Random Prize Cars - my views

  • Thread starter gumpy
  • 125 comments
  • 6,137 views
Originally posted by youth_cycler
Thing is, I'm not sure whether it's based on just the last lap, or if it's based on the cumulative time of the entire series, so I really wouldn't be able to answer you. However, it seems like 10-15 seconds is the minimum amount of time to make a relatively considerable difference, from my own experiences.
Click to expand...

Ok I think I understand what you are saying.

But it can't be based on the cumulative time of the entire series as this would mean the saved game trick would not work. My experiment above would rule this out.

If we assume the AI cars always run similar times (and they do in the Dream Car series) then how about I thrash them in one series and then just beat them in another ?

I could beat them by a total of 5x1 sec lead and then reload the saved game 20 times. And win by 5xmax lead and then reload the saved game 20 times.

Does this sound like a good experiment to compare to your theory ? I am willing to do this.

Actually doesn't my experiment above rule this out anyway as I would have registered zero in the last race and the same total time but came up with a smorgasbord of cars.

Will still experiment though.

If your theory was to hold true for the Endurance races then I assume your theory would suggest that winning by 1 sec would result in a different car than winning by 31 laps. Is it better performance = better car ? I am willing to test this at Laguna as I want the F1s and different Altezzas anyway.

Let me raise another theory that backs up the RANDOM voters.

I want the RX7 for example. I think the chances are 1/4 so I reload 4 times. I don't get it. I then think well lets try for another 4 times since it is random after all. I still don't get it so decide to try something different, (maybe reracing the series in a different car or with different margins of whatever). Guess what. The RX7 comes up. Theory proven ???? IMHO NOOOOOOO. The RX7 was always getting closer based on probability. Where random numbers/events are involved you can never predict but in general the longer you go without the RX7 coming up the closer you must be getting to it coming up.

This theory does describe my behaviour. I wanted the Red WRX from one of the series (can't remember which). Reloaded about 12 times. Got annoyed so wanted to try something different. Next time I raced about half a lap of the final race. Guess what the WRX came up. The chances of this happening on the 13th go were pretty high no matter what method I chose though.

I will do an experiment so lets wait and see.
 
Originally posted by kogashiwa_kai
It's very kind of you to go through the trouble. Personally, I do this because (as I mentioned), my 'luck' seems to be very good when I intervene ;)
Click to expand...

Cool.

I am down to only a few colour combos I want so I need to reload many times anyway.

I need:

RX7 LM Car Blue
RX7 LM Car Green
FTO LM Car Green
FTO LM Car Black
FTO LM Car White
FTO LM Car Silver
S2000 LM Car Orange
Altezza LM Car Yellow
Altezza LM Car Black

I may even go for some GT40s and Tom's X540.

And if it wins me lots of F1s in the endurance races I will be happy.

I am still voting for random but not necessarily 1/4 chance. Time will tell.
 
isnt a one in four chance the same a random?
my maths sux but they seem the same :/

btw, there is a 4/4 chance of me saying random in one form or another :P
 
Originally posted by ving
isnt a one in four chance the same a random?
my maths sux but they seem the same :/

btw, there is a 4/4 chance of me saying random in one form or another :P
Click to expand...

Not necessarily.

Random means that there is no logic or reasoning that can predict what the outcome will be.

In Russian Roulette you have 1 bullet and 5 empty chambers. If we ignore the weight of the bullet there is a 1 in 6 chance that you will get shot.

Think about the GT symbol circling around the prize cars. I bet the first time it was programmed the symbol started at the top and always ended on a certain. Then they built the randomness into it.

Some ways would have been to alter the start point, alter the speed, alter the time it takes to circle. For computers to be random (in general) they use a random number generator.

You probably need to set the boundaries eg between 1 and 4. the computer then somehow picks a number.

I have just set Microsoft Excel to pick a random number between 0 and 10. It has picked 9, 10, 10, 1, 1, 6, 2, 4, 7, 6, 4.

Even for the PS2 to start at a different point, or alter speed or time to circle it needs to decide where to start, what speed or what time. I would expect it to use a random number generator.

If I was to write a GT game I would probably want to make it hard to win the best cars. eg F094s. So I might use a 10 digit generator with 1-4 winning Clio Race car, 5-7 winning Tickford, 8-9 winning Corvette C5R and 10 winning F094s. This is random as it is not based on winning margin, number of crashes or anything. It is based on the random number generator. In this example you would have a 4/10 chance of a Clio, 3/10 chance of a Tickford, 2/10 chance of a C5R and 1/10 chance of F1.

I have no idea how a computer can pick a random number but Excel did it for me.
 
Originally posted by gumpy
Not necessarily.

In Russian Roulette you have 1 bullet and 5 empty chambers. If we ignore the weight of the bullet there is a 1 in 6 chance that you will get shot.

Think about the GT symbol circling around the prize cars. I bet the first time it was programmed the symbol started at the top and always ended on a certain. Then they built the randomness into it.

Some ways would have been to alter the start point, alter the speed, alter the time it takes to circle. For computers to be random (in general) they use a random number generator.

You probably need to set the boundaries eg between 1 and 4. the computer then somehow picks a number.
Click to expand...

Sounds like you know exactly what you're talking about. All the ways you considered are common ways to randomize for thoroughly. I think you should already have a good idea of how ramdomization works in computers.

In short, there is not such thing as a true random number that can be generated by a computer, as in the throwing of a dice. It's generated from an equation that is designed with a distribution factor.


Originally posted by gumpy
I have just set Microsoft Excel to pick a random number between 0 and 10. It has picked 9, 10, 10, 1, 1, 6, 2, 4, 7, 6, 4.

Even for the PS2 to start at a different point, or alter speed or time to circle it needs to decide where to start, what speed or what time. I would expect it to use a random number generator.

If I was to write a GT game I would probably want to make it hard to win the best cars. eg F094s. So I might use a 10 digit generator with 1-4 winning Clio Race car, 5-7 winning Tickford, 8-9 winning Corvette C5R and 10 winning F094s. This is random as it is not based on winning margin, number of crashes or anything. It is based on the random number generator. In this example you would have a 4/10 chance of a Clio, 3/10 chance of a Tickford, 2/10 chance of a C5R and 1/10 chance of F1.
Click to expand...

Yes, that's how programmers would usually do it. You randomize as much of the variables as you can. As for distributing the chances of getting any particula car, that's easy-peasy. The standard way is to distribute the items over 100%. For example, from 1 to 100, or for higher precision, increase the granularity/resolution to 1 to 1000. You get the idea.

Then, assign ranges out of a 100% within the range, and if a number falls in a range, you get that item.



Originally posted by gumpy
I have no idea how a computer can pick a random number but Excel did it for me.
Click to expand...

Here's how randomization works in a computer. Like I said, software writers define a distribution equation that generates what is defined as a 'pseudo-random number'.

This randomization function requires something known as a 'seed'. This is the number the randomization function works off of. Randomization functions in general generate a number between zero and a highest possible unsigned integer defined by the writer. So, for our computers today (32-bits), the highest possible number be 4.29 billion. However, if the PS2 development libraries used the C programming language's standard randomization function, it'll be between zero and 32767.

So, randomization functions do not operate on a flexible range. What people do, is to take the random number generated and do a modulus with a range. That is, to get a remainder. So, if I get 17, and I modulus with 4, I get 1 : 17/4 remainder 1.

As for the seed I mentioned, it's the number the equation starts working off on, and the generated number is also a factor determining the next pseudo-random number.

So, usually the seed is set only once. And the most common thing to use is time, or a timer ticker as this is guaranteed to be different at least every second, but the processor's clock can also be used for sub-second time.

Here's an illustration of how it works. Before I understood how it all worked, I did something like setting a seed, then calling a random number. I set a seed before every call for a random number. What I got was not a series of random numbers, but a series of numbers with a uniformed sequence.


Well, that's how computers make up 'random' numbers. And I do believe that the winning a car is random UNLESS there is intervention.


So, don't bother thinking that there is true real-life probability at work here. It doesn't work that way. It is a uniformed distribution and generation based on a seed.
 
Originally posted by kogashiwa_kai
So, usually the seed is set only once. And the most common thing to use is time, or a timer ticker as this is guaranteed to be different at least every second, but the processor's clock can also be used for sub-second time.
Click to expand...
So, the race time theory isn't out of range, right?
 
Originally posted by gumpy
Not necessarily.

Random means that there is no logic or reasoning that can predict what the outcome will be.

In Russian Roulette you have 1 bullet and 5 empty chambers. If we ignore the weight of the bullet there is a 1 in 6 chance that you will get shot.

Think about the GT symbol circling around the prize cars. I bet the first time it was programmed the symbol started at the top and always ended on a certain. Then they built the randomness into it.

Some ways would have been to alter the start point, alter the speed, alter the time it takes to circle. For computers to be random (in general) they use a random number generator.

You probably need to set the boundaries eg between 1 and 4. the computer then somehow picks a number.

I have just set Microsoft Excel to pick a random number between 0 and 10. It has picked 9, 10, 10, 1, 1, 6, 2, 4, 7, 6, 4....

I have no idea how a computer can pick a random number but Excel did it for me.
Click to expand...

out of 4 choices, the moment you give a weighting to the numbers it is no longer random. however if out of 4 choices the end result is random, then it is still a 1 in 4 chance of one of the numbers comming up......
no?

excel, if it picked a random number then the out come could verywell have been 6 ten times in a row.... you need to examine just how excel picked the numbers to know if there was a formular and whether in deed it was random. :)
 
Originally posted by kogashiwa_kai

So, usually the seed is set only once. And the most common thing to use is time, or a timer ticker as this is guaranteed to be different at least every second, but the processor's clock can also be used for sub-second time.
Click to expand...

Once I almost thought I would suggest something but decided against it because of the risk of getting flamed.

When I was struggling to get my red WRX I actually gave up and came back the next day. That is when I raced about 1/4 lap in the final race and presto I won the WRX.

Recently with my experiment and my experience the day before it seemed like Tuesday was RX7 day and Wednesday was Imprezza day.

I am not proposing it as a theory and suggesting if you want a certain car do it on a certain day. Gee wouldn't that get a flaming.

I did struggle to work out how a computer generates a random number because I thought it would need to start somewhere ie the seed. I didn't know that is what it was called but I thought that concept must exist.

I had thought the date may have been where it got this from. But actually it must be seconds because of the save game trick. If this is how it works.

Anyway I know I started this thread but a combination of my work getting busy and this thread is making my brain hurt.

I will probably do 1 run at Laguna tonight and 1 tomorrow night. I want more F1 colours and Altezza colours. I will probably find that I win 74 Calsonics before getting one of these though. LOL.

Over the weekend I may do the experiments mentioned earlier.

Another thing I have just thought of is this. Programmers sometimes accidentally leave some debugging code in the games (sometimes called easter eggs and the like). They would have wanted to test that all 4 prize cars come up and therefore may have had a way of manually selecting them eg pressing direction keys, X etc. I still have seen no evidence YET of this but this may explain it if it exists.

Someone said that this game has been around for that long that by now the cheats would be known. Well is there one widespread theory of how prize cars may be picked. Yes Ving I here you buddy RANDOM ! RANDOM! RANDOM!. I don't think there is common knowledge on this one though.

This game is awesome by all counts. No Playstation games comes close graphics wise and for effectiveness I would say no PC game comes close. The programmers are very good at what they do. I bet they are laughing at us all discussing this issue.
 
Originally posted by ving
out of 4 choices, the moment you give a weighting to the numbers it is no longer random. however if out of 4 choices the end result is random, then it is still a 1 in 4 chance of one of the numbers comming up......
no?
Click to expand...

No it is still random. Well as random as a computer can get. As in the recent reply from kogashiwa_kai computers do calculate their random numbers on a seed.

The weighting does not change the randomness though IMHO.

The number 1-10 are still random but the outcome based on these numbers is weighted / fixed / rigged or whatever you want to call it.

To Youth-cycler. I would suggest total series time is ruled out because otherwise the saved game trick would not work. The only way it can still be in the game is if it only determines a favorite car eg the Imprezza in my experiment and the RX7 the day before. ie the series time could only influence the outcome but not determine the outcome.

Like I said my brain hurts. I just want to win 6 more colours of the F1s and the fantasy cars listed before, get my S license, finish the game and then take some headache tablets.

So happy racing to you all (and me).
 
Originally posted by ving
*cough* random *cough* :P
Click to expand...

That is only one random.

I reckon there is a pattern to your responses. Lets see if I can work it out so I can win a different response.

Actually I have got a cough out of you this time that is a victory in itself.
 
Ving BTW do you have every prize car you want ?

Can you remember where you got each of your F1s?

Are there certain colours you cannot win in endurance races ?
 
no, i sometimes get the car i want, but after all it is random... isnt it :P
i dont have time for endurances... :(
 
Originally posted by youth_cycler
So, the race time theory isn't out of range, right?
Click to expand...

Absolutely not. In fact, anything could be a factor. We'll never know to any degree of certainty. Unless of course, dear ol' gumpy is interested enough to collect more data.

But one thing I'm fairly certain of, is that the roulette behaves differently under two conditions :

1. Without user intervention
2. With user intervention
 
Originally posted by gumpy
Once I almost thought I would suggest something but decided against it because of the risk of getting flamed.
Click to expand...
Heheh ;) Not from me.



Recently with my experiment and my experience the day before it seemed like Tuesday was RX7 day and Wednesday was Imprezza day.
I am not proposing it as a theory and suggesting if you want a certain car do it on a certain day. Gee wouldn't that get a flaming.
Click to expand...
This is seriously funny :lol: But I wouldn't put it past them to do something like this.



I had thought the date may have been where it got this from. But actually it must be seconds because of the save game trick. If this is how it works.
Click to expand...
Yes, I have also suspected the seconds or milliseconds influencing the outcome of the draw.



Another thing I have just thought of is this. Programmers sometimes accidentally leave some debugging code in the games (sometimes called easter eggs and the like). They would have wanted to test that all 4 prize cars come up and therefore may have had a way of manually selecting them eg pressing direction keys, X etc. I still have seen no evidence YET of this but this may explain it if it exists.
Click to expand...
Most of the time, if they leave debug options in, it's deliberate. And as for Easter Eggs, well, in most of the software from Microsoft anyway, it's to give themselves some credit. Some are quite amusing.



Someone said that this game has been around for that long that by now the cheats would be known. Well is there one widespread theory of how prize cars may be picked. Yes Ving I here you buddy RANDOM ! RANDOM! RANDOM!. I don't think there is common knowledge on this one though.
Click to expand...
I don't know about anyone else, but I have no doubts that the draw, by itself, is basically random. Whatever other weight distributions they use on the cars, I doubt we'll really know. Unless of course, it's like the time it kept giving the the freakin Mini instead of the supercars :D



This game is awesome by all counts. No Playstation games comes close graphics wise and for effectiveness I would say no PC game comes close. The programmers are very good at what they do. I bet they are laughing at us all discussing this issue.
Click to expand...
[/B][/QUOTE]
No doubts about it. The game is awesome. Most impressive to me is the graphics engine. As for the comparison to PC games, the graphics hardware has been well in place for quite a while, and few have used it to its capabilities. And there is more processing power than needed now. As a side note, I always preferred playing 'Tomb Raider' on the PC :) I'm growing old waiting for 'Angel Of Darkness' :banghead:
 
Originally posted by gumpy
No it is still random. Well as random as a computer can get. As in the recent reply from kogashiwa_kai computers do calculate their random numbers on a seed.
The weighting does not change the randomness though IMHO.
Click to expand...
Yes, you are absolutely correct. There might have been some confusion between the concept of 'weights' and random numbers.



No it is still random. Well as random as a computer can get. As The number 1-10 are still random but the outcome based on these numbers is weighted / fixed / rigged or whatever you want to call it.
Click to expand...
For the benefit of others, I'll give an example :

4 Cars with 'weights' out of 100% :

Nissan R33 Skyline GT-R : 1-30
Mitsubishi Lancer Evolution VI Tommi Makinnen Edition : :31-50
Toyota AE86 Trueno Sprinter GT-Apex : 51-90
Mazda RX-7 FC3S : 91-100

Generate random number : 40

Result : Evo VI TME


Of course, it's just mere speculation that the game assigns weights. If it did, it'd basically work this way...
 
I did no experiments last night. But I did do the Laguna Endurance race.

I kept falling asleep, did a lot of driving in sand, missing my pit lap so did a few laps with RED tires. Lucky it is a rumble pad as the rumble kept waking me up. Didn't win by the usual margin because of this.

Prize car = Polyphony 002 (Blue) LOL. Pity id didn't give me a different colour.

This morning before work I snuck in 3 races at the Altezza series. Tonight I will use it for an experiment.
 
I know I know. Unfortunately I don't really have the life where there is sufficient time for such an addictive game.

Beginning at 11pm is sometimes the best option.

I reckon the prize cars are allocated on an approximately random basis. Work that one out.
 
Originally posted by TokyoDrifta
I had to do the save/restart trick about 80 times before i could get the impreza LM.:banghead: :banghead: :banghead:
Click to expand...

Are you saying from the 1 saved game ?

Is this on the Amateur Dream Car Series ?

Maybe it took you this long to get to the Imprezza Day ie Wednesday. Seriously were all of these 80 reloads on the same day ?

Were you after any imprezza or a specific colour ?

Did you do anything to make it come up ?
 
Well, I was on the roulette table last night. I wanted my Arta NSX. And guess wot? I decided to let the house give me what it liked. And after 12 attempts, it just refused to give me the car I wanted. I swear, those programmers are evil. They know what you want. Then again, it could also be that the NSX was the only car there with an MR drivetrain. Maybe that was outstanding about it, and they knew people would want it, and made it less probable to come out.

So, restarted the machine, and tried again, and this time, tried the lame-o 'X' button thing. Problem was, the NSX is on the bottom right corner, and takes awhile to appear on screen. Had like 1.5 seconds to press the thing. Got it on the 2nd press attempt.

Luck? Not sure. I restarted the machine as I suspected that the game kept track of how many attempts had been made on the draw. So, if there really was a running probability, it may have been time for it to come out or something like that. I'm just happy I didn't have to wait for two hours to get the thing however it worked.
 
It could have been a combination of chaos theory, the laws of probability and randomosity.

If we are lucky I may do a long experiment tonight. The problem with these experiments is that you never know when chaos theory, the laws of probability and randomosity will change your results. If you go 20 goes without an Altezza but then get one when you press X you never know if it was you who did it or not.

However if I enjoy driving and feel I am driving quick I may get back to those @#%@#$%^ S license tests.

It is a pity that without the S license all I can do is run the enduros to try and get the other F1 colours.

I must admit that the Blue Polyphony 002 looks absolutely awful. I hope the other colours are worth it.
 
Originally posted by gumpy
If we are lucky I may do a long experiment tonight. The problem with these experiments is that you never know when chaos theory, the laws of probability and randomosity will change your results. If you go 20 goes without an Altezza but then get one when you press X you never know if it was you who did it or not.
Click to expand...

Yup, exactly. That's what I was saying too. It's really hard to say. Unless, I kept doing the same thing enough times to determine a fairly consistent pattern, if it exists. Must be a well controlled experiment.

For example, to go on the roulette table via reloads 10 times, record down the results, restart the machine and do it again. Do this 10 times, and if there are any patterns, we'll see it I'm sure.

With this as a control, then the intervention with 'X' might show some difference, or not.
 
this has got to be on the list of the most puzzling questions a video game has ever brought to us. if i were to make such a list by myself, "GT Prize Cars: Random or Not Random" would definetly be #1
 
try get the Dodge GTS-R concept in the GT-All Stars.....unless of course you use the "Get the car you want" trick...hehehe
 

Similar threads

M
Sell your prize parts
Replies
1
Views
8K
TumeK5
TumeK5
S
Update feature I like to see - random cup
Replies
0
Views
356
Steventk
S
Hiropairippu
Random Car Races
Replies
0
Views
440
Hiropairippu
Hiropairippu
Stakado
3 simple things that fix the economy of the game
Replies
17
Views
1K
Emet_PL
Emet_PL
carsonwell
Carsonwell's GT Randomizer messing around
2 3
Replies
61
Views
4K
carsonwell
carsonwell

Latest Posts

Back