Discussion about scamming

Miyuki

Administratrix
Staff member
He needed to look at the Problem as a crime problem. He also should have asked someone at AFI about the issue.
 

JoeNinety

New Member
I'm an ACFE member and my specialist field is IT security and cyber crime.

I know that it's difficult to accept that scammers can be intelligent but the evidence is there. How else would fraud (as a whole, not just internet fraud) become one of the largest world "industries" costed at $3.5 trillion per year.

You should also consider that:
- Statistics have shown that the USA and the UK, where English is the primary language, account for up to ten times more internet fraud than Nigeria.
- The errors contained within spam often include mistakes that simply do not arise from automatic translation (homophones, certain misspellings).
Neither of which support the theory that all spammers are idiots who cannot write English.

At the end of they day you can believe whatever you like; you're all doing a great job. But underestimating fraudsters is almost always counter-productive and whether they devised the scheme or not is a moot point, it's working and that only means more victims.

I also believe that by vilifying scammers and recognising them as devious and manipulative it will reduce the stigma and encourage more victims to come forward.
Everyone is vulnerable to fraud, it just depends what type.
 

basenji

Administrator
Staff member
If I've understood you correctly JoeNinety, you say there's a hypothesis commonly accepted by your peers in the field and that this hypothesis is that bad grammar and poor spelling is a technique used by scammers. You also said that are numerous studies into this and then linked to the most freely available, a paper by Cormack Herley. Unfortunately, it doesn't support the idea that bad grammar and poor spelling form part of any technique as neither bad grammar or poor spelling is mentioned in the paper. In fact, Herley states the opposite:

"[requires] considerable inventiveness and mastery of a language that is non-native for a majority of Nigerians."

The other link you provided is a link to a review of a review of Herley's paper, in which the author brings up something that isn't either of the other two; a discussion about bad grammar and poor spelling.
 

JoeNinety

New Member
First of all I didn't start this thread, it was started by a mod. I made a succinct comment to try and dissuade people from underestimating scammers, which then went off topic.
I never had the intention of discussing any hypothesis in detail and I doubt I'll add anything beyond this unless another post gets me riled. Please do not accept that as an invitation!

Herley's paper had a very narrow scope. It didn't explore the use of language beyond what was quoted (somewhat out of context) in the post above. The question it asked and attempted to answer was; why do Nigerians scammers say they are from Nigeria?

The paper demonstrates the close association between cyber crime and Nigeria. It argues that because this close association is so well known, it makes it easier for the fraud to be spotted. The scammer has already exhibited skills and proficiencies that would belie stupidity (including language skills, which is the origin of the quote in the previous post), so why would they choose to include a tell-tale sign that their email is a fraud?

The paper argues that it is included in an attempt to get the vulnerable and gullible to respond or "self-identify". Only the people who believe in the unbelievable will respond to such an email and it is this target demographic that the fraudster is most likely to be able to defraud. After all, it only requires belief in a lie to be scammed.

So in short, Herley theorises that scammers intentionally make their emails less believable, in order to selectively target a more vulnerable demographic.


Much like indicating a Nigerian origin, poor spelling and grammar is also a tell-tale sign of a fraudulent email, which is why everyone on this board sees it so often. It can likewise be used to affect the same kind of selective targeting outlined in Herley's paper. Granted this goes beyond the scope of Herley's work but it's hardly a giant leap. At the end of the day it is all just an issue of believability: the most unbelievable emails will get the most trusting of respondents. The most trusting are always the most likely to be scammed.

I only used Herley's work to demonstrate the general principle (it was freely available and sprang to mind) but I'm sure if you read through the references and search professional databases you'll find similar proponents. I work in counter-fraud for a living and I regularly attend functions across the UK. My statement regarding peers draws mainly from those experiences so I apologise if I'm over-reliant upon, and only have the time for, empirical evidence.

I have always taken the view that to underestimate a fraudster gives them the upper hand, whereas over-estimating them does no harm. It has worked well in my professional life and I'm too old to change my spots now. Feel free to disagree but I'll stick to what works for me.

I also find it difficult to attribute something that works so deviously well down to dumb luck, but I admit it does happen. Occasionally.
 

Miyuki

Administratrix
Staff member
I will agree that scammers are at various levels of intelligence but the bell curve is very skewed to the right. I would be the last person to argue that the bell curve is static. It is changing. It always changes. However, I feel that the vast majority of 419 scammers are at the low end of the ...knowledge curve. It is the same with any fraud and it takes work to stay ahead in the game. I think that parsing scammer language for anything other than humor (http://antifraudintl.org/showthread.php?62404) is a waste of time but I am willing to be proven incorrect.

I think one problem with Herley's paper is that the focus is far too narrow. Scamming is not limited to Nigeria, it's all over West Africa in particular. It's also found wherever there is a Nigerian diaspora. But Herley told me in an email that I didn't understand his paper anyway. Since it's so dense I'm not surprised. I think I still have to disagree with his hypothesis.

We work in counterfraud as a hobby but we are very good at it, especially if we've been doing it for a long time. Herley's research might be better if he had contacted more people who actually work with scammers and their victims rather than rely entirely on mathematical models.

And, JoeNinety, this is a discussion thread. I think everyone is trying to understand your point of view and Herley's while explaining our own. This isn't personal and it isn't to question your background or qualifications. Although we are not certified "professionals", this is what we do and what we do seems to work for law enforcement. I work with victims every day and I work with scammers every day and I've been doing this for 7 years and I don't get paid for it. I do this because I care.
 

basenji

Administrator
Staff member
I apologise if I'm over-reliant upon, and only have the time for, empirical evidence.

That's fine, and no need to apologise, but is it down to empirical evidence now? What happened to the hypothesis? There is a difference, as you no doubt know. And what exactly is it that we're discussing now: bad grammar and poor spelling or implausible stories?

Herley's paper had a very narrow scope. It didn't explore the use of language beyond what was quoted (somewhat out of context) in the post above.

I'm not sure I'm entirely comfortable with that remark, actually. Please explain why the quotation is out of context. To the best of my knowledge, you posted a link to Herley's paper, stating that it was one of numerous studies in support of the theory that Nigerian scammers use bad grammar and poor spelling as a technique. In my opinion, the paper doesn't support the hypothesis you say is commonly accepted by your peers in your field. I'm not sure I'm more off pist than you are.

That said, I've never thought of your replies as short in any sense of the word, and I'd like to repeat what I wrote in an earlier post.

Interesting.

I'm not scoffing at any technique or underestimating anyone. I've been sitting right next to scammers in cyber cafés in both Nairobi and Dar Es Salaam, listening to them as they chat away in either "broken" or standard Kiswahili. Uneducated and cocky yes, but not necessarily stupid. Furthermore, the English subtitles of Nollywood movies in Nigerian pidgin (I know a few things about both Nigeria and Kenya, for various reasons) are littered with spelling mistakes, and I don't think it's because any particular technique is being used, it's due to sloppiness or inferior English skills. In that respect, my empirical evidence, drawn from many years of first-hand experience of Africa, apparently differs from yours. I'm not going to discuss this any further. I've found your posts interesting (like I said) but evidently, we disagree and if it's alright with you, I'd like to suggest that we agree that we disagree and move on.
 

JoeNinety

New Member
Nothing has happened to the hypothesis. I made it clear that my comment regarding peers was drawn mainly from empirical experience. That is all. Please do not quote me out of context.

As for what we are discussing now, I've grouped tall tales, together with poor spelling/grammar under the heading of believability. All three can reduce the believability of a scam email, which in turn helps ensure the email is answered by only the most trusting, gullible and vulnerable of respondents. I've worded it several different ways and across numerous posts, but the topic has remained consistent.

As mentioned above, my original post was not intended as a discussion on this theory but instead was an attempt to point out the dangers of underestimating scammers, the only point I've made with which you appear to agree.

I'm afraid the "best of your knowledge" is flawed. I pointed out Herley's paper as being supportive of the general principle that unbelievable emails can selectively target a vulnerable demographic. Herley's paper focuses on the "tall tale" element of believability, and more specifically the reason why fraudsters say they are from Nigeria. You and the wikipedia entry Ben referenced are both correct in saying that it does not explore language or grammar.

I considered what you quoted "somewhat out of context" because the quote was not a supported or reasoned element of Herley's study and it was an aside used only to dismiss stupidity as the reason why fraudsters say they are from Nigeria. It also directly contradicts what appears to be your position: that scammers have not mastered the English language.

Your first hand experience is invaluable (even enviable) and I agree with Miyuki that consulting groups such as AFI would benefit future studies. Volunteer organisations often find different approaches and solutions because they look to help as many people as possible, as quickly as possible. Time being a factor, volunteer advice is based on the most likely scenario, whereas professionals are paid to consider the worst case scenario and work backwards; perhaps this goes someway towards explaining why Herley's paper, and my comments, have been poorly received here, yet not elsewhere.

Whilst I don't doubt that a considerable number of scammers have poor English skills, I do question whether the scammers you overhear in Internet cafes are amongst the most the successful or dangerous. Of course this doesn't mean they should be targeted any less as they still cause the same hurt and pain, but again it may go someway towards explaining our different points of view. We are only a sum of our experiences after all.

One thing I would like to clarify (I attempted to explain in my second post) is that neither Herley nor I are claiming that every scammer is purposefully using tall tales, poor spelling and bad grammar as a devious trick. There will never be a hard and fast rule (as De Master Yoda said, there are too many variables).

Likewise, I am not claiming that these elements feature within spam only as a result of this technique; for example we know many of the grammatical errors can be attributed to synonym software that alters a standard format to make spam filtering more difficult.

What I am saying is this technique is a real danger (if not ever-present one) and should be acknowledged as such, or one day you may end up underestimating a scammer and that is only a recipe for disaster.
 

Mr Humpuuki

Member
I think it's important to point out the difference between scammer 'intelligence levels' and their 'skill level'. It's possible to have a scammer who cannot spell, who is generally moronic and ill-educated and yet be highly skilled and successful at fraud. As Joe90 says they are not to be underestimated. Personally i'd suggest that any who thinks all scammers are dumb are engaging themselves routinely with idiots.

As for the spelling mistakes been used to identify the gullible victims there is certainly some truth in that.

I've been sending out emails with the subject 'nutification email' and it guarantees the respondents are of low intelligence and able to manipulated easily. However it does not then restrict the respondents to be the ones useless at fraud. Their intelligence level is not relative to the level of skill they have accrued. Someone can have little to no spelling ability and yet be doing very technical and dangerous frauds.
 

TexasBlackRose

Emeritus
Thank you to the people who matter and share and help to protect

For once I can relate, I do my hunting and such for free as I have seen the result of the scamming first hand and deal with some of the victims as well. Although it is through different avenues from the populus, I can say it damages the person that they have stolen the identity of as well.
My hat and scarf is off to all people that work hard whether paid or not to right the wrongs as best as possible as this is a global epidemic.

All I can do is hold my hand over my heart and mouth and mutter thank you to you all.


;) :D
===============================================================================



I will agree that scammers are at various levels of intelligence but the bell curve is very skewed to the right. I would be the last person to argue that the bell curve is static. It is changing. It always changes. However, I feel that the vast majority of 419 scammers are at the low end of the ...knowledge curve. It is the same with any fraud and it takes work to stay ahead in the game. I think that parsing scammer language for anything other than humor (http://antifraudintl.org/showthread.php?62404) is a waste of time but I am willing to be proven incorrect.

I think one problem with Herley's paper is that the focus is far too narrow. Scamming is not limited to Nigeria, it's all over West Africa in particular. It's also found wherever there is a Nigerian diaspora. But Herley told me in an email that I didn't understand his paper anyway. Since it's so dense I'm not surprised. I think I still have to disagree with his hypothesis.

We work in counterfraud as a hobby but we are very good at it, especially if we've been doing it for a long time. Herley's research might be better if he had contacted more people who actually work with scammers and their victims rather than rely entirely on mathematical models.

And, JoeNinety, this is a discussion thread. I think everyone is trying to understand your point of view and Herley's while explaining our own. This isn't personal and it isn't to question your background or qualifications. Although we are not certified "professionals", this is what we do and what we do seems to work for law enforcement. I work with victims every day and I work with scammers every day and I've been doing this for 7 years and I don't get paid for it. I do this because I care.
 
Top