The Cambridge Method: Struggle First, Understand Later — Transcript
Full transcript
- 0:00Everyone tells you just how brilliant
- 0:02Cambridge is for studying maths.
- 0:05Cambridge has some of the best
- 0:06researchers in the world teaching
- 0:08undergraduate students topics that
- 0:09they've spent their whole career
- 0:10studying. But they might be the best at
- 0:12research, but they're definitely not the
- 0:14best at teaching. The Cambridge lectures
- 0:16are fast, definition here, a theorem
- 0:18there, and then a proof left as an
- 0:20exercise to the reader. Then they move
- 0:21on, and they expect you to keep up. And
- 0:24not just to cope, but to be the best in
- 0:26the world at it.
- 0:28That's the teaching. That's it. So why
- 0:30do the people who survived that end up
- 0:32running maths departments and winning
- 0:33Fields Medals? Well, Cambridge have a
- 0:36different way of making the best
- 0:37mathematicians, and it's not teaching,
- 0:39but training. Now, let me be precise,
- 0:42because I'm not saying lectures are bad.
- 0:43They're probably world-class in their
- 0:45field. Some of them have wrote the thing
- 0:47that they're actually lecturing on. I'm
- 0:49saying the teaching, the actual process
- 0:51of moving information from one brain
- 0:52into another, in my opinion, is barely
- 0:56attempted. And I don't think that's an
- 0:57accident. I think it's kind of the
- 0:59point. Now, if you know anything about
- 1:01maths degree, you're already probably
- 1:02objecting to what I'm saying. 12
- 1:04lectures a week, uh end-of-year exams,
- 1:07proofs left as an exercise. That's
- 1:09Warwick, that's Imperial, that's Durham.
- 1:11That's every research-heavy maths
- 1:13department in the country. And to be
- 1:15fair, you're right. So let me be more
- 1:17specific about what makes Cambridge the
- 1:18extreme case, rather than just another
- 1:20example of the same type. It's the pace.
- 1:23It's the volume. Material that's a full
- 1:25module elsewhere is often half a module
- 1:27at Cambridge. The maths Tripos covers
- 1:29more ground faster than pretty much
- 1:31everywhere else in the world.
- 1:33And they test you on some of the hardest
- 1:35questions that you've barely been
- 1:36taught. It does it while assessing you
- 1:38exactly once in May at the end. Now, in
- 1:41what way is the teaching bad? Well,
- 1:43there's 10 to 12 lectures a week in the
- 1:45first year. You get a new definition, a
- 1:47new theorem, a proof, next, move on. No
- 1:49worked examples in the sense that you
- 1:51know from A level, no let's do one
- 1:53together. The pace just assumes that
- 1:54you've read the content ahead of time.
- 1:56It assumes you go home afterwards and
- 1:57spend 8 hours understanding the 55
- 1:59minutes they just gave you. And if you
- 2:01don't, you're behind by Thursday of week
- 2:03one. Then, secondly, [music]
- 2:05nothing you do actually counts. Example
- 2:07sheets aren't assessed, attendance isn't
- 2:09tracked, there's no coursework other
- 2:11than a bit of cat time in the second
- 2:12year.
- 2:13The entire year lands on exams in May,
- 2:15which means the system has no mechanism
- 2:17for noticing you're drowning in the work
- 2:20until it's far too late and there's
- 2:21nothing you can do about it. And then
- 2:23finally, supervisions. Supervisions are
- 2:26the thing that Cambridge sells itself
- 2:27on. It sells it the hardest. And to be
- 2:29fair, it's a lot more contact time than
- 2:30almost any other department in the
- 2:32country office. It's one or two students
- 2:34with an expert in that field for a full
- 2:36hour. And when it's good, it's the best
- 2:38teaching, in my opinion, in British
- 2:40higher education. But nothing actually
- 2:42controls that variance. Sometimes you
- 2:43get someone who transforms how you think
- 2:45about maths. Sometimes you get a second
- 2:47year PhD student who is,
- 2:49to be fair, amazing at what they're
- 2:51researching,
- 2:52but isn't really up to scratch with
- 2:54considering how a 19-year-old learns
- 2:56maths. Most of the time it's just
- 2:58students desperate not to embarrass
- 2:59themselves that's forcing them to keep
- 3:01up with the course. And even if you get
- 3:03a good supervisor, it's 1 hour once a
- 3:06fortnight. It's nowhere near enough to
- 3:07carry you through an entire course of
- 3:09something you're struggling with.
- 3:11That hour is the only thing that's
- 3:12standing between a really struggling
- 3:14student and a year that they just can't
- 3:16recover from. And just nobody actually
- 3:18checks whether it's any good for the
- 3:19student, whether they're actually making
- 3:21any progress.
- 3:22So, with all this going wrong, why does
- 3:25Cambridge end up with such good
- 3:26mathematicians even though the teaching
- 3:28seems to be absolutely non-existent?
- 3:30Well, from what you'd probably expect
- 3:32and what I did expect when I went here,
- 3:34non-existent teaching should lead to the
- 3:36worst students. However, something
- 3:39happens when you get no help. When
- 3:40nobody explains it to you, then you have
- 3:43to explain it to yourself. And
- 3:44explaining something to yourself is not
- 3:46the same activity as being explained to.
- 3:48It's a completely different process and
- 3:50a different result.
- 3:52With good teaching, the kind we're all
- 3:53trained to want and to get from A
- 3:55levels,
- 3:56it hands you a clean path through the
- 3:58material. You follow it, it makes sense,
- 4:00you feel the click. The problem is that
- 4:02by following somebody else's path, you
- 4:03don't actually stray from that path.
- 4:06You don't understand why things work and
- 4:08why they don't work.
- 4:10And don't find alternate paths that
- 4:12might just work.
- 4:13Change the question slightly and the
- 4:15path that you knew and loved completely
- 4:16disappears cuz you never built it, you
- 4:18just walked down a path that somebody
- 4:20else taught you. Now, the reason why
- 4:22Cambridge doesn't like teaching is
- 4:23because they don't give you the path.
- 4:25So, you work out where that theorem came
- 4:27from, what breaks if you drop a
- 4:29hypothesis, why the proof goes that way
- 4:31and not the obvious way.
- 4:32When you explain it to yourself, you
- 4:34have to critique and understand every
- 4:36single angle to truly understand it.
- 4:39And the way the example sheets work, you
- 4:40can be sat with a question for like 6
- 4:42hours. On one specific case, I sat with
- 4:44a problem for 4 days, at one point
- 4:46contemplating how much gravity will
- 4:47affect me if I jump out of the second
- 4:49story of the library.
- 4:51But, the point is you get absolutely
- 4:53nowhere. You learn everything that
- 4:54doesn't work. And when you do finally
- 4:56get there, you truly understand why it
- 4:59does work.
- 5:01That 6 hours of getting absolutely
- 5:02nowhere is entirely the point.
- 5:04Everything else is just scaffolding to
- 5:06help you build towards the answer. And
- 5:08I'm trying not to be romantic about
- 5:10suffering. I'm not a masochist, but
- 5:12there is a name for it in science. It's
- 5:13called desirable difficulties. But, the
- 5:15short version is conditions that make
- 5:17learning feel slower and worse while
- 5:20it's actually happening, produce
- 5:21knowledge that lasts longer and
- 5:23transfers to more things.
- 5:26Struggling to understand something
- 5:27before you're told it beats just being
- 5:29told. Failing at a problem before you
- 5:31see the solution always beats just
- 5:34looking at the solution and going, "Ah,
- 5:35that's the answer." Now, on the back of
- 5:37this, there's probably two questions you
- 5:38might ask. One, are students failing in
- 5:41school because of bad teaching, whereas
- 5:42I've just claimed that bad teaching
- 5:43leads to good students? And secondly,
- 5:46does Cambridge actually produce good
- 5:47mathematicians or they just have such
- 5:49high standards that they only let in
- 5:52good mathematicians and then just take
- 5:53the credit? And to be fair, you're
- 5:55asking the right questions. But the
- 5:58second question actually answers the
- 5:59first. Cambridge does have pretty much
- 6:01the highest requirements to get in. I
- 6:03mark STEP papers. I I know it is a
- 6:05really hard exam. It's the hardest one
- 6:07A-level students have to go through. It
- 6:09really pushes students to think
- 6:10differently and understand the math that
- 6:11they're doing. So yes, Cambridge does
- 6:14admit some of the best. Well, that's
- 6:15exactly why this method works. Put a few
- 6:17hundred students like that in one place
- 6:19and they start trying to outwork each
- 6:21other. Most of their working hours go
- 6:23into the course because everyone around
- 6:25them is doing the same. That peer
- 6:27pressure doesn't exist if you don't
- 6:28admit those people in the first place.
- 6:31Does that mean it's selection rather
- 6:32than training? Well, partly, yes. I'm
- 6:35not going to pretend otherwise. But I
- 6:37think the honest version is that it's
- 6:38both. The system selects the best
- 6:39students and then it forces those best
- 6:41students to develop and really learn as
- 6:44quickly as a skill that nobody would
- 6:46choose to do voluntarily.
- 6:48A lot of people put themselves through
- 6:49quite a lot at Cambridge. A lot of them
- 6:51arrive able to sit with a hard problem
- 6:52for maybe an hour. But when they leave,
- 6:54they're able to sit with a hard problem
- 6:55for like over a week. That's really
- 6:57important because that's what research
- 6:59math is. You sit in front of a problem
- 7:01nobody has ever solved before. There are
- 7:03no lecture notes, there's no supervisor
- 7:04who knows the answer because there is no
- 7:06answer. You just sit with a problem for
- 7:08days and days trying to understand it.
- 7:10And the math Tripos is essentially a
- 7:12three-year rehearsal for exactly that.
- 7:14It isn't badly designed teaching, it's
- 7:16well-designed training that happens to
- 7:18look identical to badly designed
- 7:20teaching from the outside. The people
- 7:21who come out at the end have spent three
- 7:23years building [music] the specific
- 7:24muscle that a researching job requires.
- 7:27Tolerance for not knowing and the habit
- 7:29of sticking with a problem to find out
- 7:30that solution.
- 7:32You can't get that from being spoon-fed
- 7:33answers. There is no version of being
- 7:35taught well that also teaches you to
- 7:37work without knowing anything.
- 7:39Okay, so getting on to probably the most
- 7:41important part of this video. You don't
- 7:43need to be at Cambridge to benefit from
- 7:45this system. You can incorporate these
- 7:47habits into your own studies. So, before
- 7:50you seek for the solution, sit on it for
- 7:51a while and fail properly. Don't just
- 7:54give up and then look for the answer.
- 7:56>> [music]
- 7:56>> You need to give yourself time to make
- 7:57mistakes, try different methods, and
- 8:00struggle. That's the important part. 10
- 8:02minutes, half an hour, on paper actually
- 8:05attempting it. The failure is what makes
- 8:07you fully understand. The instinct to go
- 8:09and check the answer as soon as
- 8:10something goes wrong is the exact
- 8:11instinct that's just costing you.
- 8:14Cuz it feels like efficiency. It feels
- 8:15like you're being super efficient,
- 8:16finding the answer, understanding really
- 8:18quickly.
- 8:19But you're just memorizing their
- 8:20solution. You're not understanding why
- 8:22the solution works.
- 8:23>> [music]
- 8:23>> And secondly, a bit more of a
- 8:24controversial one, hold yourself to a
- 8:26very high standard. And I'm not saying
- 8:28do loads and loads and loads of problems
- 8:30just so that you you've done everything
- 8:31and you've you've worked really really
- 8:33hard. I'm saying hold yourself to a high
- 8:35standard when it comes to your
- 8:36understanding. See yourself as a
- 8:38researcher, not just a student. When you
- 8:41start to learn something new, don't just
- 8:43accept it because you've been taught it.
- 8:45Push yourself to develop a deep
- 8:46understanding of why things work and why
- 8:49they don't.
- 8:51That's the entire Cambridge method. It
- 8:53costs you nothing to implement yourself
- 8:55and nobody has to teach it to you, which
- 8:57is kind of more or less the point of
- 8:58this video.
- 9:00But this is the exact method I used back
- 9:02when I was in A level to get the highest
- 9:04grades, five A stars. I did every module
- 9:07on the further maths and back then the
- 9:09further additional maths back. And my
- 9:11school at the time just didn't have
- 9:12anyone who knew the content to be able
- 9:14to teach it to me. So, I had to learn it
- 9:15myself and I had to go through this
- 9:17exact method because I just didn't have
- 9:19the solutions. I couldn't look anything
- 9:21up. I just had to struggle with the
- 9:23problem until I understood them and
- 9:25until I got the answers right. So, yeah,
- 9:27Cambridge might not have the best
- 9:28teaching. But the Cambridge method of
- 9:30training, rather than simply teaching,
- 9:32produces not only the best
- 9:33mathematicians in the world, but also of
- 9:35best thinkers. People who go on to not
- 9:37just study maths, but maybe philosophy
- 9:40or law.
- 9:41Now, if you're applying for Cambridge or
- 9:43any other university for maths, or even
- 9:44if you're just sitting in first year
- 9:46right now, quietly wondering whether
- 9:47everyone else is finding it as hard as
- 9:49you, they are. It's just nobody wants to
- 9:51say so. Try and implement some of these
- 9:53ideas. See if they can help you level up
- 9:55your maths game. See if it helps you
- 9:56with your university maths. And if
- 9:58you've already been through the Tripos
- 9:59and have a different view than me, put
- 10:01it in the comments. I'd actually really
- 10:02like to read it. See what you thought of
- 10:03being at Cambridge. But yeah, I'll see
- 10:05you in the next one. Bye.
About this transcript
This page contains the full transcript of The Cambridge Method: Struggle First, Understand Later by Ryan Ward, generated from the public captions YouTube serves with the video. The transcript has 2,133 words across 331 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.
What you can do with it
Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.
Free YouTube transcript tool
YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.