Our ReVolt Pub

Posted: **23 Feb 2015, 07:08**

Not sure if I'm in the correct topic ?

I don't know if this has been reported and fixed, but I get the feeling that Newton's 3rd law "For every action there is an equal and opposite reaction", just doesn't ring true for me.
Using Current version Alpha 15.0125, using Simulation mode.

AI are very aggressive and at every opportunity run me off the racing line, yes that's what the AI have always done, but these guy's hit me with real force as if they were 3-4 times heavier than me, and if I try to punt one of them they barely move off line. In a virtual head to head at a crossroads I always come off the worse.
Is it just me and how I drive ?

Off topic, is there a topic/thread for surface properties and their physics ?

Posted: **23 Feb 2015, 07:51**

Manmountain @ 23 Feb 2015, 02:38 AM wrote: AI are very aggressive and at every opportunity run me off the racing line, yes that's what the AI have always done, but these guy's hit me with real force as if they were 3-4 times heavier than me, and if I try to punt one of them they barely move off line. In a virtual head to head at a crossroads I always come off the worse.
Is it just me and how I drive ?

no, this has been noticed before, and there are other symptoms.

it appears to be a gravity bug?

seems the players car is lighter and the bots are heavier. player cars can 'fly' farther while airborne and they also slide around a lot more as if they were lighter too.

we are being handicapped again

Posted: **23 Feb 2015, 08:05**

I have to agree that they would run you off, though I thought it was the normal way of this 15.0125 to cater for the physics as for what thugsrook is mentioning Idk but he's shown that he's very perceptive

Posted: **23 Feb 2015, 08:18**

This thread's topic is a total different story to what you're describing and there's no other about it apart from vague complaints about new AI being more difficult to defeat than the original.

So, Manmountain would you make a new thread and explain the problem further?

Posted: **23 Feb 2015, 08:30**

Phantom @ 23 Feb 2015, 03:48 AM wrote: So, Manmountain would you make a new thread and explain the problem further?

Done, but what more can I explain ?

Posted: **23 Feb 2015, 08:50**

For example if you think that the problem comes from A.I behaviour or from a physics change.

You probably know this already, but some time in 2012 many changes in the A.I. were introduced.

One of the patches from 2012 introduced an "improved" AI which was smarter, stronger and more intelligent than original AI which, from the developer's point of view, contained "bugs".

Later, people suggested that Re-Volt should keep both the original AI released by Acclaim and the new improved AI because the new AI was excellent for Pro players but impossible to defeat by new or amateur players, even by veterans like you.
I remember we even made a Poll about it and we won by a massive amount but I guess 2 difficulties was difficult to implement by the developers.

The developer's choice was to simply lower down the strenght and the intelligence in the new AI trying to make something acceptable instead of adding 2 difficulties. So basically the current A.I. is the new A.I. from 2012 but already reduced in strenght by Huki.

On the other hand, if this comes from a Physics change may god help you because I've no luck trying to fix physics changes.

Posted: **23 Feb 2015, 08:58**

Hmm, AI is kinda better nowdays, its still average beat.

well, not quite, but its a better challenge for sure: gold championship is hard and tough against Adeon.

yeah, physics is pretty buggy and odd: random "springs" may happen, omgrockets, superbombs or a bomb that doesnt affect you much.

yeah, since 13.0815 ai is in effect (pretty much), 0820 introduces rubberbanding.

Posted: **23 Feb 2015, 09:25**

I always thought the AI not being knocked far off course was less because of physics nonsense and more because it can instantly react and calculate, with precision, how best to realign itself.

Meanwhile the player has to deal with reaction times, human error etc in a game where things tend to happen pretty damn fast.

Posted: **23 Feb 2015, 09:38**

Yes, I admit it is better for me too because it's more challenging: being more aggressive, making less mistakes, reacting faster after being attacked and taking the best racelines.

Though everytime I've tried to get a friend of mine to play the game I have to do it with an old version which includes the original AI before moving on to the current one. The current AI even with its lowered down strenght is yet too strong that for an amateur player it gets impossible to compete, even adults. I can't imagine children. They get bored easily fighting an impossible fight and RC Junior changes collission mode and speed, not difficulty, intelligence or aggression.

Posted: **23 Feb 2015, 19:23**

@Manmountain
Does it occurs with any cars, I mean can we notice it with any stock level in a stock-car-only race?

ThugsRook @ 23 Feb 2015, 03:21 AM wrote:no, this has been noticed before, and there are other symptoms.
[...]
seems the players car is lighter and the bots are heavier. player cars can 'fly' farther while airborne and they also slide around a lot more as if they were lighter too.

As far as I'm concerned that's the first time I hear about that... Car pushing in online race, yes, but offline against the AI, I can't remember.

Manmountain wrote:Off topic, is there a topic/thread for surface properties and their physics ?

What are you asking for actually? The properties of materials that can be applied to a track?

Posted: **23 Feb 2015, 20:55**

jigebren @ 23 Feb 2015, 02:53 PM wrote:@Manmountain
Does it occurs with any cars, I mean can we notice it with any stock level in a stock-car-only race?

Most certainly, well, it always seem's that way to me.

Phantom @ 23 Feb 2015, 04:20 AM wrote:For example if you think that the problem comes from A.I behaviour or from a physics change...

...On the other hand, if this comes from a Physics change may god help you because I've no luck trying to fix physics changes.

jigebren wrote:
Manmountain wrote:Off topic, is there a topic/thread for surface properties and their physics ?
What are you asking for actually? The properties of materials that can be applied to a track?

What I'm asking is, are the AI car physics alone the reason for their seemed extra aggressiveness or is it in combination with possibly altered surface physic's as I feel that some surface's don't give the same reaction's as they use to, IMO.

Code: Select all

;====================
; Car AI details
;====================

AI { &nbsp; &nbsp; &nbsp; &nbsp;	&#59;Start AI
...
Aggression &nbsp; &nbsp; &nbsp;0
} &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &#59; End AI

Does this value have any input/impact on how the AI cars behave, has anyone got definitive evidence ?

Posted: **23 Feb 2015, 22:40**

Manmountain @ 23 Feb 2015, 12:25 PM wrote: What I'm asking is, are the AI car physics alone the reason for their seemed extra aggressiveness or is it in combination with possibly altered surface physic's as I feel that some surface's don't give the same reaction's as they use to, IMO.

Both A.I. and physics have been changed, so it could be caused by the combination of both things which leads to a whole different sensation when you race.

If you are sure that some surfaces don't give the same reaction in an offline race this starts to be interesting.

1. Did you have any computer upgrade that could be triggering the different feeling or you think that this comes after updating the game?

2. Can you compare the current version with an old version in offline race?
Here I provide you with a ready-to-use past build before any physics or A.I. changes.
Just extract it and run it as admin if you are in Windows 7. Do not need to use Win98 mode.

Please play with it the same way you do with the current version and let us know if you feel it better or not.

About the aggression in Parameters, this is from Parameters Notes 7.0.

15. Aggression wrote:This setting may correspond to PickupBias, BlockBias, and OvertakeBias. It is possible that this setting could tell the car how often it should use these setting. Assuming, higher numbers make the car more aggressive, and lower numbers make the car more passive. This setting needs further testing.

Posted: **23 Feb 2015, 23:05**

WOW !
I nearly won on TH1 even after taking the time to configure controller during the race.
Trust me, I very rarely win.

The surface physics are much more grippy and the AI interaction is there but with a more even and less aggressive feel.

I'll test with a few custom tracks and custom AI to see if I can still notice a difference.
...be back soon.

Posted: **23 Feb 2015, 23:10**

^ all versions of Alpha have a problem ~ test/compare with Beta or original Re-Volt, do not test with any Alphas.

these problems are easy for me to notice in Alpha... because i usually use Beta so im very used to it.

Posted: **23 Feb 2015, 23:14**

ThugsRook @ 23 Feb 2015, 02:40 PM wrote: ^ all versions of Alpha have a problem ~ test/compare with Beta or original Re-Volt, do not test with any Alphas.

these problems are easy for me to notice in Alpha... because i usually use Beta so im very used to it.

Dont worry Thugs, the one I sent him is the oldest alpha available. Im sure it doesnt have any physics or AI changes. Its the one we use for online races.

Posted: **23 Feb 2015, 23:22**

Phantom @ 23 Feb 2015, 02:44 PM wrote:
ThugsRook @ 23 Feb 2015, 02:40 PM wrote: ^ all versions of Alpha have a problem ~ test/compare with Beta or original Re-Volt, do not test with any Alphas.

these problems are easy for me to notice in Alpha... because i usually use Beta so im very used to it.
Dont worry Thugs, the one I sent him is the oldest alpha available. Im sure it doesnt have any physics or AI changes. Its the one we use for online races.

Oh really?
I thought we use 0820.....
Isn't the beta better? lol

Posted: **23 Feb 2015, 23:52**

Well, I tested/raced other tracks and with my own extreme AI cars, and there is definitely a surface physics difference and the AI are not as aggressive, and the so called 'Rubberband Effect' does not seem as harsh.

Are the physics of this version based on or the same as the original 1207 patch version ?

Posted: **23 Feb 2015, 23:53**

Very interesting Manmountain. Thanks for the test and review.

Since you have the precious gift to sense the difference maybe you can help in fixing this. Only if you have the time and interest of course.

Look at this link
http://rv12.revoltzone.net/oldbuilds.php

The build I sent you is the 11.0421. The physics and the A.I. of this version are still the same as the 1207. But the builds released after it obviusly contain changes in different aspects of ReVolt. Ive tried to identify from which build this strange feeling starts but it ended up confusing me. Could you help me if you have the time. I need someone with experience in offline races that can offer a different point of view about this.

Posted: **23 Feb 2015, 23:54**

Above you sunshine

Posted: **24 Feb 2015, 01:53**

Phantom @ 23 Feb 2015, 06:10 PM wrote: About the aggression in Parameters, this is from Parameters Notes 7.0.

15. Aggression wrote:This setting may correspond to PickupBias, BlockBias, and OvertakeBias. It is possible that this setting could tell the car how often it should use these setting. Assuming, higher numbers make the car more aggressive, and lower numbers make the car more passive. This setting needs further testing.

I could be wrong (or it could have been changed in 1.2) but as far as I can tell by looking at the source code it seems that this parameter doesn't have any effect at all.

edit: As well as OvertakeBias and BlockBias.

Posted: **24 Feb 2015, 02:41**

Kenny @ 23 Feb 2015, 09:23 PM wrote:I could be wrong (or it could have been changed in 1.2) but as far as I can tell by looking at the source code it seems that this parameter doesn't have any effect at all.

It doesn't indeed.

Manmountain, if you feel like testing different builds, here an (unfinished) summary I made about the builds available on the v1.2 site. It would be interesting to know more precisely where between 11.0525 and 11.1215 the differences appeared.

rv1.2a11.1215 | known to be buggy (some or most bugs have been fixed later)
rv1.2a11.0825 | not sure if this one is already affected
rv1.2a11.0810 |
rv1.2a11.0731 |
rv1.2a11.0717 |
rv1.2a11.0710 |
rv1.2a11.0625 | maybe ok but no sure
rv1.2a11.0610 |
rv1.2a11.0525 |
rv1.2a11.0510 |
rv1.2a11.0501 | same v1.1 behavior
rv1.2a11.0426 | same v1.1 behavior
rv1.2a11.0421 |
rv1.2b11.0208 | same v1.1 behavior

Trying the latest 15.0211 test build in comparison would be very interesting too. This one is apparently agreed to be better (and should be very close to the upcoming v1.2 Alpha release), and you can even type OLDSTEER blindly as a cheat to disable the last steering modification.

Phantom wrote:Both A.I. and physics have been changed

Physics have been changed AND restored... so if any differences remain they have an other root than direct physics changes.

v1.2 changelog wrote:---------------
Alpha 13.0815
---------------
- Disabled "little step collision fix".
---------------
Alpha 12.0815
---------------
- Limited "car sliding in horizontal surface" fix to Frontend preview screen, disabled in-game (should fix reported issues with car handling).

Posted: **24 Feb 2015, 05:52**

jigebren @ 23 Feb 2015, 10:11 PM wrote:Manmountain, if you feel like testing different builds...

Not really

jigebren wrote:Trying the latest 15.0211 test build in comparison would be very interesting too.

OK, I'll give this one a look see.

Posted: **24 Feb 2015, 06:32**

Firstly, it shows Alpha 15.0131 is that right ?
The AI aggression and car weight seem more even.
But the surface physic's seem not as grippy.

Posted: **24 Feb 2015, 06:42**

Manmountain @ 23 Feb 2015, 10:02 PM wrote: Firstly, it shows Alpha 15.0131 is that right ?
The AI aggression and car weight seem more even.
But the surface physic's seem not as grippy.

yep, seems that he forgot to change versioning

Posted: **25 Feb 2015, 03:26**

I compared this latest 15.0211 test build in Alpha 15.0131 and yes it is not as grippy as 15.0131 I get the felling of flying faster and further when jumping, still I don't now if I prefer the adherence (I'm probably duping myself so I need more testing) I just made a timetrial in both and by that I would go for the latest 15.0211

Our ReVolt Pub

Newton's 3rd Law