AI in the wet - all circuits

[Faya 1,189]

Hi all,

Please note I've removed a number of posts that were either off topic or arguments. The only reason this thread hasn't been outright closed is because I know that this is a hot topic at the moment and I want to continue to hear your thoughts and findings on the matter. 

Please stay civil. If you can't get on with someone, stick them on your ignore list and keep them there. 

 

[joserdcosta 52]

I think we shouldn't discuss if people are good or not in the wet. The whole point for me is: if I can keep up with the AI in the dry at a certain level (in my case 100%), I think I should also be able to do the same in the wet. People can say that maybe we are not good enough in the wet, but having to drop the AI level 10 clicks is not correct IMO. It wasn't like this in previous games.

That's the whole point. Adjustments. 

Cheers

[1henryw 113]

I have to drop the AI from 100 to 70 to stay competitive in the wet. Partly because of the wet and partly because I lose WAY more grip on a half dry track as the AI do so I get messed up by changing conditions either way.

[joserdcosta 52]

Right. I think you and the vast majority of players here have to do the same. I'm sure codies will have a look. As last resource I think it's better to leave AI as it was before the 1.06 patch.

Cheers

Edited August 13, 2019 by joserdcosta

[SIMRACER123 355]

1 hour ago, Faya said:

Hi all,

Please note I've removed a number of posts that were either off topic or arguments. The only reason this thread hasn't been outright closed is because I know that this is a hot topic at the moment and I want to continue to hear your thoughts and findings on the matter. 

Please stay civil. If you can't get on with someone, stick them on your ignore list and keep them there. 

 

I'm not entirely sure why this thread would be considered for closure.

Its a simple case of one person trying to undermine your entire product based on a poorly thought out test.

People on here are merely pointing out his test is not carried out in a conclusive way and how he needs to go about finding a more conclusive result.

I would have thought it would be in your companies best interests for @Striker_703 to understand this to then complete a proper test to gather data for your company.

Closing threads isn't ever going to assist Codemasters solve any issues with their game that may or may not exist with their game.

Simply closing the thread really does send the message codemasters really aren't bothered about any potential issues.

Personally, i don't think this thread had ever desended into inappropriate arguments. It was just strong opposing views.

[Worntoathread 865]

@Striker_703 

I did some testing at 110 AI around Spain, heavy rain, short qualifying.  

The AI did a 1:30.197

Want to know the world record time trial Lap for Spain in the wet?  It's a 1:30.511 

So yeah, you could set a world record and still be nearly half a second off pole on the highest difficulty, or drive the 15th best lap time a human has ever done on TT this game (out of thousands of laps) and you'd still qualify P6 in a Mercedes, against slower cars. So please stop insisting we are just 'slow' because you beat them a couple of times on 80 AI in multiplayer practice rather than an actual qualifying session.

  

Edited August 13, 2019 by Worntoathread

[Ho3n3r 118]

5 minutes ago, Worntoathread said:

@Striker_703 

I did some testing at 110 AI around Spain, heavy rain, short qualifying.  

The AI did a 1:30.197

Want to know the world record time trial Lap for Spain in the wet?  It's a 1:30.511 

So yeah, you could set a world record and still be nearly half a second off pole on the highest difficulty, or drive the 10th best lap time a human has ever done on TT in this game and you'd qualify P6. So please stop insisting we are just 'slow' because you beat them a couple of times on 80 AI in multiplayer practice rather than an actual qualifying session.

  

Also consider that time trials are not affected by tyre temperature or wear, and the grip level is 100% for the conditions, not 98/99% as it would be for qualifying. Considering that, I'd say you could easily add another second to those time trial times.

But when I said Spain is too OP in the wet, I was met with "Moar practice" responses from the fanboy mob. I did time trial and I was in P29 in the wet at Spain, so I firmly disagree with that, but left it alone as I stopped career mode after Canada due to pure boredom.

Edited August 13, 2019 by Ho3n3r

[SIMRACER123 355]

The most frustrating part of this entire thing is actually if @Striker_703  opened his mind somewhat it could actually be a good thing to test.

Problem is you need a huge number of people to participate and im not really convinced anyone would want to work with @Striker_703 as he doesn't understand the principles of testing.

[caiocrq 26]

1 hour ago, SIMRACER123 said:

Eu não estou totalmente certo porque este segmento seria considerado para o encerramento.

É um caso simples de uma pessoa tentando prejudicar todo o seu produto com base em um teste mal pensado.

As pessoas aqui estão meramente apontando que seu teste não é realizado de forma conclusiva e como ele precisa encontrar um resultado mais conclusivo.

Eu teria pensado que seria das melhores empresas para o @ Striker_703  entender isso para, então, completar um teste apropriado para coletar dados para a sua empresa.

Tópicos de encerramento não vão ajudar a Codemasters a resolver qualquer problema com o jogo que pode ou não existir com o jogo deles.

Simplesmente fechar o thread realmente envia a mensagem. Os codemasters realmente não estão preocupados com possíveis problemas.

Pessoalmente, eu não acho que esse segmento já despendeu em argumentos inadequados. Foi apenas fortes visões opostas.

Exactly! This answer from him really makes me think that Codemasters will not solve this problem. Perhaps it is the biggest problem of this game and they are dealing with much neglect. I feel this will be the last F1 I get as they are not concerned about the players. I feel this problem is related to traction control. And it seems to me that every year they care less about who plays with this assistance. I say this because besides all that is being discussed in this topic, I feel very harmed also in the starts. It is impossible to start well with full traction control, it is no use doing pole-position because we will always lose 2 or more positions. I think for online game modes, it is only fair that there will be a performance loss as we are playing against other players without assistance. But in career mode, we are playing against the machine and it makes no sense to lose, it just discourages us. Anyway, I'm just raising this starting point, because I believe both problems are linked to full traction control, which Codemasters seems to be not concerned with solving.

[Striker_703 133]

3 hours ago, Worntoathread said:

@Striker_703 

I did some testing at 110 AI around Spain, heavy rain, short qualifying.  

The AI did a 1:30.197

Want to know the world record time trial Lap for Spain in the wet?  It's a 1:30.511 

So yeah, you could set a world record and still be nearly half a second off pole on the highest difficulty, or drive the 15th best lap time a human has ever done on TT this game (out of thousands of laps) and you'd still qualify P6 in a Mercedes, against slower cars. So please stop insisting we are just 'slow' because you beat them a couple of times on 80 AI in multiplayer practice rather than an actual qualifying session.

  

Can you do something for me, everyone one as well. Look at these two GUI's pay attention to the numbers on the wheel and graph on the tire temps. Now go into TT's a push hard and do the same thing in in ever other game mode and even online and tell me what you see. Then you'll notice if you've been paying attention as you have to the times. Have a nice day. 🏎️

Edited August 13, 2019 by Striker_703

[seVVen 14]

- Wheel
- abs on, rest off
-105 ai

I had some decent races with McLaren, scoring points regularely, but in Mexico and USA, i was stone dead last in wet conditions.
The AI always have perfect traction and never have a little sideways moment. There's the occasional lock up, but even that should happen more often in the rain.

[Worntoathread 865]

5 minutes ago, Striker_703 said:

Can you do something for me, everyone one as well. Look at these two GUI's pay attention to the numbers on the wheel and graph on the tire temps. Now go into TT's a push hard and do the same thing in in ever other game mode and even online and tell me what you see. Then you'll notice if you've been paying attention as you have to the times. Have a nice day. 🏎️

What is it you're trying to say exactly? 

[Striker_703 133]

If you do the test @Worntoathreadyou'll notice. I not showing you. Everyone here is this big bigger on this and that. And they're not even paying attention to what the game is showing them. Those screens shots with the GUI menu's along with each game mode should tell you ALL OF YOU what you haven't been paying attention to and missing. When it comes to comparing your times to the AI and to the times on Time Trails. 

Edited August 13, 2019 by Striker_703
Typo

[Worntoathread 865]

Just now, Striker_703 said:

If you do the test @Worntoathreadyou'll notice. I not showing you. Everyone here is this big bigger on this and that. And they're not even paying attention to what the game is showing them. Those screens shots with the GUI menu's along with each game mode should tell you ALL OF YOU have been missing. When it comes to comparing your times to the AI and to the times on Time Trails. 

If you're talking about tyre temperatures, you just proved my point. TT temps are always ideal. Other game modes are not, yet, TT record times don't come close to the AI. 

[Striker_703 133]

THERE YOU GO @Worntoathread!!! GIVE THIS PERSON A GOLDEN DOLLAR!! Now where do people get post of there set-ups from. Times Trails. What is the key factor in Time Trails. THERE IS NO SIMULATION MODEL!! No matter what you do to the tires, they will stay at the same TEMPERATURE. When you driving other game mode the tires react differently, hence the simulation model. What else do you got? No one asked CM to have the simulation model to work in Time Trails did they now?

Edited August 13, 2019 by Striker_703
Forgot about CM

[Worntoathread 865]

7 minutes ago, Striker_703 said:

THERE YOU GO @Worntoathread!!! GIVE THIS PERSON A GOLDEN DOLLAR!! Now where do people get post of there set-ups from. Times Trails. What is the key factor in Time Trails. THERE IS NO SIMULATION MODEL!! No matter what you do to the tires, they will stay at the same TEMPERATURE. When you driving other game mode the tires react differently, hence the simulation model.   

Mate, I've been league racing on a high level for years, I know how the game works. The ideal temps in TT make you FASTER. This only backs up the point that AI are too quick in the rain, because the TT times don't come close. I've no idea what you are even thinking, I'm guessing you're actually trolling at this point. Have you even tried a qualifying session in the wet on your normal AI level? 

Edited August 13, 2019 by Worntoathread

[Akkan74 250]

6 minutes ago, Striker_703 said:

THERE YOU GO @Worntoathread!!! GIVE THIS PERSON A GOLDEN DOLLAR!! Now where do people get post of there set-ups from. Times Trails. What is the key factor in Time Trails. THERE IS NO SIMULATION MODEL!! No matter what you do to the tires, they will stay at the same TEMPERATURE. When you driving other game mode the tires react differently, hence the simulation model. What else do you got? No one asked CM to have the simulation model to work in Time Trails did they now?

I don't get it what you are trying to say. I think i have to leave this discussion.

[Striker_703 133]

I'm done lol. Yes I have @Worntoathread ran against the AI in qualifying I did pretty good. You guys deal with. I'm helping more. If you can prove to them that the AI is BROKE. Then they'll fix. If it's not it will, and you all will still complain, LOL. Have a great day lol 

[Ho3n3r 118]

37 minutes ago, Striker_703 said:

THERE YOU GO @Worntoathread!!! GIVE THIS PERSON A GOLDEN DOLLAR!! Now where do people get post of there set-ups from. Times Trails. What is the key factor in Time Trails. THERE IS NO SIMULATION MODEL!! No matter what you do to the tires, they will stay at the same TEMPERATURE. When you driving other game mode the tires react differently, hence the simulation model. What else do you got? No one asked CM to have the simulation model to work in Time Trails did they now?

That's exactly the point. I'm confused as to why you're arguing against yourself.

[Ho3n3r 118]

27 minutes ago, Striker_703 said:

If you can prove to them that the AI is BROKE. Then they'll fix.

Hi. I see this is your first rodeo. Feel free to call if you need any assistance.

[SIMRACER123 355]

30 minutes ago, Striker_703 said:

I'm done lol. Yes I have @Worntoathread ran against the AI in qualifying I did pretty good. You guys deal with. I'm helping more. If you can prove to them that the AI is BROKE. Then they'll fix. If it's not it will, and you all will still complain, LOL. Have a great day lol 

I don't really know how you can claim anything is fine when you admit you race at a level your "comfortable" at. 

You need to be at a level where its challenging. Not comfortable.

Personally i think instead of focusing on tests you need to step out of your comfort zone and drive on harder levels to see what the game is like.

Then after this you can learn about tests.

[Striker_703 133]

You all are the experts. You all know how to code, read UI and GUI's and know AI code like the back of your hand. So @Faya and CM should hear you out and get the AI fix on the next patch to 1.08 was released today.   

[gtfclad82 49]

The test your trying to do on your own will not work your talking I'd say atleast 10 testers for each Al level which there are 110 levels of which adds upto 1,100 testers to even come close to completing your test.

[Striker_703 133]

2 hours ago, gtfclad82 said:

The test your trying to do on your own will not work your talking I'd say atleast 10 testers for each Al level which there are 110 levels of which adds upto 1,100 testers to even come close to completing your test.

So far I and @Worntoathreadhas done any testing. I'm using a G29 with no assist. In my costume home built rig. I don't know what he's using. He may be on the pad. And the majority are using assist which is another factor. As of now I'm the only person that's on the wheel. That could be a factor as well. Streaming the testing as well helps a lot to. Just taking screen shots doesn't prove. If you can stream it record it some how. And comparing the AI times to the TT times is not given a fair test. You're putting the another human element in as a variable. The only human variable should be the end user doing the test. You're comparing your times and speed at the AI level of your choice time and speed in adverse conditions. Along with the fact the simulation model comes into effect. It's not in effect in TT's so that's another variable that eliminate.   

Edited August 13, 2019 by Striker_703
Had to add another remark

[ChasteWand 178]

2 hours ago, Striker_703 said:

You all are the experts. You all know how to code, read UI and GUI's and know AI code like the back of your hand. So @Faya and CM should hear you out and get the AI fix on the next patch to 1.08 was released today.   

Sorry striker but the phrase pot calling kettle black springs to mind!

Nobody else has posted a degree certificate or claimed to have worked on AI more advanced than that of the game. It appears the degree post was one removed by Faya for being argumentative or off topic. However that seems nonsensical to me, but then who am I to judge.

43 minutes ago, Striker_703 said:

So far I and @Worntoathreadhas done any testing. I'm using a G29 with no assist. In my costume home built rig. I don't know what he's using. He may be on the pad. And the majority are using assist which is another factor. As of now I'm the only person that's on the wheel. That could be a factor as well. Streaming the testing as well helps a lot to. Just taking screen shots doesn't prove. If you can stream it record it some how.   

I'm sure many folk appreciate the testing you've done even if it's methodology is seriously flawed. Assists vs no assists, pad vs wheel, along with differing game modes and difficulties, all need to be tested to come to a meaningful conclusion. I simply wanted to help improve your testing method and move things forward. However it appears your ego is to big to listen to others. How can you possibly claim to be the only end user with a wheel. That beggars belief! Likewise the claim that your the only gamer that can keep pace with the AI in all conditions. I very much doubt that is the case!

Go take a look at the project cars forum and search for AI pace in wet/changeable conditions. You'll see codemasters are not the only developers to struggle with balancing. It will always be an issue whilst the AI run on a different physics model, not to mention the fact that AI is still in it's infancy and yet to be anything like truly intelligent. Today's computers are still lacking in power. Programmers are also lacking in understanding how the human brain works in terms  of problem solving and intelligence/cognition.

Hell I don't even own the game, so chances are this post will be removed for being argumentative!