How to get better

You can "catch up" but only because you weren't going very fast

She’s way ahead in points, but will he catch her in Double Jeopardy?

On May 6th, 1954 Roger Bannister broke the 4-minute mile. It is common knowledge at this point that after doing what many people thought was impossible, a tsunami of other athletes broke the barrier soon after. According to Wikipedia there are now more than 1,700 athletes who have broken four minutes, including Sam Ruthe from New Zealand who did it when 15-years old, and Hicham El Guerrouj from Morocco who has the current record at 3m43s (No woman has broken the mile yet, the fastest is Faith Kipyegn from Kenya at 4m07s).

The general explanation for why something that was previously “impossible” for anyone, is now very possible for many people (even talented high school students), is that the fact once people see that it was done, that knowledge inspires others. The fact they know it is not impossible helps them strive to achieve it. The hypothesis is that there was a mental, rather than physical, barrier to the achievement. But that does not explain why no woman has broken the “barrier”. Women clearly know as well as men that breaking the four-minute mile is possible, and women have been told for generations that they can do anything men can do. If the barrier was purely mental, then women would have broken it long ago.

I have a different explanation for why things like the four-minute mile, once broken, tend be broken more than once. To understand where my idea comes from, we should examine less famous, but possibly more impressive runner, Paavo Nurmi.

Paavo Nurmi was a Finish runner 30-years before Bannister. He dominated running in the 1920s. He set 22 world records in distances between 1500m to the marathon. That itself is remarkable. No world class runner today would be competitive at distances more than an order of magnitude apart, let alone win world records in all of them. At one point Nurmi held the world record in the 1 mile, 5 km and 10 km at the same time. Two of those world records were set on the same day with an hour rest between attempts. He competed in three Summer Olympics (1920, 24 and 28) in a total of 12 events. He won gold in nine and silver in the other three. In his career he was undefeated in the 10 km and all cross country events.

While Nurmi did not break a “barrier” like Bannister — which is likely why he is less famous (that and being Finish rather than British) — he still inspired others to get much faster. But the big thing Nurmi did for the sport was not the example he set in races, but the example he set in training. Nurmi trained DIFFERENTLY than everyone else of his era, and when he started winning, the other competitive runners decided to duplicate his techniques.

What were those techniques?

Like most great innovations of the past they are all things that seems obvious to anyone with even a passing familiarity with the sport today:

• Race at an even pace

• Use a stopwatch in training (and while racing to keep pace)

• Year round training

• Interval training

• Long distance training (longer than the races you expect to run)

Once others figured out what worked, they copied it. Humans did not get genetically better at long distance running, they got smarter at how to train to be a long distance runner. What was once the secret of one man who was the best in the world, is now common knowledge for the weekend athlete.

Beeline

The winning words at the National Spelling Bee championship last three years were ‘abseil’, ‘psammophile’ and ‘moorhen’. The winner words from 1939-1941 were ‘canonical’, ‘therapy’ and ‘initials’.

I’m not sure if there is a quantitative measurement for the difficult in word spelling, but I can tell you I did not need to look up how to spell ‘initials’, but I have no idea what a ‘moorhen’ and a ‘psammophile’ are (Even my spell checked thinks I am misspelling psammophile; I did know how to spell abseil. I did it once when I was in New Zealand — Americans call it rappelling).

The National Spelling Bee is like competitive running in that the best in the world are a much better now than they were half a century ago. And, like running, it is not due to genetic improvements. It is because of changes in how spellers “train”. The improvements come in two dimensions:

1. More time spent training

2. More effective training per hour spent

Shalini Shankar’s book “Beeline: What Spelling Bees Reveal About Generation Z’s New Path to Success” chronicles, among other things, how the best spellers in the country train to compete at the national level. Some of these kids (all 8th grade and under) practice seven hours per day, seven days per week. The traditional way to get better at spelling was reading the dictionary and rote memorization. Now the top players study linguistics, do focused studying based on language of origin, and use spaced repetition tools like Anki. They spend more hours preparing but they also learn more per hour they spend.

Increased quantity and quality of training is why the best kids in America today are better spellers than ever before in history. And the reason those kids are training as much as they are, and that the quality of that training has improved is because of the Scripps National Spelling Bee. Just like in long distance running, competition drives performance in a way that nothing else does.

Competition increases the motivation to learn (increasing the hours spent training) and increases the legibility of the learning (increasing the quality of the training).

The National History Bee

Everest’s first History Bee Regional was last November. She had been preparing for about six weeks. The competition was small. She was the only 4th grader (She won!), so she competed against a 5th grader (Ashray) and a 6th grader. She beat the 6th grader, but lost to 5th grade Ashray who had come in 4th at nationals the previous spring. Ashray was amazing. Everest got some questions on him, but he walked away with the competition. In regionals you “score out” when you get five questions correct. He got his five correct answers in the first 8-9 questions every round. Everest took 22, 12, and 16 questions to “score out”.

I spoke with Ashray afterwards to understand how he trained and he told me he just likes to read. His dad backed that up. He said his son particularly enjoys reading about presidents and has read a biography on each of them (He scored on all the president questions in the bee). He also shared some insights on his buzzer strategy at nationals (he said at nationals you get penalties if you buzz in and are incorrect. He said many kids would end rounds with negative scores. He got through to finals by getting questions right, but more importantly, by not getting questions wrong)

My though on Ashray’s training strategy was: “This kid is good, but he is no Paavo Nurmi”

Reading presidential biographies can’t be the most effective way to prepare for these competitions. It was like reading the dictionary to prepare for the spelling bee. And yet, Ashray was good enough to come in 4th in the country in the 2024 National Championships.

The results of last year’s nationals were available online, so I downloaded them and did some digging. There was Ashray among the best 4th grade History Kids in the country: 4th place. At nationals kids can get up to 6 points per question based on how early they buzz in. Once a kid gets 40 points they "win the round” (like how one wins the round with five correct questions in regionals). Here were the scores in finals:

1. 44 (19 questions)

2. 33

3. 30

4. 8 (Ashray)

5. 3

6. -2

Clearly Ashray was very good. But equally clearly there were three other kids who were much much better. Were they using the equivalent of Nurmi training methods? Were they the only three 4th graders in the country using Nurmi training methods?

My current hypothesis is that is exactly what happened.

History Bee Regionals, Winter 2025

Since that first Regional Bee in November, Everest has competed in the following additional events:

1. History Bee Regionals, Houston (Red)

2. History Bowl Regionals, Online “B”

3. History Bowl Regionals, Online “C”

4. History Bee Regionals, San Antonio (Blue)

5. History Bee Regionals, Online “A”

Some nomenclature:

History Bee is the 30-question quiz with up to ten kids competing to see who will buzz in first.

History Bowl is a team based competition with up to four kids competing per side (it has four rounds, each of which is slightly different, but all are basically history quizzes with buzzing in one form or another)

The color events (Blue, White, Red) are all grade-based. Kids compete against others in the same grade. They are all the same level of difficulty.

The letter events (C, B, A) are all middle-school and younger events — so 8th graders compete against 3rd graders, and everything in-between. The letter events increase in difficulty from C→B→A, with A being similar to National level difficulty.

The first event of 2025 was the Houston Bee. Everest dominated. In preliminaries she got five questions correct in ten, six, and eight questions. She scored out in finals in eight questions, before any other kids got a single correct response. Her performance earned her 39 points total across the three prelim rounds - the same score Ashray earned when he seemed unbeatable a few months earlier. I was feeling pretty good about her chances at nationals. Did she have the best performance of any 4th grader in the country at any regional?

I checked.

She did not.

I went through every history bee regional across the country (there are a lot of them) and pulled out the champions. A score of 39 across three prelims would give her the 7th best score in the country. At nationals the top six kids in each grade make finals. Her “unbelievable performance” I had just witnessed wouldn’t even give her a lock on getting into finals, let alone winning. In the Houston finals Everest “Scored out” in eight questions, which I thought was very impressive, but in three other tournaments there was a kid who scored out in five questions. Seven kids scored out faster than Everest did. Neither the prelim nor the final metric put her in the top-6 in the country.

Everest went to her third regionals this past weekend in San Antonio. This time she went in with a goal. The goal was not to win the competition. That was assumed. Her goal was to beat her Houston score of “39 points in prelims” and “out in under 8 questions” in finals. She did it. Her prelim score was 40 and she was out in six questions in finals (her only miss in finals was mixing up the youngest person who won a Nobel prize with the youngest person to be Time Magazine’s Person of the Year).

Her combined score from her prelim and finals performance in San Antonio would be 54 points, which would tie her for third place in the country.

My take-away from this is that, unlike last year, I think there are more than three 4th graders this year using “Paavo Nurmi preparation techniques”.

Everest is now dominate at the regional level (I had multiple parents approaching me after the San Antonio competition asking how she knows so much and how does she prepare. One made a point to tell me that his son was at nationals last year and he is convinced Everest will be in finals. Another family asked if they could get a photo with her). But regionals are not nationals.

There is still a big gap between most competitors and the top competitors. And it looks to me like the number of top competitors is increasing rapidly. The four-minute mile on History Bees has been broken and the tsunami is starting.

Nationals Is Different

Clearly the level of competition will be different at nationals. Instead of the kids Everest has been competing against locally, she will be up against kids like Hridhaan from Celina Texas (he came in 3rd at nationals last year and had a combined regionals score of 58 this year) and Keaton from Washington (Regional score 54, homeschooled with his older brothers, and competes in high school level quiz bowl). The other difference is that the questions at Nationals will be much much harder than regionals.

This past weekend Everest competed in two History Bees. The first was San Antonio on Saturday. The second was the Online “A” History Bee. Recall that in the “Letter Bees” kids 8th grade an under all compete in one pool. The questions in the letter bees are also more difficult, and the most difficult of the letter-bees is the “A” question set. The “A-set” questions are meant to be at the same level of difficult as Nationals.

I am not going to share any of the questions or answers (there are still kids out there doing these question sets), but I can give you a rough idea by pulling what I think are representative answers from previous years.

Here are some example answers from regionals:

• Genghis Khan

• Winston Churchill

• Andrew Jackson

• Alexander the Great

• Vietnam War

A normal college educated adult knows who those people and events are. They might not get the answer on the more difficult parts of the clue, but most people could get them correct on the “last line” I would think.

Here are some example answers from nationals (and the types of answers we saw in the A-bee last weekend):

• Grandma Moses

• Fritz Harber

• Pomp and Circumstance

• Pullman Strike

• Creek War

I consider myself stronger at history than most college educated adults and while all recognize all of those people and events, before this year I could not have told you anything about any of them. I would not have been able to answer “last line” clues on any of them (let alone earlier clues in the question).

In the “A-bee” kids scored out after eight correct questions (instead of five).

Everest did well. In the first round she scored out by the end but it took her to 26 questions. She made finals, but she finished finals with a score of zero (she got one correct, but got a -1 penalty for buzzing in incorrectly later. She got another question “almost correct” but used the old name for the country when the question specifically asked for the current name. There were other questions she knew the answer to, but the other kids in finals were faster on the buzzers).

The good news is that the only other 4th grader, Nivedh, who competed against her did not do much better. He beat her by one point in the prelim round they were head-to-head on (getting to eight questions two questions before she did), but he also scored zero in the finals. The bad news is that Nivedh, when he was competing in the “color regionals” earlier this year only scored 43 points across the four rounds. If a kid every kid who scored 43 points in regionals is as good as Nivedh, then nationals will be a LOT more competitive than my earlier hypothesis. There are 18 kids so far this year who scored 43 points or more in a regionals. Either there are a lot of kids training effectively, or Nivedh is on an exceptional upward trajectory.

One more piece of good news:

Everest is more motivated than ever before.

Last week the public schools in Texas had winter break (her school did not). That mean that all the afterschool activities were cancelled. We used that as an opportunity to do some fun family stuff (we saw a professional magic show, we took Everest to Jeopardy Bar League, we watched a movie), but we also used it as an opportunity to increase the training velocity of History Bee studying.

Now that activities are starting again this week, Everest still wants to keep the velocity. She saw the impact this past weekend when three questions she scored on on Sunday came from stuff she learned only a couple of days prior (partially this is my getting better at predicting what they are likely to ask about, and better focusing study on that, but mostly it is because we were able to learn more stuff last week than we normally do. And she got lucky).

Catching Up

We have less than nine weeks to nationals. Even with our increased velocity, and improved study techniques she will not be able to cover everything I think she needs to know. But she will be as prepared as I know how.

I think she will do well. She is smart and she is working hard and she is working efficiently.

But I also think that she is doing well as she is in comparison to others because we are in the very beginning of the “Paavo Nurmi stage” of the History Bee. There are some kids training efficiently — a lot more than last year — but overall, unlike say the spalling bee, that number is relatively low.

The lack of competition is why there are still many kids out there that compete and are competitive in the History Bee, Geography Bee, and Science Bee at the same time (and do science fair, and math olympiad and spelling bee). Organizers have reached out to me personally and suggested Everest do the Geography Bee as well. So far we have resisted. While Geography and History have high overlap, time focused on Geography is, by definition, time not spent on History. Knowing all the capitals in Africa would be helpful for history, but not as helpful as knowing all the wars that happened in Africa. And she doesn’t have time to do it all.

I used to scoff at parents worried about taking their kids out of school during the year for a family vacation, “Don’t worry about it. They will catch up. And the family experience is a valuable learning opportunity”. I still believe that, but not because I still think it is easy to catch up generally. I now believe it because I think schools go so slowly that it is easy to catch up based on how fast the kids are learning at the school based on how it is structured.

If Everest misses a week preparing for the National History Bee there is no way she will catch up. Sure she can try and work harder for a few weeks afterwards or something, but that week that she lost does not magically mean that her future weeks will accelerate. She is learning fast. And every day she is both learning more, and locking in previous days learnings. Stop doing that for a week and she loses both the lock in from the previous days and the new content she would have otherwise learned.

When she starts again after the break, her base is lower than it would have been, but her velocity is also lower — she needs to work extra hard to lock in the things she learned from before the break — and that is a lot harder than if she reviewed things a day later than a week later. She not only did not accelerate, she actually dug herself into a small hole.

That doesn’t mean that breaks are not good, or that kids should be studying seven hours per day. It just means that breaks have real costs and more hours of study does mean you will learn more than fewer hours of study. We have been fooled into believing otherwise by a school system that doesn’t teach very much per hour of attendance.

Taking a year off school to travel the world with your family is a valuable experience. And if the alternative is being locked in a classroom for a year not learning very much (nothing you couldn’t learn on your own in ten hours of focus per year), then of course the travel is a better choice if given the option. But that doesn’t mean traveling the world is the best and fastest way to learn things, or that doing that does not have an opportunity cost.

---

One experience I did not write about today were the History Bowls Everest competed in. Both involved 4th grader Everest against teams of 8th graders. She did not win either bowl. I hope to write about that experience next, and how to managed competing and LOSING in competitions — which is, in practice, what happens to most kids when they compete (almost by definition) and the reason most parents and educators give for why competition is “bad”.

Until then, keep learning,

Edward (and Everest)