My $1M challenge to ANYONE: Can you show the COVID vaccine is safe using the Japan record level data?

The Japanese were heavily vaccinated against COVID. Their data shows it was a net harmful intervention. I'm offering up to $1M to anyone who can show the opposite. Offer details in the post.

Executive summary

I just got over 100,000 records with over 4,000 deaths from Japan.

I’ve obfuscated the data for privacy reasons and posted it publicly here. It has age as of May, 2025 (essentially a DOB), DOD, and date of each vaccination.

Rules: You can win up to $1M. To enter you deposit 20% of the reward you are seeking (minimum $10K). So I’m giving you 5:1 odds. We submit our arguments to ChatGPT at the same time (neutral account). Loser of the determination goes first and tries to rebut. If fails, it’s over. If succeeds, it goes to the next person. This continues for up to 5 rounds. If no winner, then it’s a push. 24 hours per iteration max.

It’s 5:1 odds. If the vaccine is safe, everyone should jump on this with a $200K entry fee to make the $1M. It’s easy money. No longer limited to epidemiologists.

Offer expires end of July 2025.

If you accept, the rules on this post, as of the acceptance date, govern the agreement (since I may clarify things in the interim).

5:1 odds. Why is nobody accepting?

If the vaccines are safe, everyone should want to make 5X their money.

I’m using the Japan data because the Japanese were super compliant with 75% of the elderly getting 5 shots. This created a massive harm signal I’ve never seen before. It’s the most heavily vaccinated record level data I have. That’s why I’m using it.

If the vaccines saved lives, the net benefit signal would be super large as well.

So Japan is the PERFECT testing ground for the hypothesis as to whether the vaccine saved lives.

Note that:

1. No epidemiologist in the world wants to touch this data, incentive or not

2. Despite a $1M incentive, nobody in the entire world wants to show me I’m wrong and make 5X their money.

You don’t need a lot of critical thinking skills on that one, do you?

KDOR analysis results are quite stunning

Risk climbs only for 1 year after the shot and levels off. This is not a safe vaccine. This is a disaster. This is likely why I have no takers on my offer.

This is after 2 shots compared to the unvaccinated. But it keep rising because people kept getting the shots until around Jan 2023 which is when they started realizing their friends were dying. So fewer people got more shots so the mortality climb leveled off at that point.

Here’s a later enrollment date after 90% had dose 2:

Here is KCOR on Dose 3 compared with all people who didn’t get Dose 3. Looks similar doesn’t it? It rises for a year and stops rising.

ChatGPT analysis

Note that the Japanese were super-compliant.

A high % of the elderly went for the subsequent shots. This is why the mortality in Japan kept on increasing to be much higher by mid-2023 (over an 80% increase in mortality).

This is why I’m using this for the challenge because the signal is so big and obvious due to the high uptake of subsequent vaccine doses among the elderly (75% had shot 5).

Virtually everyone got Dose 2. This is what the cohorts look (below).

The Dose 2 people above track the people who got Dose 2 as of the enrollment date. Most of those people later went on to get more doses and KCOR reflects this. So this is NOT per dose harm. This is the harm level in people who got at least 2 doses and went on to get more.

Each dose drives the harm higher. This is why Japan has a MUCH higher vax signal in KCOR than the Czech Republic. Had the Japanese stopped at dose 2, we would have seen smaller numbers.

Here are the # of people who got each dose were:

Here are the % of each group who got one or more vaccine shots by age:

ChatGPT analyzed the Japan data independently

I don’t trust any of the other methods, but here they are for comparison.

The following can be validated with my ChatGPT discussion.

The mortality rate of the cohorts getting each shot was very similar, even though those getting 7 shots had WAY LESS time to die than those getting 1 shot.

You don’t need to be a rocket scientist to figure that one out, do you?

When you take time to die into account, you get this:

See? This is a guaranteed win for any skilled epidemiologist. This is exactly what “safe and effective” looks like, isn’t it?

Here’s what the Cox analysis shows. ChatGPT ran it on the data.

See how safe the shots are? Only a 39% increase in mortality per shot. What could be safer than that??? Gosh, I can’t understand why people aren’t rushing to get the shots.

This should be a piece of cake for epidemiologists to show the opposite. After all, epidemiologists are skilled in the art of data deception.

We input both papers and see what our neutral judge thinks as to which is more likely… that the COVID vaccines were net harm or net benefit.

Any takers? Reply in the comments if you want to accept the challenge.

Deadline for submissions: 7/25/2025.

The Japanese dataset

Available here.

Also open to infectious disease experts

If Paul Offit, Peter Hotez, etc. want to enter, I’m find with that.

ZDoggMD: you can enter too.

Neil deGrasse Tyson: You can enter too. I heard you say the COVID shots have saved 10M lives.

Dr. Oliver J. Watson: Your Lancet paper must be right since it wasn’t retracted. Now you can prove it with this data that 10M lives were saved. How about it?

This is a test of epidemiology

One person wrote that you can interpret this data any way you want.

If that is true, then epidemiology is useless, no better than throwing darts.

Either the vaccine is safe or it isn’t. I’m using the Japanese data here because it’s a crystal clear sign of harm. If epidemiologist think this case is ambiguous, it’s time to ignore all epidemiologists.

One “epic fail” so far

This is from Henjin. No, the deaths were all there the whole time. Henjin is one of the data geeks those on the other side rely on. But he made a rookie mistake in not looking at the whole database. All the deaths were there, he just didn’t see them. And his favorite “go to” method is ASMR which is very inferior to KCOR.

Summary

I predict no entries.

I predict that nobody will use the Japan data to submit a paper to the peer-reviewed literature showing that this dataset shows the vaccines are more likely than not, safe and effective.

Nobody will enter because the dataset is so obviously showing that the vaccines are unsafe that they can’t credibly use this data to show it shows the opposite.

I’m doing this to prove a point. When you make the record level data public for the COVID vaccine, all the world’s epidemiologists crawl under a rock and hide from it. EVERY. SINGLE. ONE. This is why all the world’s health authorities keep the data SECRET.

One thing is for sure… the shots worked or they didn’t. We have the data. We’re about to find out what it really says.

Oh, and if people accept my challenge and LOSE, boy, they are going to look REALLY bad now, aren’t they? That’s why nobody will accept.