Ai

  • Thread starter all4golf
  • 38 comments
  • 5,781 views
Yes the AI pit but they don't pit as often as most of us humans have pit. They seem to be able to go further on the same tires but as you would expect, that is negated by our ability to drive far faster than them. :mischievous:

If you are having trouble winning races where the AI don't pit or don't seem to pit then let us know and we will all help as well as possible. :cheers:
 
But isn't the car the AI are using also have an effect whether they pit or not, or after how many laps they pit?

For example, a bunch of Ford Ka's don't have too much wheelspin and have little power, but if the last AI competitor was using a Shelby Cobra, there's more power, wheelsin, and "drifts" maybe.
 
ImprezaAddict
But isn't the car the AI are using also have an effect whether they pit or not, or after how many laps they pit?

For example, a bunch of Ford Ka's don't have too much wheelspin and have little power, but if the last AI competitor was using a Shelby Cobra, there's more power, wheelsin, and "drifts" maybe.
Click to expand...


This is true, But It works the same exact way for humans so...
 
The AI's tyre wear characteristics for the car they're driving is exactly the same as it is for us - that is to say, if a certain car uses tyres twice as quickly for us, it will do the same for an AI driver. The difference is that the AI's driving style results in a very low level of actual tyre wear. It is possible to make a set of tyres last as long as the AI do, but then you'd have to go as slowly as them... :P

DE
 
In my experience the tyre wear difference between myself and my friendly AI driver - b-spec Bob, seems to depend on what car we're in. In some cars I get more distance as well as more speed than him, in other cars I get less distance but more speed. Also, when competing in the, um, that FGT world championship series, I found I made the tyres last longer than the AI competitors could. However, in the GTWC, when I was in a Jag XJR-9, there was an AI jag in the race as well, and he seemed able to make his tyres last a little bit longer than I could. Unlike the FGT cars though, I did not know if the AI in the Jag was on harder tyres than me, or the same compound.
 
The reason the AI tires dont wear as first is simple, they drive like sissy's! When I drive, I come very close to a corner before braking, HARD!Then I only give it half gas in the middle of the corner, then put my foot hard down on the exit. Now from what I know, this causes more tire wear as the tyres take more load. But the AI is slower. They lift off the gas, then only brake half strength before the coner. So as you can tell, their sissy driving goes easy on the tires. Someone should teach them to drive like real racing drivers, FAST!
 
They also have traction control set at the default levels, which extends tire life at the cost of speed.
 
Dark Elite
The AI's tyre wear characteristics for the car they're driving is exactly the same as it is for us - that is to say, if a certain car uses tyres twice as quickly for us, it will do the same for an AI driver. The difference is that the AI's driving style results in a very low level of actual tyre wear. It is possible to make a set of tyres last as long as the AI do, but then you'd have to go as slowly as them... :P

DE
Click to expand...
So a tuned Corvette Grand Sport can do 52 laps???
 
But, if you get to the big races in the game, the pit strategies of the AI are weird. For instance, the GT championship, the R92cp and R89c have the craziest of them all. Ive seen them pit on the last lap, or the lap before. The rest of the competition will pit on lap 7 and they will pit on 5 and 9. I don't get it.:confused: :boggled:
 
The AI has no idea at all how long the race is. They just know when the tires are cooked, and they will pit in even if it's the next to last lap. I don't think I've ever seen them pit on the very last lap - yards before the finish line - so there must be some code that stops them from being that stupid at least.

If it was smart, they'd do a quick calculation at the beginning to decide if the tires would last the whole race, and if not, they'd pit at the halfway point. But that takes more coding rather than just monitoring the tire condition by lap.
 
I never have an idea how long the race is because the picture seems to be to tall for the screen, so all the time the top of the picture goes of the screen and you can not see it. It does not bug me on GT4, because I do not play it because the disc is screwed up, well actually I have 2 of them and they are both screwed. Anyway, when I play the other GT's, I just have to trust my judgement!
 
Dark Elite
It is possible to make a set of tyres last as long as the AI do, but then you'd have to go as slowly as them... :P

DE
Click to expand...


Even longer while still driving a faster pace then the AI cars, at least from my experiences. My last Le Sarthe enduro was one I did where I kept it easy, no seriously fast lap times (3'12 and sorts of that). I used an Audi R8, the AI R8 pits every 8 or 9 laps, if I remember correctly. Somehow, I managed to stay out 10 laps while still be faster on both the in- and outlap, even it were only 0.3 - 0.5 seconds :cheers:
 
Dark Elite
If driven slowly enough? Yes.

DE
Click to expand...
:confused: Then why doesn't bob do 52 laps, but the AI could? They drive the same if they have the same car, and pace. He wasn't driving that slow either. :confused:

The thing weights 1500Kg or so, with approxamately 450HP under the hood, since the AI modified it. I can try everything on the same tires while still trying to win, and end up pitting in less than a quarter laps that it took him to pit. There must be a mistake PD made with this, or didn't program right. GT3 had the ability to show the AI's tire wear, so they programmed it right there. Some cars had grippier tires, meaning more pits. That is why the Altezza LM Race Car on RSS tires pits in 2 laps on Deep Forest. If we can't see the tire wear on the AI, they must had figured that they didn't have to slap on that much work into it. :guilty:
 
Dark Elite
Bob works differently to the opponent AI, as he develops and they don't.
Click to expand...
True.

In addition, when using B-spec, both tire wear and fuel mileage will change rather significantly depending on what level of aggression you have your B-spec driver set at.

If you have it set at 4, you'll find the tire wear and fuel usage will increase significantly then if you have it set at 1.

Tire type of course plays a huge roll in tire wear, and the AI uses different tire types for different races. So if you are not using the same tire type as the AI, obviously you will see a difference in tire wear. If you are not aware already, the way to check to see what tire type the AI is using in a race is to first select the "Preview Race" option. At the start it will tell you the HP and tire type of the last AI car in the lineup. All the AI cars in the same race will be using the same tire type.

In addition, if your car is modified, tuned, and set-up differently then the AI, then you'll also more apt to see differences in tire wear and fuel consumption.

As has been mentioned, as long as the cars, mods, tires, tuning, set-up, and driving style is the same, then regardless of whether you, your B-spec driver, or the AI is driving the car, you'll all have the same rate of tire wear and fuel consumption. 👍
 
The AI drive the same as Bob with 0 skill and the pace set at 3. No car can do 52 laps while racing on El Capitan in GT4. (Or was it Laguna Seca?)
 
When you find out what race it was where an AI driven Corvette Grand Sport drove 52 laps without pitting let us know. The AI will have nearly identical race performance (+/-1%), so with that race info we can run it ourselves to see how it managed to go 52 laps.
 
racerx510
No car can do 52 laps while racing on El Capitan in GT4.
Click to expand...
You should stop making generic statements like this, I'll bet the Diahatsu Midget I won't wear out its tyres in that time... :P

DE
 
I believe racerx510 is discovering he was very much mistaken about his claim that he saw an AI driven Corvette Grand Sport on either El Capitan or Laguna Seca go 52 laps before having to pit.

In the event he has not, I'll help him out by saving him some time...

The AI can use the Corvette Grand Sport in the three 200 mile endurance events:
  • Laguna Seca 200 Miles
  • El Capitan 200 Miles
  • New York 200 Miles
  • In the Laguna Seca race, the Corvette Grand Sport the AI uses is modified and tuned to 501 HP and equipped with RH (R2) Racing: Hard tires.
    On average it will pit every 40 miles due to tire wear. Each lap is 2.24 miles, thus it will pit four times, once every 18 laps.

  • In the El Capitan race, the Corvette Grand Sport the AI uses is modified and tuned to 496 HP and equipped with SM (S2) Sport: Medium tires.
    On average it will pit every 39 miles due to tire wear. Each lap is 2.97 miles, thus it will pit four times, once every 13 laps.

  • In the New York race, the Corvette Grand Sport the AI uses is modified and tuned to 501 HP and equipped with SH (S1) Sport: Hard tires.
    On average it will pit every 57 miles due to tire wear. Each lap is 2.59 miles, thus it will pit three times, once every 22 laps.
Seeing as the AI pits for the fourth time on lap 52 in the El Capitan race, it appears you simply didn't notice it had pitted three times before.

BTW: This is a good example of how having a highly modified and tuned car can strongly effect tire wear, as well as fuel consumption.

For instance, if you give your B-spec driver a stock Corvette Grand Sport which comes with SM tires, the same as the AI in the El Capitan race, enter that race, and keep the aggressive setting to 3 and turn on "Overtake", you'll find that even though you are using the same tire type as the AI, your B-spec driver will be able to go 19 laps, nearly 50% longer than the AI, before having to pit... of course seeing as it will also have nearly 200HP less than the AI, your B-spec driver will be well back of the AI's Grand Sport, and the other cars in the race. :)

The point is, GT4 is very detailed when it comes to tire wear and fuel consumption, and that besides just the difference between tire types. Mods, tuning set-ups, driving style, and the type of track will also impact both tire wear and fuel consumption. This is one of the features in GT4 that really impressed me. 👍👍

This is especially true when compared to other similar style driving games where tire wear and fuel consumption is not affected in a realistic way, as it is in GT4.

Unfortunately, this topic brings up one of the things about GT4 that I am most disappointed in, and that is the nearly complete lack of specs for the AI cars you are up against. All they tell you is the HP and the tire type! Using this as an example, try and figure out how the AI modded the Corvette Grand Sport to get 501 and 496 HP. Trust me, it isn’t easy. 👎

In addition, there is no way of telling with any certainty what other modifications the AI have added to their cars that have no impact on HP but do impact performance, like brakes, drivetrain, suspension, weight reduction, etc. 👎

Adding to this problem is the fact that even the HP of the AI cars is hard to find out because they only list it for the last car in the field when you select “Preview”. This means it could take several dozen times of starting and stopping a preview before finally getting all the HP stats for a race with only 15 or so cars that the AI can use in that race. 👎
 
What car was it that took 52 laps though, if you can tell me, I don't have GT4, since I sold it for GT1, 2, and 3. I know there was that car that took only one pit stop in an endurance race, on lap 52. I could've sworn it was a corvette.
 
racerx510
What car was it that took 52 laps though, if you can tell me, I don't have GT4, since I sold it for GT1, 2, and 3. I know there was that car that took only one pit stop in an endurance race, on lap 52.
Click to expand...
I'm afraid you are definitely mistaken. [EDIT] With the exception of the 150 Mile Super-Speedway race[/EDIT] There is no AI car that pits only once in any of the endurance races, let alone on lap 52 during the El Capitan race.

I could've sworn it was a corvette.
Click to expand...
You may have sworn, but you were still mistaken.

It very likely did pit on lap 52... but it was it's fourth pit stop... 13 + 13 + 13 + 13 = 52



And it's not just the Grand Sport.

Here is the performance data of all the AI cars in those 200 mile endurance races:
You can find the following data and more within those links:
  • AI Ranking
  • Best Race Time
  • Average Race Time
  • Best Lap Time
  • Average Lap Time
  • Total Number of Pit Stops
  • Average Laps per Pit Stop
  • Average Miles per Pit Stop
  • Fuel Used per Pit Stop
  • Total Fuel Used in Race
  • Fuel Mileage
The Pagani used by the AI in the New York race comes the closest thanks to using SH (S1) Sport: Hard tires, but even then it can only go 30 laps (78 miles) before pitting, and there are absolutely no AI cars that can finish these races or any of the endurance races with just one pit stop... [EDIT]except for the 150 Mile Super-Speedway race thanks to the use of Super-Hard tires, and banked oval course.[/EDIT]

The fact that you strongly believe it was the Grand Sport, in the El Capitan race, and the stop was on lap 52 suggests that you were simply mistaken about it being its first stop, and when it pitted on lap 52 it was in fact it's fourth pit stop of the race. 👍


BTW: if you want to see some of the data I have collected from some of the other enduro races, or the 1000 Miles event, you can find links to them HERE.
 

Similar threads

Fearpinkmist
Pit stops and missing rewards
Replies
4
Views
679
jrbabbitt
jrbabbitt
B
How do I make my pit crew faster?
Replies
17
Views
5K
TUSMBOX
TUSMBOX
N
Le Mans 82 AI
Replies
0
Views
2K
Nosferatu2022
N
UnkaD
AI in the rain
Replies
22
Views
8K
05XR8
05XR8
Rikir98
Doubt about b spec mode
Replies
1
Views
2K
BobK
BobK

Latest Posts

Back