Open letter to Russ re/engine use

Originally posted by Kepler
You obviously missed the bit where I pointed out that I had been deliberately controversial in that first post of mine. I was after some information and I got it. Sometimes one just has to take an extreme stance before people will engage in any debate. Unfortunate and I am fully aware I have ruffled some feathers but I got the information I needed. I knew tha as "Low match up is not a guarantee that the suspect is not using an engine".

I am not trying to prove or disprove the methods the mods use.

Content of your posts in this thread (and also in earlier threads) does not match with this claim - it`s obvious that you are trying to disprove the methods the mods use.

It`s the same as you would say "I am not trying to prove or disprove the methods the mods use, but they are wrong"

Originally posted by Korch
[b]I am not trying to prove or disprove the methods the mods use.

Content of your posts in this thread (and also in earlier threads) does not match with this claim - it`s obvious that you are trying to disprove the methods the mods use.

It`s the same as you would say "I am not trying to prove or disprove the methods the mods use, but they are wrong"[/b]

I repeat. I don't know what methods the mods use. On that basis I cannot comment on the accuracy or correctness of those methods. You were a game mod Korch so you know what methods were used. Are you confident they are accurate and/or correct? If you are then that is good enough for me. What would worry me is if you (or another ex game mod) was saying that there was a problem with the methods.

Originally posted by Kepler
Low match up rates evidence for non-engine use? I prefer to take the statistician's view that low match up rates show there is no evidence to reject the idea that someone is a human. It doesn't actually give us evidence they are not something else. You can interpret that as "Low match up is not a guarantee that the suspect is not using an engine".

I see that you refuse to address my individual points. AGAIN.

I was asked if it is possible to distinguish between the play of engines and the play of humans.
And you devised the WRONG test to investigate this. I already explained why.

If player x matches over n% with an engine then he is using an engine". And? That is not an answer to the question asked.
Seriously, are you stupid? How is this not an answer to the above question?

There are many knowledgeable people in here even if suspicion and paranoia sometimes gets the better of them.
Including me. So why does your ego prevent you from learning something instead of sticking your head in the sand?

Low match up rates evidence for non-engine use?
Seriously, learn how to read. I never even implied that. 🙄

Originally posted by Palynka
I see that you refuse to address my individual points. AGAIN.

[b]I was asked if it is possible to distinguish between the play of engines and the play of humans.
And you devised the WRONG test to investigate this. I already explained why.

If player x matches over n% with an engine then he is using an engine". And? That is not an answer to the ...[text shortened]... evidence for non-engine use?
Seriously, learn how to read. I never even implied that. 🙄[/b]

lol, pathetic baseless assertion!

Originally posted by robbie carrobie
lol, pathetic baseless assertion!

Which one?

Originally posted by Korch
[b]I am not trying to prove or disprove the methods the mods use.

Content of your posts in this thread (and also in earlier threads) does not match with this claim - it`s obvious that you are trying to disprove the methods the mods use.

It`s the same as you would say "I am not trying to prove or disprove the methods the mods use, but they are wrong"[/b]

Well, even if Kepler were trying to disprove the current methods, he wouldn't be able to, right? So what's the fuss?

Originally posted by Palynka
Which one?

lol, none actually, sorry but i coudnt resist, i just like the saying and if you don't mind i am going to use it as my own - regards Robbie.

really i will refrain from intruding in this serious discussion again, sorry for interlope!

Originally posted by robbie carrobie
lol, none actually, sorry but i coudnt resist, i just like the saying and if you don't mind i am going to use it as my own - regards Robbie.

really i will refrain from intruding in this serious discussion again, sorry for interlope!

Feisty. I like it. 😉

Originally posted by Palynka
I see that you refuse to address my individual points. AGAIN.

[b]I was asked if it is possible to distinguish between the play of engines and the play of humans.
And you devised the WRONG test to investigate this. I already explained why.

If player x matches over n% with an engine then he is using an engine". And? That is not an answer to the ...[text shortened]... evidence for non-engine use?
Seriously, learn how to read. I never even implied that. 🙄[/b]

As far as I know I have addressed your points but I am not averse to trying again.

I was not asked to distinguish between the play of a particular human and a particular engine which is what match up rates for individuals will do. I was asked if there is any evidence for a difference in match up rates between humans and engines in general.

"If player x matches over n% with an engine then he is using an engine" is an answer to a question specifically about player x, it does not provide any evidence of a difference between humans and engines. That statement about match up rates is hopefully based on some evidence that engines do indeed achieve higher match ups than humans. Unfortunately it is not evidence in itself. The alternative is that I am, as you say stupid. That's an interesting idea and maybe it is true.

Just being paranoid and suspicious does not guarantee you can teach me anything or that I am capable of learning. Nor does it tell you anything about my ego.

I am actually trying to convince a group of people that cheat detection is viable but I have run into an obstruction in the form of doubts that those people have. Just running analysis on a known cheat and saying "There, he matches n% so he is guilty" will not help here. I know the questions I will be asked. Is n% high? Low? How do we know? What would an engine score? How about another engine? How about a GM? Oh, right, how about a correspondence GM taking a really long time over his moves? Does the engine used for analysis affect things? Hardware? These are all questions that have been asked about the cheat detection process here or in other places from time to time by those who doubt its efficacy or accuracy. If I can show that there is a significant difference between engines and humans and that it is actually not too dependent on factors such as engine used, hardware and so then another site may take action against engine users.

Actually, I have learned something but maybe not what you want me to! That's the thing with learning, it can't be forced.

Originally posted by Kepler
As far as I know I have addressed your points but I am not averse to trying again.

I was not asked to distinguish between the play of a particular human and a particular engine which is what match up rates for individuals will do. I was asked if there is any evidence for a difference in match up rates between humans and engines in general.

"If player x ...[text shortened]... ng but maybe not what you want me to! That's the thing with learning, it can't be forced.

That statement about match up rates is hopefully based on some evidence that engines do indeed achieve higher match ups than humans.

Wrong again. It needs only to be based on the painfully obvious statement that an engine matches higher with itself than with a human not using an engine. It is true irrespective of the match-up rates between engines.

The alternative is that I am, as you say stupid. That's an interesting idea and maybe it is true.
I don't think you're stupid, but I do think you're blinded by pride. You don't want your work to be irrelevant, but it is. Because of the fact I mention above.

I know the questions I will be asked. Is n% high? Low? How do we know? What would an engine score? How about another engine? How about a GM? Oh, right, how about a correspondence GM taking a really long time over his moves? Does the engine used for analysis affect things? Hardware?

As a statistician, you should know how to correctly address the specific question of how to test the null. Let me show you how:

- n% should be the n% of the null, which is the expected match-up of a human against an engine + a security buffer. To make the test more stringent, we assume the player is as good as top chess players. This reduces the type I error.

There is some subjectivity here, but it is unrelated to the match-up rates (MUR) of engine C with engine A. The presence of subjectivity is why we stack the odds against guilty by choosing not only top players but also by adding the security buffer. Again, unrelated with the MUR of engine C against A.

Originally posted by Kepler
I repeat. I don't know what methods the mods use. On that basis I cannot comment on the accuracy or correctness of those methods. You were a game mod Korch so you know what methods were used. Are you confident they are accurate and/or correct? If you are then that is good enough for me. What would worry me is if you (or another ex game mod) was saying that there was a problem with the methods.

I repeat. I don't know what methods the mods use.

You (like other readers) are informed that checking suspects matchup with engine is (or at least was) used as one of the most important indicators to detect engine use. So please don`t pretend to know nothing.

You were a game mod Korch so you know what methods were used. Are you confident they are accurate and/or correct?

I`m not sure which methods are used by current secret mod team, but I`m for 100% sure about methods which I have been used to detect cheats when I was mod.

If you are then that is good enough for me. What would worry me is if you (or another ex game mod) was saying that there was a problem with the methods.

Quoting your text from page 1
This also suggests a reason why it has taken so long to ban some alleged cheats, match up rates are no indicator of engine use!

Another your quote from page 1
If anyone has been banned on the basis of match up rates alone I consider that there is at least a 50% chance that they were wrongly banned.

Originally posted by Palynka
[b]That statement about match up rates is hopefully based on some evidence that engines do indeed achieve higher match ups than humans.

Wrong again. It needs only to be based on the painfully obvious statement that an engine matches higher with itself than with a human not using an engine. It is true irrespective of the match-up rates between e ...[text shortened]... ut also by adding the security buffer. Again, unrelated with the MUR of engine C against A.[/b]

It is indeed obvious that an engine will get a high match with itself. No debate there, that is exactly the reason I did not use HIARCS for my own analysis. Now, if we are saying that a human cheating by using an engine gets a high match up rate irrespective of the engine he is using and the engine used to analyse his games then surely the match up rate between engines is also high? After all, a human using an engine is essentially an engine. I myself have no doubt that engine users should achieve a higher match up rate than ordinary mortals but it is not me that needs convincing.

The test you are proposing will provide evidence whether a human is likely to be making the moves or not. That is not what I am testing. I am testing for equality of means between two samples which is why I used a two sample t-test, it was designed to do just that. The null hypothesis is that the sample means are equal. The t-test provides no evidence to reject that hypothesis. In effect, the probability of the means being equal is high. If the test staistic were low then this would indicate that the probability of the sample means being equal is also low. In other words, I would have evidence for the rejection of the null hypothesis and one could say that it is likely the two samples are drawn from distinct populations, namely games played by humans and games played by engines.

This is only the first step along the way. Once I can show that, yes, there is a distinct difference between engines and humans I have to then convince these people that it is possible to determine with reasonable ease and accuracy that an individual player is using an engine. At present, the result I did obtain is what concerns me. I have to be able to show that it is anomalous in some way. Just saying I did the wrong test will not work since I performed exactly the test I should be performing to do what I wanted to do.

Originally posted by luctruc
Well, even if Kepler were trying to disprove the current methods, he wouldn't be able to, right? So what's the fuss?

Disproving methods (which were used to detect many obvious and not so obvious cheats) is advantageous for banned cheats, complaining that they were banned for nothing, defaming RHP.

Originally posted by Korch
[b]I repeat. I don't know what methods the mods use.

You (like other readers) are informed that checking suspects matchup with engine is (or at least was) used as one of the most important indicators to detect engine use. So please don`t pretend to know nothing.

You were a game mod Korch so you know what methods were used. Are you confident they ...[text shortened]... rates alone I consider that there is at least a 50% chance that they were wrongly banned. [/i]

[/b]Another person who managed to miss the bit about poking a hornet's nest. That was in a reply to one of your posts actually as well as one of Palynka's. If I had quietly asked for the information I needed I would have got nothing. I deliberately made my post controversial in order to provoke a reaction since I was sure that the information would then follow. It worked.

We have been assured that checking match up is important but then we all know the details of that method. It is not just used to check a suspect but to identify the suspects in the first place. If it doesn't work then someone would have spotted it long ago and it would be discredited already. I am happy to test the match up thing if you want but I suspect i will add nothing new to what is known already. We also know that match up is not the final arbiter here, if it were then the player talked about in the original post in this thread would have been banned. There are other, more important factors that I am unaware of. I cannot disprove that which I do not know about.

Originally posted by Korch
Disproving methods (which were used to detect many obvious and not so obvious cheats) is advantageous for banned cheats, complaining that they were banned for nothing, defaming RHP.

Which is why I am not trying to disprove those methods. Apart that is from the pointlessness of disproving something that I don't know about.