This is why we don't have leader boards and proably never will

R

Runner50783i

6
runner50783i
Hi every one!

I wanted to take some time to share with you my thoughts on the leader boards situation that is plaguing the forum, particularly after the inclusion of the seasonal events.

First of all I'm new to this forum but a long term GT (Played all the series) and racing simulators in general fan, and while I've some experience on data base management due to work and studies, I don't know the history of Gran Turismo or the Playstation Network as deeply as some of you probably do.

When I read comments about people demanding leader boards for the game and comparing GT's network service with something like Forza 3, I realize that people don't really understand the implications of this type of service and what's needed to bring a such a service to the masses.

GT's network service is and will be (as far as I know) "free" of charges, meaning that PD pays the service infrastructure out of the sales of GT5. That means that out of every penny that PD gains selling Gran Turismo, an unknown fraction of it is used to pay for the on-line service. This is a money that PD receives only once and needs to be used to support the service until they decide to shoot it down (hopefully, in many years).

Using this model, it is my guess that PD can only sustain a modest data center, one that provides basic functionality for it users.

The data center PD would have to build to support things like per car - per track - per user leader boards as people is ingeniously demanding, will need to be a huge one, I'm talking about multi million dollar infrastructure, one that also needs to be sustained for the years to come and also sustain exponential growth of users and data, all offering performance good enough not to interfere with the end user experience.

Using the most simplistic model that comes to my mind, you can realize the enormous amount of data and thus, amount of resources PD will need to put in place to support this type of feature.

An exercise that one does when calculating the gross amount of data a data base will hold, you simply cross multiply the amount of records for the tables it contains.

Doing a naive assumption that the leader boards data base will have a the tables:

- users (5 000 000 + records)
- Cars (1000 + records)
- Tracks variations (71 records)

To know the potential amount of records that It will have I just have to multiply each table distinct row count that 5 000 000 x 1000 x 71 = 355 000 000 000.

Doing this simplistic approach you can have a sense of the dimension of the problem that PD will be facing. And while is true that not all people plays on line and not all that play on line set lap records on every track and every course, chances are that there are other factors (tire rule, aids, controller type, etc...) that will increase this number a lot more.

To support a global infrastructure able to provide fast, reliable service with the amount of data that PD will be managing, they need to invest several million dollars, I'm guessing 5 millions + based on some real world examples I've worked with, and that's just for starters, this is a service that needs to be supported overtime and despite the good sales figures of the title, I don't find economically feasible to sustain such a service at least until they start having some heavy return of investment something that will happen in the long run.

Forza's online service is sustained by it users who monthly pay MS for the on-line service, so, it is not fair to compare it against GT expecting it to be on par or even better.

I believe PD is doing their best and is doing it right for what we paid for. I admire Kaz and its team for the tremendous effort they are doing on improving the game, that while is not perfect, its the best racing game on the PS3 and arguably the best in the market.

I hope they keep been creative doing the best they can with what they have and continue providing challenging and varied events to enjoy for "free", if you don't pay your on-line bill with Forza you get 0-nada-nothing.
 
This is why I would happily pay for a decent online experience. Voice chat and a faster service would be worth it.

When you pay for it you have the right to complain too.
PS is free which is nice but if the servers are down or removed you can't really moan can you.
 
Sony already has the infrastructure in place to do this. And if the presence of leaderboards helps fuel further sales then that would offset what little cost there would be in creating these things.
 
You honestly think that almost all the people who have GT5 play it online? I'd say about half the people who have it play it online. Also, we are complaining about leaderboards because it says on the GT5 box that the game has leaderboards, so we have a right to be complaining about it.

Turn 10 doesn't get the money for people subscribing to Xbox live, that's Microsoft.
 
If all they stored was times, then assuming 8 bytes to store the time, that's only 2.84 Terabytes... not very much at all.

Edit: If they also wanted to store replays for #1 for each track and car, that's 1000 * 71 * 1MB = 71GB
 
This is all kind of moot considering that the purchases alone of Gran Turismo are generating far more income than purchases + subscriptions for Forza.

BTW, 355 000 000 000 entries - assuming each entry is just a time and a car id, that's about 64 bits per entry, that adds up to about 2,500 gigabytes in total - you can get regular family PCs with that amount of disk space these days.

It's really not that unachievable at all.
 
Uh prologue had it why cant the full game have it? That is the question i've been asking myself about a lot of the features in this game. I know GT5 has more players and content but it also generated more revenue so it only makes sense to put what we had in the PREVIEW version of the game into the FINAL one.
 
Runner50783i
Hi every one!

I wanted to take some time to share with you my thoughts on the leader boards situation that is plaguing the forum, particularly after the inclusion of the seasonal events.

First of all I'm new to this forum but a long term GT (Played all the series) and racing simulators in general fan, and while I've some experience on data base management due to work and studies, I don't know the history of Gran Turismo or the Playstation Network as deeply as some of you probably do.

When I read comments about people demanding leader boards for the game and comparing GT's network service with something like Forza 3, I realize that people don't really understand the implications of this type of service and what's needed to bring a such a service to the masses.

GT's network service is and will be (as far as I know) "free" of charges, meaning that PD pays the service infrastructure out of the sales of GT5. That means that out of every penny that PD gains selling Gran Turismo, an unknown fraction of it is used to pay for the on-line service. This is a money that PD receives only once and needs to be used to support the service until they decide to shoot it down (hopefully, in many years).

Using this model, it is my guess that PD can only sustain a modest data center, one that provides basic functionality for it users.

The data center PD would have to build to support things like per car - per track - per user leader boards as people is ingeniously demanding, will need to be a huge one, I'm talking about multi million dollar infrastructure, one that also needs to be sustained for the years to come and also sustain exponential growth of users and data, all offering performance good enough not to interfere with the end user experience.

Using the most simplistic model that comes to my mind, you can realize the enormous amount of data and thus, amount of resources PD will need to put in place to support this type of feature.

An exercise that one does when calculating the gross amount of data a data base will hold, you simply cross multiply the amount of records for the tables it contains.

Doing a naive assumption that the leader boards data base will have a the tables:

- users (5 000 000 + records)
- Cars (1000 + records)
- Tracks variations (71 records)

To know the potential amount of records that It will have I just have to multiply each table distinct row count that 5 000 000 x 1000 x 71 = 355 000 000 000.

Doing this simplistic approach you can have a sense of the dimension of the problem that PD will be facing. And while is true that not all people plays on line and not all that play on line set lap records on every track and every course, chances are that there are other factors (tire rule, aids, controller type, etc...) that will increase this number a lot more.

To support a global infrastructure able to provide fast, reliable service with the amount of data that PD will be managing, they need to invest several million dollars, I'm guessing 5 millions + based on some real world examples I've worked with, and that's just for starters, this is a service that needs to be supported overtime and despite the good sales figures of the title, I don't find economically feasible to sustain such a service at least until they start having some heavy return of investment something that will happen in the long run.

Forza's online service is sustained by it users who monthly pay MS for the on-line service, so, it is not fair to compare it against GT expecting it to be on par or even better.

I believe PD is doing their best and is doing it right for what we paid for. I admire Kaz and its team for the tremendous effort they are doing on improving the game, that while is not perfect, its the best racing game on the PS3 and arguably the best in the market.

I hope they keep been creative doing the best they can with what they have and continue providing challenging and varied events to enjoy for "free", if you don't pay your on-line bill with Forza you get 0-nada-nothing.
Click to expand...
OR - they could do what they're doing with GT academy, and limit the number of users kept track of.
Change your 5,000,000 (although only 4m have been sold) to the top 50,000 users.
1,000 X 71 x 50,000 = 3,550,000,000
Now, only allow records to be set "with" or "without" skid recovery, as in 2 boards, one for each.
7,100,000,000 Total.
Pretty far short of 355 billion, wouldn't you say?

The truth is, it took 6 years to make this game, and they are having to patch it to match the promised content on the box.
I say within 2 years we'll have leader boards.
 
Sorry OP, but i think you're exxagerating, they could do it for Prologue, why not now? atleast the Premium Cars ffs. Might sell the game if i hear an official confirmation there will never be leaderboards..
 
I am sure the game will have leaderboards in the near future. The leaderboards in GT5 Prologue were huge considering it was just a big demo so I really can't see what holds them back. They could just limit the leaderboards to premium cars if the complexity of 1000 cars would be a problem to maintain.

Let's wait and see how the game expands and PD obviously like to tease our patience and reward us with big patches. However, it would be a disaster if they never add this to GT5.
 
Nielsen
I am sure the game will have leaderboards in the near future. The leaderboards in GT5 Prologue were huge considering it was just a big demo so I really can't see what holds them back. They could just limit the leaderboards to premium cars if the complexity of 1000 cars would be a problem to maintain.

Let's wait and see how the game expands and PD obviously like to tease our patience and reward us with big patches. However, it would be a disaster if they never add this to GT5.
Click to expand...

I think they're waiting for a few players to die off, but maybe that's just me.
less players = equals less data.
Of course, we're talking an amount of data 3 new standard hard drives can hold. (ahhhh)
 
Apollo33
If all they stored was times, then assuming 8 bytes to store the time, that's only 2.84 Terabytes... not very much at all.

Edit: If they also wanted to store replays for #1 for each track and car, that's 1000 * 71 * 1MB = 71GB
Click to expand...

A data base does not care much about storage space. It is the amount of records what matters the most, you are also just assuming 8 bytes for the actual lap time field, you are not considering the rest of the fields that need to complement it to make it useful, I'd say multiply that number for 5 or 10 ...
 
Yes it might be costly, but not too costly because many games have online leaderboards. Just look at Black Ops. They have all kinds of different stats that are kept and are viewable in online leaderboards. This is despite the fact that they have to support a ton more users since the game has sold far more copies. And on top of that they have to set it up multiple consoles separately.
 
OP you are doing a lot of math but just assuming associated cost with the work you are suggesting.

It's very easy to write server side script to accept the lap times and update a database of massive proportions and the traffic involved with sending a few bytes of data is miniscule (look how often GT5 already connects home and consider that in terms of leaderboard info the amount of data a user would send in YEARS of time trialing would probably never amount to the just the last patch in terms of bandwidth useage).

There are plenty of webpages that do have huge numbers of users adding tons of data and changing on the fly and do it for little or not cost... such is the way of the internet these days and the bigger the bulk you buy in the cheaper it becomes per unit of service.

Just look at GTP (or any number of free forums or websites) who have thousand of people logged in all the time submitting posts like these that take FAR more bandwidth than a lap time, host all the info in a queriable and linked manner for anyone to access freely in seconds.

Sorry, your logic does not hold up as being a daunting or even particularly large burden on PD.

This is not directed so much at the OP but just in general, I am kind of tired of all the posts where people "figure out why" something is and the logic behind it is:

Thing A is like this because:

This thing B that I don't know about. I don't know about this thing B and therefore it must be hard/expensive

The proof that this thing B I don't know about is hard/expensive becuase if it wasn't, then Thing A wouldn't be like this.

It's empty circular logic.
 
Last edited:
Spagetti69
This is why I would happily pay for a decent online experience. Voice chat and a faster service would be worth it.

When you pay for it you have the right to complain too.
PS is free which is nice but if the servers are down or removed you can't really moan can you.
Click to expand...

LOL, Unless you pay for live. Then complaint = cut account. :P
 
71,071 leaderboards? Includes tracks and car combos, and the total track bests.
Small time flash games do leaderboards fairly easily especially if all it is send-receiving of data records. An almost bear bones txt file with maybe a paragraph in it is nothing more than 141 Bytes.

assuming every user gives a record for every car and every track
141 * (5,000,000 * 71,071) = 50,105,055,000,000 bytes
47783904.08 Megabytes
46663.97 Gigabytes
45.57 Terabytes

of course that is totally ridiculous that it would all be in individual txt files for every record, but this is more of an example of the amount of data transfer.
 
vaneyk
LOL, Unless you pay for live. Then complaint = cut account. :P
Click to expand...

I own both consoles. Hardware aside I pay for live but the experience is better. All I am stating is that if the experience was comparable I'd pay for PSN too.

The op stated unless your online Forza gives you nothing. Let me ask what GT gives you offline, straight from the box, no patches.

Quite similar.
 
Also I thought of this now...
4 Million copies sold, we'll assume 8 total in the long run.
8 million users.
Are you going to hotlap every car on every track variation?
I know I sure as hell won't, not even close.
So online players equals (magical number less than copies sold)
Amount of car/track combos average players actually completes = An incredibly low number compared to anything posted here.

But I guess we could naively assume every player drives every car around every track......?
But reality, says probably 90% or more won't hotlap half the cars, and will pick and choose favorites of the tracks, and probably hotlap 1/4 of the tracks.
I can't give you even a roundabout figure, but even if our worst case scenario comes true, we're probably talking 200 cars, 25 tracks for 1/2 the users.
(and that's assuming all 200 cars on each of the 25 tracks.)
Happy math :)
 
They will implement leaderboards for each track (you are kidding yourself if they think they will do one for each car). Then a one car circle jerk will commence and everyone will cry about it. True story.
 
  • Sony subsidizes at least part of the game's server needs.
  • PD rents or leases servers, they do not own their own (wouldn't make sense to).
  • It's not that much data to manage and handle; not for corporations like these.
  • We will see all of this happen.

:drool:
 
Sony's Net Worth $45.6 billion

Polyphony Digital Inc. have exceeded 56.36 million units worldwide (as of Sept 2010) across the “Gran Turismo” series. ( note number doesn't include GT5 sales )


I don't think we have a problem here.....
 
When I read comments about people demanding leader boards for the game and comparing GT's network service with something like Forza 3, I realize that people don't really understand the implications of this type of service and what's needed to bring a such a service to the masses.
Click to expand...

OP needs to get GT5P.

I don't see a problem in supporting more tracks and cars. People don't seem to understand that GTHD and GT5P had it done pretty well. They even had tables for each tyre type but later cut it down to S2 tyres as it was spreading the users thinly. I think PD are just wondering if they're going to do 200 premiums only or are how many standard cars they should include, what tyre types etc.

I don't think people realize that GT5 was released in a hurry. Just look at what the patches have contained, take a look at trial mountain, observe how many events and tracks are missing, look at damage and you'll see PD are still working hard to get the game right. They had to release it.

I'm sure PD think people should be busy with single player and online racing and chose to look at it later.

People have been crying like freaking babies for 3 years over GT5P not having private rooms. Now we have 200+ premiums and 20 tracks, proper online and GT5 offline. Seriously, play the game.

Another I'm puzzled at is "I would prefer 400-500 Premiums instead of 800 standard cars" as if PD could magically model another 300 cars instead of porting GT4PSP cars. PD have just about scraped GT5 together and couldn't model anymore unless they get more employees and a bigger setup or outsource.
 
Last edited:
I don't know why people think a database require millions of dollars. It's just harddrive space and a bit of bandwidth. With leaderboards it doesn't even need to have a good response time. Just upload it when you can.

I've got oracle 11g running on a 1.3ghz single core with maybe 20 tables around 30 columns each. More than 10 million rows on each table. I wrote a windows service on a 2.4ghz single core machine to provide a middle tier to the web front end. I don't see why a leader board would require anything more. My total cost - about $300.
 
To be honest, I only care about the top 100 players. And whether my times are in the top 5%, 10%, etc.

Store my data locally so I can keep track of my progress.

I only want to download ghosts from the best players and know where I rank.

But I don't think this is the issue at all because we already had leader boards in GT5P. I think it's the 1000 cars x tires x assists x tracks that is slowing them down.
 
They could very easily reduce the number of required tables and records by using a performance system for the leader boards. Rather than a seperate listing for every car a listing for a range of performance points. They had a system for the points in GT5P that was suitable. It would be doable to do a board for each 50 points or so in performance per track which would drop the number of boards per track from 1000 to maybe 20 or so. Each board could have 1 record per user of course and 1 board per track so the muber is still large but in the millions rather than several billion as stated in the OP.

Also the Op is only taking into account a system which has a seperate board for each car but this is not what I would be looking for. I as well as I'm sure others want to have a board to track time of our tuned cars hopefully geared toward the performance range of the tuned car else the car would onyl appear either stock or fully modded where as it may actually be really good in the mid range on some cars and high range on others.

Forzas leaderboards are far from perfect but they are pretty good.
 
Another apologist OP on PD's behalf. Do you think PD is an independent developer? They've got one of the two biggest companies in gaming publishing their game. They are the ones who will pay for the online infrastructure. Have you ever bothered to read the disclaimer before COD: MW2 starts? They basically say that they can pull the plug on online anytime they want and that is Activision's decision when that will be. And tell me, doesn't COD with all its loathsome market strategy and terrible gameplay far surpass PD's online attempt?

Hell, Vanquish (a single player game) with Sega as publisher has leaderboards. They could just have said "What's the need for leaderboards? Our game is perfect anyhow." And they would be right. I've found japanese games to be usually complete and perfect in what they set out to do. PD's the exception in my eyes.
 
Well PD have done the PP system for online time trials in GT5P, this could and has worked just the same as Forza 3 for tuned boards but without the part upgrades. It was just sliders for BHP. Tuning side in GT5P was more indepth than GT5 unfortunately, and much easier to navigate.

Not a big fan of tuned boards as they just end up with 1 car mostly and the leaderboard car scenario of Forza. GT5 tuning menus are bad enough.

Pretty sure most who want leaderboards want them stock. PD will have to introduce the class system that PP was in prologue for tuned cars. Forza has lots of boards but its mainly tuned stuff. Whatever, be nice to have both.

I don't think PD even know about how Forza has boards for the career mode too. PD could use your friends list and make a clean easy to use interface (stumbling block right there already) and it would post that your friend did such and such an event in this time when you load GT mode and also include Time trial boards as well. NFS does this but this is western devs leaving PD behind.
 
Last edited:

Similar threads

half_sourly
Time Trial - Results and Community Leader Boards
37 38 39
Replies
1K
Views
135K
Lomic
Lomic
ScottPuss20
Will PD Even Bother Making Another Game?
5 6 7
Replies
200
Views
22K
Raf997
Raf997
Skyrocket44
Where is the data logger? No, seriously - where is it?
Replies
0
Views
736
Skyrocket44
Skyrocket44
John_Marsten
So it looks like Nenkai found some more events/championships that were cut
2
Replies
49
Views
8K
Raf997
Raf997
J
Data Transfer GT Sport to GT7
Replies
6
Views
2K
jrbabbitt
jrbabbitt

Latest Posts

Back