New Tuning Method?

[barumba]

That reminds me of when I (we?) first got into Forza racing. Before we got into tuning. Buy as much power as possible (remember the horrifying U999 class) and blow the doors off the competition in the straights, and wiggle waggle through the turns, only to be caught up in the twisty bits. Then blow by again in the straights. Wash, rinse, repeat. This worked on tracks with enough straight (or oval) line driving to take advantage of the power, but failed miserably on more technical tracks.
After a bit of astute tuning, even the dreaded U999's could be tamed, with better handling at the expense of raw power the pay-off.

[donpost2]

Tuning is never simple is it!! 

I like drift's idea of starting with pure handling then adding power. The question is, do you add power for top end speed (ie no weight reduction), or do you add power-to-weight (with a mix of power and weight reduction upgrades). These are two distinct build paths that would give different results on different tracks. I guess you would look to the length of the straights to judge?

The simplified way I look at is it that you are looking to balance top speed (power), acceleration (power-to-weight), and handling (handling ). I guess I'm looking for a simple, quick, methodical method to find the correct balance for a specific track, class, and car. When I tried using the AI method to evaluate tunes, I took data like the average laptimes of different builds, and the three stats of the builds and plotter graphs to search for any relationships. There were clear patterns to be seen. A bit too time consuming though on a time spent per car basis, so I won't be repeating that 

[Drift2XL]

Depends on the track you plan on running it at.

I have a slew of A Class cars just for the 'Ring. That's a power track. So they run at full weight (Maybe not the SUV's). And have minimal tires. (BTW, my fav is a RWD Gallardo S with nothing more than aero, diff, roll bars, and Dubs)

If you're looking at another track you want to match it if you're looking to do some public room racing. You certainly wouldn't built a Shelby for Tsukuba and expect to win.

In most cases, I run as high a tire as I can think I can get away with. Then move to weight redux. If it has quirks I don't like, Adjustable parts. Aero depends on balance, and grip. If a car has a good natural turn in, and has good bite coming out of a turn I tend not to run aero. Some of the 911's are rather pushy so the front aero can help them turn in. Muscle cars have a ton of power and need more rear weight (DF) to balance out the nose heavy stature.

I think there is still a track breakdown on this site on which "type" of track they are. I believe we broke them down to power, twisty, and flowing.

There are some "cheater" cars that have low drag and have better top ends than more brick like cars. Mainly, mid engined, the Ferraris, Lambos, and Vettes. And the NSX to that list too. I have an NSX in B Class (with aero parts installed) and it can hold its own on Road Atlanta versus some more powerful cars. It has a great top end. And can get there with less power. You really can build up one of these cars for a class as full grip and once you get comfortable, start to take away tire width and compound.

[Blooze]

I guess I'll scatter my two cents worth in here.  

You all know my penchant for spread sheets and record keeping - if it can be tracked, I most likely have a spreadsheet tracking it.  One of my sheets is a collection of all the stat and benchmark data for each car in a class.  The only thing not there is that silly 7.7 Speed/Handling/Launch crap.  And there is the QR system.  When the world is perfect, I have a full set of times for every car in the class that I have built, with specific emfasis on the three QR tracks, Atlanta Club, Sebring Short, and Tsukuba.  Over the years and three generations of the game this one particular item has showed up to be true.

The list of cars sorted by their QR is almost identical to the list of cars sorted by their Gs @ 120 value.

That little ditty has served me well.  When I build a car, my first priority is to get that G number up.  I add the power after I have the handling.  That proves to be valid in all but the occassional loosey goosey top build that I may put together for an Oval run - Remember the 250 mph GT40s from Forza II - that sort of thing.

Another "rule", if you will, that has pretty well worked for me is...  

Build the very best car you can, and then let it tell you where it is best to race it.

I will occassionaly bend a car toward a certain behaviour, but that is usually when I am working on a car that has particularly strong Handling numbers, or looks favorable for running on the power tracks.  But my goal is always to make that car be its very best.

So, I build a transmission to work with a particular engine/weight package.  The car will accellerate as fast as it can, no matter which track it is on.  It will reach its top end as quickly as it can.  I don't mess with it for this track, or that track.  I have found that gearing for a track almost always lends itself to gearing problems.  Watch those numbers sometime, see what happens.  You can start inching the FD up, and the Top End will improve.   You are getting more torque, the car can power through the wind...  but pretty soon, the Top Speed starts dropping, and now you are redlining before you can reach an optimum speed.  Go the other way, let the gears out.  The speed will increase to a point, and then it just sort of stops changing - no matter what you do to gears.  In fact you find it is getting slower - it no longer has the power to pull it through the wind.   You mess around like this, up and down with the FD and pretty soon you will find the optimum spot, where terminal velocity is reached as the car redlines out in 6th gear.  Now, that is the best you are going to do for that car.  That engine, that handling package.  That is the best you'll get for that.

So, that is sort of how I do it.  I go for the G's, then tune to that result.  Then it is out to the test tracks, run my QR times and find out if this is a Power rig, or if it is hell on wheels in the Twisties.  It may not be any good any where, but by golly...  it is as good as it can be.

There for a while I used to try to manipulate a car for Speed and Grip.  A little gearing and aero tuning.  But I always ended up abandoning the "Speed" version because everyplace but on an Oval, the grip build was the better car.

So, give it some thought.  How many times have you tried to shoe horn a car into a job it just can't handle?  I see those posts on the Daught Gnet...  Some kid begging for help getting his Bently into the Top 100 on Tsukuba.  "It just won't turn." he wails...

$

Oh... and to the original post.   The AI will never drive a car like I do.  Many times I wind up with a transmission that does its main work in the red, operating at the very top just clear of the Rev Limiter.  The AI will never drive a car into the red.  The AI will make a gear change when powering through in a higher gear is faster - (up the hill on Atlanta Short for example).   So using the AI to validate choices I've made for my style of driving doesn't make much sense.  However, using the AI on an endurance run is a dandy place to get out the telemetry and see how the car is working in a less stressful environment - especially with tire temps...

[chargercrazy]

When you are talking about adding grip, how do you quantify it?  For your normal "grip" builds, what are the G-forces it gives you in the benchmark menu or in the upgrade menu when you add different tires?

Do you shoot for > 1 G?  I guess I'm trying to find that minimum value that your builds must meet before you start adding horsepower or reducing weight.

[Blooze]

To the first question, it doesn't matter.  If I decide to buy slicks, all that counts is can I afford the Pi points they will cost.  The slicks improve grip, how much?  Does it matter?  I can't see myself thinking that since these slicks only improve grip by 2 tenths, I will not buy them.

That pretty well goes with all the rest of the parts as well.  If the part improves grip and I can afford it, then I buy it.  How much it improves the grip is more or less out of my hands.

How much grip?  All I can get.  Sometimes I make decisions on history - this 3000 lb car has 1.12 Gs and handles well, then another 3000 lb car should do okay with the same amount.

My personal regimine for almost all cars is...

• The Tuner's Seven
• Tire Width
• Tire Type
• Weight Reduction
• Torque engine upgrades

Now, temper it all with common sense and a gut feeling.

Sorry I can't be more "Do this and then that"  about it...

$

[chargercrazy]

So most of your cars have more than 1 G in the benchmark menu?  I guess what I'm asking is does the benchmark stats on G-Forces correlate to better grip in the game?

G's at 60 > 1.0 and G's at 120> 1.0 has more grip than G's at 60 = .75, etc.

[bimmerlovere39]

G's are an indication of lateral force when turning: a car that can pull 1.00G in a turn has one gravity of force (if you wanted to be technical, I suppose it's actually acceleration; specifically, 1g = 9.8m/s/s) towards the outside of the turn.  The higher the benchmark puts the G force at a given speed, the more cornering force the car can muster at that speed.  The two different speeds are largely there to showcase any loss or gain in grip that comes out of aerodynamics (lift/downforce).

[Blooze]

So most of your cars have more than 1 G in the benchmark menu? 

I have to look all the way up to C Class (38 builds) to find the first car with a G @ 120 higher than 1.00, and it is 1.01.  Everything else there and below is in the .9x's.   B Class is the class where the number is 1.00 or higher.  It is relative, more or less.  The car with the higher G number will be better than a car with a lower number, generally.

I guess what I'm asking is does the benchmark stats on G-Forces correlate to better grip in the game?
G's at 60 > 1.0 and G's at 120> 1.0 has more grip than G's at 60 = .75, etc.

Again, GENERALLY this is true.

Sorry to keep dancing on this, but you are looking for absolutism where none exists.

$

[feuerdog]

You guys are too analytical about tuning.

 

Seriously though, some of your methodiolgies look far deeper in to building than mine do.

For FM3 my general tunes(because I don't tune track specific very often) are built around a rough framework of simple relationships. Almost every tune is built around this generic mental formula.
1. Approx. 100mm tire sidewall height.
2. Speed rating should be approx. 20% higher than grip rating for E thru A.
3. Weight reduction and handling is a priority over power("missile" build exceptions apply).
4. Accentuate the strengths, minimize the weaknesses, ...usually in that order.
5. Let the tune/chassis decide where it performs.
6. Tire compound upgrades are a last resort for grip.

Unlike some of you I think, I enjoy the drive more than the competition to find or be the fastest.

I primarily enjoy the improvement of myself, cars, and tunes over my previous bests. Comparing myself to you guys just lets me know if I can keep up with the high standards.