Jump to content

The trainer cheats


PrideSP

Recommended Posts

I had over 20 coin flips and all was tails. The cpu had over 10 coin flips and all was heads. Obviously many many battles I did, not all in 1 battle. The cpu cheats.

Link to comment
Share on other sites

I had over 20 coin flips and all was tails. The cpu had over 10 coin flips and all was heads. Obviously many many battles I did, not all in 1 battle. The cpu cheats.

Dude pls...

 

I really cant be bothered to give you the long explanation right now, but the TL;DR version is that your idea of probabilities and/or statistics is completely wrong. There is this law of mathematics known as the "Law of the Large Numbers" that states more than less the following:

 

"As the sample size grows the experimental values approach the average or expected value"

 

Inversely, the smaller the sample size the greater is the chance that the experimental value will deviate from the average.

 

30 coin flips cannot be considered as a "large" sample size by any stretch of the word.

 

You got a streak of bad luck that's all. And if you can't accept that, then you're playing the wrong game. TCGs in general have a very significant RNG component, so if you play a long enough time you WILL have streaks of both good and bad luck.

Link to comment
Share on other sites

You're wrong. I took statistic class. 30 coin flips is considered a large sample size and I said over. So that's about 40. And 30 itself is well enough. The cpu cheats.

Link to comment
Share on other sites

   While I do think the RNG is a bit off, I do agree with gonzo that getting 20 tails is quite probable, but still quite unfortunate. I have had games where I will get 0-2 head out of 10 flips and also 9-10 heads out 10 flips, so no, no cheating is going on, just happened to have a bad game, happens to everyone.

Link to comment
Share on other sites

While I do think the RNG is a bit off, I do agree with gonzo that getting 20 tails is quite probable, but still quite unfortunate. I have had games where I will get 0-2 head out of 10 flips and also 9-10 heads out 10 flips, so no, no cheating is going on, just happened to have a bad game, happens to everyone.

The other day I had an NPC flip 10 heads on RS Electrode. The funniest thing is that I only had 20hp left, so 1 head would have killed me xd Oh, and I had weakness. So I took 400 damage at 20hp xd

Link to comment
Share on other sites

The odds are the same, IRL and in game... Believe it or not, I have had a run of 15 heads while using M Kangaskhan-EX IRL. I know the odds of that are ridiculously low, and so did my opponent. But I was using my opponents coin flip die, and he knows that it wasn't rigged. My point is, runs of bad luck on your side and runs of good luck on your opponents does not constitute cheating. If there was an attack that made you flip 30+ coins, and all were tails, I would be suspicious of the RNG, but as it is, the coin flips probably all occurred independently...

 

The odds of getting a coin flip tails on one flip: 1:2

The odds of flipping 30 consecutive tails: 1:30000000

 

So it is low, but it can happen.

 

Bottom line, I have beaten the Trainer Challenge, and in no way found itnto be cheating

Link to comment
Share on other sites

Bottom line, I have beaten the Trainer Challenge, and in no way found it to be cheating

I believe they fixed it, but when you play a Tropical Beach in the testing VS the AI, it would do nothing but play the beach and never attack or anything! So you could cheat it, but not the other way. :P

Link to comment
Share on other sites

Of course it's probable. Getting a trillion heads in a row is probable. But that's just ridiculous. This is stretching too far. It's totally fixed.

Link to comment
Share on other sites

I believe they fixed it, but when you play a Tropical Beach in the testing VS the AI, it would do nothing but play the beach and never attack or anything! So you could cheat it, but not the other way. :P

Well, I don't have any beaches, so i wouldn't know.. I agree that the AI is a little bit underwhelming for playtesting though...
Link to comment
Share on other sites

Of course it's probable. Getting a trillion heads in a row is probable. But that's just ridiculous. This is stretching too far. It's totally fixed.

I'm detecting a lost cause...

 

Good luck convincing OP guys...

 

 

Adios,

seminc

Link to comment
Share on other sites

I believe they fixed it, but when you play a Tropical Beach in the testing VS the AI, it would do nothing but play the beach and never attack or anything! So you could cheat it, but not the other way. :P

 

The AI still makes a lot of questionable decitions xd I just went to grind what remained of the Trainer Challenges (maxed out all of them now) and they would sometimes do really funny things. I realized that sometimes, when you end your turn without attacking, they don't attack either. So if you're stuck in a bad situation, just pass turns untill you get something good xd (only works sometimes though). Sometimes, they also insist in spamming their first attack even though they have enough energies for the second one (given the second one is the better option), and sometimes, they spam their last attack even when the first one is more ideal xd The list go on and on. The AI is far from challening, even in hard mode. But I think it does its job properly. It's still a great place to grind tradelocked packs regardless, and a good tool to teach the ropes of the game mechanics to the new players.

 

 

I'm detecting a lost cause...

Good luck convincing OP guys...

 

Adios,

seminc

 

Yeah this guy is one tough nut to crack, but don't worry I'll give it a few more tries.

 

Of course it's probable. Getting a trillion heads in a row is probable. But that's just ridiculous. This is stretching too far. It's totally fixed.

 

As per custom, I'll attempt to convince you to the best of my ability up to thrice. If that doesn't work I'll have to agree with Seminc and decide you're a lost cause.

 

First, compare your hypothetical scenario (1,000,000,000,000 heads in a row) with your real scenario (I'm gonna be generous and say 30 heads in a row):

 

Here's the probability of rolling 30 heads in a row: 

 

(0,5)^30 = 0,00000000009

 

That's roughly1 in every 10,000,000,000 games. Very unlikely, I agree. It probably won't happen more than once or twice in your entire gaming history, but it's possible. Now lets analyze your hypothetical scenario:

 

(0,5)^1,000,000,000,000 = 0

 

The result is so incredibly small that my calculator outright refuses to give me any decimals. I calculated this using a Texas Instrument Graphing Calculator, mind you. I was honestly expecting my TI to spit out a ludicrously small power of 10 (like 10^-100 or something like that), but it actually just straight up gave me 0 as a result. The chance of that is so small that it borders on impossibility (as in, it's almost completely impossible no matter how many times you repeat the experiment). You could literally sit there flipping coins untill you die of old age, and you would probably not get a trillion heads in a row even once in your entire life.

 

 

Still not convinced? Then do the following experiment. Go grab a fair coin and flip it 30 times in a row until you've acumulated a couple thousands of flips. You will realize something really interesting: Most of the times the spread will be aproximately 15 and 15; however, you WILL have at least a couple of sets where the spread is 10/20 and even maybe one close to 30/0. And when you average ALL of those, the end result will be around 15/15 or otherwise within a very reasonable margin of error.

 

 

I don't know who taught you mathematics, but whoever told you 30 can be considered as a significant sample size was mistaken. A significant sample size by definition numbers in the thousands at least, and ideally in the millions. Just go google it. I can't post external links here, but if you ask anyone who has even the vaguest idea of what probability means they will tell you 30 is to small to mean jack. If you have a hard time believing people, go search for any math textbook and look at the chapter of probabilities. 

 

Just go to the library of your school/college/local library and pick up ANY math textbook. No matter who you ask or what source you consult, in every corner you will run into the concept of the Law of the Large Numbers, and they will all agree that 30 is way, way, waaaaaaaaaaaaaaaaaaaaaay to small to be considered a "large number".

Link to comment
Share on other sites

As per custom, I'll attempt to convince you to the best of my ability up to thrice. If that doesn't work I'll have to agree with Seminc and decide you're a lost cause.

 

First, compare your hypothetical scenario (1,000,000,000,000 heads in a row) with your real scenario (I'm gonna be generous and say 30 heads in a row):

 

Here's the probability of rolling 30 heads in a row: 

 

(0,5)^30 = 0,00000000009

 

That's roughly1 in every 10,000,000,000 games. Very unlikely, I agree. It probably won't happen more than once or twice in your entire gaming history, but it's possible. Now lets analyze your hypothetical scenario:

 

(0,5)^1,000,000,000,000 = 0

 

The result is so incredibly small that my calculator outright refuses to give me any decimals. I calculated this using a Texas Instrument Graphing Calculator, mind you. I was honestly expecting my TI to spit out a ludicrously small power of 10 (like 10^-100 or something like that), but it actually just straight up gave me 0 as a result. The chance of that is so small that it borders on impossibility (as in, it's almost completely impossible no matter how many times you repeat the experiment). You could literally sit there flipping coins untill you die of old age, and you would probably not get a trillion heads in a row even once in your entire life.

 

 

Still not convinced? Then do the following experiment. Go grab a fair coin and flip it 30 times in a row until you've acumulated a couple thousands of flips. You will realize something really interesting: Most of the times the spread will be aproximately 15 and 15; however, you WILL have at least a couple of sets where the spread is 10/20 and even maybe one close to 30/0. And when you average ALL of those, the end result will be around 15/15 or otherwise within a very reasonable margin of error.

 

 

I don't know who taught you mathematics, but whoever told you 30 can be considered as a significant sample size was mistaken. A significant sample size by definition numbers in the thousands at least, and ideally in the millions. Just go google it. I can't post external links here, but if you ask anyone who has even the vaguest idea of what probability means they will tell you 30 is to small to mean jack. If you have a hard time believing people, go search for any math textbook and look at the chapter of probabilities. 

 

Just go to the library of your school/college/local library and pick up ANY math textbook. No matter who you ask or what source you consult, in every corner you will run into the concept of the Law of the Large Numbers, and they will all agree that 30 is way, way, waaaaaaaaaaaaaaaaaaaaaay to small to be considered a "large number".

*mind implodes* XD

Link to comment
Share on other sites

My mind Imploded and then my body imploded and the earth flipped upside down :P

 

with what imploded appendage did you type this message, o great inverted one?
Link to comment
Share on other sites

with what imploded appendage did you type this message, o great inverted one?

My brain :P inb4 not appendage and i look like dummy 

Link to comment
Share on other sites

At least you're having bad luck against the AI. Having bad luck against a real player in the final round of a tournament would be far more catastrophic.

Link to comment
Share on other sites

*mind implodes* XD

 

The AI still makes a lot of questionable decitions xd I just went to grind what remained of the Trainer Challenges (maxed out all of them now) and they would sometimes do really funny things. I realized that sometimes, when you end your turn without attacking, they don't attack either. So if you're stuck in a bad situation, just pass turns untill you get something good xd (only works sometimes though). Sometimes, they also insist in spamming their first attack even though they have enough energies for the second one (given the second one is the better option), and sometimes, they spam their last attack even when the first one is more ideal xd The list go on and on. The AI is far from challening, even in hard mode. But I think it does its job properly. It's still a great place to grind tradelocked packs regardless, and a good tool to teach the ropes of the game mechanics to the new players.

 

 

 

Yeah this guy is one tough nut to crack, but don't worry I'll give it a few more tries.

 

 

As per custom, I'll attempt to convince you to the best of my ability up to thrice. If that doesn't work I'll have to agree with Seminc and decide you're a lost cause.

 

First, compare your hypothetical scenario (1,000,000,000,000 heads in a row) with your real scenario (I'm gonna be generous and say 30 heads in a row):

 

Here's the probability of rolling 30 heads in a row: 

 

(0,5)^30 = 0,00000000009

 

That's roughly1 in every 10,000,000,000 games. Very unlikely, I agree. It probably won't happen more than once or twice in your entire gaming history, but it's possible. Now lets analyze your hypothetical scenario:

 

(0,5)^1,000,000,000,000 = 0

 

The result is so incredibly small that my calculator outright refuses to give me any decimals. I calculated this using a Texas Instrument Graphing Calculator, mind you. I was honestly expecting my TI to spit out a ludicrously small power of 10 (like 10^-100 or something like that), but it actually just straight up gave me 0 as a result. The chance of that is so small that it borders on impossibility (as in, it's almost completely impossible no matter how many times you repeat the experiment). You could literally sit there flipping coins untill you die of old age, and you would probably not get a trillion heads in a row even once in your entire life.

 

 

Still not convinced? Then do the following experiment. Go grab a fair coin and flip it 30 times in a row until you've acumulated a couple thousands of flips. You will realize something really interesting: Most of the times the spread will be aproximately 15 and 15; however, you WILL have at least a couple of sets where the spread is 10/20 and even maybe one close to 30/0. And when you average ALL of those, the end result will be around 15/15 or otherwise within a very reasonable margin of error.

 

 

I don't know who taught you mathematics, but whoever told you 30 can be considered as a significant sample size was mistaken. A significant sample size by definition numbers in the thousands at least, and ideally in the millions. Just go google it. I can't post external links here, but if you ask anyone who has even the vaguest idea of what probability means they will tell you 30 is to small to mean jack. If you have a hard time believing people, go search for any math textbook and look at the chapter of probabilities. 

 

Just go to the library of your school/college/local library and pick up ANY math textbook. No matter who you ask or what source you consult, in every corner you will run into the concept of the Law of the Large Numbers, and they will all agree that 30 is way, way, waaaaaaaaaaaaaaaaaaaaaay to small to be considered a "large number".

Nope, you're wrong. 30 is enough. If you flip a coin 20 times, about 10 times those will be head and the other 10 times will be tails. Not, 20 in a row will be tails. Just like if your opponent flips a coin 10 times, 5 times will be heads and 5 times will be tails. Of course I'm talking about on average and in general. Therefore, 30 is enough and therefore, the cpu is cheating, thus you're wrong.

Link to comment
Share on other sites

Nope, you're wrong. 30 is enough. If you flip a coin 20 times, about 10 times those will be head and the other 10 times will be tails. Not, 20 in a row will be tails. Just like if your opponent flips a coin 10 times, 5 times will be heads and 5 times will be tails. Of course I'm talking about on average and in general. Therefore, 30 is enough and therefore, the cpu is cheating, thus you're wrong.

 

I was gonna give you one more try but I'm too tired xd OP is a lost cause. I'm outta here.

Link to comment
Share on other sites

 

 

I was gonna give you one more try but I'm too tired xd OP is a lost cause. I'm outta here.

"******* guys, I'm going home." XD That is the first thing that came to my mind. :D (South Park reference btw)

 

Really censors that? Y Tho?

Link to comment
Share on other sites

hello pridesp, 30 flips as a lot of people mentioned isn't a very big sample size.

 

on first look it may appear so, but here's a few reasons why it shouldn't be:

 

 

1. bias about starting:

 

you started counting when? on your first flip ever? or your first tails?

 

it's much more common to notice when something frustrates you than when it doesn't

 

 

2. ptcgo has thousands of coinflips daily. Next to that 30 is nothing. Nada.

 

If you really want to prove to someone that the ai is messed up, then take at least 200-500 flips. MUCH more reliable.

 

 

3. mental or written.

 

When you take a sample size like 20, you end up noting it mentally- which means you are more likely to make mistakes (miss a flip or alter outcome)

 

if you were to save game logs, that would be much more reliable.

 

 

And before you say you have taken statistics classes, i'd like to let you know that I have as well- and NOWHERE would any statistics professor count 30 as a "large" sample size

Link to comment
Share on other sites

"******* guys, I'm going home." XD That is the first thing that came to my mind. :D (South Park reference btw)

 

Really censors that? Y Tho?

 

Pretty much what went through my head when I posted that xd

Link to comment
Share on other sites

hello pridesp, 30 flips as a lot of people mentioned isn't a very big sample size.

 

on first look it may appear so, but here's a few reasons why it shouldn't be:

 

 

1. bias about starting:

 

you started counting when? on your first flip ever? or your first tails?

 

it's much more common to notice when something frustrates you than when it doesn't

 

 

2. ptcgo has thousands of coinflips daily. Next to that 30 is nothing. Nada.

 

If you really want to prove to someone that the ai is messed up, then take at least 200-500 flips. MUCH more reliable.

 

 

3. mental or written.

 

When you take a sample size like 20, you end up noting it mentally- which means you are more likely to make mistakes (miss a flip or alter outcome)

 

if you were to save game logs, that would be much more reliable.

 

 

And before you say you have taken statistics classes, i'd like to let you know that I have as well- and NOWHERE would any statistics professor count 30 as a "large" sample size

30 flip is a big sample size. It doesn't matter how many coins get flips it doesn't change the fact that 20 tails in a row and 10 heads in a row for the npc is highly unlikely. And I started counting when I got heads, hence the 20 tails in a row I said sarcastically. What do you think I start counting? 20 tails in a row and then the npc gets 10 heads in a row at the same time. So you're wrong, 30 coin flip is big enough.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...