[WIP]AI Mod v0,1b (Patch 1.2!)

[zlep]

/edit:
@zlep

I have no Idea what got changed in the files you tested, but if you used orignal trackfiles and rmCM=1.6, then I might see a way to get rid of the out of sight, at least for dry conditions.
If you used different settings then my way probably does not work.
I don't have the time, but could you test the following: rmCM=1.6, original trackfiles, and write down the time gap between in and out of sight.

I used the original track files and rmCM=1.6. I don't know what kristiann changed to be honest, I had no time to compare the files. Just tested with his files.
To be honest, it seems not too bad on most tracks. round about a half second gap. If it works like he explained it above, we could get it working. But I still doubt it, but hope is alive. And I still love this game, that's why I still try it and try it and try it

 

[zlep]

Yes, it has an effect but I think he is right to leave it untouched. We saw that it fixed Out of sight bug pre patch but it messed up other things. So if he found a better way to fix it, we should give it a try.

Btw. Very important: I tested his files with original track files and with a edited track file (braking points). Lap times got 2 seconds faster, but out of sight gap didn't got bigger.

 

[Kartik Sri Harsha]

Yes, it has an effect but I think he is right to leave it untouched. We saw that it fixed Out of sight bug pre patch but it messed up other things. So if he found a better way to fix it, we should give it a try.

Btw. Very important: I tested his files with original track files and with a edited track file (braking points). Lap times got 2 seconds faster, but out of sight gap didn't got bigger.

So it seems the problem is getting lighter now i think now we can find a optimal value to fix this

 

[zlep]

So it seems the problem is getting lighter now i think now we can find a optimal value to fix this

I still have not too high hopes.

Depends also on what your test results were. You tested his "group1" I think?

 

[Kristian Nenov]

So it seems the problem is getting lighter now i think now we can find a optimal value to fix this

It's not getting lighter. The bug has nothing to do with the breaking points. It's always been like this. I'll still search for other fixes till all groups are tested from you so zlep, you still need to do abu dhabi. I think Abu Dhabi will be a problem. India is a big problem already... =/

 

[zlep]

There is Abu Dhabi: #117
Gap is not too big.

 

[Kartik Sri Harsha]

India is a big problem the AI was lot slower in that track you can start from 24th from any mid field car and you can win the track so easily
i believe that we can fix this

 

[Kartik Sri Harsha]

@Kristiann do you got the points you need from my test results???

 

[Kristian Nenov]

@Kristiann do you got the points you need from my test results???

I got your results, don't worry. You still have to do: hockenheim, hungaroring, brazil, nurburgring and you didn't have to do some of the tracks because zlep did 'em. As for Abu Dhabi, I didn't see it, sorry.

I thought it's going to be a big problem as it has a big straight. Every track with a big straight makes bigger difference and is problematic. China, India... So I thought Abu Dhabi is going to be like them too but it's slow corners balance it.

 

[Kartik Sri Harsha]

SO whats your conclusion after my test results???

 

[zlep]

Hockenheim is not in 2011

 

[Kartik Sri Harsha]

Is there will be any effect if we delete the Catchup.xml file just like we did the AI toolkit for F1 2010???

 

[Kristian Nenov]

Is there will be any effect if we delete the Catchup.xml file just like we did the AI toolkit for F1 2010???

Nope. The catchup.xml in my mod doesn't give any effect at all..

SO whats your conclusion after my test results???

I don't need you to do tests for a conclusion, I need you to do tests so I can make an average value.

Australia - 1:25.765 In sight
1:26.015 Out of sight
Turkey - 1:26.850
1:28.114
Spain- 1:20.775
1:21.854
Montreal - 1:14.286
1:15.174
Valencia- 1:38.954
1:40.121

I think it's pretty obvious that I should make the lap times out of sight faster but I wanted to have all tracks's times.

Hockenheim is not in 2011

Shame on CM for adding it in the list! They're embarrassing me!




I found 2 more values which change the out of sight lap times but it seems they all work the same way which is strange!

 

[zlep]

Ok, here are some new test results with version 0.76 from your mod you send me tonight.

Abu Dhabi
Vettel 1st
In sight (Lap 2): 1:42:023
Out of sight (Lap 3): 1:41:998
Out of sight (Lap 4): 1:41:959

Alonso 2nd
In sight (Lap 2): 1:42:904
Out of sight (Lap 3): 1:42:900
Out of sight (Lap 4): 1:43:061

Webber 3rd
In sight (Lap 2): 1:43:025
Out of sight (Lap 3): 1:43:298
Out of sight (Lap 4): 1:42:954

Malaysia
In sight (Lap 2): 1:35:839
Out of sight (Lap 3): 1:36:337

China
In sight (Lap 2): 1:36:598
Out of sight (Lap 3): 1:37:173

Canada
In sight (Lap 2): 1:15:116
Out of sight (Lap 3): 1:14:907

Valencia
In sight (Lap 2): 1:41:917
Out of sight (Lap 3): 1:41:481

So, Abu Dhabi and Canda looking good.
Malaysia and China +0,6 seconds
Valencia -0,5 seconds

But it's looking not so bad. Maybe we can really get this working.

 

[mertbabaakar]

Ok, here are some new test results with version 0.76 from your mod you send me tonight.

Abu Dhabi
Vettel 1st
In sight (Lap 2): 1:42:023
Out of sight (Lap 3): 1:41:998
Out of sight (Lap 4): 1:41:959

Alonso 2nd
In sight (Lap 2): 1:42:904
Out of sight (Lap 3): 1:42:900
Out of sight (Lap 4): 1:43:061

Webber 3rd
In sight (Lap 2): 1:43:025
Out of sight (Lap 3): 1:43:298
Out of sight (Lap 4): 1:42:954

Malaysia
In sight (Lap 2): 1:35:839
Out of sight (Lap 3): 1:36:337

China
In sight (Lap 2): 1:36:598
Out of sight (Lap 3): 1:37:173

Canada
In sight (Lap 2): 1:15:116
Out of sight (Lap 3): 1:14:907

Valencia
In sight (Lap 2): 1:41:917
Out of sight (Lap 3): 1:41:481

So, Abu Dhabi and Canda looking good.
Malaysia and China +0,6 seconds
Valencia -0,5 seconds

But it's looking not so bad. Maybe we can really get this working.

Zlep todays time is better than other so this times looking good and which difficulty is it ? and if its legend this so bad becuase I am going to canada 1.11.xxx 1.12.xxx like this.

 

[RowingAce]

yeah, in and out of sight looks quite okay, but I have to agree with mert... (shocking, isn't it? ) the speed on all tracks is quite slow.
At least if you were testing anything shorter than 50% ...

I guess if you make them faster again, the problem comes back ..

 

[zlep]

It was 50%, legend AI.

Later I can test it with an edited track file which makes the AI 2 seconds faster. Will see if the gap gets bigger. But I already tested it at Malaysia and the gap didn't got bigger. But it was an older version of his mod, so I test it again. We will see.

 

[zlep]

I can now confirm, out of sight bug does not return when editing braking points in ai_vehicle_track.xml:

Malaysia
In sight (Lap 2): 1:33:554
Out of sight (Lap 3): 1:33:921

Gap even got little bit smaller in this test. Seems I have to apologize. He is really on the right way to get it fixed! Good work so far.

 

[RowingAce]

did you adjust ONLY braking points or curve modifiers as well?

 

[Kartik Sri Harsha]

So we are on right track after all the time
Zlep can you make the test in India with edited breaking zones????