Testing the Performance Index System (PI Rating) 🚙🏁

I figured I might as well share this in case it’s of interest to anyone.

With no more news on the next Motorsport and an itch for some track action that needs to be scratched, I decided to start lapping the Horizon Mexico Circuit in a bunch of different vehicles. What started out as a bit of fun has now become a little side-project to lightly test the accuracy of the in-game Performance Index rating.

It remains a light test as the game’s PI rating takes into account a wide range of performance parameters. The Horizon Mexico Circuit is but one of many race routes in the game and a single flying lap of the relatively short tarmac circuit doesn’t take into account performance aspects such as launch, top-end speed and acceleration, and dirt or cross-country performance. That having been said, the vehicles on my project list are all road-going performance models so the test remains somewhat relevant.

I’ve created a basic spreadsheet to record my results. Here’s a snapshot:

Working from left to right, the first three columns record the PI rating, make and model for each vehicle. The next column details the tyre compound that each vehicle is running. For the most part this column will be blank as I’ll be running the default compound (for each vehicle). (Different vehicles do, of course, come with different default compounds but that’s all part of the game.)

There are some vehicles, however, that I want to compare more directly so I’ll also be testing certain vehicles with the racing semi-slicks fitted as a control tyre (as indicated in the column). As an example of why this interests me, take a look at how these two Honda models compare:

In standard form, the 2007 models appears to be the faster car. With semi-slicks however, it’s the 1997 model that comes out on top (according to the Performance Index rating). I’m interested to see if the lap time backs this up.

Returning to my spreadsheet, the next four columns record the fastest lap (for each vehicle), the lap time split and difference in Performance Index rating (both compared to the vehicle below), and the Performance Index ratio (for each vehicle). Of these it’s the Performance Index ratio which is the most interesting. It’s a simple calculation (lap time split over difference in PI rating) but the end result shows what the value of each PI point relative to lap time. (Obviously, bigger is better here.)

The working copy of my spreadsheet is currently sorted by lap time (as I haven’t tested all the vehicles yet and I like to see where each car slots in). My results however are sorted by PI (as this gives a more natural result in the PI ratio column). As it is I’ve already started to get some interesting results. To keep things tidy, I’ll keep all the results in this first post (and update things as and when required).

1 Like

The problem with something like this is finding a representative track that covers as many car and available track traits as possible without becoming too long which then increases variance.
HMC isn’t that good imho because it favours acceleration by a huge margin. A lot of the top tuners have abandoned it for this reason.
Also for a stock vehicle testing some parameters have to be defined. Cars with shift time penalties have to be driven with MC for example.
Further the question if stock adjustable things like suspension, aero or brakes should be tuned. Imho yes because the stock car already paid the PI cost to have stock adjustable suspension for example.
I’ve tried a stock PI accuracy test in the past as well but it quickly turned into a documentation hell due to all the “maybe” parameters. Full-tuned testing was less effort and gave more reliable results.
But always glad to see the testing spirit here. So, big thumbs up for your ambition.

1 Like

Would it be ok to have this thread removed? It’s now a duplicate.

I’ve created a new one as (due to the forum update) it looks like I can’t edit this one to update the results.

Thank you. :+1:t2: