximize bounce, it's
not enough to just combine two really
bouncy things. But what I found really
surprising was that when I tried to find
the optimum bounce using math, I ended
up generating fractals. First, I want to
be clear about what we're doing in this
video. Like I've shown you really bouncy
things before, like the atomic
trampoline and well, maybe you've seen
an advert for the world's bounciest ball
or whatever, but that type of
optimization isn't what we're dealing
with here. Like those examples are about
avoiding energy loss during the bounce.
So, the atomic trampoline minimizes
plastic deformation, the super bouncy
balls minimize internal friction during
the collision. But the thing is, this
bouncy ball and this bouncy surface are
both really good at retaining energy.
So,
what went wrong here? Well, look at this
ball bearing on this rubber sheet. The
ball is really hard and the surface is
really flexible and that gives you a
really good bounce. And with the super
bouncy ball, the ground is really hard,
but the ball is really flexible and that
gives you a good bounce as well. The
problem seems to arise when you have a
ball and a surface that are similarly
flexible. For example, Orbeez are really
flexible and so is this rubber sheet.
And right now, the combination is super
bouncy. But if I add a little bit of
weight to the rubber sheet, well,
suddenly it all goes wrong. And that's
why I built this contraption to try and
figure out what's going on. See, it's
when I change the weight that it goes
from a good bounce to a bad bounce. It's
weird, isn't it? Actually, this
particular configuration is quite fun
because when gravity pulls the ball back
down, it gets a kind of second kick,
which is such an odd behavior. By the
way, this is somewhat related to how
children typically injure themselves on
a trampoline. A child can often get an
unexpected second kick if there are
other people on the trampoline. Before I
show you what I figured out with this
device, I also had a good chat with the
professor of the author of this paper.
In the paper, they're thinking about it
in terms of a golf ball and a golf club.
And basically, they do a whole lot of
simulations where they simplify it down
to four things: the mass of the ball,
the stiffness of the ball, the mass of
the golf club, and the stiffness of the
golf club. And when I say stiffness, I
mean in the spring sense. So, like, if a
spring is hard to stretch, then it's a
stiff spring. If it's easy to stretch,
then it's a less stiff spring. And a
golf ball and a golf club are in some
sense springy. So, we talk about
stiffness. Anyway, when they run a whole
lot of simulations where they kept three
things constant and varied the fourth
thing, they discovered something really
surprising. See, in this graph, they're
varying the mass of the golf club while
keeping everything else fixed. And the
height of the curve is how good the
bounce was. Specifically, it's the
velocity of the ball when it leaves the
club. And you can see how the quality of
the bounce well, it goes up and down as
the mass varies. So, you get an optimal
bounce for these masses. But, if you
change the mass of the ball to one of
these masses, well, you'd actually get a
terrible bounce. To see what's going on,
I built this simulation that represents
the model that they were using in the
paper, and it means we can see actually
what's physically going on with
different examples. So, this is the mass
of the ball. This spring represents the
stiffness of the ball. This is the mass
of the club. And this spring represents
the stiffness of the club. Let's see
what happens if we set the mass of the
club to be about twice the mass of the
ball. So, the ball engages with the
club, it starts to squish, but also the
club is being pushed now, so that starts
to squish as well. Crucially, look, the
ball disengages from the club when the
club is at almost maximum squish, but
keep watching because but maybe the club
will catch up with the ball and it'll
give it a second kick.
No, it just missed.
Crucially though, look, see how much
vibrational energy there is in the club.
That vibration is energy locked away in
the club that could have been given to
the ball but wasn't. So, this represents
a bad bounce. Let's compare that to when
the mass of the club is about 2/3 of the
mass of the ball. So, again, the ball
starts to squish, and then the club
starts to squish as well. Right, the
club is now at maximum squish. The ball
didn't disengage this time because the
spring is still compressed. So, as the
club spring is re-expanding, it's
actually working hard against the mass
of the ball because the spring is still
compressed, and that's slowing it down
at just the right rate so that when the
ball finally disengages, look, the club
mass is at rest at its rest length. That
means there's no vibrational energy left
in the club. So, all of the energy went
into the kinetic energy of the ball.
This is a perfect bounce. But if you
look at the graph in the paper, you can
see that actually there are lots of
points of maximum bounce. So, let's also
look at when the club mass is about 1/6
of the mass of the ball. So, really it's
the same situation except if you watch
the club mass, actually it goes through
multiple oscillations before that
perfect separation that we saw before.
And it turns out that when you do the
math, you get quite a simple result that
actually, hopefully, makes some
intuitive sense. So, it's all about the
natural oscillation of the club. See,
the club naturally oscillates like this
over time. But because it's being pushed
by the ball throughout the whole
collision, this graph of position gets
skewed like this. Now, what I'm showing
you is a perfect bounce. Because, look,
the club is back at its rest position at
the end of the collision. That tells us
that there's no energy stored in the
spring at the end of the collision. But
also notice that the graph of position
is flat at the end. So, we know there's
no kinetic energy in the club mass,
either. Now, you'll notice that the club
mass has gone through two and a half
oscillations. But let's stiffen the
spring so it oscillates faster.
It's still a perfect bounce, but now
it's gone through three and a half
oscillations.
And there's a perfect bounce at four and
a half oscillations, and so on. And
look, if you go through a whole number
of oscillations, that's when you get a
perfectly bad bounce. How does that
compare to the natural oscillation of
the ball? Well, the ball always goes
through exactly half an oscillation
during the collision because
well, it's not sticky. So, here's the
simple rule. The club needs to go
through a whole number of oscillations
plus a half in the time it takes the
ball to go through half an oscillation
if you want to get a good bounce. If you
want a bad bounce, just go for a whole
number of oscillations. In general,
though, the surface needs to jiggle
faster than the ball for interesting
things to happen. But in reality, it's
hard to find a combination of ball and
surface where the surface vibrates
faster than the ball, but not loads
faster. In my setup, the ball always
vibrates faster than the surface.
Doesn't matter what I tried in terms of
springs and everything, but clearly
interesting things still happen. I think
that's because we made some simplifying
assumptions, but actually there is
another issue. See, in the simulation
some energy can be left behind in the
oscillating surface as we've seen, but
you can see in this footage that you can
also end up with oscillation energy left
in the ball. But actually because of the
way this simulation is put together,
there's no way to account for
oscillation energy left behind in the
ball because the spring doesn't have any
mass, so it can only oscillate when it's
in contact with another mass. In other
words, during the collision. So I
decided to make my own simulation. In
this version, the mass of the ball is
split in two. So the ball can now
oscillate on its own. Honestly, I'm not
sure whether it's a reasonable model of
reality, but I just want to show you
what happens when you explore it because
some of it's really interesting. So
yeah, I can vary the stiffness of the
springs and the masses or I can sweep
through one of those values to get a
plot. So look,
you can see here this is a lot like the
graph from the paper. So look, I can
select a maximum and then play that.
Oh. [sighs]
That's satisfying, isn't it? Let's look
at a bad one. Actually, look at the
worst one over here.
Oh, that's awful.
In the plot, you can also see the number
of collisions. So it's interesting,
isn't it? The collisions go up and up.
This is the internal vibrational energy
of the ball after the collision. It
doesn't actually get that high. So what
we can do is sweep two things at once if
we want to try and say optimize the
amount of jiggle in the ball after the
collision. We'll vary the mass of the
golf club and the spring constant of the
golf club. So, here's a heat map showing
how good the bounce is. If I go here,
this is a good bounce. See what that
looks like.
That's getting it. And then here's a
really bad bounce, let's say.
And if we switch to the internal energy,
there's a high point right here. So,
this is where we end up with lots of
internal energy in the ball at the end.
Look at that.
That is a terrible bounce. Most of the
energy is in the ball. But
interestingly, there's some structure
down here. Let's do a different sweep
with the mass of the club and the
stiffness of the club.
Look at that. Almost like a fractal.
Maybe it is a fractal. Like if I zoom in
here,
isn't that weird?
But anyway, an unexpected fractal
appears. Something else I discovered
that matches real life. Like if there
are several collisions within a bounce,
are you only looking at what happens
after the first collision, or do you
count all the collisions? This is a
perfect example.
It stops dead. It's a terrible bounce.
But then on the second kick, it's an
amazing bounce. And I got something like
this in my studio. Look at that. It's a
terrible bounce. The ball stops dead,
but moments later it gets a second kick.
So, how does this compare to reality? To
compare to the model, I don't need to
worry about weighing things or
calculating the stiffness, because I can
just measure the period of oscillation
from the video. Okay, so we're
interested in how many times does the
surface oscillate during half an
oscillation of the ball. The model from
the paper predicts good bounces at 1 and
1/2, 2 and 1/2, 3 and 1/2, and so on.
But I was never able to get the surface
to oscillate fast enough to explore
that, though I did get quite close. In
our data, the surface only ever goes
through a fraction of an oscillation
during the collision. This is where the
model from the paper predicts the good
and bad bounces will be, and these are
our good and bad bounces marked in green
and red. We never get into the
interesting region because my surface is
too slow. But clearly the data doesn't
match what's going on at the boring end
either. But here's the cool thing. In my
version of the simulation, interesting
things can happen when the surface
oscillates slower than the ball. In
fact, it's basically a mirror image when
you look at it on a log scale so that
you get bad bounces when the surface
oscillates faster than the ball and good
and bad bounces when the ball oscillates
faster than the surface. In reality,
it's more complicated than that for
several reasons, and our data doesn't
really fit this model either. But what's
clear is the model is very spiky. Like,
you don't have to change very much for
the quality of the bounce to change
drastically. And our data is also very
spiky. So our data has roughly the right
character, but I'm not sure we can claim
much more than that. One cool thing
about both of these models is they
predict what should happen for a really
bouncy ball on concrete or a steel ball
bearing on a rubber sheet. See, this is
a plot of how good the bounce is as the
stiffness of the surface increases. And
the line still goes up and down in that
counterintuitive way, but as the
stiffness increases, actually that
effect gets less and less until it
eventually just flattens out. In other
words, if you've got a squishy ball and
a concrete floor, you don't need to
worry about comparing their periods of
oscillation. And then you're back to
just worrying about dissipation through
internal friction and things like that.
And the same is true if you've got a
stiff ball and a squishy surface. So if
you want to find the bounciest
combination of ball and surface, you
can't just combine the two bounciest
things you own. You have to tune their
vibrations as well. Otherwise, you could
end up with a dud. I messed up while
making this video and I want to share
the story with you as a cautionary tale.
So, I wanted this surface to vibrate
faster than the ball vibrates. So, I end
up buying like stiffer and stiffer
springs, but nothing was working. The
frequency of this thing barely shifted
at all. It got to the point where I had
to print this part in a stronger plastic
and I was using these nuts to tension
the springs, which just took ages. I
eventually realized the problem. The
stiffer springs were heavier and so
adding mass to the oscillator, which was
counteracting the additional stiffness
of the oscillator. The thing that
eventually worked was ungluing this part
so I could slide it up and down and then
clipping it in place at different spring
extensions. But, the surprising thing is
that this is an almost perfect analogy
for cartridge razors. I want to be clear
that's a segue, not a crowbar. The
difference being that the link is
actually really strong. It's a good
link. It's a strong link. I mean, you be
the judge. Okay, here's the thing. The
problem with plastic cartridges is
there's always going to be a little bit
of give in the blade, so you get an
inconsistent shave. To counteract that,
manufacturers just add more blades in
the hope of catching more hairs during
each pass. But, the first blade will cut
some hairs, so the subsequent blades
will be scraping against skin in certain
places leading to irritation. So, then
they have to make the blades
deliberately more springy, which makes
the shave worse. So, they add even more
blades. Gets to the point where there
are so many blades, you have to add a
lubricating strip to overcome the
friction and people over time learn to
press really hard and so you also have
to add a post-shave balm and things like
that. Complexity piled on top of
complexity just because the
manufacturers weren't able to implement
the simple fix at the very beginning of
just holding a single blade really
family because you can never do that
with plastic. The solution is to not buy
cartridge razors. Instead, buy a
precision engineered aluminum handle
that holds a single blade firmly in
exactly the right position at exactly
the right angle. And if you're going to
do that, may I humbly suggest the
sponsor of this video, Henson Shaving.
With a Henson AL13, the training wheels
are off. You don't press hard. You can
learn the shape of your face because you
can actually feel it.
And you end up with an actually
enjoyable shave. No lubricating strip,
no aftershave balm. The handle is more
expensive, but you only ever buy it
once. And the blades are pennies. So, it
actually ends up being cheaper after
just a few months. Go to
hensonshaving.com/steve
mold and use code steve mold at checkout
to get 100 free blades with your
purchase of a Henson AL13. That's
equivalent to about 3 or 4 years of
irritation-free shaving. The link is
also in the description, so check out
Henson Shaving today. I hope you enjoyed
this video. If you did, don't forget to
hit subscribe. And the algorithm thinks
you'll enjoy this video next.