GT Player Handicap Experiment. Test Participants Required.

[W3H5]

Experiment: Player Handicap in GT5.​

Introduction.​

Since an early update we've had the ability to match car performance using the PP (Performance Points) system and it has played a big role in matching up cars relatively fairly for online and offline racing alike. The GTP WRS has a ranking system which ranks drivers on their racing skills in categories.

Is there though, a specific way to rank a players skill overall in a system similar to PP for cars? Would it be useful for racing series or online match ups, even for setting time trials perhaps? This is what I intend to investigate.

The Plan.​

A technique used in sport, handicapping, is a way of setting an advantage or disadvantage to higher or lower skilled players, respectively. Applied in horse racing with the use (or not) of a ballast applied to the horse to level competition, handicapping is also a rank of how skilled a sports participant is.

How would the system work?​

To set a fair benchmark which can be used by all comers as a guide a series of tests must be done with as limited variables as possible to keep statistics stable.
It's use would be dependant on users to whether the system is implemented or not. This is just an experiment to work out how a handicap would be attained.
By way of a weight ballast the handicap would be applied.

The Experiment.​

A selection of 10 skilled drivers will be needed to complete several time trial tests to work out an overall benchmark for each driver. The results (times) from each test will be added together and an average time worked out. Once each average driver's time is worked out a grid in order of fastest first will be compiled. Form that list the fastest drivers will be allocated a ballast weight and asked to rerun the test series and report back with their times to once again be calculated and from that an appropriate handicap ballast setting can be established.

The Test Series.​

Each driver will be performing the exact same test with no variables other than their preferred choice of control (wheel / DS3). All trials will be completed in arcade mode where car set up will remain stock from the arcade garage. Tyre choices for each car will be stipulated along with the track they are to be tested on and the number of laps to be performed for the test.

Car/ Track combo 1.
HIGH SPEED RING - PAGANI ZONDA R '09 (Tyres RH) - 10 laps.

Car / Track combo 2.
COTE D'AZUR - VW GOLF GTI ' 05 (Tyres CS) - 5 laps.

Car / Track combo 3.
NURBURGRING NORDSCHLEIFE - MCLAREN MP4-12C '10 (Tyres SM) - 2 laps.

Data of the fastest lap of each C/T combo within the lap limit from each driver will be used to calculate a final, overall handicap score. Once these stage of testing is complete, stage II of finding the handicap / ballast ratio will begin.

Driver Requirements.​

Each driver must have a fair level of skill to eligible to participate in this test series. Drivers must be honest about their lap times and use of car/settings which must remain controlled throughout the test to provide accurate data.

If you feel you have the ability, time, honesty and dedication to participate in this experiment please sign up in this thread. Please also make mention of any championships, series or GTP WRS results you may have that would prove you abilities to participate. Rewards will be awarded to those who help with the test, though I'm yet to work out what they will be.

Driver data will be sent and received via PM and all test subjects will be listed as a reference number from #1 - #10 rather than user name to eliminate any possible embarrassment/ cheating/ boasting. Names of test drivers will only be revealed after the tests are finalised and all data has been calculated, so long as the participant is not against having their user name published.




Feel free to add your questions and or driver recommendations. If you have any recommendations on how to improve the test please also feel free to comment.

 

[W3H5]

My first attempt at a trial run to play with some figures.
​

Q&As
​

What of the use of aids for the testing, and track grip?
Grip will be set to real as a good enough driver should be on the track, not off.
Aids will be set to all off with the exception of ABS which will be the drivers choice.

Why arcade mode and not online?
The main reason is data consistency and limiting variables. On-line testing offers to many variables and options outside of the test parameters.

Why arcade mode cars and not personal cars?
The same as above. The arcade mode cars will be in exactly the same stock form for every tester, eliminating tuning variants. They are also accessible to everyone who would wish to attain their own handicap once the benchmark has been set.

Why not use PP alone?
Using PP is a car restriction, not a driver ranking. Two racers with the same PP, same car, same settings could race and one could be lots faster because of personal skill. This isn't so much about the cars as it is the drivers.

Why not use Bob to test run the trials?
Mainly because they aren't human and are programmed to respond in a certain way. Bob runs wide into a corner, he may have one line of code on how to recover. A human makes the same error; a whole brain to judge the best course of action.

What sort of time scale will this run over?
It's an experiment. It could take a few months. I'm not pushing drivers to give me results day in, day out, just to check in once a week with their data for me to calculate.

What effort does this require if I am to be given a driver position?
A few hours a week of you time and of your racing skills, and obviously the patience to stick out the experiment and cooperate with me.

What are driver requirement?
Well, if you've read this far, you probably know by now.

Why are you doing this Shem?
I'm a bit nuts. No, I kid. I can upon the idea one night an wondered if it was a logical thing to proceed with. Computer said "yes", and here I am. Plus (serious answer coming up) I'd like to give something back to a game, and also a community that has given so much to me over the course of my time here. Dry your eye, now.

What do you want to know?
Something like this is apt.

Was in the intercontinental GT500 championship organised on here for 3 seasons as the only DS3 user, I had a win and and numerous podiums, the races were 1 hour normal and 1.5 to 2 hour endurances, so I have the experience.

I regularly win online in open lobbies, all aids off, grip real rooms!

I am also fair, clean and quick!

Potential Driver Candidates
VEXD
rumblina
1241penguin
phil_75
Greycap
uMadson?
ZohsixGT5
homeforsummer
Apokalipse
tomhart9
TheBook


Chosen drivers will be notified by PM and be allocated a number by which they will be referenced in the time chart to avoid envy / arguments / despondency. Essentially successful drivers will be anonymous until the end of testing.

 

[amar212]

IMO, you should try to do this experiment with the LV40 B-Spec Bobs who reached their peak through ordinary B-Spec play (not Remote Races) and with Career-gauge maximised.

Also, the best type would be Neutral-driver or slightly aggressive.

Support for the idea 👍

 

[crowhop]

Is there a requirement for signing up? 60 Gold Licenses? Gold Special Events?

 

[rumblina]

What sort of benchmark time will be the entry requirements? Also, I guess this will be offline, not online? I probably wouldn't meet online timings due to being +10 GMT.

I am interested pending 👍

 

[W3H5]

IMO, you should try to do this experiment with the LV40 B-Spec Bobs who reached their peak through ordinary B-Spec play (not Remote Races) and with Career-gauge maximised.

Also, the best type would be Neutral-driver or slightly aggressive.

Support for the idea 👍

Great call, amar!

It would help with driver consistency but there isn't a way to set Bobs to do time trials with specific cars and tracks. I chose these cars and tracks so they cover a wide spectrum of racing techniques.

Another thing would be that although Bob seems to drive well with the high powered cars (better than me sometimes!) they tend to be slightly robotic even at L40, since I've spent a lot of time racing Bobs in tournaments.

I appreciate the input, really. 👍

 

[1241Penguin]

I'm interested! When will this happen? I'm free until January 3rd.

Also, can you please specific the driving options (e.g. ASM, TCS, Grip Reduction, Tire Wear, etc).

 

[W3H5]

Is there a requirement for signing up? 60 Gold Licenses? Gold Special Events?

Just proof enough that a driver is quick enough. It's not necessary to have all golds or special events. Just a good command of race craft and preferably somebody to vouch from their abilities.

What sort of benchmark time will be the entry requirements? Also, I guess this will be offline, not online? I probably wouldn't meet online timings due to being +10 GMT.

I am interested pending 👍

The benchmark will be established once the drivers have finished stage I of testing. Offline in arcade mode is my preference for the tests as it narrow down variables greatly.

I'm interested! When will this happen? I'm free until January 3rd.

Also, can you please specific the driving options (e.g. ASM, TCS, Grip Reduction, Tire Wear, etc).

There will be no set time for this test. Once the drivers have been finalised they have as long as is needed up to around a month to compile their best lap times and submit them. There is no race to get the data in first.

All aids will be off except ABS and as far as I know tyre wear isn't a factor in arcade mode and grip reduction shouldn't matter to a good enough driver as the track is where the car should be.
But to save arguments grip will be set to real.


Thanks to everyone for showing support and interest. Keep checking in to see how things develop and when the drivers will be picked. 👍

 

[Wardez]

Yeah good idea. The great, although just a cool early use for results, thing would be that you could then do online "1 lap magic" events between participating drivers to really prove every one's average. You start them off and give the faster drivers a handicap in order by rank going back, then start the race. Goal would be to have every one nearly finish at the same time.

But a different approach to rating driver skill objectively is always welcomed and I think if enough interest is here, and good work is put in, we could really see something awesome come out of it.

Always good ideas shem 👍

 

[berty1979]

Great idea Shem, ill be keeping an eye on this. I wont drive though as I am a horrible driver of anything that isnt a formula car and regularly finish off podium in 'normal' cars
Thanks for the invite though

 

[rumblina]

The benchmark will be established once the drivers have finished stage I of testing. Offline in arcade mode is my preference for the tests as it narrow down variables greatly.

Keep me in mind then, I am not super fast but can hold my own; Penguin can vouch for me (I hope)

 

[VEXD]

Good luck with this Shem, it's a great idea, that if given enough exposure, will hopefully yield some interesting and useful results.

I dismay somewhat at what may happen when the GT5 community get their hands on it and think of it as some sort of 'bestest driverz, hottest laptimez' competition, but other than that I think it's great.

I would drive for you if you wanted, but I am not so quick these days.

I am honest when it comes to GT though.

Just GT.

👍

 

[rumblina]

I see it as a new model on the old school 'tuners challenge' once run by Adrenaline

 

[Mike Rotch]

but I am not so quick these days.

Why must you lies.

 

[phil_75]

I'll give it a go!

Was in the intercontinental GT500 championship organised on here for 3 seasons as the only DS3 user, I had a win and and numerous podiums, the races were 1 hour normal and 1.5 to 2 hour endurances, so I have the experience.

I regularly win online in open lobbies, all aids off, grip real rooms!

I am also fair, clean and quick!

 

[Vol Jbolaz]

I don't know why you would want only really good drivers. I would imagine you'd want a large sampling across skill levels in an effort to see just how well your ballast idea works. If you ease up off of the elite-only drivers, then I would be more than happy to run those times and submit the totals to you.

I would also imagine that you are already aware of racing leagues that use ballast. I have seen two. For the next season of the one I'm running, we are going to use PP and not ballast because we can set limits on it and not on weight. Top drivers will get a couple PP knocked off, bottom drivers will get a couple PP added. Lower the PP for the room and let them enter. Then raise the PP limit and let the next batch enter, and so on. This is meant to not only compensate for different driver skill levels but allow drivers to pick cars that may normally shy away from.

We did some testing and it was fantastic fun, making much tighter racing than we had before.

 

[W3H5]

The great, although just a cool early use for results, thing would be that you could then do online "1 lap magic" events between participating drivers to really prove every one's average. You start them off and give the faster drivers a handicap in order by rank going back, then start the race. Goal would be to have every one nearly finish at the same time.

But a different approach to rating driver skill objectively is always welcomed and I think if enough interest is here, and good work is put in, we could really see something awesome come out of it.

Always good ideas shem 👍

There is scope in the future to bring an online element to the testing but at this stage I want to give all test drivers a level playing field by using AM (Arcade Mode). Perhaps that could be brought in at a later phase of the experiment.

I'm really trying to keep this as scientifically accurate as possible so that a players handicap can be gauged against the benchmark from the tests.

Great idea Shem, ill be keeping an eye on this. I wont drive though as I am a horrible driver of anything that isnt a formula car and regularly finish off podium in 'normal' cars
Thanks for the invite though

Cheers, perhaps after normal testing has been thoroughly run through a special Formula class handicap could be looked into.

Keep me in mind then, I am not super fast but can hold my own; Penguin can vouch for me (I hope)

I'll put you on a potential list of drivers. Please keep in mind that this testing series will be very thorough with times and commitments so that we can all benefit from the system. 👍

I dismay somewhat at what may happen when the GT5 community get their hands on it and think of it as some sort of 'bestest driverz, hottest laptimez' competition, but other than that I think it's great.

I would drive for you if you wanted, but I am not so quick these days.

I am honest when it comes to GT though.

Just GT.

👍

I can put you on a potential drivers list if you've got a few hours a week spare, VEXD?

The last thing I want is my "experiment" to be taken out of context and the whole race community thinking I've got a time trial comp going on! It really is testing and experimentation. 👍

I see it as a new model on the old school 'tuners challenge' once run by Adrenaline

It's by no means a challenge apart from the drivers challenging themselves to push for the best times in the lap allowance. It's part of the reason that drivers will be numbered and not tagged by their username, to avoid it turning into a competition and increase the likelihood of correct data.

Why must you lies.

I'll take that as a vouch for VEXD, then?

I'll give it a go!

Was in the intercontinental GT500 championship organised on here for 3 seasons as the only DS3 user, I had a win and and numerous podiums, the races were 1 hour normal and 1.5 to 2 hour endurances, so I have the experience.

I regularly win online in open lobbies, all aids off, grip real rooms!

I am also fair, clean and quick!

Ace candidate in the making, I'll put you on the potential list! Your commitment to racing series is just what's needed for this kind of thing. 👍

I don't know why you would want only really good drivers. I would imagine you'd want a large sampling across skill levels in an effort to see just how well your ballast idea works. If you ease up off of the elite-only drivers, then I would be more than happy to run those times and submit the totals to you.

I would also imagine that you are already aware of racing leagues that use ballast. I have seen two. For the next season of the one I'm running, we are going to use PP and not ballast because we can set limits on it and not on weight. Top drivers will get a couple PP knocked off, bottom drivers will get a couple PP added. Lower the PP for the room and let them enter. Then raise the PP limit and let the next batch enter, and so on. This is meant to not only compensate for different driver skill levels but allow drivers to pick cars that may normally shy away from.

We did some testing and it was fantastic fun, making much tighter racing than we had before.

In response to why skilled drivers are needed; because we're trying to set a kind of benchmark for a handicap to be judged by. Nobody compares their golfing skill to their friend's dad's, but to a legendary golfer.
A fine racing manoeuvre is referred to as "Sena-esk" because of the high levels of skill attributed to him.
This is the reason for needing highly skilled driver. Top find a high level to compare against, not a low one.

I do know of leagues that use ballasts to level the competition. They are awarded for wins and such to keep the league table even generally. They do not, however, provide a point of recognition to where I, racing perhaps another GTP member for the first time, could display my handicap and we could regulate ballast according.

The real aim, for me, is in your last sentence. Close racing through even competition.

 

[Vol Jbolaz]

The real aim, for me, is in your last sentence. Close racing through even competition.

Yes... I'm still not sure how you expect it to work? You take uber elite drivers and handicap them against what?

Unlike in golf, this handicap system isn't against an fixed point. And it is not only relative to the driver to which you are comparing yourself, but it has to factor in the environment (30 Kg may be too much when ballasting Kei cars but not enough when talking GT500).

I do wish you luck, but... I still don't see how you intend to get to a finished product.

Speaking of, since PD didn't get us a finished product, it makes any sort of regulation enforcement in a league setting impossible.

 

[uMadson?]

I'm also interested but there's nothing to back me up so I'm not entirely sure about joining

I know how to drive clean, people who tried running with me such as C-ZETA and Onboy123 would vouch for me. I was able to hold my own against a tuner too.

 

[W3H5]

Yes... I'm still not sure how you expect it to work? You take uber elite drivers and handicap them against what?

Unlike in golf, this handicap system isn't against an fixed point. And it is not only relative to the driver to which you are comparing yourself, but it has to factor in the environment (30 Kg may be too much when ballasting Kei cars but not enough when talking GT500).

I do wish you luck, but... I still don't see how you intend to get to a finished product.

Speaking of, since PD didn't get us a finished product, it makes any sort of regulation enforcement in a league setting impossible.

I like your thinking.

We many have good, skilled drivers set times in specific cars on specific tracks with the same regulations across the board. From this we calculate a benchmark by which the handicap level/line is established. With me so far? Good.

After the 'benchmark', if you like, is set the second stage of testing begins. This will then be to determine what handicap/ ballast ratio should be given where and how. Following? Good. (Beta phase, FYI)

The fixed point will become the 'benchmark' created by those that participate in the test series, hence the need for good drivers.

Once the 'benchmark' is in place the same exact test can be trialled by the/ any individual for them to see where they stand against the 'benchmark'. From that point they will either have a higher or lower handicap (still to be attributed a value) and thus be able to set a ballast or not against other driver handicaps, as the case my be.

I'm glad you brought up this point because it took this question for me to really understand the answer myself. Thanks. 👍

uMadson?
I'm also interested but there's nothing to back me up so I'm not entirely sure about joining

I know how to drive clean, people who tried running with me such as C-ZETA and Onboy123 would vouch for me. I was able to hold my own against a tuner too.

I see no reason why you cannot go on the candidates list. Welcome aboard!
(If C-ZETA and Onboy can vouch that's a brilliant start.)

 

[Greycap]

Sounds like a plan that might work really well, I'm interested. Not many people have seen me race but I made a replay archive some time back so you can take a look yourself.

 

[Vol Jbolaz]

Ah, okay, now that makes sense.

I'll be glad to help out in any of the future rounds of testing.

 

[uMadson?]

Is the driving line allowed? I'm not so sure if it's also considered as a driving aid to you, shem.

 

[W3H5]

Sounds like a plan that might work really well, I'm interested. Not many people have seen me race but I made a replay archive some time back so you can take a look yourself.

Potential list, then? 👍

Ah, okay, now that makes sense.

I'll be glad to help out in any of the future rounds of testing.

Cheers. 👍

Is the driving line allowed? I'm not so sure if it's also considered as a driving aid to you, shem.

Err......driving line........drivers discretion? (Gray Area)

I don't consider it an aid because it doesn't always allow for the best line for a drivers style. Let's say for now it's OK, but there'll be feedback on that point, for sure.

 

[ZohsixGT5]

I'd like to help with this. I have no real backup to account for my skill but on a scale of one to 10 I'd say I'm a 6.8. 1 being a 12 year old noob and 10 being Aryton Senna.

I can lap quickly without crashing, never use aids, mostly race on sports tires, and use a G27. I am not the best but I can do well.

 

[Morgoth_666]

Some input on the 'skilled drivers' issue, from a not-so-skilled driver :

It would seem to me that for the first stages of this you would want to get two things from the tests: 1 - a clear, accurate representation of driver skill, and 2 - a field where small differences in handicap can be analyzed to calculate its efficacy.

1 - Elite drivers tend to give a good indication of what they can do within a few laps in any car, whereas others may sometimes spend an hour driving before they "really nail it." Means extra variables.

2 - If this is used broadly it may often be to eliminate slight but clear differences between drivers, be it a room full of experts or an average-skilled group. I'd think it's important to get accurate data for normalizing a consistent .5 - 1.0 second difference before expanding it to cover 3-5 second variances. Drivers who's skill level is 4 seconds faster than the room often tend to get fed up quickly with the driving of the others anyway, regardless of the actual lap times.

 

[homeforsummer]

I'd be happy to participate. I hope my UKGTP showings are sufficient proof of my speed

 

[Techy]

This experimentation should be interesting. And so I would like to join

 

[W3H5]

I'd like to help with this. I have no real backup to account for my skill but on a scale of one to 10 I'd say I'm a 6.8. 1 being a 12 year old noob and 10 being Aryton Senna.

I can lap quickly without crashing, never use aids, mostly race on sports tires, and use a G27. I am not the best but I can do well.

That sounds good to me. Once a fairly extensive list of drivers is compiled I'll select the drivers who seem to have the most skill. 👍

I'd be happy to participate. I hope my UKGTP showings are sufficient proof of my speed

Great stuff hfs, once we get rolling I'll be filling in applicants with the details.

Some input on the 'skilled drivers' issue, from a not-so-skilled driver :

It would seem to me that for the first stages of this you would want to get two things from the tests: 1 - a clear, accurate representation of driver skill, and 2 - a field where small differences in handicap can be analyzed to calculate its efficacy.

1 - Elite drivers tend to give a good indication of what they can do within a few laps in any car, whereas others may sometimes spend an hour driving before they "really nail it." Means extra variables.

2 - If this is used broadly it may often be to eliminate slight but clear differences between drivers, be it a room full of experts or an average-skilled group. I'd think it's important to get accurate data for normalizing a consistent .5 - 1.0 second difference before expanding it to cover 3-5 second variances. Drivers who's skill level is 4 seconds faster than the room often tend to get fed up quickly with the driving of the others anyway, regardless of the actual lap times.

Part of the reason that test runs will be carried out over a set number of laps on each track is to establish a collection of times from that track. The HSR/Zonda combo trial is set over 10 laps so as to allow an average time from the total of laps.

A fast driver will set a fast lap almost every time round, this is the reason for the limit on laps, but there is no limit on how many sets of runs can be done on each track. I hope that makes sense, it does in my mind anyway.

This experimentation should be interesting. And so I would like to join

You guys all rock, and thanks for the support.

 

[Apok]

I'd like to participate. I did the WRS qualifier that put me in D3 gold and then did one WRS combination (yeah just one).
I don't know who would vouch for me especially because I don't do very well online usually (though I totally blame that on my lack of tuning skill).