Flawed analysis of fossil record casts doubt on a decade's worth of evolution research

A decade of research on the evolution of ancient life including the dinosaurs has now been called into question after a fatal flaw in the way fossil data is analysed was exposed. The study, recently published in Methods in Ecology and Evolution by Dr Manabu Sakamoto and Dr Chris Venditti, from Reading, and Professor Michael Benton, from Bristol, says a technique used to 'correct' records of diversity in fossils is actually giving misleading results.

Dr Sakamoto, Evolutionary Biologist at the University of Reading, said:

"Our work calls into question nearly a decade's worth of scientific reports and interpretations on the way life on Earth has evolved.

"The method has been criticised by many scientists for producing misleading and surprising results. So I looked closely at each step of the procedure, and couldn’t believe my eyes."

From the abstract of the paper:

1. It is widely accepted that the fossil record suffers from various sampling biases – diversity signals through time may partly or largely reflect the rock record – and many methods have been devised to deal with this problem. One widely used method, the ‘residual diversity’ method, uses residuals from a modelled relationship between palaeodiversity and sampling (sampling-driven diversity model) as ‘corrected’ diversity estimates, but the unorthodox way in which these residuals are generated presents serious statistical problems; the response and predictor variables are decoupled through independent sorting, rendering the new bivariate relationship meaningless.
2. Here, we use simple simulations to demonstrate the detrimental consequences of independent sorting, through assessing error rates and biases in regression model coefficients.
3. Regression models based on independently sorted data result in unacceptably high rates of incorrect and systematically, directionally biased estimates, when the true parameter values are known. The large number of recent papers that used the method are likely to have produced misleading results and their implications should be reassessed.
4. We note that the ‘residuals’ approach based on the sampling-driven diversity model cannot be used to ‘correct’ for sampling bias, and instead advocate the use of phylogenetic multiple regression models that can include various confounding factors, including sampling bias, while simultaneously accounting for statistical non-independence owing to shared ancestry. Evolutionary dynamics such as speciation are inherently a phylogenetic process, and only an explicitly phylogenetic approach will correctly model this process.

Source Article: www.bris.ac.uk...

The researchers ran thousands of simulations to test the data correction method, but found it failed to return correct results in as much as 100 percent of the simulated cases.

Professor Mike Benton, from the School of Earth Sciences at the University of Bristol, said: "The core assumption is that any portion of fossil diversity that can be explained by variations in rock volume should be explained by variations in rock volume. This assumption is based on no evidence.

"At the extreme, if you have no rock you get no fossils. However, there are many cases where two time intervals are represented by the same amount of rock worldwide, and yet fossil diversity varies massively. Explain that."

I was trying to find free access to the full paper, but no luck yet. Anyone who might be able to help would be greatly appreciated, as I'd like to understand more about what this all means. The article cites over 150 research papers came to their conclusions based on this flawed model, which seems pretty significant if true.

Carry on...

Are they saying there is far less bio-diversity than previously believed? Sorry... do you mind dumbing it down a bit for me?


ETA: I didn't have luck finding the full article, but it was interesting to see that Dr Sakamoto led research that showed the dinosaurs were facing extinction even before the asteroid strike: phys.org...

Is it any wonder that as we come closer to these end times truth is spilling out and knowledge is being increased? I'm sure I've read that somewhere before. Funnily enough he same book tells us of the people who profess to be wise but are instead fools....... Hmmmm...... Wonder what book that was.


Once there is confirmation on giants existing we'll have the missing piece for the Biblical history perspective.

What I think they are saying is that extinction rates in the geological past may have been more extreme than our current interpretation of that fossil record appears to indicate.

Bearing in mind we are talking geological timescales. So think millions, not tens, of years


So maybe the current mass extinction rate is not so unususal, even if its cause is .....

a reply to: AndyMayhew

Hmm I interpret it a bit differently. What I can remember about fossil diversity, to bring things a bit down to earth is, for starters time, the timeframe being sampled has to be consistent, otherwise you extend or shorten the timeframe, you obviously get a higher or lower sampling count, which skews the extrapolated data of the geological record which a summation of the taxa - or the species, genera, family.

Now fast-forward to the problem, the basic assumption was that if you had no rock to sample then you get no fossils from it, ok sounds fair BUT upon sampling identical time-size rock worldwide, they found that perhaps in sampling # 1 there may have been X number of fossil diversity and in sampling # 2 there may have been Y number of fossil diversity...according to the conventional wisdom, this should not be, so now they may have to reclassify decades worth of samplings of fossils.

All of this can be attributed to the simulation hypothesis. Relying on simulations to prove anything one way or the other is erroneous at best and downright lazy. ie."oh,let's fit it into the simulation to prove our findings", but that only works when the simulation is correct. Obviously theirs wasn't.
As has been proven with another simulation of a more controversial problem a simulation is only as good as the person setting up the parameters of the simulation model.
I really think that the vast variety of the fossil biodiversity one would find it would by nigh on impossible to fit it all in one simulation thus giving rise to erroneous data.
Anyway a simulation should only be used as a rough guide and not a perfect black or white conclusion.

So they are basically saying that evolution is a hoax?

Just kidding. This is just a team noticing that the formula needs a bit of tweeking. Science at it's best. Noticing a problem and figuring out the proper way of addressing it.

Unfortunately a lot of work has been done in the last decade. It will all need to be done again with the proper procedure now.

a reply to: PhotonEffect

a reply to: PhotonEffect

I did not come here to argue but evolution is full of holes.... now in order to defend this it would takes volumes of info so anyone truly interested in evolution (and I am) should start from the beginning.

This bone ...... not that one but this one......

Science better yet evolution has NOT shown a direct link that man comes from an ape. NOW there have been a few (in the past) that have tried to pass off the skull of "LUCY" as a link but it was found out later that part of th skull had been ADDED TO with a plaster/sheet rock powder meaning ......fraud. I know there are other examples but they too are nothing more than theories and these same theories are challenged by creationist and there are no winners.

BUT if I may add to which will blow some minds....... When God created man and woman in his image...... there were already human beings walking on this rock......hmmmm........ now how can that be?

Dont want to crash the party folks but for decades evolution main goal and the REASON why evolution is so important.....? Destroy God so that he can be removed from all publications and listed as MYTH. Why this would be a way to remove religion from the planet.... BUT we all know evolution (Pro and Con) will go on until times stops! NEVER will evolution CONCERNING MAN be a fact because that would make God a liar......

So I challenge the world to prove evolution with 100% accuracy otherwise..... I will simply ignore and turn the page...

a reply to: DeathSlayer

Pretty much everything in your post is factually incorrect. Congratulations.

originally posted by: DeathSlayer
a reply to: PhotonEffect

I did not come here to argue but evolution is full of holes.... now in order to defend this it would takes volumes of info so anyone truly interested in evolution (and I am) should start from the beginning.

This bone ...... not that one but this one......

Science better yet evolution has NOT shown a direct link that man comes from an ape. NOW there have been a few (in the past) that have tried to pass off the skull of "LUCY" as a link but it was found out later that part of th skull had been ADDED TO with a plaster/sheet rock powder meaning ......fraud. I know there are other examples but they too are nothing more than theories and these same theories are challenged by creationist and there are no winners.

BUT if I may add to which will blow some minds....... When God created man and woman in his image...... there were already human beings walking on this rock......hmmmm........ now how can that be?

Dont want to crash the party folks but for decades evolution main goal and the REASON why evolution is so important.....? Destroy God so that he can be removed from all publications and listed as MYTH. Why this would be a way to remove religion from the planet.... BUT we all know evolution (Pro and Con) will go on until times stops! NEVER will evolution CONCERNING MAN be a fact because that would make God a liar......

So I challenge the world to prove evolution with 100% accuracy otherwise..... I will simply ignore and turn the page...

So, in other words, you lied. You DID come here to argue, with complete fallacies, dogmatic rantings, and what's tantamount to putting your fingers in your ears and stomping your foot.

Why are you bringing God into a scientific discussion forum OTHER than to argue anyway. No no no no no...you're not allowed to put your non-reason up here on the shelf with our reason.

a reply to: DeathSlayer

Evolution says we have a common ancestor, nothing close to coming from apes. After the split we evolved in modern man (with the inclusion of all of it's offshoots) and they evolved to become apes. See, it works perfectly there.

If you want really want to make people upset, tell them some dinosaurs evolved to become birds.

Nature is awesome eh?

a reply to: DeathSlayer

Our DNA actually does prove that we do, and you are here to attempt to argue.

You are fully aware of the Religious forums here on ATS, that is the area for your, ahem, debunking of sciencey stuffs.

Humans interbred with many species to come to what we are today, with the proof of Denisovan, Neanderthal etc DNA complete throughout our own in the %s of 2-4 in each human and in the case of some only Denisovan DNA present.

You can state what you like however the proof in yours and every ones blood (DNA) on this planet says you are incredibly 100%, without even a 0.01% chance of being right, wrong.

originally posted by: DeathSlayer
a reply to: PhotonEffect

I did not come here to argue but evolution is full of holes....

Your argument is full of holes, inaccuracies, and misunderstood information.

Science better yet evolution has NOT shown a direct link that man comes from an ape.

This is not accurate. No evolutionist scientist says that Humans evolved from apes. What they do say is that humans and apes come from a common ancestor. That ancestor was not an ape, nor was it a human.

NOW there have been a few (in the past) that have tried to pass off the skull of "LUCY" as a link but it was found out later that part of th skull had been ADDED TO with a plaster/sheet rock powder meaning ......fraud.

There have been several specimens of Australopithecus found, not just "Lucy". From the other remains found, a large percentage of the Australopithecus skull could be recreated. But its isn't only the skull that leads paleoanthropologists to believe that Lucy belongs on the family tree of Humans; the rest of the skeleton also suggests that Lucy/Australopithecus was a early hominin.

It is unknown for sure whether Lucy's species (Australopithecus afarensis) is a direct ancestor of modern humans (Homo sapiens), or if her species was a branch on the human family tree that died out. The going evolution theory does NOT say that she was necessarily a direct ancestor, but rather maybe Australopithecus afarensis and Homo sapiens share a common ancestor -- making Australopithecus afarensis a cousin rather than being a great great great (many more greats) grandparent to modern humans.

However, even if Australopithecus afarensis is not a direct ancestor and rather just a cousin of Homo sapiens, that does not change the idea of how evolution works. In fact, it backs up the idea.


By the way, calling Lucy a "Missing Link" would imply that her species were a grandparent to both Apes and Humans, and that just isn't true, nor has any reputable scientist said it's true. People who entertain the notion the Lucy is a missing link between apes and humans don't understand what they're talking about. That link between apes and humans (the "common ancestor" of apes an humans I mentioned above in my reply to your second quote) lived up to twice as long ago as Lucy's species.

a reply to: alphabetaone

Yes, this is the way I interpreted it as well.

Now to better understand where exactly their modelling went off the rails.

a reply to: crayzeed

Yes, I agree with this premise. I don't work with simulations at all so I could definitely be off base with this, but it does seem that certain modelling is designed to achieve or find certain results. If one variable is an assumption due to a gap in knowledge for that variable, then results can be skewed from reality. I believe there are measures in place to account for these statistical errors, but in this case it seemed far off enough to render the outcomes flat out wrong.

originally posted by: Joneselius
Is it any wonder that as we come closer to these end times truth is spilling out and knowledge is being increased? I'm sure I've read that somewhere before. Funnily enough he same book tells us of the people who profess to be wise but are instead fools....... Hmmmm...... Wonder what book that was.


Once there is confirmation on giants existing we'll have the missing piece for the Biblical history perspective.

Science self correcting itself isn't the sign of the end times... Science does that all the time.

originally posted by: PhotonEffect
a reply to: crayzeed

Yes, I agree with this premise. I don't work with simulations at all so I could definitely be off base with this, but it does seem that certain modelling is designed to achieve or find certain results. If one variable is an assumption due to a gap in knowledge for that variable, then results can be skewed from reality. I believe there are measures in place to account for these statistical errors but in this case it seemed off enough to render the outcomes flat out wrong.

Well, to be fair, that is exactly a models raison d'etre, to achieve or find certain results. I think the key to the entire anamoly, based upon what they had indicated in the article, were the predictor variables being decoupled through independent sorting. Predictor variables are often coupled with response variables.

Let me provide an example....baking a cake:

The topic is the Cakes recipe
The predictor variable(s) would be the baking time and oven temperature
The response variable(s) would be the cakes moistness, cakes firmness, cakes depth or "fluffiness"

Now, imagine decoupling the predictor variables in the above scenario from the response variables (through independent sorting)....you would be attempting to change a tire based upon the cakes moistness, cakes firmness and cakes depth or "fluffiness"


Do you see what I'm getting at?

At least this is my interpretation.

originally posted by: alphabetaone
Well, to be fair, that is exactly a models raison d'etre, to achieve or find certain results. I think the key to the entire anamoly, based upon what they had indicated in the article, were the predictor variables being decoupled through independent sorting. Predictor variables are often coupled with response variables.


Let me provide an example....baking a cake:

The topic is the Cakes recipe
The predictor variable(s) would be the baking time and oven temperature
The response variable(s) would be the cakes moistness, cakes firmness, cakes depth or "fluffiness"

Now, imagine decoupling the predictor variables in the above scenario from the response variables (through independent sorting)....you would be attempting to change a tire based upon the cakes moistness, cakes firmness and cakes depth or "fluffiness"


Do you see what I'm getting at?
At least this is my interpretation.

If I am getting it as you've laid it out, then this seems like a pretty significant decoupling. Is the ultimate result a misclassification of time periods for some fossils?

originally posted by: PhotonEffect

originally posted by: alphabetaone
Well, to be fair, that is exactly a models raison d'etre, to achieve or find certain results. I think the key to the entire anamoly, based upon what they had indicated in the article, were the predictor variables being decoupled through independent sorting. Predictor variables are often coupled with response variables.


Let me provide an example....baking a cake:

The topic is the Cakes recipe
The predictor variable(s) would be the baking time and oven temperature
The response variable(s) would be the cakes moistness, cakes firmness, cakes depth or "fluffiness"

Now, imagine decoupling the predictor variables in the above scenario from the response variables (through independent sorting)....you would be attempting to change a tire based upon the cakes moistness, cakes firmness and cakes depth or "fluffiness"


Do you see what I'm getting at?
At least this is my interpretation.

If I am getting it as you've laid it out, then this seems like a pretty significant decoupling. Is the ultimate result a misclassification of time periods for some fossils?

Worse than that. That would actually be easy to fix. It's more misclassification by way of variables that should never have been brought together in the first place. In laymans terms, classifications that make absolutely no sense.

ETA: it's an extremely significant decoupling as it has the potential for 100% error rate.