Tag Archives: technology

Flappy Bird Physics Is Real Life?

If you don’t already know, Flappy Bird is the hot new mobile game right now. The premise is simple: navigate the bird through the gaps between the green pipes. Tapping the screen gives a slight upward impulse to the bird. Stop tapping and the bird plummets to the ground. Timing and reflexes are the key to Flappy Bird success.

This game is HARD. It took me at least 10 minutes before I even made it past the first pair of pipes. And it’s not just me who finds the game difficult. Other folks have taken to Twitter to complain about Flappy Bird. They say the game is so difficult, that the physics must be WRONG.




So, is the physics unrealistic in Flappy Bird?

Sounds like a job for Logger Pro video analysis! I used my phone to take a video of Flappy Bird on my iPad. To keep the phone steady, I placed it on top of a ring stand with the iPad underneath.


(I’ve uploaded several of the videos here if you’d like to use them yourself or with students: Flappy Bird Videos.)

Then I imported the videos into Logger Pro and did a typical video analysis by tracking Flappy’s vertical position in the video. Sure enough, the upside-down parabolic curves indicate Flappy is undergoing downward acceleration.


But do the numerical values represent normal Earth-like gravity or insanely hard Jupiter gravity? In order to do this, we need to (1) set a scale in the video so that Logger Pro knows how big each pixel is in real life and (2) determine the slope of Flappy’s velocity-time graph while in free fall, which is equal to the gravitational acceleration.

The only thing we could realistically assume is the size of Flappy Bird. If we assume he’s as long as a robin (24 cm), then the slope of the velocity-time graph is 9.75 m/s/s, which is really close to Earth’s gravitational acceleration of 9.8 m/s/s. Flappy Bird is REAL LIFE.


So then why is everyone complaining that the game is unrealistic when, in fact, it is very realistic? I blame Angry Birds and lots of other video games. Repeating the same video analysis on Angry Birds and assuming the red bird is the size of a robin (24 cm), we get a gravitational acceleration of 2.5 m/s/s, which only 25% of Earth’s gravitational pull.


In order to make Angry Birds more fun to play, the programmers had to make the physics less realistic. People have gotten  used to it, and when a game like Flappy Bird comes along with realistic physics, people exclaim that it must be wrong. As one of my students notes:


UPDATE 31 Jan 2014:
Inspired by a tweet from John Burk,

we made a video showing Flappy Bird falling at the same rate as a basketball:

Here’s what I did: We determined from the analysis above that Flappy Bird is about 24 cm across. Conveniently, basketballs are also about 24 cm across. So I had my physics teacher colleague Dan Longhurst drop a basketball so I could video it with my iPad. Dan just needed to be the right distance away from the camera so that the size of the basketball on the iPad screen was the same size as Flappy Bird on the screen (1.5 cm). Next, I played the basketball drop video and Flappy Bird on side-by-side iPads and recorded that with my phone’s camera. Once I got the timing right, I uploaded the video to YouTube, trimmed it, made a slow motion version in YouTube editor, then stitched the real-time and slow motion videos together to create the final video you see above.

UPDATE 1 Feb 2014: While the gravitational acceleration in Flappy Bird is realistic, the impulse provided by the taps are NOT realistic. Here’s a velocity-time graph showing many taps. When a tap happens, the velocity graph rises upward:


As you can see, no matter what the pre-tap velocity (the velocity right before the graph rises up), the post-tap velocity is always the same (a bit more than 2 m/s on this scale). This means that the impulses are not constant. In real life, the taps should produce equal impulses, which means that we would see that the differences between pre- and post-tap velocities are constant.

TL;DR: Is the physics in Flappy Bird realistic? Yes AND no.
YES: The gravitational pull is constant, producing a constant downward acceleration of 9.8 m/s/s (if we scale the bird to the size of a robin).
NO: The impulse provided by each tap is variable in order to produce the same post-tap velocity. In real life, the impulse from each tap would be constant and produce the same change in velocity.

UPDATE 1 Feb 2014 (2): Fellow physics teacher Jared Keester did his own independent analysis and shares his findings in this video:


Edtech PR Tips

I’m not a PR guy. I’m just a teacher. But they say that if you want to be a disruptor, the best experience is no experience. So here goes…

1. It’s not about the technology. It’s about what students are empowered to do because of your technology. Show us how you take students beyond what they could do previously. Show student work (“Hey, look what this kid can do!”). Stop focusing on checkmarks, badges, data, dashboards, and slick UI.

2. Learning is social. Show students interacting with each other, questioning, helping, constructing — all as a result of using your technology. Don’t show kids glued to screens, headphones on, working en masse and in isolation. It’s creepy.


The Learning Lab at a Rocketship school, where students spend 2 hours each day.

3. Don’t use phrases that signal you have simplistic views about teaching and learning. In particular: Learning stylesdigital nativesindividualized instruction, and content delivery.

4. Practices are equally as important as content. Show how you enable students to engage and grow in the core practices in math, science, and ELA.


Credit: Tina Cheuk, tcheuk@stanford.edu [PDF (scroll to bottom)]

5. Show how you implement/compliment research-based practices about how students learn. Study up on these characteristics of effective teaching methods. Otherwise…

6. Run controlled, peer-reviewed experiments that use conceptual diagnostic tests to measure growth. We know most anything works better than (or as well as) passive lecture instruction. But how does implementation of your technology stack up to other evidence-based teaching methods? And be sure to use conceptual diagnostic tests, not final exams or standardized tests or failure rates. CDTs have been painstakingly researched and designed to measure true conceptual understanding rather than algorithm memorization. Without strong evidence, we’re just skeptical of your claims.


Hake’s analysis of 62 different physics courses as determined by gain on a physics conceptual diagnostic test.

7. Don’t contradict yourself. Your words should match your actions.

8. Show feedback and testimonials from students. In particular, have students demonstrate their deeper understanding and expert thinking that has been enhanced by using your product. Or perhaps your technology has decreased student anxiety and contributed to a positive classroom climate. However, don’t have students talk about shallow things such as raising grades and doing well on tests.


Testimonials from Pearson/Knewton’s MyEconLab

9. There’s nothing revolutionary about old wine in new bottles. A digital textbook is still a textbook. A video lecture is still a lecture.

10. Read everything Audrey Watters writes. Everything.

Do you have any more edtech PR tips to share? Any more examples of bad PR? Any good examples? Thanks!

Video Analysis of a Bouncing Ball


Nothing earth-shattering here. I just wanted to share the activity we worked on today, which was an introduction to quantitative energy conservation by doing a video analysis of a bouncing ball. (Up until now, we were only doing qualitative energy pie charts.) Here are the handouts and the video:

The graphs from the analysis are just beautiful:

HeightTime VelocityTIme EnergyTime

Lots to talk about in those graphs!

Feel free to edit and reuse the handouts as you see fit. They’re not perfect, but I figure it’s better to share them than having them collect dust on my flash drive.

PS: I’ll sheepishly admit that I don’t do the whole suite of paradigm labs in the Modeling unit to mathematically derive the energy equations from experiments. But we do some simple qualitative demos/experiments to discover what variables would be in those energy equations. We start by talking about how the further a rubber band is stretched, the more energy it stores. Then we launch carts into a rubber band “bumper” (i.e., big rubber bands from Staples and two C-clamps) to qualitatively see the energy stored.

In doing so, we see that the cart’s kinetic energy depends on its speed and its mass. (Or is it weight? What would happen if we repeated the experiment on the moon?)

For gravitational energy, we can repeat the experiment, but have carts rolling down an incline. Or use the rubber band to launch the cart up the incline. I’ve also dropped balls into sand and looked at the depth to which they get buried. Either way, we see that gravitational energy depends on height and weight. (Or is it simply mass? What would happen on the moon?)

For elastic energy, we already know it depends on the distance the rubber band is stretched. Then, we can swap out the rubber band in the bumper with a stiffer/looser one to see the effects of the spring constant on energy stored.

Then, after we predict what the energy equations might look like, I just give them the actual energy equations, or have them look them up. (Gasp! See Schwartz’s A Time for Telling, aka Preparation for Future Learning.)

So, modelers, what am I missing by not doing the full-blown energy paradigm labs? How do you introduce the quantitative energy equations?

Going Beyond the Physics Textbook

I have the honor of being invited by Discovery Education to attend their second “Beyond the Textbook” forum to be held this Wednesday and Thursday at their headquarters in Silver Spring, Maryland. The event is spearheaded by Steve Dembo and, in exchange for travel expenses, he gets to pick my brain about digital textbooks, resources, and curriculum. There will be 18 other outstanding educators as well, including my edu-heroes  Christopher DanielsonMichael DoyleKarl Fisch, and Tom Woodward.

In preparation for the event, I’m updating/remixing an old blog post I wrote called “My Vision for a Physics iBook” ….


I keep thinking about what a physics iBook would look like. Not a book for consumption, as with a traditional text, but rather a book to enable exploration. So what would a student see when they first opened such a book?

It’s blank.

No content. No classical references like Feynman’s Lectures on Physics. No integration with Khan Academy’s video library.  Nothing.


Students should be learning to do science, not simply learning about science. They should be making observations, posing questions, conducting experiments, finding patterns, analyzing data, and sharing their conclusions.

In this sense, the iBook would function more like an electronic lab notebook. As with curricula like Modeling Instruction and ISLE, students would create the physics content from their own investigations and evidence, rather than deferring to authority.

Actually, the iBook wouldn’t be completely blank. While it would initially be empty of content, it would be chock-full of tools to help students collect and analyze experimental data. Software like Tracker for video analysis, VPython and GlowScript for computation and visualization, LoggerPro for graphing and electronic data collection, along with PhET simulations and Direct Measurement Physics Videos for conducting virtual experiments.

In the realm of traditional physics textbooks, only a few make it a priority to incorporate experiments into their storylines. Three that come to mind are:

The Manga Guide to Physics


Understanding Physics

FIGURE P-2  Electronic temperature sensors reveal that if equal amounts of hot and cold water mix the final temperature is the average of the initial temperatures.

and PSSC Physics.

Eugenia Etkina‘s upcoming College Physics text gets a step closer to my iBook vision. The text incorporates her work with video experiments in her ISLE and Physics Union Mathematics curriculula. In the text, there are QR codes which link to videos of the experiments to be analyzed.

For example, here’s a video of a momentum experiment, followed by the corresponding section of the text.


But, as you can see, the text does the analysis for the student. In my opinion, this would make a good reference only after the student has completed a similar activity on their own. Fortunately, her text also comes with a workbook that asks students to do this sort of scientific reasoning on their own:


Also taking the “experiments first” approach is Live Photo Physics Interactive Video Vignettes, a collaborative project by well-known physics education researchers Robert Teese, Priscilla Laws, and David Jackson. During a vignette, students are asked to make predictions and do video analysis on-the-fly. Here’s a preview:

Science is never done in isolation, however, so the iBook would come equipped with tools for sharing data, content, photos, videos, and resources among students and between teacher-student.

For me, going beyond the textbook means giving students a toolbox rather than an instruction manual.

What’s your vision for the future of textbooks?

You can follow along with us at the Beyond the Textbook forum this week by searching for the Twitter hashtag #BeyondTextbooks.

Bonus: 5 reasons why iPads won’t replace textbooks in science class.


Projectile Motion Assessment Task

You are a game designer for Rovio Entertainment, the company that makes Angry Birds.  The human resources department wants your input. They are hiring several programmers to build the physics engine for Rovio’s newest game. Here are the demo videos from the top four applicants. Which applicant(s) would you recommend for hire?

Applicant A

Applicant B

Applicant C

Applicant D

Download the original video files for analysis in Logger Pro or Tracker.

These videos were not created by me. I found them online several years ago, but I can’t remember where. If anyone knows, please tell me so I can give the creator proper credit. Thanks!

VPython Screencasts

This year I’ve decided to have my AP Physics C students (15) make screencasts explaining the workings of and reasonings behind their VPython programs. I got the idea from college physics professor Andy Runquist, who makes his students do similar screencasts for their Mathematica assignments. What I like about screencasting is that it gives added insight into which students understand the physics and the coding of their programs and which do not.

We’ll be using Screencast-o-matic because it is easy to use and it’s web-based (no software to download and install). Another reason is because Screencast-o-matic allows for “open submissions” — i.e., students can record and submit their screencasts directly to a designated channel without having to create an account or upload their video to YouTube. Which is great because all the screencasts will be in one place and I don’t have to worry about getting/managing links from students.

To help students with screencasting, I’ve made a tutorial video, along with examples of good and bad screencasts.

Screencast-o-matic Tutorial

Low Quality Screencast

High Quality Screencast

Happy Screencasting!

My TEDxNYED Session: Learning Science by Doing Science

Many thanks to the TEDxNYED 2012 crew, especially True Life Media, Basil Kolani, Karen Blumberg, and Matthew Moran for an awesome event. Be sure to check out the rest of the TEDxNYED 2012 talks.

Learn more about Modeling Instruction in Science.

Disrupt This: My Challenge to Silicon Valley

Over the past few months, Audrey WattersDan Meyer, and Keith Devlin have been critical of Silicon Valley, edtech startups, and iPad textbooks which hope to “disrupt” education. In my opinion, the real stumbling block to meaningful change is students’ formal reasoning skills — analytical thinking that cannot be cultivated by pausing and rewinding video or playing Math Blasters.

Here are my 5 points:

  1. Many of our students are transitioning from concrete to formal reasoning.
  2. A significant barrier to learning for understanding is students’ own formal reasoning skills.
  3. Formal reasoning skills (and thus learning for understanding) can be developing when instruction is structured around the Learning Cycle.
  4. Silicon Valley and edtech startups have been focusing on (often inappropriately) just a small fraction of the learning cycle.
  5. My Challenge to Silicon Valley: Help students learn for understanding by innovating around the rest of the learning cycle.

1. Many of our students are transitioning from concrete to formal reasoning.

Below are 3 reasoning puzzles, each followed by a video of college students attempting to solve the puzzle while explaining and discussing their logic. It’s a highly illuminating look at students’ reasoning processes.

I. The Algae Puzzle (Combinatorial Reasoning)

II. The Frog Puzzle (Proportional Reasoning)

III. The Mealworm Puzzle (Scientific Reasoning)

2. A significant barrier to learning for understanding is students’ own formal reasoning skills.

You’re probably thinking, “So, what? Just because Johnny can’t figure out all the possible combinations of algae doesn’t mean he can’t learn physics.” But the research strongly suggests that it does, even in interactive engagement classes.

In a previous post, I presented this graph from Hake’s famous six thousand student study:

As you can see, interactive engagement course outperformed traditional courses in learning gains as measured by the Force Concept Inventory (FCI). The FCI is the most widely used test of physics understanding. But why is there such a wide range of FCI gains among the IE courses and (not shown) among the individual students within a particular course? A study entitled “Why You Should Measure Your Students’ Reasoning Ability” (Coletta, Phillips, and Steiner) suggests reasoning ability is strongly correlated with physics success.

In the study, several different physics courses administered both the FCI (to measure physics gains) and the Lawson Test of Classroom Reasoning Skills (to measure formal reasoning ability). The Lawson test contains several items very similar the three puzzles above. Here’s what they found:

The data were split into quartiles based on the Lawson scores. The light green bars represent the average Lawson test score for each quartile and the dark green bars represent the average FCI gain for each quartile. There is clear correlation between reasoning ability and learning gains in physics. I’d wager this correlation extends to other subjects as well.

3. Formal reasoning skills (and thus learning for understanding) can be developed when instruction is structured around the Learning Cycle.

According to Piaget, intellectual growth happens through self-regulation — a process in which a person actively searches for relationships and patterns to resolve contradictions and to bring coherence to a new set of experiences.

In order to get students to experience self-regulation and further develop their reasoning skills, classroom experiences should be constructed around the Karplus learning cycle, which contains the the stages of EXPLORATION, INVENTION, and APPLICIATION. From Karplus’s workshop materials on the learning cycle:

EXPLORATION: The students learn through their own actions and reactions in a new situation. In this phase they explore new materials and new ideas with minimal guidance or expectation of specific accomplishments. The new experience should raise questions that they cannot answer with their accustomed patterns of reasoning. Having made an effort that was not completely successful, the students will be ready for self-regulation.

INVENTION: Starts with the invention of a new concept or principle that leads the students to apply new patterns of reasoning to their experiences. The concept can be invented in class discussion, based on the exploration activity and later re-emphasized by the teacher, the textbook, a film, or another medium. This step, which aids in self-regulation, should always follow EXPLORATION and relate to the EXPLORATION activities.  Students should be encouraged to develop as much of a new reasoning pattern as possible before it is explained to the class.

APPLICATION: The students apply the new concept and/or reasoning pattern to additional examples. The APPLICATION phase is necessary to extend the range of applicability of the new concept. APPLICATION provides additional time and experiences for self-regulation and stabilizing the new reasoning patterns. Without a number and variety of APPLICATIONs, the concept’s meaning will remain restricted to the examples used during its definition. Many students may fail to abstract it from its concrete examples or generalize it to other situations. In addition, APPLICATION activities aid students whose conceptual reorganization takes place more slowly than average, or who did not adequately relate the teacher’s original explanation to their experiences. Individual conferences with these students to help identify and resolve their difficulties are especially helpful.

4. Silicon Valley and edtech startups have been focusing on (often inappropriately) just a small fraction of the learning cycle.

Unfortunately, Silicon Valley has been dumping its disruptive dollars almost solely into the INVENTION phase and on the tail-end of the phase at that. It views education purely as a content consumption process and ignores the development of formal thinking and reasoning.

Remember, in the invention phase, “The concept can be invented in class discussion, based on the exploration activity and later re-emphasized by the teacher, the textbook, film, or another medium.” That’s Khan Academy videos, flipclass videos, iBooks, an similar technologies designed to present content via direct instruction. However, “Students should be encouraged to develop as much of a new reasoning pattern as possible before it is explained to the class.” Which means that this type of direct instruction should be as minimal as possible, because it robs kids from reasoning and making meaning. In other words, Silicon Valley is putting its energy into the portion of the invention phase that should be as small as possible!

Now let’s look at the application phase. There has been some development here as well, most notably in apps and exercise software which seek to gamify the classroom. But the application phase isn’t about getting 10 right answers in a row or solving problems to shoot aliens. Remember, Without a number and variety of APPLICATIONs, the concept’s meaning will remain restricted to the examples used during its definition. Real learning with understanding means students can reason about the concepts well enough to use them in new and unique concepts (aka transfer). Applications should require students to examine their own thinking, make comparisons, and raise questions. Great applications examples are open-ended problems, problems which present a paradox, and student reflection on both successful and unsuccessful problem-solving methods. Deep learning does not end when the Application phase begins.

5. My Challenge to Silicon Valley: Help students learn for understanding by innovating around the rest of the learning cycle.

Real disruption isn’t going to come from skill and drill apps, self-paced learning, badges, YouTube videos, socially-infused learning management systems, or electronic textbooks. Students must be continuously engaged in the learning cycle. We need to equip our students with the reasoning skills to learn how to learn anything. Focus on experiences in the exploration phase, meaningful sense making in the invention phase, and worthy problems in the application phase.

But, in reality, we only have ourselves to blame. It shouldn’t come as a surprise to us when students can’t think — the status-quo in education has been to spend most of our time on content delivery while robbing students of exploring and reasoning opportunities. And current edtech trends aren’t fixing this problem; rather, they are making it easier to make the problem worse.

To be fair, a few “good disrutptions” have occurred in the other phases of the learning cycle. Motion detectors allow students to “walk a graph” so they can easily explore position-time and velocity-time graphs. GeoGebra allows students to explore and play with geometry and functions quickly and easily. PhET simulations allows students to conduct open-ended planetary orbit experiments that would be impossible in real life. And VPython programming gets students to apply what they learned to write their own simulations and visualizations.

So when presented with the next great edtech “disruption,” ask yourself: has this innovation actually changed how student think about math and science concepts? Or has it just allowed students to get a few more questions correct on the state exam?

For further reading:

The next two articles:

  • “Promoting Intellectual Development Through Science Teaching” (Renner and Lawson)
  • “Physics Problems and the Process of Self-Regulation” (Lawson and Wollman)

are found here: Module 11: Suggested Reading (Workshop Materials for Physics Teaching and the Development of Reasoning)

My Vision for a Physics iBook

UPDATE 1/22/2012: Now with links and Apple’s iBook video!

Warning: This post is a brain dump of all thoughts and conversations I’ve been having about next generation textbooks since Apple’s iBook textbook announcement. Sorry this isn’t polished.

I keep thinking about what a physics iBook would look like. Not a book for consumption (like a traditional text) but rather a book to enable exploration. More like a text-journal-workbook-lab notebook combo, where students would create content from investigations (also pooling created content/data from classmates, etc) and also have reference text for afterwards in the same vein as the Minds-On-Physics text, the first edition of M&I, and the Physics by Inquiry texts.

Stuck on a problem? An intelligent tutor would be able to re-direct you back to a video or animation or even your own data from an exploration where you initially encountered the concept.

There would be these components:

It’d be like an electronic version of the PSSC/Modeling/ISLE /PUM curricula on steroids. And I see this more as the teacher having these tools to deploy to the students, rather than the students following a linear path through text and activities. The class actually builds the text together, and each year the text is different.

The capabilities and content for this iBook already exist. No one has put them together in one package yet. I think it could even be web/cloud-based and platform independent if done with the proper tools.

What am I missing? What’s your vision?

Physics of Angry Birds Lesson on CUNY-TV

Many thanks to Ernabel Demillo and the crew of Science and U!

You can read more about how we use Angry Birds in class here:
Angry Birds in the Physics Classroom