Performance parts demystified

[Amaratha]

Courtesy of VLTEd 2.0. and an hour of free time

In performanceparts, 0xa9a278a9 field contains the name of the part manufacturer (check them under partmanufacturer). Nodes with "gearhead" in their node name are Gromlen/Medion/Transtek/etc., i.e. yellow (balanced). Nodes with "pistonhead" in their name are Norset/Ursa/etc. i.e. green (speed). Nodes with "rails" are Richter/Ventura, i.e. red (handling). Nodes with "smokeshow" are Dawndraft/Velocicom, i.e. blue (accel)

pvehicle node for every car has transmission, brakes, chassis, tires, induction, engine. Every one of these looks the same way, [0] is the default, then [1] is the default entry with "_h" appended, [2] has "_a" appended, [3] has "_t" appended. These must point to the real performance effects of the performance parts, as opposed to merely increasing the HAT values by a percentage (which seems to be later modified)

These must be related to the fact that almost every car-related node has three variants in addition to the parent node. (I suppose the ones that don't have the variants are Carbon/testing leftovers)

The variants are "_a", "_h" and "_t". I guess the parent node is either untuned or balanced (or maybe both), while "_a" is for blue parts, "_h" for red and "_t" for green.

Therefore we can lay to rest the overall tuning theory, the color tuning theory was correct!

Let's go with the game order.
1) engine/default node determines engine performance part's effect. The variants (see above) only differ in TORQUE.
2) induction/default. They usually differ in HIGH_BOOST and LOW_BOOST (a-t-h in descending order). Some cars have identical values, however, for all three variants.
3) transmission/default. They only differ in FINAL_GEAR - smallest value for "_t", medium for "_h" and highest for "_a"
4) chassis/default. For some cars, the "_h" variant is significantly lighter (for instance, BFH SUV). They also differ in the following:
SHOCK_STIFFNESS
SHOCK_EXT_STIFFNESS (which always seems to be higher than the other, just FYI)
SWAYBAR_STIFFNESS
SPRING_STIFFNESS
(The "_h" versions have higher stiffness)
Some cars (e.g. BFH, Viper) also differ in the following:
SPRING_PROGRESSION
AERO_COEFFICIENT (higher for the "_h" version)
SHOCK_DIGRESSION (lower for "_h" version)
5) brakes/default. They differ in BRAKES and EBRAKE. I guess EBRAKE is the handbrake (the one when you press Space). Anyhow, in most cases "_a" and "_t" are identical, and only "_h" differs. Sometimes (e.g. 300c) the game's programmers got lazy and didn't change anything.
6) tires/default retains the Carbon categories of muscle/exotic/tuner. Once again, every car has "_a", "_t", "_h" versions in addition to parent node. They differ in:
DYNAMIC_GRIP - slightly higher for "_h" variant
STATIC_GRIP - lowest for "_t", medium for "_a", highest for "_h"

---

For completeness's sake:
/This is the part I don't understand/
In partcollections, "title" field cycles every four nodes, GEARHEAD-PISTONHEAD-SMOKESHOW-RAILS. If you looked under performanceparts, they will seem familiar
stage_name field
0001-0008 STREET
0009-0015 RACE
0016-0024 PRO
0025-0032 UBER
Under hat_bonus field, every node has three values, 1, 2 and 3, but VLT only shows them as strings of numbers, which look like hex values to my untrained eyes. My guess is that they are bonuses added on top of what is listed for every car under pvehicle, but whether to individual H/A/T or to overall is anyone's guess.

[VilleroyBoch]

"Therefore we can lay to rest the overall tuning theory, the color tuning theory was correct!"

So you mean the World Garage guide and such were correct?

[Amaratha]

Originally Posted by VilleroyBoch

"Therefore we can lay to rest the overall tuning theory, the color tuning theory was correct!"

So you mean the World Garage guide and such were correct?

Well, they were not correct (and they differed quite a lot between authors), but they were better than people tuning on basis of the overall/HAT scores.

For instance, LamboGTR recommended the following for Mazda RX-08
R/G engine, R/Y induction, B transmission, B suspension, Y brakes, Y tires
Nexuo recommended
R engine, Y induction, G/R transmission, B suspension, R brakes, R tires

Based on the data, I would recommend:
engine - whichever you want, but is gonna have the most effect on the performance. I don't understand the torque data, but B is for accel, R seems the middle way and G is for top speed
R induction (Y does nothing) - curious, for this car the relationship is exactly the other way round than for the 370z I checked yesterday (h being superior to t and a)
transmission based on what you want (B for accel, R for middle ground, G for top speed)
R suspension (G/B are identical), R brakes (no listing for Y parts suggests they do nothing and G/B are identical), R tires (G/B are identical)

Actually R-R-R seems the way to go for suspension-brakes-tires for all cars (unless you've got a lazybones devs car like 300c, where it matters less)

---
I am going to do some testing today, but everything points to the parent node being untuned, and yellow parts not affecting real performance in any way.

Part stage (street/race/pro/uber/elite) seems not to matter either, based on the data in the database (or rather complete absence of any references to them apart from partcollections, which only refer to HAT). I guess their only function is to increase HAT further, bumping you into higher car classes.

---

So, I picked a VW Scirocco as the test car.

Edited the commerce.bin to make almost all perf parts have 0 influence on H/A/T hence overall. Or at least I tried to, I seem to have missed some entries
The car's overall is 213.

Untuned Scirocco = 211 km/h top
Yellow * engine = 211 km/h top
Green * engine = 213 km/h top
Blue * engine = 211 km/h (but noticeably better acceleration)
Green * engine yellow rarity = 213 km/h top
Yellow *** engine = 211 km/h
Green **** engine = 213 km/h top

EDIT: Actually that 213 km/h seems to have been a dip in the terrain.
B engine, G induction, B transmission, B suspension, B brakes, B tires come out to the same 211 km/h.
Driving is affected, as is gearbox, but top speed remains the same.

When I accidentally used red parts, which retain the HAT effects, the overall went up and so did the top speed.

So to sum up, top speed seems determined by the overall score (as someone theorized when the game was still online)
... But all the rest (gearbox, grip, stiffness) works according to the database entries. Part rarity or stage do not matter when it comes to the performance.

And yellow parts seem to be the ultimate dummies when you take out the HAT=overall=top speed effect.