GPLrank is a way of recording and comparing your Grand Prix Legends times with those of other drivers. It is also a way of organizing the many racing venues now available into orderly collections so that a driver can have a straightforward method to learning the world's racing circuits. The basic principle is a handicap system similar to the one used in golf. This means that there's a certain "benchmark" that your achievements can be compared to. The fundamental benchmark for GPLRank is the combined lap time of the replays that come with Grand Prix Legends which add up to 25 minutes and 20.22 seconds. Your GPLRank handicap is calculated as the sum of all your best lap times minus the benchmark time.
GPLRank consists of the 11 tracks that are included with Grand Prix Legends. With the introduction of other tracks several other ranks have been added.
In order to get your GPLRank rating, you must upload your laptimes to the GPLRank server. For the 11 original tracks those times are stored in the player.ini file which you can find in your player directory of the GPL installation of your local hard drive. The GPLrank system then parses your player.ini for your lap times and does the appropriate calculations.
For the addon-tracks you need to use the excellent tool GPL Replay Analyser (GPLRA). This tools allows you to analyse replays, so you can figure out why you aren't as fast as you would like to be, but it also contains a feature to scan your replays and generate an overview of your Personal Best times. On the window with your personal best times there's a button "Upload laptimes to GPLRank".
GPLRA must have the replays because there is no other record of what times were recorded. Yes, this method can consume a huge amount of disk space. However, you don't necessarily need to keep the replays indefinitely. GPLRA has a CD-burner feature that stores times in a database and lets you delete your replays from hard disk. Furthermore, you need not keep the entire replay fileonly a tiny snippet is required for GPLRA to save the times. Of course, saving those replays and using GPLRA helps you figure out where you are fast or slow, so this is probably a good thing if you're interested in GPLrank at all.
Thanks to tireless efforts of the GPLSecrets group we can now also drive the high-revving 1500cc cars of the 1965 season. It's only natural that GPLRank can process the laptimes that are set in these cars.
For the standard tracks, the only thing that is different is that the laptimes for the 65-chassis are stored in a file called plac65.ini instead of player.ini, so you must upload the plac65.ini file. For addon-tracks you should use GPLRA version 7.51 or higher.
The great GPL Online Tracks Database contains just about all tracks that have seen the light of day, including obscure Alpha versions and beta versions. GPLRank only contains the tracks that have been officially released. This is because an official track is not likely to change, so laptimes set on it remain valid, whereas pre-release tracks tend to change, which invalidates laptimes set on them.
In addition, there are a few special tracks that GPLRank doesn't contain. One is the Goodwood Hillclimb, another is the Skidpad (because you cannot actually set a time on that one).
If you have to reinstall GPL you naturally want your original laptimes back. Luckily, there's a feature in GPLRank to restore the laptimes that you already uploaded to GPLRank. Just choose the "Generate .ini file" option and you'll be given the choice for which GPL version you want to restore the times. There's even an option to retrieve the laptimes of an earlier date. You can use it in case you accidentally uploaded newer (and slower!) laptimes already.
Please use the "Self Service" feature to delete any laptimes that are incorrect. You can also clean up any spikes in the history of your handicap that a wrongfull upload has caused.
Here's how Don Scurlock explains this: "This is how I view the overall concept. Imagine all 11 tracks hooked end to end to make one long 48.484 mile long track. You have to master this track. How can you then take a 14 mile section out of the lap, and discount it? The Ring is not only 29% of the lap, it's the most difficult 29%. It doesn't matter how fast you are on the other 71%, if you're not just as competent at the Ring, your lap time, and your GPLRank score, is going to suffer, and you have not mastered GPL, overall." The same rationale applies to tracks in other ranks as well. Note that relative ratings are provided on the Show My Ranks pages, in the form of the "% of Benchmark" column, and on the '107%' pages.
No, not by employing standard html stuff. Roger Walker contributed the following solution: "I made a shortcut to my player.ini on my desktop. So instead of browsing for the sierra/gpl/....blah...blah, I just click it, and IE automatically puts the path into the upload file text box." Note: This also works with Netscape, Opera, Konqueror and Mozilla.
You must set at least one lap time on each of the standard tracks to get a handicap rating. So, even if you hate Mexico like I do, if you want a GPLrank handicap, you'll have to do at least one fast lap there in whatever chassis you choose. Then, resubmit your player.ini in order to have your handicap rating assigned to you.
If you are submitting a time to GPLRank
that is faster (i.e. better than) the current world record as seen
on this site, you will see a notification that requests you to upload the
replay to GPLRank. You must use the Upload Replay page. This page will only
allow you to upload replays for laptimes that are faster than the current world record.
A GPLRank admin will inspect the replay to ensure that railriding, warp, shortcuts, engine swappers and the like are not used to set the laptime. During the entire lap the entirety of at least two wheels must be on (or above - think of Flugplatz at the Nurburgring) the pavement at all times (every single frameyes, we look at every frame).
Every now and then we will check if there are laptimes that are faster than the current world record but the user has failed to upload the replay for. We will delete these laptimes and you will be notified via an email of the deletion.
Although GPLRank by its nature must be very liberal in terms of what's acceptable, A World Record lap time is special, and should be held to a higher standard. Far to many people are submitting lap times that are below the current World Record. We can only assume these are due to online clock smash's and kiddy/arcade style, no damage railriding. We really don't want to start dealing with replays, what we want is for people to pay attention to the validity of their Best Laps, to hold themselves to a higher standard, and delete bogus times before submitting them.
The player.ini file is to be edited ONLY TO DELETE bogus times from online clock smashes, or times acquired using unacceptable techniques such as no damage railriding, bouncing off barriers, etc., any lap time you feel you haven't driven cleanly. It is NOT to be edited to add times for any reason.
There's a saying to the effect that "Never attribute to malice what can be equally attributed to stupidity." Most weird and unbelievable lap times are caused by clock smashes during online play and much less by people actually trying to cheat. After all, what good would it be? Believe it or not, sometimes we even get mail from people who think their ranking is too high and kindly ask us to wipe the times from the database. With the huge majority of GPL players, you'll find Gentlemen both in on and off-track behavior.
Railriding, shortcuts etc, you get the idea. If you think your new personal best has been achieved to unclean driving, please do not submit it to GPLrank. If your times are suspiciously good we might even choose to delete them.
Using the engine swapper to put a F2 or F3 engine into an F1 chassis and calling the car F1-2 and F1-3 respectively is allowed. Going further and putting a Lotus motor into the Cooper chassis, etc, ect, is not allowed as far as GPLRank is concerned. Unfortunately this is difficult to detect (but definitely not impossible!), and we don't care to spend our time trying to enforce it. If the F2, F3, F1-2, F1-3, users of GPLRank are going to do this anyway, they are simply making the whole ranking system less meaningful to themselves, as they are not comparing themselves on a level playing field.
Bottom line: If you think someone's not playing fair, send a mail to email@example.com and tell us what you think is going wrong.
The handicap is the difference between the benchmark times and a driver's best times. Like golf, lower--and negative!--scores are better. As this is written, Greger Huttu holds the best GPLrank handicap at -90.25. This means that Greger's fastest laps at each of the Papyrus tracks, added together, is an aggregate 90.25 seconds faster than the Papyrus benchmark laps. Another user has a handicap of -0.008, faster at some tracks than the benchmark and slower at others.
Once the handicap is computed, it becomes possible to rank drivers. Greger Huttu's rank is 1: he is currently the fastest driver at the original 11 tracks. The author's handicap of +2.15 yields a rank of 1212; there are 1211 faster drivers.
Handicaps are computed for each set of tracks; handicaps and ranks for each set of tracks are independent.
A regular rank uses your best time from any chassis. It is common, for example, to set a best time at Spa in an Eagle but a best time at Monaco in a Ferrari or Brabham. A Monster rank handicap is the sum of your best times at each track for each chassis! (See, it's called Monster for a reason!) Monsters of GPLrank was implemented somewhat after the original GPLrank, and ChallengeRank now has a companion Monsters of ChallengeRank! Much fewer peoplehistorically only about 10%complete a Monster handicap than an equivalent regular Rank handicap, but it adds interest and a further challenge over and above the already difficult Rank goals.
The only difference is the source of the tracks and benchmarks that make them up. The classic GPL handicap and rank are based on the original eleven tracks that shipped with Grand Prix Legends itself. The benchmark times are exactly those of the reference replays provided for each track in the original distribution. For example, the replay at the Nurburgring shows a lap taking 8:21.930 seconds, so that is the benchmark time for the Nurburgring. In general the Papyrus benchmark times have proven to be about 104% of the world records.
For the other ranks, the source of the benchmark times is quite different. Rather than taking a single replay lap from each of the constituent tracks, the benchmark times were set based on the 80th percentile of already recorded best laps in the database. In other words, the benchmark time for each track is set at approximately the time that was faster than 80% of the previously recorded best laps.
The other ranks are here precisely because GPLrank is so successful. After hanging around the GPL community for a year or more, it became obvious that a lot of drivers were organizing their GPL activities specifically around GPLrank. Drivers allocated their time to the eleven Papyrus tracks specifically because additional practice there improved their GPLrank handicaps. Many leagues have organized their schedules around this as well. This is unfair to the many track editors who have built truly masterful worksand yet very few people drive on them, at least in part because GPLrank provides a strong motivation to stick with the Papyrus tracks. The last straw was a discussion about a league schedule that went along the lines of this:
We need to put Mosport and the Monaco on the schedule, so Le Sarthe and Solitude don't fit.
Why do we need Mosport and Monaco?
Because they're in GPLrank.
So what? Le Sarthe and Solitude are more fun.
Yeah, but more people will drive the GPLrank tracks.
OK, enough. The point isn't whether or not one agrees with the choices, but the rationale that drove the decision. To give an idea of how overwhelming this disparity is, at the time that we decided to create the first rank for addon-tracks (which was Challengerank), there were about 40,000 best laps recorded for Monza and over 22,000 at the Nurburgring, compared with less than 1,000 at the most popular of all add-on tracks. We hope that the extra ranks will motivate people to drive addon-tracks as well as the standard tracks.
It's a hack--but a very useful one. It's actually not a "Rank" in the strictest sense, because there is no benchmark time for the Rank as a whole, nor are there benchmark times for most of the member tracks. What TotalRank does do is provide a convenient means for seeing all tracks at the same time. This is especially useful for comparing back-to-back with your buddies, because it lets you compare your times against your buddies' even if the track in question is not a member of a traditional Rank. A number of drivers asked for this so that they can evaluate their chances when their league goes to a non-Rank track.
Although it shows up as a Rank in the displays, it probably always shows up as "Incomplete". That is, unless you have at least one laptime at each of the tracks that are in the GPLRank database, in which case your TotalRank will be the sum of your fastest laptime at each track. Mind you that TotalRank will change at a moment's notice, simply by someone releasing a new track. As such, "a TotalRank handicap" is pretty meaningless, even if we would have a benchmark time for every track.
TaintRank was a companion site to GPLrank that was devoted to add-on tracks. It never had quite the following that GPLrank did, for a variety of reasons. When ChallengeRank was implemented, we used the TaintRank database and merged the TaintRank and GPLrank sites together. There was never a Rank in the sense of being a set of tracks with a defined benchmark.
The version numbers are there to accommodate the future. The original Papyrus tracks don't change, nor are there more of them. However, at the time of this writing, there are in excess of 140 add-on tracks with another 120 or so in various stages of development. Versioning the benchmarks permits an orderly transition in the event that something must be modified. For example, we might set a benchmark time that is either impossible or so easy that it distorts the overall handicap scores by making everyone negative. It is also possible that an add-on track might become unavailable, or that it might otherwise become unsuitable for widespread use (for example if it became incompatible with new hardware). Another scenario is the arrival of an updated version of a track that is already in a Rank. For example, ChallengeRank originally would have included the Osterriechring. However, Zeltweg 1970 is the same basic track but with different distances, cambers and grip levels. Now that the much superior Zeltweg has been released, nobody drives the original Osterrichring; had it been included, a ChallengeRank v2 probably would have been required just to keep up with the track release state.
Although GPLrank is shown with a version number, we guarantee that GPLrank itself will not change.
League Central is a way of indicating your league affiliation in such a way that GPLrank can provide interesting information. The latest version of League Central, written by Ed Hut, permits management of leagues without the intervention of the (limited, and distinctly non-scalable) group of GPLrank administrators. A league administrator can set up the schedules, add or delete drivers, etc. This is a useful function on its own, but it is even more useful when League Central and GPLrank cooperate on a feature called Big Brother.
Many GPL servers automatically store replays of all races run on their systems. Some of the more popular dedicated servers export these replays to LC/BB. Big Brother is a facility that might be likened to an automated pit reporter that visits all of the testing sites. This is similar to what you read in the news about the real F1 guys testing at Fiorano or Barcelonayou get an indication of how many laps the driver ran in each chassis, and their fast lap in each session. This way, league members get a chance to see how the competition is stacking up.
Actually, quite a few sick minds thought this up! Uwe Schuerkamp first implemented GPLrank, League Central and TaintRank in early 2000. Don Scurlock originally conceived the notion of a benchmark consisting of the sum of references times at a set of tracks that made GPLrank a success. Edo provided the first GPLrank host server. Ed Hut completely re-implemented League Central and built Big Brother, as well as lending expert advice on the operation and optimization of SQL. Brian Wong extended the notion of a Rank to the add-on tracks and implemented the GPLrank/TaintRank merge, as well provides the current GPLrank hosting facilities.