GP3 Calculator

Hi all. This might be of interest for some. It's a spreadsheet calculator that gives values that you can enter into GP3Edit, based on laptimes from real qualifying and race sessions.

Use the built-in mini-calculator to insert the fastest lap for any individual driver and the best time in the session, then put the result to the correct place in the light orange fields. The yellow fields will then give the recommended value if there is enough information.

Race performance is influenced by the qualifying performance as the fastest laps in the race have much more variance than those of qualifying. The calculator assumes uneven results are mostly the fault of the car and will influence that team's reliability. I used median instead of average for most calculations which gives far better results as one bad result doesn't drag the performance down too much.

The file is in Excel's .xls format for best compatibility (I used OpenOffice). Here is the download link for the 2009 season (copy the perf file from the sheet and save it in GP3Edit immediately after loading it there to make it work with GPxPatch):
[www.mediafire.com]



Edited 43 time(s). Last edit at 07/12/2009 04:51PM by Pete79.

Hi Pete,
your tool is interesting. Right now I am testing it.
Few comments:
- given that now Q3 in qualifying is influenced by fuel load and limited to 10 drivers, I use times from Q1. Do you do the same?
- unfortunately Q1 is close, but not accurate in terms of real order
- I guess you don't use race laps times from wet races, as they can be not representative?
- do you use race laps times of drivers that were DNF after few laps? Same as above

Pep

peppekk Wrote:
-------------------------------------------------------
> Hi Pete,

Hello. Thanks for your comments.

> your tool is interesting. Right now I am testing
> it.
> Few comments:
> - given that now Q3 in qualifying is influenced by
> fuel load and limited to 10 drivers, I use times
> from Q1. Do you do the same?
> - unfortunately Q1 is close, but not accurate in
> terms of real order

I put in all performances from Q1, Q2 and Q3. If there was only one event and a driver participated in all three qual sessions, the worst and best performance would be eliminated and the middle one used for calculating the performance values. In a typical qualifying event, for most drivers in the top 10 the worse result between Q1 and Q2 is used and Q3 ignored, whereas for the light runners the better result between Q1 and Q2 is used while Q3 is now eliminated as the best result.

After Canada, drivers who have participated in a total of 21 qual sessions get their best 10 and worst 10 results nuked and the middle one is used to determine their performance.

> - I guess you don't use race laps times from wet
> races, as they can be not representative?
> - do you use race laps times of drivers that were
> DNF after few laps? Same as above

I use both of those as really bad results are eliminated, the penalty being getting a strong performance negated with no massive hit to the overall performance of that driver (unless there are only very few results in total in which case I ignore the bad result for the time being but put it back later when he gets more decent results). For example, Sebastien Bourdais has these race performances after Canada: 2,30 2,42 5,30 2,59 31,90 1,59

The effective performance is the average between 2,42 and 2,59 after the two best and two worst are eliminated, which is then compared to his qualifying performance (+1,89% after Canada). His real race performance is closer to his qual performance than what his absolute lap times in the races would imply instead of just the 2,505 to reduce the huge variance races produce. In addition, there are some more complex calculations to adjust the race performances between team-mates.

Basically, I just go to the official F1 website, get the qual and race laptimes and chuck them in the calculator, having adjusted it to produce sensible values despite a wild weekend in real F1.



Edited 14 time(s). Last edit at 06/19/2008 01:50PM by Pete79.

you should post this in the gp4 forum too, nice tool

You could create a sheet where it posts all the data in the performance file so all you have to do is just save that sheet as a txt file rather than punching in each value.

Original file updated, use the link in the first message.

fongu Wrote:
-------------------------------------------------------
> You could create a sheet where it posts all the
> data in the performance file so all you have to do
> is just save that sheet as a txt file rather than
> punching in each value.

It is now possible, thanks for the suggestion. I haven't used performance files at all, just saving things straight to one file in GP3Edit and then to the game, but this does save some work.

Copy the relevant part from the bottom of the sheet and save to a text file, then load into GP3Edit. If you use performance files through GPxPatch, remember to save the perf file again to remove spaces around the values (GPxPatch doesn't seem to like them and I couldn't remove them in the sheet).



Edited 4 time(s). Last edit at 06/21/2008 11:34PM by Pete79.

would this work for GP4?

I have edited the download link in the first post, with the season 2009 drivers and teams in place. Slight updates have been made to some of the calculations in case you downloaded this last year. In qualifying, I now input Q1 results for all drivers who didn't get to Q2 and Q2 results for all others, as opposed to all results from all sessions last year. Should a driver fail to get a time in Q2, I will use their Q1 performance.

As more data is being made available, results will be more and more accurate. The performance file can then be copied from the sheet. For those interested in this, download link in the first post.



Edited 6 time(s). Last edit at 03/31/2009 06:27PM by Pete79.

Wrong driver numbers for Force India and Brawn GP, the performance file
is a bit obscured IMO, Fisichella 16351 and Vettel only 16265, not realistic
at all for me.



Happy racing from Germany !!!

>Wrong driver numbers for Force India and Brawn GP

Fixed. I had earlier info.

>the performance file
>is a bit obscured IMO, Fisichella 16351 and Vettel only 16265, not realistic
>at all for me.

They are compared to their team-mates, not to each other. The performances are accurate according to GP3Edit within 0,01% when looking at the whole field. It would be completely illogical to do this any other way. I see little reason to try to estimate who is better than who if they drive for different teams.

Hope this helps.



Edited 1 time(s). Last edit at 03/28/2009 02:07PM by Pete79.

It makes no sense for me to compare only the team-mates, if you want
a REALISTIC performance file, you MUST compare drivers to each others,
look at other performance files and you will see they do so.



Happy racing from Germany !!!

why brawn gp has lot more bph than mclaren? button is going to win the race lapping everybody

-------------------------------




Edited 2 time(s). Last edit at 03/28/2009 01:46PM by R_Scandura.

>It makes no sense for me to compare only the team-mates, if you want
>a REALISTIC performance file, you MUST compare drivers to each others,
>look at other performance files and you will see they do so.

You seem to misunderstand what it is I'm presenting here. This is a calculator, not a performance file. You can calculate any season if you put in the values and change the names in the sheet.

>why forse india has lot more bph than mclaren? Fisichella is going to win the race
>lapping everybody

They have a lot less BHP, actually.

EDIT: I see you meant Brawn. Yes, they are a lot better than McLaren based on one day of qualifying. GP3Edit has completely identical relative lap times to what happened today, therefore there's nothing to change. Is McLaren actually better than that? Maybe, but try to put it in a formula. Again, this is a calculator. It crunches numbers, nothing else.



Edited 3 time(s). Last edit at 04/17/2009 02:16PM by Pete79.

meant to say Brawn Gp, maybe for the performance you should wait the real race and combine the data, aerodynamic problems could be simulated giving lewis less grip but that bunch of bph makes the whole thing weird

-------------------------------

Excuse me, Pete79, a little misunderstanding.


Happy racing from Germany !!!

No problem. The real effect of this calculator is usually seen only after a couple of races has passed and the perf file has evolved a bit.

For now, it is not possible yet to use anything else than the results from today's qualifying session as it's the only serious event completed this season. Therefore Brawn is well in front until they prove to not belong there (they seem extremely impressive though). Same for McLaren, except that they need to prove the opposite.

I really like this idea, but regardless of the state of the season now, there are clearly some flaws. I hope that once I get over my beer intake, enjoy the race this morning, and then sleep for a bit I will remember to come back here!

----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
The test of a first-rate intelligence is the ability to hold two opposed ideas in the mind at the same time, and still retain the ability to function. -- F.Scott Fitzgerald

2008 data now available for an example of what a full season inserted into the calculator looks like. See first post.

I don't understand this... How can I geth this to work in GPxPatc? on GP3

The performance file is now inside the archive. Check the download links in the first post.



Edited 1 time(s). Last edit at 04/17/2009 02:17PM by Pete79.

I never did reply, which probably means I have spent the past two weeks permanently drunk!

However, my main concern would be with regards to BHP, which I believe shouldn't play too much of a factor in overall performance. Assuming we are working with 1.13 in defeault form, the most powerful team should have 800bhp, and the rest relatively lower depending on whatever genuine figures can be attained. Perhaps it is just a personal prefernce, but I like to have, for example, all Toyota powered cars with the same bhp, all Mercedes cars with the same bhp and so on. I don't see any sense in the example that was shown above where Brawn Mercedes is much more powerful than McLaren Mercedes.

By standardising bhp values (if you know what I mean) you are left with less variables to mess up. The performance values for the driver would then simply be a combination of driver skill and chassis efficiency. Ability range is a whole other thing and I confess not something I have ever really been fully comfortable with. Right now I have all drivers set to 128 in my 2009 perf file, which is decent enough at this early stage of the season.

----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
The test of a first-rate intelligence is the ability to hold two opposed ideas in the mind at the same time, and still retain the ability to function. -- F.Scott Fitzgerald