PI Rating vs. Vehicles Statistics vs. Actual In-Game Performance

Forza Horizon Forza Horizon 5 FH5 Discussion
1

These are two tuning setups which I’ve created for the same vehicle.

Both have been created for road-racing events. The setup on the left has been converted to AWD and uses the rally compound. The setup on the right remains FWD and uses the semi-slick compound. Both setups use front aero and both setups use the standard engine (albeit with different mods due to the PI cap). The left setup does use race brakes whilst the right setup does not.

The mechanical statistics (i.e. the power, torque and weight) are very similar.

The power to weight ratio for the left setup is 317.24 hp/tonne.
The power to weight ratio for the right setup is 308.27 hp/tonne.

Yet the resulting in-game statistics are totally different.

With everything added up, the left setup is rated more highly by 10.4 points.
Side-by-side, the left setup looks far superior.

But before we jump to conclusions, take a look at the race results.

Around Horizon Mexico Circuit (a 2-mile tarmac race track), the FWD setup was nearly a full second quicker (on a 1 minute lap). I also ran a Sprint layout to test both cars from a standing start and the FWD setup took the win again (by nearly 4 full seconds on a 4 minute sprint).

This is by no means a comprehensive test but something is clearly wrong here.

For the in-game statistics to come to such conclusions, for those conclusions to be so wildly different, and for those conclusions to then be in total contradiction to how things actually play out…

…it might be time to re-think the PI system (and other associated ratings). :+1:t2:

Screenshot 2022-06-02 at 21.55.11
Screenshot 2022-06-02 at 21.55.111548×1520 1.35 MB

2 Likes
2

Check my post on all things being equal.

Yes, you’re right. There’s definitely a failing in the algorithms used to determine vehicle performance. Just from simple observations of engine building (I need to actually gather numbers but it wouldn’t make any difference) it would almost seem as if the numbers applied to engine build items (ie cams, valves, etc) are just randomly or arbitrarily applied. A camshaft applied to one l4 engine doesn’t have the same impact on performance as the same level camshaft applied to another l4 engine. There’s definitely inconsistencies in the game.

1 Like
3

Further to my post above, I’ve now tested both setups more fully.

I chose to ignore Rivals mode this time and, instead, test things in Online Adventure. I wanted to see how the setups compare in the heat of the moment, when you can’t always take your ideal line and when you need to adapt for other vehicles on the route. In reality this does stack things in the AWD setup’s favour. For circuit races I was interested in both the lap time and the overall race time. For sprints it’s just the race time and I made sure to record comparable runs for each setup in the interests of fairness and accuracy.

As it is, the FWD build was quicker on 16 of the 18 routes that I tested. (There are one or two more routes in the Online Adventure rotation but they just didn’t appear for me.)

That’s in real races against real players and taking into account the difference in launch. The AWD setup does get off the line quicker but at some point the FWD build is catching it up and beating it to the finish line.

Which leads to a bit of a dilemma…

Going by the PI rating, the setups should be even. They’re both A800.

Going by the in-game statistics, the AWD setup should be much quicker. The stats say it should launch, go and stop more quickly without any drop in cornering ability or top speed. There’s no argument here in favour of the FWD setup.

Going by actual performance though… it’s a totally different story. The FWD setup is quicker on 16 of the 18 routes presented in Online Adventure.

So…

… the actual performance then is at odds with the shared PI rating.

…the actual performance is at odds with the game-generated statistics.

And even if you remove the player and in-race performance from the equation, the difference in statistics is completely at odds with giving both setups the same PI rating. The statistics say one car should be much faster. The PI rating indicates they should be the same.

Nothing here matches up.

Which points, in my opinion, to a clear need to revise and update the in-game rating system. You simply cannot expect accurate balancing of performance from a system which, fundamentally, appears to be at odds with it’s own rating system and with how in-game vehicles actually perform. This extends from the balancing of completely stock vehicles all the way up to meta-tuning and vehicles which exceed the performance expectation.

#BetterForza :+1:

2 Likes
4

I appreciate the time and thought you put into these studies, however, I don’t see the PI assignments changing any time soon. That isn’t to say they shouldn’t, but it would render every tune made in the game less useful or even downright useless. Hopefully they’ll learn though and adapt/adopt a more “equal” value to everything. I can’t imagine this is an easy process, given the infinite number of variables. Still though, I’m optimistic that a change will be made by no later than FH 324, (perhaps FH 323, (fingers crossed).

2 Likes
5

Another serious shortfall of the PI system is vehicle type and size. I just completed the German Quality seasonal. There was the M-B truck competing in with purpose built sports coupes. And coming in 2nd to 3rd place. I don’t care if it is A800, there is no physical way it can compete on a level with the vehicles in that race.

So there is no way for the PI system to account for vehicle size and type. A truck, 4x4, sports coupe, sedan, sports car, or purpose built racecar all have the same chances given the same PI.

It’s not a bug in the program it’s just a nature of the beast. Unless they redesign games to include VI for vehicle index, and SI for size index. Or maybe WI for a weight index class.

With new consoles having more power it’s possible for game developers to take these games to the next level. And we could actually start seeing them become truly sims rather than just games.

1 Like
6

Class based racing. Pre-built kits but with tweakable parts. You bought an old BMW Coupe. Cool. Now choose the touring car or rally package. No crazy drift transmission or ferris wheels here. xD

1 Like
7

I always thought the tuning was messed up anyway. The programmers aren’t good enough to get it right.

1 Like
8

Every almost every new version of NFS that comes out, the tuning, the painting, is different. Sometimes better, sometimes not. Sometimes change is good.

I horizon, FH4, the paint shop and the tune shop had problems. As they just did a cut and paste and tweek to make FH5 from FH4 so we could import our paints and tunes they also imported the shortcomings of FH4 rather than trying to make improvements. I would much rather have lost all my FH4 work, started all over in FH5 with a newer, better, system.