rec.autos.simulators

NR2003: AI still inconsistent from one track to the next

Marc Collin

NR2003: AI still inconsistent from one track to the next

by Marc Collin » Thu, 13 Feb 2003 02:32:21

These are the "voodoo" settings that Papy tweaks to come up with the
"balanced" playability at each track, which unfortunately has been seriously
imbalanced in N3, N4, NR2002 and now, finally, NR2003.

The rest of the sim is so damned good that it annoys me that something like
this is overlooked year after year.  Oval racing isn't my cup of tea to
start with, but if I have to go to a track and practise and twist my brain
to drive a car that has a twisted to the side unnatural set-up, I get really
annoyed when my sense of accomplishment or competence (as measured by how
well I do against the only benchmark available--the AI at that track) is as
much of a ***shoot as surviving an online pick-up race.

Marc



> > So I rest my case...each and every user needing to do a detailed
research
> > project just to enjoy the main premise of the game--a championship
> > season--is ridiculous.  No wonder so many don't bother and just play
online.
> > Unfortunately, 90% of the people playing online drive like they should
have
> > been forced to complete a few races offline and actually finish higher
than
> > 43rd before being allowed online.

> > When the average experienced user can take the fast set-up at California
and
> > Michigan and achieve roughly the same results (since the skill level is
the
> > same), the game will have been designed properly (at least for those two
> > tracks).  When one track is easy to place in the top 3 and the other you
> > have to struggle to make it to the bottom 3, something is seriously
wrong.

> > Marc

>      [serious snippage]

> California.ini
> -----------------
> [ai track]
> ai_accel_modifier = 1.00                        ; acceleration grip
efficiency
> ai_decel_modifier = 0.90                        ; braking grip efficiency
> ai_fuel_use = 0.98                              ; > 1.0 = more fuel
consumed
> ai_grip_modifier = 1.05                         ; > 1.0 = more grip
> ai_drag_modifier = 1.08                         ; > 1.0 = more drag, which
is
> slower

> Michigan.ini
> -----------------
> [ai track]
> ai_accel_modifier = 1.20                        ; acceleration grip
efficiency
> ai_decel_modifier = 0.88                        ; braking grip efficiency
> ai_fuel_use = 0.96                              ; > 1.0 = more fuel
consumed
> ai_grip_modifier = 1.06                         ; > 1.0 = more grip
> ai_drag_modifier = 1.04                         ; > 1.0 = more drag, which
is
> slower

> The biggest difference is in the accel modifier (which I suppose
> means less slippage when accelerating out of the corners).  For the
> others there are small advantages for the Michigan AI (and why the
> hey should fuel use be a variable from track to track?  From trial
> and error I found that .93 will ensure the AI pits when you do-
> and I will note that fuel mileage for _me_ seems better than before).

>     John DiFool

> --
> ============================================
> Reach heaven far too high
> ============================================

Jay Taylo

NR2003: AI still inconsistent from one track to the next

by Jay Taylo » Thu, 13 Feb 2003 02:51:24

Over looked?
If you had any idea the amount of time that went into play balancing NR2003
I doubt you'd say it was over looked.  Perhaps the settings don't work
perfectly for everyone, I'll agree thats possible.  However they are
certainly close, and should provide enjoyable racing.  I'm sorry if you feel
otherwise.
The effort was certainly put into it, and the beta team felt the tracks were
balanced pretty well.  I'm not sure what you mean by "Voodoo settings" but I
dont remember the use of shaman totems, shrunken heads, or bat wings being
involved in the balancing of NR2003 <G>

Jay Taylor


> These are the "voodoo" settings that Papy tweaks to come up with the
> "balanced" playability at each track, which unfortunately has been
seriously
> imbalanced in N3, N4, NR2002 and now, finally, NR2003.

> The rest of the sim is so damned good that it annoys me that something
like
> this is overlooked year after year.  Oval racing isn't my cup of tea to
> start with, but if I have to go to a track and practise and twist my brain
> to drive a car that has a twisted to the side unnatural set-up, I get
really
> annoyed when my sense of accomplishment or competence (as measured by how
> well I do against the only benchmark available--the AI at that track) is
as
> much of a ***shoot as surviving an online pick-up race.

> Marc




> > > So I rest my case...each and every user needing to do a detailed
> research
> > > project just to enjoy the main premise of the game--a championship
> > > season--is ridiculous.  No wonder so many don't bother and just play
> online.
> > > Unfortunately, 90% of the people playing online drive like they should
> have
> > > been forced to complete a few races offline and actually finish higher
> than
> > > 43rd before being allowed online.

> > > When the average experienced user can take the fast set-up at
California
> and
> > > Michigan and achieve roughly the same results (since the skill level
is
> the
> > > same), the game will have been designed properly (at least for those
two
> > > tracks).  When one track is easy to place in the top 3 and the other
you
> > > have to struggle to make it to the bottom 3, something is seriously
> wrong.

> > > Marc

> >      [serious snippage]

> > California.ini
> > -----------------
> > [ai track]
> > ai_accel_modifier = 1.00                        ; acceleration grip
> efficiency
> > ai_decel_modifier = 0.90                        ; braking grip
efficiency
> > ai_fuel_use = 0.98                              ; > 1.0 = more fuel
> consumed
> > ai_grip_modifier = 1.05                         ; > 1.0 = more grip
> > ai_drag_modifier = 1.08                         ; > 1.0 = more drag,
which
> is
> > slower

> > Michigan.ini
> > -----------------
> > [ai track]
> > ai_accel_modifier = 1.20                        ; acceleration grip
> efficiency
> > ai_decel_modifier = 0.88                        ; braking grip
efficiency
> > ai_fuel_use = 0.96                              ; > 1.0 = more fuel
> consumed
> > ai_grip_modifier = 1.06                         ; > 1.0 = more grip
> > ai_drag_modifier = 1.04                         ; > 1.0 = more drag,
which
> is
> > slower

> > The biggest difference is in the accel modifier (which I suppose
> > means less slippage when accelerating out of the corners).  For the
> > others there are small advantages for the Michigan AI (and why the
> > hey should fuel use be a variable from track to track?  From trial
> > and error I found that .93 will ensure the AI pits when you do-
> > and I will note that fuel mileage for _me_ seems better than before).

> >     John DiFool

> > --
> > ============================================
> > Reach heaven far too high
> > ============================================

Marc Collin

NR2003: AI still inconsistent from one track to the next

by Marc Collin » Thu, 13 Feb 2003 08:10:37

What I mean by Voodoo is things like ai_accel_modifier...the layman would
ask why would that vary from on track to the next?  Same goes for
ai_grip_modifier...if the track surface has a certain grip, varying from
track to track of course, why would you need a modifier for this?

I have no doubt that many, many hours, sweat and tears were poured into the
effort.  But I think I explained why the result is less than satisfactory.
And I will eat my hat and shut up about this forever as soon as you or
someone else provides two "fast" set-ups, one for Michigan and one for
California, that permit an experienced user/driver/racer to hit the top 10
in each with approximately the same effort...since that would be realistic
and also demonstrate balanced gameplay as a bonus.

If I can't do a single top-10 lap at Michigan in practice mode with the AI
at 100% when I can beat the AI at 100% in GPL at every track (except the
'Ring) and easily do top 5's at the sister track of California, then
something is wrong...and it ain't me.  It's either the game or the set-up
and I am more than willing to believe you that it is the set-up since you
are intimately involved with the title and I am a lowly purchaser/user.

As I asked in the other part of this thread, please attach your two set-ups
and we'll see what results.

Thanks,

Marc


> Over looked?
> If you had any idea the amount of time that went into play balancing
NR2003
> I doubt you'd say it was over looked.  Perhaps the settings don't work
> perfectly for everyone, I'll agree thats possible.  However they are
> certainly close, and should provide enjoyable racing.  I'm sorry if you
feel
> otherwise.
> The effort was certainly put into it, and the beta team felt the tracks
were
> balanced pretty well.  I'm not sure what you mean by "Voodoo settings" but
I
> dont remember the use of shaman totems, shrunken heads, or bat wings being
> involved in the balancing of NR2003 <G>

> Jay Taylor



> > These are the "voodoo" settings that Papy tweaks to come up with the
> > "balanced" playability at each track, which unfortunately has been
> seriously
> > imbalanced in N3, N4, NR2002 and now, finally, NR2003.

> > The rest of the sim is so damned good that it annoys me that something
> like
> > this is overlooked year after year.  Oval racing isn't my cup of tea to
> > start with, but if I have to go to a track and practise and twist my
brain
> > to drive a car that has a twisted to the side unnatural set-up, I get
> really
> > annoyed when my sense of accomplishment or competence (as measured by
how
> > well I do against the only benchmark available--the AI at that track) is
> as
> > much of a ***shoot as surviving an online pick-up race.

> > Marc




> > > > So I rest my case...each and every user needing to do a detailed
> > research
> > > > project just to enjoy the main premise of the game--a championship
> > > > season--is ridiculous.  No wonder so many don't bother and just play
> > online.
> > > > Unfortunately, 90% of the people playing online drive like they
should
> > have
> > > > been forced to complete a few races offline and actually finish
higher
> > than
> > > > 43rd before being allowed online.

> > > > When the average experienced user can take the fast set-up at
> California
> > and
> > > > Michigan and achieve roughly the same results (since the skill level
> is
> > the
> > > > same), the game will have been designed properly (at least for those
> two
> > > > tracks).  When one track is easy to place in the top 3 and the other
> you
> > > > have to struggle to make it to the bottom 3, something is seriously
> > wrong.

> > > > Marc

> > >      [serious snippage]

> > > California.ini
> > > -----------------
> > > [ai track]
> > > ai_accel_modifier = 1.00                        ; acceleration grip
> > efficiency
> > > ai_decel_modifier = 0.90                        ; braking grip
> efficiency
> > > ai_fuel_use = 0.98                              ; > 1.0 = more fuel
> > consumed
> > > ai_grip_modifier = 1.05                         ; > 1.0 = more grip
> > > ai_drag_modifier = 1.08                         ; > 1.0 = more drag,
> which
> > is
> > > slower

> > > Michigan.ini
> > > -----------------
> > > [ai track]
> > > ai_accel_modifier = 1.20                        ; acceleration grip
> > efficiency
> > > ai_decel_modifier = 0.88                        ; braking grip
> efficiency
> > > ai_fuel_use = 0.96                              ; > 1.0 = more fuel
> > consumed
> > > ai_grip_modifier = 1.06                         ; > 1.0 = more grip
> > > ai_drag_modifier = 1.04                         ; > 1.0 = more drag,
> which
> > is
> > > slower

> > > The biggest difference is in the accel modifier (which I suppose
> > > means less slippage when accelerating out of the corners).  For the
> > > others there are small advantages for the Michigan AI (and why the
> > > hey should fuel use be a variable from track to track?  From trial
> > > and error I found that .93 will ensure the AI pits when you do-
> > > and I will note that fuel mileage for _me_ seems better than before).

> > >     John DiFool

> > > --
> > > ============================================
> > > Reach heaven far too high
> > > ============================================

Jay Taylo

NR2003: AI still inconsistent from one track to the next

by Jay Taylo » Thu, 13 Feb 2003 09:07:17

First of all if you want to match up verse the AI in practice you better
load  a qual setup not a race, as through out testing you would always see a
few AI doing more or less qual runs, while others were more in race trim.
Practice is just that, you wont find a hard line on exactly what they will
run for the race.

As for ai_grip_modifier, and other settings the reason is simple.  The AI do
not run on exactly the same physics the player does. If they did you might
be able to run with a dozen cars max on the fastest of systems.  So they
have a somewhat simplifed physics model to keep CPU overhead reasonable.
Therefore after you have lines run for a track that work the way you want
them to, and dont have any choke points that mess up the AI, you use these
values to fine tune the AI to get them to drive the track as close to the
way the typical player would.  These are  crucial in smoothing out the AI,
and keeping them in line with the player.  If everything was equal physics
wise, and the AI had the same knowledge the player does about getting around
the track then these values wouldn't be needed.  However until someone finds
away to do all this and make it run well on a normal system, you have to
make adjustments, and try to dial things in with what you have to work with.

Jay Taylor

"Marc Collins" <marc_coll...@sympatico.ca> wrote in message

news:Lpf2a.30455$ns3.436218@news20.bellglobal.com...
> What I mean by Voodoo is things like ai_accel_modifier...the layman would
> ask why would that vary from on track to the next?  Same goes for
> ai_grip_modifier...if the track surface has a certain grip, varying from
> track to track of course, why would you need a modifier for this?

> I have no doubt that many, many hours, sweat and tears were poured into
the
> effort.  But I think I explained why the result is less than satisfactory.
> And I will eat my hat and shut up about this forever as soon as you or
> someone else provides two "fast" set-ups, one for Michigan and one for
> California, that permit an experienced user/driver/racer to hit the top 10
> in each with approximately the same effort...since that would be realistic
> and also demonstrate balanced gameplay as a bonus.

> If I can't do a single top-10 lap at Michigan in practice mode with the AI
> at 100% when I can beat the AI at 100% in GPL at every track (except the
> 'Ring) and easily do top 5's at the sister track of California, then
> something is wrong...and it ain't me.  It's either the game or the set-up
> and I am more than willing to believe you that it is the set-up since you
> are intimately involved with the title and I am a lowly purchaser/user.

> As I asked in the other part of this thread, please attach your two
set-ups
> and we'll see what results.

> Thanks,

> Marc

> "Jay Taylor" <jma...@prodigy.net> wrote in message
> news:wKa2a.3765$Ly6.1187@newssvr16.news.prodigy.com...
> > Over looked?
> > If you had any idea the amount of time that went into play balancing
> NR2003
> > I doubt you'd say it was over looked.  Perhaps the settings don't work
> > perfectly for everyone, I'll agree thats possible.  However they are
> > certainly close, and should provide enjoyable racing.  I'm sorry if you
> feel
> > otherwise.
> > The effort was certainly put into it, and the beta team felt the tracks
> were
> > balanced pretty well.  I'm not sure what you mean by "Voodoo settings"
but
> I
> > dont remember the use of shaman totems, shrunken heads, or bat wings
being
> > involved in the balancing of NR2003 <G>

> > Jay Taylor

> > "Marc Collins" <marc_coll...@sympatico.ca> wrote in message
> > news:Gsa2a.358$3g1.52683@news20.bellglobal.com...
> > > These are the "voodoo" settings that Papy tweaks to come up with the
> > > "balanced" playability at each track, which unfortunately has been
> > seriously
> > > imbalanced in N3, N4, NR2002 and now, finally, NR2003.

> > > The rest of the sim is so damned good that it annoys me that something
> > like
> > > this is overlooked year after year.  Oval racing isn't my cup of tea
to
> > > start with, but if I have to go to a track and practise and twist my
> brain
> > > to drive a car that has a twisted to the side unnatural set-up, I get
> > really
> > > annoyed when my sense of accomplishment or competence (as measured by
> how
> > > well I do against the only benchmark available--the AI at that track)
is
> > as
> > > much of a crap shoot as surviving an online pick-up race.

> > > Marc

> > > "John DiFool" <jdif...@earthlink.net> wrote in message
> > > news:3E491723.9B016299@earthlink.net...
> > > > Marc Collins wrote:

> > > > > So I rest my case...each and every user needing to do a detailed
> > > research
> > > > > project just to enjoy the main premise of the game--a championship
> > > > > season--is ridiculous.  No wonder so many don't bother and just
play
> > > online.
> > > > > Unfortunately, 90% of the people playing online drive like they
> should
> > > have
> > > > > been forced to complete a few races offline and actually finish
> higher
> > > than
> > > > > 43rd before being allowed online.

> > > > > When the average experienced user can take the fast set-up at
> > California
> > > and
> > > > > Michigan and achieve roughly the same results (since the skill
level
> > is
> > > the
> > > > > same), the game will have been designed properly (at least for
those
> > two
> > > > > tracks).  When one track is easy to place in the top 3 and the
other
> > you
> > > > > have to struggle to make it to the bottom 3, something is
seriously
> > > wrong.

> > > > > Marc

> > > >      [serious snippage]

> > > > California.ini
> > > > -----------------
> > > > [ai track]
> > > > ai_accel_modifier = 1.00                        ; acceleration grip
> > > efficiency
> > > > ai_decel_modifier = 0.90                        ; braking grip
> > efficiency
> > > > ai_fuel_use = 0.98                              ; > 1.0 = more fuel
> > > consumed
> > > > ai_grip_modifier = 1.05                         ; > 1.0 = more grip
> > > > ai_drag_modifier = 1.08                         ; > 1.0 = more drag,
> > which
> > > is
> > > > slower

> > > > Michigan.ini
> > > > -----------------
> > > > [ai track]
> > > > ai_accel_modifier = 1.20                        ; acceleration grip
> > > efficiency
> > > > ai_decel_modifier = 0.88                        ; braking grip
> > efficiency
> > > > ai_fuel_use = 0.96                              ; > 1.0 = more fuel
> > > consumed
> > > > ai_grip_modifier = 1.06                         ; > 1.0 = more grip
> > > > ai_drag_modifier = 1.04                         ; > 1.0 = more drag,
> > which
> > > is
> > > > slower

> > > > The biggest difference is in the accel modifier (which I suppose
> > > > means less slippage when accelerating out of the corners).  For the
> > > > others there are small advantages for the Michigan AI (and why the
> > > > hey should fuel use be a variable from track to track?  From trial
> > > > and error I found that .93 will ensure the AI pits when you do-
> > > > and I will note that fuel mileage for _me_ seems better than
before).

> > > >     John DiFool

> > > > --
> > > > ============================================
> > > > Reach heaven far too high
> > > > ============================================

Dave Henri

NR2003: AI still inconsistent from one track to the next

by Dave Henri » Thu, 13 Feb 2003 11:44:25

"Marc Collins"I get really annoyed when my sense of accomplishment or

  GPLrank has a sister site for N2k2,  I would assume they also have or
plan for the same with N2k3.  Instant benchmark.
dave henrie

Marc Collin

NR2003: AI still inconsistent from one track to the next

by Marc Collin » Sat, 15 Feb 2003 13:34:56

Oh well, I guess no one is willing to put their money where there mouth is
by ponying up a couple of set-ups.

Marc

"Jay Taylor" <jma...@prodigy.net> wrote in message

news:Veg2a.3869$dl.1715@newssvr16.news.prodigy.com...
> First of all if you want to match up verse the AI in practice you better
> load  a qual setup not a race, as through out testing you would always see
a
> few AI doing more or less qual runs, while others were more in race trim.
> Practice is just that, you wont find a hard line on exactly what they will
> run for the race.

> As for ai_grip_modifier, and other settings the reason is simple.  The AI
do
> not run on exactly the same physics the player does. If they did you might
> be able to run with a dozen cars max on the fastest of systems.  So they
> have a somewhat simplifed physics model to keep CPU overhead reasonable.
> Therefore after you have lines run for a track that work the way you want
> them to, and dont have any choke points that mess up the AI, you use these
> values to fine tune the AI to get them to drive the track as close to the
> way the typical player would.  These are  crucial in smoothing out the AI,
> and keeping them in line with the player.  If everything was equal physics
> wise, and the AI had the same knowledge the player does about getting
around
> the track then these values wouldn't be needed.  However until someone
finds
> away to do all this and make it run well on a normal system, you have to
> make adjustments, and try to dial things in with what you have to work
with.

> Jay Taylor

> "Marc Collins" <marc_coll...@sympatico.ca> wrote in message
> news:Lpf2a.30455$ns3.436218@news20.bellglobal.com...
> > What I mean by Voodoo is things like ai_accel_modifier...the layman
would
> > ask why would that vary from on track to the next?  Same goes for
> > ai_grip_modifier...if the track surface has a certain grip, varying from
> > track to track of course, why would you need a modifier for this?

> > I have no doubt that many, many hours, sweat and tears were poured into
> the
> > effort.  But I think I explained why the result is less than
satisfactory.
> > And I will eat my hat and shut up about this forever as soon as you or
> > someone else provides two "fast" set-ups, one for Michigan and one for
> > California, that permit an experienced user/driver/racer to hit the top
10
> > in each with approximately the same effort...since that would be
realistic
> > and also demonstrate balanced gameplay as a bonus.

> > If I can't do a single top-10 lap at Michigan in practice mode with the
AI
> > at 100% when I can beat the AI at 100% in GPL at every track (except the
> > 'Ring) and easily do top 5's at the sister track of California, then
> > something is wrong...and it ain't me.  It's either the game or the
set-up
> > and I am more than willing to believe you that it is the set-up since
you
> > are intimately involved with the title and I am a lowly purchaser/user.

> > As I asked in the other part of this thread, please attach your two
> set-ups
> > and we'll see what results.

> > Thanks,

> > Marc

> > "Jay Taylor" <jma...@prodigy.net> wrote in message
> > news:wKa2a.3765$Ly6.1187@newssvr16.news.prodigy.com...
> > > Over looked?
> > > If you had any idea the amount of time that went into play balancing
> > NR2003
> > > I doubt you'd say it was over looked.  Perhaps the settings don't work
> > > perfectly for everyone, I'll agree thats possible.  However they are
> > > certainly close, and should provide enjoyable racing.  I'm sorry if
you
> > feel
> > > otherwise.
> > > The effort was certainly put into it, and the beta team felt the
tracks
> > were
> > > balanced pretty well.  I'm not sure what you mean by "Voodoo settings"
> but
> > I
> > > dont remember the use of shaman totems, shrunken heads, or bat wings
> being
> > > involved in the balancing of NR2003 <G>

> > > Jay Taylor

> > > "Marc Collins" <marc_coll...@sympatico.ca> wrote in message
> > > news:Gsa2a.358$3g1.52683@news20.bellglobal.com...
> > > > These are the "voodoo" settings that Papy tweaks to come up with the
> > > > "balanced" playability at each track, which unfortunately has been
> > > seriously
> > > > imbalanced in N3, N4, NR2002 and now, finally, NR2003.

> > > > The rest of the sim is so damned good that it annoys me that
something
> > > like
> > > > this is overlooked year after year.  Oval racing isn't my cup of tea
> to
> > > > start with, but if I have to go to a track and practise and twist my
> > brain
> > > > to drive a car that has a twisted to the side unnatural set-up, I
get
> > > really
> > > > annoyed when my sense of accomplishment or competence (as measured
by
> > how
> > > > well I do against the only benchmark available--the AI at that
track)
> is
> > > as
> > > > much of a crap shoot as surviving an online pick-up race.

> > > > Marc

> > > > "John DiFool" <jdif...@earthlink.net> wrote in message
> > > > news:3E491723.9B016299@earthlink.net...
> > > > > Marc Collins wrote:

> > > > > > So I rest my case...each and every user needing to do a detailed
> > > > research
> > > > > > project just to enjoy the main premise of the game--a
championship
> > > > > > season--is ridiculous.  No wonder so many don't bother and just
> play
> > > > online.
> > > > > > Unfortunately, 90% of the people playing online drive like they
> > should
> > > > have
> > > > > > been forced to complete a few races offline and actually finish
> > higher
> > > > than
> > > > > > 43rd before being allowed online.

> > > > > > When the average experienced user can take the fast set-up at
> > > California
> > > > and
> > > > > > Michigan and achieve roughly the same results (since the skill
> level
> > > is
> > > > the
> > > > > > same), the game will have been designed properly (at least for
> those
> > > two
> > > > > > tracks).  When one track is easy to place in the top 3 and the
> other
> > > you
> > > > > > have to struggle to make it to the bottom 3, something is
> seriously
> > > > wrong.

> > > > > > Marc

> > > > >      [serious snippage]

> > > > > California.ini
> > > > > -----------------
> > > > > [ai track]
> > > > > ai_accel_modifier = 1.00                        ; acceleration
grip
> > > > efficiency
> > > > > ai_decel_modifier = 0.90                        ; braking grip
> > > efficiency
> > > > > ai_fuel_use = 0.98                              ; > 1.0 = more
fuel
> > > > consumed
> > > > > ai_grip_modifier = 1.05                         ; > 1.0 = more
grip
> > > > > ai_drag_modifier = 1.08                         ; > 1.0 = more
drag,
> > > which
> > > > is
> > > > > slower

> > > > > Michigan.ini
> > > > > -----------------
> > > > > [ai track]
> > > > > ai_accel_modifier = 1.20                        ; acceleration
grip
> > > > efficiency
> > > > > ai_decel_modifier = 0.88                        ; braking grip
> > > efficiency
> > > > > ai_fuel_use = 0.96                              ; > 1.0 = more
fuel
> > > > consumed
> > > > > ai_grip_modifier = 1.06                         ; > 1.0 = more
grip
> > > > > ai_drag_modifier = 1.04                         ; > 1.0 = more
drag,
> > > which
> > > > is
> > > > > slower

> > > > > The biggest difference is in the accel modifier (which I suppose
> > > > > means less slippage when accelerating out of the corners).  For
the
> > > > > others there are small advantages for the Michigan AI (and why the
> > > > > hey should fuel use be a variable from track to track?  From trial
> > > > > and error I found that .93 will ensure the AI pits when you do-
> > > > > and I will note that fuel mileage for _me_ seems better than
> before).

> > > > >     John DiFool

> > > > > --
> > > > > ============================================
> > > > > Reach heaven far too high
> > > > > ============================================

Jay Taylo

NR2003: AI still inconsistent from one track to the next

by Jay Taylo » Sat, 15 Feb 2003 14:00:09

So this all comes down to looking for setups?

To be honest I really dont feel like just tossing out setups, after the
amount of time I've put into them.  I'm not saying their are anything
special, but they work great for me.  I happen to run in a highly
competitive open setup league, and so I guess I tend not to want to just
give them.  Agree with it or not, this is my main reasoning.   Beyond that I
really dont feel any need to "prove" anything.  I'll discuss the AI, and do
my best to explain what I can. If you dont agree with me, thats your right
I'm just trying to offer what I know.  Its seems so far that I've tried my
best to answer your questions, and all we are down to is you asking for my
setups.  So save the "put your money where your mouth is" crap for someone
else.

Jay

"Marc Collins" <marc_coll...@sympatico.ca> wrote in message

news:Sl_2a.32947$Ww1.800055@news20.bellglobal.com...
> Oh well, I guess no one is willing to put their money where there mouth is
> by ponying up a couple of set-ups.

> Marc

> "Jay Taylor" <jma...@prodigy.net> wrote in message
> news:Veg2a.3869$dl.1715@newssvr16.news.prodigy.com...
> > First of all if you want to match up verse the AI in practice you better
> > load  a qual setup not a race, as through out testing you would always
see
> a
> > few AI doing more or less qual runs, while others were more in race
trim.
> > Practice is just that, you wont find a hard line on exactly what they
will
> > run for the race.

> > As for ai_grip_modifier, and other settings the reason is simple.  The
AI
> do
> > not run on exactly the same physics the player does. If they did you
might
> > be able to run with a dozen cars max on the fastest of systems.  So they
> > have a somewhat simplifed physics model to keep CPU overhead reasonable.
> > Therefore after you have lines run for a track that work the way you
want
> > them to, and dont have any choke points that mess up the AI, you use
these
> > values to fine tune the AI to get them to drive the track as close to
the
> > way the typical player would.  These are  crucial in smoothing out the
AI,
> > and keeping them in line with the player.  If everything was equal
physics
> > wise, and the AI had the same knowledge the player does about getting
> around
> > the track then these values wouldn't be needed.  However until someone
> finds
> > away to do all this and make it run well on a normal system, you have to
> > make adjustments, and try to dial things in with what you have to work
> with.

> > Jay Taylor

> > "Marc Collins" <marc_coll...@sympatico.ca> wrote in message
> > news:Lpf2a.30455$ns3.436218@news20.bellglobal.com...
> > > What I mean by Voodoo is things like ai_accel_modifier...the layman
> would
> > > ask why would that vary from on track to the next?  Same goes for
> > > ai_grip_modifier...if the track surface has a certain grip, varying
from
> > > track to track of course, why would you need a modifier for this?

> > > I have no doubt that many, many hours, sweat and tears were poured
into
> > the
> > > effort.  But I think I explained why the result is less than
> satisfactory.
> > > And I will eat my hat and shut up about this forever as soon as you or
> > > someone else provides two "fast" set-ups, one for Michigan and one for
> > > California, that permit an experienced user/driver/racer to hit the
top
> 10
> > > in each with approximately the same effort...since that would be
> realistic
> > > and also demonstrate balanced gameplay as a bonus.

> > > If I can't do a single top-10 lap at Michigan in practice mode with
the
> AI
> > > at 100% when I can beat the AI at 100% in GPL at every track (except
the
> > > 'Ring) and easily do top 5's at the sister track of California, then
> > > something is wrong...and it ain't me.  It's either the game or the
> set-up
> > > and I am more than willing to believe you that it is the set-up since
> you
> > > are intimately involved with the title and I am a lowly
purchaser/user.

> > > As I asked in the other part of this thread, please attach your two
> > set-ups
> > > and we'll see what results.

> > > Thanks,

> > > Marc

> > > "Jay Taylor" <jma...@prodigy.net> wrote in message
> > > news:wKa2a.3765$Ly6.1187@newssvr16.news.prodigy.com...
> > > > Over looked?
> > > > If you had any idea the amount of time that went into play balancing
> > > NR2003
> > > > I doubt you'd say it was over looked.  Perhaps the settings don't
work
> > > > perfectly for everyone, I'll agree thats possible.  However they are
> > > > certainly close, and should provide enjoyable racing.  I'm sorry if
> you
> > > feel
> > > > otherwise.
> > > > The effort was certainly put into it, and the beta team felt the
> tracks
> > > were
> > > > balanced pretty well.  I'm not sure what you mean by "Voodoo
settings"
> > but
> > > I
> > > > dont remember the use of shaman totems, shrunken heads, or bat wings
> > being
> > > > involved in the balancing of NR2003 <G>

> > > > Jay Taylor

> > > > "Marc Collins" <marc_coll...@sympatico.ca> wrote in message
> > > > news:Gsa2a.358$3g1.52683@news20.bellglobal.com...
> > > > > These are the "voodoo" settings that Papy tweaks to come up with
the
> > > > > "balanced" playability at each track, which unfortunately has been
> > > > seriously
> > > > > imbalanced in N3, N4, NR2002 and now, finally, NR2003.

> > > > > The rest of the sim is so damned good that it annoys me that
> something
> > > > like
> > > > > this is overlooked year after year.  Oval racing isn't my cup of
tea
> > to
> > > > > start with, but if I have to go to a track and practise and twist
my
> > > brain
> > > > > to drive a car that has a twisted to the side unnatural set-up, I
> get
> > > > really
> > > > > annoyed when my sense of accomplishment or competence (as measured
> by
> > > how
> > > > > well I do against the only benchmark available--the AI at that
> track)
> > is
> > > > as
> > > > > much of a crap shoot as surviving an online pick-up race.

> > > > > Marc

> > > > > "John DiFool" <jdif...@earthlink.net> wrote in message
> > > > > news:3E491723.9B016299@earthlink.net...
> > > > > > Marc Collins wrote:

> > > > > > > So I rest my case...each and every user needing to do a
detailed
> > > > > research
> > > > > > > project just to enjoy the main premise of the game--a
> championship
> > > > > > > season--is ridiculous.  No wonder so many don't bother and
just
> > play
> > > > > online.
> > > > > > > Unfortunately, 90% of the people playing online drive like
they
> > > should
> > > > > have
> > > > > > > been forced to complete a few races offline and actually
finish
> > > higher
> > > > > than
> > > > > > > 43rd before being allowed online.

> > > > > > > When the average experienced user can take the fast set-up at
> > > > California
> > > > > and
> > > > > > > Michigan and achieve roughly the same results (since the skill
> > level
> > > > is
> > > > > the
> > > > > > > same), the game will have been designed properly (at least for
> > those
> > > > two
> > > > > > > tracks).  When one track is easy to place in the top 3 and the
> > other
> > > > you
> > > > > > > have to struggle to make it to the bottom 3, something is
> > seriously
> > > > > wrong.

> > > > > > > Marc

> > > > > >      [serious snippage]

> > > > > > California.ini
> > > > > > -----------------
> > > > > > [ai track]
> > > > > > ai_accel_modifier = 1.00                        ; acceleration
> grip
> > > > > efficiency
> > > > > > ai_decel_modifier = 0.90                        ; braking grip
> > > > efficiency
> > > > > > ai_fuel_use = 0.98                              ; > 1.0 = more
> fuel
> > > > > consumed
> > > > > > ai_grip_modifier = 1.05                         ; > 1.0 = more
> grip
> > > > > > ai_drag_modifier = 1.08                         ; > 1.0 = more
> drag,
> > > > which
> > > > > is
> > > > > > slower

> > > > > > Michigan.ini
> > > > > > -----------------
> > > > > > [ai track]
> > > > > > ai_accel_modifier = 1.20                        ; acceleration
> grip
> > > > > efficiency
> > > > > > ai_decel_modifier = 0.88                        ; braking grip
> > > > efficiency
> > > > > > ai_fuel_use = 0.96                              ; > 1.0 = more
> fuel
> > > > > consumed
> > > > > > ai_grip_modifier = 1.06                         ; > 1.0 = more
> grip
> > > > > > ai_drag_modifier = 1.04                         ; > 1.0 = more
> drag,
> > > > which
> > > > > is
> > > > > > slower

> > > > > > The biggest difference is in the accel modifier (which I suppose
> > > > > > means less slippage when accelerating out of the corners).  For
> the
> > > > > > others there are small advantages for the Michigan AI (and why
the
> > > > > > hey should fuel use be a variable from track to track?  From
trial
> > > > > > and error I found that .93 will ensure the AI pits when you do-
> > > > > > and I will note that fuel mileage for _me_ seems better than
> > before).

> > > > > >     John DiFool

> > > > > > --
> > > > > > ============================================
> > > > > > Reach heaven far too high
> > > > > > ============================================

Dave Boyl

NR2003: AI still inconsistent from one track to the next

by Dave Boyl » Sat, 15 Feb 2003 14:33:36


Why don't you just turn on the Auto AI like was suggested earlier? Isn't
this specifically what it's for?

db

Marc Collin

NR2003: AI still inconsistent from one track to the next

by Marc Collin » Sun, 16 Feb 2003 12:52:02

Well first off Jay this is a public newsgroup, so I am addressing my request
to everyone.  I said from the outset that if it simply that Bob Stanley
doesn't know how to set-up a car for NR2003 and others do have set-ups that
you can run competitively at Michigan and California, then so be it.
However, you claimed you can run more or less equally well against the AI at
these two tracks--as everyone should be able to if the gameplay was tweaked
properly.  Let's be honest, they are probably the two easiest tracks and
they are so similar that you cannot argue that there is some special trick
or skill required at one that isn't required at the other.

I have never found set-ups in N4, NR2002 or so far in NR2003 that permit
this.  Bob Stanley has publicly stated that the AI speeds are unrealistic
(this was in the past referring to N4 and NR2002) at these two tracks and
that he doesn't even bother trying to adjust his set-up to duel with them at
100%--he just sets it up to be the best it can to the best of his ability.
It appears to me that NR2003 is exactly the same in this regard, but you
deny that.  I am sincerely deferring to your expertise and experience, but
in the absence of any evidence, I am asking you to contribute just two
set-ups, for the two easiest tracks.

If you don't want to provide that, I understand, especially if you are in a
league.  But perhaps you could find ANY set-up wiz out there who has a
set-up that works equally well at Michigan and California against 100% AI.
I haven't been able to find one, but I am sure you will know more than me.
ANY person's set-ups that work equally well at these two tracks will prove
you are right and I am wrong.

Thanks,

Marc

"Jay Taylor" <jma...@prodigy.net> wrote in message

news:tJ_2a.1637$8s5.1000@newssvr19.news.prodigy.com...
> So this all comes down to looking for setups?

> To be honest I really dont feel like just tossing out setups, after the
> amount of time I've put into them.  I'm not saying their are anything
> special, but they work great for me.  I happen to run in a highly
> competitive open setup league, and so I guess I tend not to want to just
> give them.  Agree with it or not, this is my main reasoning.   Beyond that
I
> really dont feel any need to "prove" anything.  I'll discuss the AI, and
do
> my best to explain what I can. If you dont agree with me, thats your right
> I'm just trying to offer what I know.  Its seems so far that I've tried my
> best to answer your questions, and all we are down to is you asking for my
> setups.  So save the "put your money where your mouth is" crap for someone
> else.

> Jay

> "Marc Collins" <marc_coll...@sympatico.ca> wrote in message
> news:Sl_2a.32947$Ww1.800055@news20.bellglobal.com...
> > Oh well, I guess no one is willing to put their money where there mouth
is
> > by ponying up a couple of set-ups.

> > Marc

> > "Jay Taylor" <jma...@prodigy.net> wrote in message
> > news:Veg2a.3869$dl.1715@newssvr16.news.prodigy.com...
> > > First of all if you want to match up verse the AI in practice you
better
> > > load  a qual setup not a race, as through out testing you would always
> see
> > a
> > > few AI doing more or less qual runs, while others were more in race
> trim.
> > > Practice is just that, you wont find a hard line on exactly what they
> will
> > > run for the race.

> > > As for ai_grip_modifier, and other settings the reason is simple.  The
> AI
> > do
> > > not run on exactly the same physics the player does. If they did you
> might
> > > be able to run with a dozen cars max on the fastest of systems.  So
they
> > > have a somewhat simplifed physics model to keep CPU overhead
reasonable.
> > > Therefore after you have lines run for a track that work the way you
> want
> > > them to, and dont have any choke points that mess up the AI, you use
> these
> > > values to fine tune the AI to get them to drive the track as close to
> the
> > > way the typical player would.  These are  crucial in smoothing out the
> AI,
> > > and keeping them in line with the player.  If everything was equal
> physics
> > > wise, and the AI had the same knowledge the player does about getting
> > around
> > > the track then these values wouldn't be needed.  However until someone
> > finds
> > > away to do all this and make it run well on a normal system, you have
to
> > > make adjustments, and try to dial things in with what you have to work
> > with.

> > > Jay Taylor

> > > "Marc Collins" <marc_coll...@sympatico.ca> wrote in message
> > > news:Lpf2a.30455$ns3.436218@news20.bellglobal.com...
> > > > What I mean by Voodoo is things like ai_accel_modifier...the layman
> > would
> > > > ask why would that vary from on track to the next?  Same goes for
> > > > ai_grip_modifier...if the track surface has a certain grip, varying
> from
> > > > track to track of course, why would you need a modifier for this?

> > > > I have no doubt that many, many hours, sweat and tears were poured
> into
> > > the
> > > > effort.  But I think I explained why the result is less than
> > satisfactory.
> > > > And I will eat my hat and shut up about this forever as soon as you
or
> > > > someone else provides two "fast" set-ups, one for Michigan and one
for
> > > > California, that permit an experienced user/driver/racer to hit the
> top
> > 10
> > > > in each with approximately the same effort...since that would be
> > realistic
> > > > and also demonstrate balanced gameplay as a bonus.

> > > > If I can't do a single top-10 lap at Michigan in practice mode with
> the
> > AI
> > > > at 100% when I can beat the AI at 100% in GPL at every track (except
> the
> > > > 'Ring) and easily do top 5's at the sister track of California, then
> > > > something is wrong...and it ain't me.  It's either the game or the
> > set-up
> > > > and I am more than willing to believe you that it is the set-up
since
> > you
> > > > are intimately involved with the title and I am a lowly
> purchaser/user.

> > > > As I asked in the other part of this thread, please attach your two
> > > set-ups
> > > > and we'll see what results.

> > > > Thanks,

> > > > Marc

> > > > "Jay Taylor" <jma...@prodigy.net> wrote in message
> > > > news:wKa2a.3765$Ly6.1187@newssvr16.news.prodigy.com...
> > > > > Over looked?
> > > > > If you had any idea the amount of time that went into play
balancing
> > > > NR2003
> > > > > I doubt you'd say it was over looked.  Perhaps the settings don't
> work
> > > > > perfectly for everyone, I'll agree thats possible.  However they
are
> > > > > certainly close, and should provide enjoyable racing.  I'm sorry
if
> > you
> > > > feel
> > > > > otherwise.
> > > > > The effort was certainly put into it, and the beta team felt the
> > tracks
> > > > were
> > > > > balanced pretty well.  I'm not sure what you mean by "Voodoo
> settings"
> > > but
> > > > I
> > > > > dont remember the use of shaman totems, shrunken heads, or bat
wings
> > > being
> > > > > involved in the balancing of NR2003 <G>

> > > > > Jay Taylor

> > > > > "Marc Collins" <marc_coll...@sympatico.ca> wrote in message
> > > > > news:Gsa2a.358$3g1.52683@news20.bellglobal.com...
> > > > > > These are the "voodoo" settings that Papy tweaks to come up with
> the
> > > > > > "balanced" playability at each track, which unfortunately has
been
> > > > > seriously
> > > > > > imbalanced in N3, N4, NR2002 and now, finally, NR2003.

> > > > > > The rest of the sim is so damned good that it annoys me that
> > something
> > > > > like
> > > > > > this is overlooked year after year.  Oval racing isn't my cup of
> tea
> > > to
> > > > > > start with, but if I have to go to a track and practise and
twist
> my
> > > > brain
> > > > > > to drive a car that has a twisted to the side unnatural set-up,
I
> > get
> > > > > really
> > > > > > annoyed when my sense of accomplishment or competence (as
measured
> > by
> > > > how
> > > > > > well I do against the only benchmark available--the AI at that
> > track)
> > > is
> > > > > as
> > > > > > much of a crap shoot as surviving an online pick-up race.

> > > > > > Marc

> > > > > > "John DiFool" <jdif...@earthlink.net> wrote in message
> > > > > > news:3E491723.9B016299@earthlink.net...
> > > > > > > Marc Collins wrote:

> > > > > > > > So I rest my case...each and every user needing to do a
> detailed
> > > > > > research
> > > > > > > > project just to enjoy the main premise of the game--a
> > championship
> > > > > > > > season--is ridiculous.  No wonder so many don't bother and
> just
> > > play
> > > > > > online.
> > > > > > > > Unfortunately, 90% of the people playing online drive like
> they
> > > > should
> > > > > > have
> > > > > > > > been forced to complete a few races offline and actually
> finish
> > > > higher
> > > > > > than
> > > > > > > > 43rd before being allowed online.

> > > > > > > > When the average experienced user can take the fast set-up
at
> > > > > California
> > > > > > and
> > > > > > > > Michigan and achieve roughly the same results (since the
skill
> > > level
> > > > > is
> > > > > > the
> > > > > > > > same), the game will have been designed properly (at least
for
> > > those
> > > > > two
> > > > > > > > tracks).  When one track is easy to place in the top 3 and
the
> > > other
> > > > > you
> > > > > > > > have to struggle to make it to the bottom 3, something is
> > > seriously
> > > > > > wrong.

> > > > > > > > Marc

> > > > > > >      [serious snippage]

> > > > > > > California.ini
> > > > > > > -----------------
> > > > > > > [ai track]
> > > > > > > ai_accel_modifier = 1.00                        ; acceleration
> > grip
> > > > > > efficiency
> > > > > > > ai_decel_modifier = 0.90                        ; braking grip
> > > > > efficiency

...

read more »

Marc Collin

NR2003: AI still inconsistent from one track to the next

by Marc Collin » Sun, 16 Feb 2003 12:53:23

You have completely missed the point if that is your serious suggestion.
Using the Auto AI is the opposite of running a realistic/accurate simulated
season.

Marc




> > Oh well, I guess no one is willing to put their money where there mouth
is
> > by ponying up a couple of set-ups.

> > Marc

> Why don't you just turn on the Auto AI like was suggested earlier? Isn't
> this specifically what it's for?

> db

Dave Boyl

NR2003: AI still inconsistent from one track to the next

by Dave Boyl » Sun, 16 Feb 2003 14:34:57

Well, I think you are trying to make your point based on a false assumption.
That being that you are equally adept at Michigan as you are at California
(and that you should be at all the other tracks.) I'm not saying you are
wrong, but you can't seriously expect me to believe that you have gotten
every bit of speed possible out of the car at *both* Michigan and California
in less than a week. Especially when you are driving setups made for someone
else. Regardless of how similar each track might seem, they all are
different and some people happen to be better at some tracks than others.
You don't know how good Bob is at either of these tracks compared to the
rest of the beta testers or the racing public in general. You also don't
know at what stage of the beta testing that Bob made each of the setups and
how long he spent optimizing each of them after the last of the many physics
changes. There are just way too many unknowns to take this seriously.
Especially when there is a solution you could implement yourself.

How is using the Auto AI the opposite of running what you deem to be a
realistic/accurate simulated season? Why don't you just set the
AutoAIStrength value for each track in your playrecs.ini to what you feel
the AI strength should be, then run your season on Auto?

db


Bill Bollinge

NR2003: AI still inconsistent from one track to the next

by Bill Bollinge » Sun, 16 Feb 2003 22:56:43

Marc, 90% of oval racing is about setups.  90% of road racing is the driver.
If you can't beat the AI at 100% at Michigan, it should tell you that you
need to spend more time making/learning setups.  You need to get in there
and learn how to do it yourself.


> What I mean by Voodoo is things like ai_accel_modifier...the layman would
> ask why would that vary from on track to the next?  Same goes for
> ai_grip_modifier...if the track surface has a certain grip, varying from
> track to track of course, why would you need a modifier for this?

> I have no doubt that many, many hours, sweat and tears were poured into
the
> effort.  But I think I explained why the result is less than satisfactory.
> And I will eat my hat and shut up about this forever as soon as you or
> someone else provides two "fast" set-ups, one for Michigan and one for
> California, that permit an experienced user/driver/racer to hit the top 10
> in each with approximately the same effort...since that would be realistic
> and also demonstrate balanced gameplay as a bonus.

> If I can't do a single top-10 lap at Michigan in practice mode with the AI
> at 100% when I can beat the AI at 100% in GPL at every track (except the
> 'Ring) and easily do top 5's at the sister track of California, then
> something is wrong...and it ain't me.  It's either the game or the set-up
> and I am more than willing to believe you that it is the set-up since you
> are intimately involved with the title and I am a lowly purchaser/user.

> As I asked in the other part of this thread, please attach your two
set-ups
> and we'll see what results.

> Thanks,

> Marc



> > Over looked?
> > If you had any idea the amount of time that went into play balancing
> NR2003
> > I doubt you'd say it was over looked.  Perhaps the settings don't work
> > perfectly for everyone, I'll agree thats possible.  However they are
> > certainly close, and should provide enjoyable racing.  I'm sorry if you
> feel
> > otherwise.
> > The effort was certainly put into it, and the beta team felt the tracks
> were
> > balanced pretty well.  I'm not sure what you mean by "Voodoo settings"
but
> I
> > dont remember the use of shaman totems, shrunken heads, or bat wings
being
> > involved in the balancing of NR2003 <G>

> > Jay Taylor



> > > These are the "voodoo" settings that Papy tweaks to come up with the
> > > "balanced" playability at each track, which unfortunately has been
> > seriously
> > > imbalanced in N3, N4, NR2002 and now, finally, NR2003.

> > > The rest of the sim is so damned good that it annoys me that something
> > like
> > > this is overlooked year after year.  Oval racing isn't my cup of tea
to
> > > start with, but if I have to go to a track and practise and twist my
> brain
> > > to drive a car that has a twisted to the side unnatural set-up, I get
> > really
> > > annoyed when my sense of accomplishment or competence (as measured by
> how
> > > well I do against the only benchmark available--the AI at that track)
is
> > as
> > > much of a ***shoot as surviving an online pick-up race.

> > > Marc




> > > > > So I rest my case...each and every user needing to do a detailed
> > > research
> > > > > project just to enjoy the main premise of the game--a championship
> > > > > season--is ridiculous.  No wonder so many don't bother and just
play
> > > online.
> > > > > Unfortunately, 90% of the people playing online drive like they
> should
> > > have
> > > > > been forced to complete a few races offline and actually finish
> higher
> > > than
> > > > > 43rd before being allowed online.

> > > > > When the average experienced user can take the fast set-up at
> > California
> > > and
> > > > > Michigan and achieve roughly the same results (since the skill
level
> > is
> > > the
> > > > > same), the game will have been designed properly (at least for
those
> > two
> > > > > tracks).  When one track is easy to place in the top 3 and the
other
> > you
> > > > > have to struggle to make it to the bottom 3, something is
seriously
> > > wrong.

> > > > > Marc

> > > >      [serious snippage]

> > > > California.ini
> > > > -----------------
> > > > [ai track]
> > > > ai_accel_modifier = 1.00                        ; acceleration grip
> > > efficiency
> > > > ai_decel_modifier = 0.90                        ; braking grip
> > efficiency
> > > > ai_fuel_use = 0.98                              ; > 1.0 = more fuel
> > > consumed
> > > > ai_grip_modifier = 1.05                         ; > 1.0 = more grip
> > > > ai_drag_modifier = 1.08                         ; > 1.0 = more drag,
> > which
> > > is
> > > > slower

> > > > Michigan.ini
> > > > -----------------
> > > > [ai track]
> > > > ai_accel_modifier = 1.20                        ; acceleration grip
> > > efficiency
> > > > ai_decel_modifier = 0.88                        ; braking grip
> > efficiency
> > > > ai_fuel_use = 0.96                              ; > 1.0 = more fuel
> > > consumed
> > > > ai_grip_modifier = 1.06                         ; > 1.0 = more grip
> > > > ai_drag_modifier = 1.04                         ; > 1.0 = more drag,
> > which
> > > is
> > > > slower

> > > > The biggest difference is in the accel modifier (which I suppose
> > > > means less slippage when accelerating out of the corners).  For the
> > > > others there are small advantages for the Michigan AI (and why the
> > > > hey should fuel use be a variable from track to track?  From trial
> > > > and error I found that .93 will ensure the AI pits when you do-
> > > > and I will note that fuel mileage for _me_ seems better than
before).

> > > >     John DiFool

> > > > --
> > > > ============================================
> > > > Reach heaven far too high
> > > > ============================================

Bill Bollinge

NR2003: AI still inconsistent from one track to the next

by Bill Bollinge » Sun, 16 Feb 2003 22:59:05

I can BEAT the AI at 102% at Michigan.  Havn't worked at Cali.  Just get in
there and work on your setups.

BTW, this is from starting at the back of the field in a 20 car race.

"Marc Collins" <marc_coll...@sympatico.ca> wrote in message

news:FPi3a.1089$bU6.146636@news20.bellglobal.com...
> Well first off Jay this is a public newsgroup, so I am addressing my
request
> to everyone.  I said from the outset that if it simply that Bob Stanley
> doesn't know how to set-up a car for NR2003 and others do have set-ups
that
> you can run competitively at Michigan and California, then so be it.
> However, you claimed you can run more or less equally well against the AI
at
> these two tracks--as everyone should be able to if the gameplay was
tweaked
> properly.  Let's be honest, they are probably the two easiest tracks and
> they are so similar that you cannot argue that there is some special trick
> or skill required at one that isn't required at the other.

> I have never found set-ups in N4, NR2002 or so far in NR2003 that permit
> this.  Bob Stanley has publicly stated that the AI speeds are unrealistic
> (this was in the past referring to N4 and NR2002) at these two tracks and
> that he doesn't even bother trying to adjust his set-up to duel with them
at
> 100%--he just sets it up to be the best it can to the best of his ability.
> It appears to me that NR2003 is exactly the same in this regard, but you
> deny that.  I am sincerely deferring to your expertise and experience, but
> in the absence of any evidence, I am asking you to contribute just two
> set-ups, for the two easiest tracks.

> If you don't want to provide that, I understand, especially if you are in
a
> league.  But perhaps you could find ANY set-up wiz out there who has a
> set-up that works equally well at Michigan and California against 100% AI.
> I haven't been able to find one, but I am sure you will know more than me.
> ANY person's set-ups that work equally well at these two tracks will prove
> you are right and I am wrong.

> Thanks,

> Marc

> "Jay Taylor" <jma...@prodigy.net> wrote in message
> news:tJ_2a.1637$8s5.1000@newssvr19.news.prodigy.com...
> > So this all comes down to looking for setups?

> > To be honest I really dont feel like just tossing out setups, after the
> > amount of time I've put into them.  I'm not saying their are anything
> > special, but they work great for me.  I happen to run in a highly
> > competitive open setup league, and so I guess I tend not to want to just
> > give them.  Agree with it or not, this is my main reasoning.   Beyond
that
> I
> > really dont feel any need to "prove" anything.  I'll discuss the AI, and
> do
> > my best to explain what I can. If you dont agree with me, thats your
right
> > I'm just trying to offer what I know.  Its seems so far that I've tried
my
> > best to answer your questions, and all we are down to is you asking for
my
> > setups.  So save the "put your money where your mouth is" crap for
someone
> > else.

> > Jay

> > "Marc Collins" <marc_coll...@sympatico.ca> wrote in message
> > news:Sl_2a.32947$Ww1.800055@news20.bellglobal.com...
> > > Oh well, I guess no one is willing to put their money where there
mouth
> is
> > > by ponying up a couple of set-ups.

> > > Marc

> > > "Jay Taylor" <jma...@prodigy.net> wrote in message
> > > news:Veg2a.3869$dl.1715@newssvr16.news.prodigy.com...
> > > > First of all if you want to match up verse the AI in practice you
> better
> > > > load  a qual setup not a race, as through out testing you would
always
> > see
> > > a
> > > > few AI doing more or less qual runs, while others were more in race
> > trim.
> > > > Practice is just that, you wont find a hard line on exactly what
they
> > will
> > > > run for the race.

> > > > As for ai_grip_modifier, and other settings the reason is simple.
The
> > AI
> > > do
> > > > not run on exactly the same physics the player does. If they did you
> > might
> > > > be able to run with a dozen cars max on the fastest of systems.  So
> they
> > > > have a somewhat simplifed physics model to keep CPU overhead
> reasonable.
> > > > Therefore after you have lines run for a track that work the way you
> > want
> > > > them to, and dont have any choke points that mess up the AI, you use
> > these
> > > > values to fine tune the AI to get them to drive the track as close
to
> > the
> > > > way the typical player would.  These are  crucial in smoothing out
the
> > AI,
> > > > and keeping them in line with the player.  If everything was equal
> > physics
> > > > wise, and the AI had the same knowledge the player does about
getting
> > > around
> > > > the track then these values wouldn't be needed.  However until
someone
> > > finds
> > > > away to do all this and make it run well on a normal system, you
have
> to
> > > > make adjustments, and try to dial things in with what you have to
work
> > > with.

> > > > Jay Taylor

> > > > "Marc Collins" <marc_coll...@sympatico.ca> wrote in message
> > > > news:Lpf2a.30455$ns3.436218@news20.bellglobal.com...
> > > > > What I mean by Voodoo is things like ai_accel_modifier...the
layman
> > > would
> > > > > ask why would that vary from on track to the next?  Same goes for
> > > > > ai_grip_modifier...if the track surface has a certain grip,
varying
> > from
> > > > > track to track of course, why would you need a modifier for this?

> > > > > I have no doubt that many, many hours, sweat and tears were poured
> > into
> > > > the
> > > > > effort.  But I think I explained why the result is less than
> > > satisfactory.
> > > > > And I will eat my hat and shut up about this forever as soon as
you
> or
> > > > > someone else provides two "fast" set-ups, one for Michigan and one
> for
> > > > > California, that permit an experienced user/driver/racer to hit
the
> > top
> > > 10
> > > > > in each with approximately the same effort...since that would be
> > > realistic
> > > > > and also demonstrate balanced gameplay as a bonus.

> > > > > If I can't do a single top-10 lap at Michigan in practice mode
with
> > the
> > > AI
> > > > > at 100% when I can beat the AI at 100% in GPL at every track
(except
> > the
> > > > > 'Ring) and easily do top 5's at the sister track of California,
then
> > > > > something is wrong...and it ain't me.  It's either the game or the
> > > set-up
> > > > > and I am more than willing to believe you that it is the set-up
> since
> > > you
> > > > > are intimately involved with the title and I am a lowly
> > purchaser/user.

> > > > > As I asked in the other part of this thread, please attach your
two
> > > > set-ups
> > > > > and we'll see what results.

> > > > > Thanks,

> > > > > Marc

> > > > > "Jay Taylor" <jma...@prodigy.net> wrote in message
> > > > > news:wKa2a.3765$Ly6.1187@newssvr16.news.prodigy.com...
> > > > > > Over looked?
> > > > > > If you had any idea the amount of time that went into play
> balancing
> > > > > NR2003
> > > > > > I doubt you'd say it was over looked.  Perhaps the settings
don't
> > work
> > > > > > perfectly for everyone, I'll agree thats possible.  However they
> are
> > > > > > certainly close, and should provide enjoyable racing.  I'm sorry
> if
> > > you
> > > > > feel
> > > > > > otherwise.
> > > > > > The effort was certainly put into it, and the beta team felt the
> > > tracks
> > > > > were
> > > > > > balanced pretty well.  I'm not sure what you mean by "Voodoo
> > settings"
> > > > but
> > > > > I
> > > > > > dont remember the use of shaman totems, shrunken heads, or bat
> wings
> > > > being
> > > > > > involved in the balancing of NR2003 <G>

> > > > > > Jay Taylor

> > > > > > "Marc Collins" <marc_coll...@sympatico.ca> wrote in message
> > > > > > news:Gsa2a.358$3g1.52683@news20.bellglobal.com...
> > > > > > > These are the "voodoo" settings that Papy tweaks to come up
with
> > the
> > > > > > > "balanced" playability at each track, which unfortunately has
> been
> > > > > > seriously
> > > > > > > imbalanced in N3, N4, NR2002 and now, finally, NR2003.

> > > > > > > The rest of the sim is so damned good that it annoys me that
> > > something
> > > > > > like
> > > > > > > this is overlooked year after year.  Oval racing isn't my cup
of
> > tea
> > > > to
> > > > > > > start with, but if I have to go to a track and practise and
> twist
> > my
> > > > > brain
> > > > > > > to drive a car that has a twisted to the side unnatural
set-up,
> I
> > > get
> > > > > > really
> > > > > > > annoyed when my sense of accomplishment or competence (as
> measured
> > > by
> > > > > how
> > > > > > > well I do against the only benchmark available--the AI at that
> > > track)
> > > > is
> > > > > > as
> > > > > > > much of a crap shoot as surviving an online pick-up race.

> > > > > > > Marc

> > > > > > > "John DiFool" <jdif...@earthlink.net> wrote in message
> > > > > > > news:3E491723.9B016299@earthlink.net...
> > > > > > > > Marc Collins wrote:

> > > > > > > > > So I rest my case...each and every user needing to do a
> > detailed
> > > > > > > research
> > > > > > > > > project just to enjoy the main premise of the game--a
> > > championship
> > > > > > > > > season--is ridiculous.  No wonder so many don't bother and
> > just
> > > > play
> > > > > > > online.
> > > > > > > > > Unfortunately, 90% of the people playing online drive like
> > they
> > > > > should
> > > > > > > have
> > > > > > > > > been forced to complete a few races offline and actually
> > finish
> > > > > higher
> > > > > > > than
> > > > > > > > > 43rd before being allowed online.

> > > > > > > > > When the average experienced user can take the fast set-up
> at
> > > > > > California
> > > > > > > and
> > > > > > > > > Michigan and achieve roughly the same results

...

read more »

Marc Collin

NR2003: AI still inconsistent from one track to the next

by Marc Collin » Mon, 17 Feb 2003 01:09:44

What I am asking is whether you find the AI at Calif. and Michigan at 100%
similarly difficult.  They should be.  Try it out and let us know.  I'll bet
you can run circles around the AI at California if you have a set-up and
skills to beat them easily at 102% at Michigan.  Without belabouring the
point again, each track shouldn't be a gamble as to whether the AI are
smart, stupid or somewhere in between.  They should be similarly fast
(hopefully related as closely as possible to real world speeds) at all
tracks.  They aren't, in my experience.

Marc

"Bill Bollinger" <teambricky...@bluemarble.net> wrote in message

news:v4sht7jem1a241@corp.supernews.com...
> I can BEAT the AI at 102% at Michigan.  Havn't worked at Cali.  Just get
in
> there and work on your setups.

> BTW, this is from starting at the back of the field in a 20 car race.

> "Marc Collins" <marc_coll...@sympatico.ca> wrote in message
> news:FPi3a.1089$bU6.146636@news20.bellglobal.com...
> > Well first off Jay this is a public newsgroup, so I am addressing my
> request
> > to everyone.  I said from the outset that if it simply that Bob Stanley
> > doesn't know how to set-up a car for NR2003 and others do have set-ups
> that
> > you can run competitively at Michigan and California, then so be it.
> > However, you claimed you can run more or less equally well against the
AI
> at
> > these two tracks--as everyone should be able to if the gameplay was
> tweaked
> > properly.  Let's be honest, they are probably the two easiest tracks and
> > they are so similar that you cannot argue that there is some special
trick
> > or skill required at one that isn't required at the other.

> > I have never found set-ups in N4, NR2002 or so far in NR2003 that permit
> > this.  Bob Stanley has publicly stated that the AI speeds are
unrealistic
> > (this was in the past referring to N4 and NR2002) at these two tracks
and
> > that he doesn't even bother trying to adjust his set-up to duel with
them
> at
> > 100%--he just sets it up to be the best it can to the best of his
ability.
> > It appears to me that NR2003 is exactly the same in this regard, but you
> > deny that.  I am sincerely deferring to your expertise and experience,
but
> > in the absence of any evidence, I am asking you to contribute just two
> > set-ups, for the two easiest tracks.

> > If you don't want to provide that, I understand, especially if you are
in
> a
> > league.  But perhaps you could find ANY set-up wiz out there who has a
> > set-up that works equally well at Michigan and California against 100%
AI.
> > I haven't been able to find one, but I am sure you will know more than
me.
> > ANY person's set-ups that work equally well at these two tracks will
prove
> > you are right and I am wrong.

> > Thanks,

> > Marc

> > "Jay Taylor" <jma...@prodigy.net> wrote in message
> > news:tJ_2a.1637$8s5.1000@newssvr19.news.prodigy.com...
> > > So this all comes down to looking for setups?

> > > To be honest I really dont feel like just tossing out setups, after
the
> > > amount of time I've put into them.  I'm not saying their are anything
> > > special, but they work great for me.  I happen to run in a highly
> > > competitive open setup league, and so I guess I tend not to want to
just
> > > give them.  Agree with it or not, this is my main reasoning.   Beyond
> that
> > I
> > > really dont feel any need to "prove" anything.  I'll discuss the AI,
and
> > do
> > > my best to explain what I can. If you dont agree with me, thats your
> right
> > > I'm just trying to offer what I know.  Its seems so far that I've
tried
> my
> > > best to answer your questions, and all we are down to is you asking
for
> my
> > > setups.  So save the "put your money where your mouth is" crap for
> someone
> > > else.

> > > Jay

> > > "Marc Collins" <marc_coll...@sympatico.ca> wrote in message
> > > news:Sl_2a.32947$Ww1.800055@news20.bellglobal.com...
> > > > Oh well, I guess no one is willing to put their money where there
> mouth
> > is
> > > > by ponying up a couple of set-ups.

> > > > Marc

> > > > "Jay Taylor" <jma...@prodigy.net> wrote in message
> > > > news:Veg2a.3869$dl.1715@newssvr16.news.prodigy.com...
> > > > > First of all if you want to match up verse the AI in practice you
> > better
> > > > > load  a qual setup not a race, as through out testing you would
> always
> > > see
> > > > a
> > > > > few AI doing more or less qual runs, while others were more in
race
> > > trim.
> > > > > Practice is just that, you wont find a hard line on exactly what
> they
> > > will
> > > > > run for the race.

> > > > > As for ai_grip_modifier, and other settings the reason is simple.
> The
> > > AI
> > > > do
> > > > > not run on exactly the same physics the player does. If they did
you
> > > might
> > > > > be able to run with a dozen cars max on the fastest of systems.
So
> > they
> > > > > have a somewhat simplifed physics model to keep CPU overhead
> > reasonable.
> > > > > Therefore after you have lines run for a track that work the way
you
> > > want
> > > > > them to, and dont have any choke points that mess up the AI, you
use
> > > these
> > > > > values to fine tune the AI to get them to drive the track as close
> to
> > > the
> > > > > way the typical player would.  These are  crucial in smoothing out
> the
> > > AI,
> > > > > and keeping them in line with the player.  If everything was equal
> > > physics
> > > > > wise, and the AI had the same knowledge the player does about
> getting
> > > > around
> > > > > the track then these values wouldn't be needed.  However until
> someone
> > > > finds
> > > > > away to do all this and make it run well on a normal system, you
> have
> > to
> > > > > make adjustments, and try to dial things in with what you have to
> work
> > > > with.

> > > > > Jay Taylor

> > > > > "Marc Collins" <marc_coll...@sympatico.ca> wrote in message
> > > > > news:Lpf2a.30455$ns3.436218@news20.bellglobal.com...
> > > > > > What I mean by Voodoo is things like ai_accel_modifier...the
> layman
> > > > would
> > > > > > ask why would that vary from on track to the next?  Same goes
for
> > > > > > ai_grip_modifier...if the track surface has a certain grip,
> varying
> > > from
> > > > > > track to track of course, why would you need a modifier for
this?

> > > > > > I have no doubt that many, many hours, sweat and tears were
poured
> > > into
> > > > > the
> > > > > > effort.  But I think I explained why the result is less than
> > > > satisfactory.
> > > > > > And I will eat my hat and shut up about this forever as soon as
> you
> > or
> > > > > > someone else provides two "fast" set-ups, one for Michigan and
one
> > for
> > > > > > California, that permit an experienced user/driver/racer to hit
> the
> > > top
> > > > 10
> > > > > > in each with approximately the same effort...since that would be
> > > > realistic
> > > > > > and also demonstrate balanced gameplay as a bonus.

> > > > > > If I can't do a single top-10 lap at Michigan in practice mode
> with
> > > the
> > > > AI
> > > > > > at 100% when I can beat the AI at 100% in GPL at every track
> (except
> > > the
> > > > > > 'Ring) and easily do top 5's at the sister track of California,
> then
> > > > > > something is wrong...and it ain't me.  It's either the game or
the
> > > > set-up
> > > > > > and I am more than willing to believe you that it is the set-up
> > since
> > > > you
> > > > > > are intimately involved with the title and I am a lowly
> > > purchaser/user.

> > > > > > As I asked in the other part of this thread, please attach your
> two
> > > > > set-ups
> > > > > > and we'll see what results.

> > > > > > Thanks,

> > > > > > Marc

> > > > > > "Jay Taylor" <jma...@prodigy.net> wrote in message
> > > > > > news:wKa2a.3765$Ly6.1187@newssvr16.news.prodigy.com...
> > > > > > > Over looked?
> > > > > > > If you had any idea the amount of time that went into play
> > balancing
> > > > > > NR2003
> > > > > > > I doubt you'd say it was over looked.  Perhaps the settings
> don't
> > > work
> > > > > > > perfectly for everyone, I'll agree thats possible.  However
they
> > are
> > > > > > > certainly close, and should provide enjoyable racing.  I'm
sorry
> > if
> > > > you
> > > > > > feel
> > > > > > > otherwise.
> > > > > > > The effort was certainly put into it, and the beta team felt
the
> > > > tracks
> > > > > > were
> > > > > > > balanced pretty well.  I'm not sure what you mean by "Voodoo
> > > settings"
> > > > > but
> > > > > > I
> > > > > > > dont remember the use of shaman totems, shrunken heads, or bat
> > wings
> > > > > being
> > > > > > > involved in the balancing of NR2003 <G>

> > > > > > > Jay Taylor

> > > > > > > "Marc Collins" <marc_coll...@sympatico.ca> wrote in message
> > > > > > > news:Gsa2a.358$3g1.52683@news20.bellglobal.com...
> > > > > > > > These are the "voodoo" settings that Papy tweaks to come up
> with
> > > the
> > > > > > > > "balanced" playability at each track, which unfortunately
has
> > been
> > > > > > > seriously
> > > > > > > > imbalanced in N3, N4, NR2002 and now, finally, NR2003.

> > > > > > > > The rest of the sim is so damned good that it annoys me that
> > > > something
> > > > > > > like
> > > > > > > > this is overlooked year after year.  Oval racing isn't my
cup
> of
> > > tea
> > > > > to
> > > > > > > > start with, but if I have to go to a track and practise and
> > twist
> > > my
> > > > > > brain
> > > > > > > > to drive a car that has a twisted to the side unnatural
> set-up,
> > I
> > > > get
> > > > > > > really
> > > > > > > > annoyed when my sense of accomplishment or competence (as
> > measured
> > > > by
> > > > > > how
> > > > > > > > well I do against the only benchmark available--the AI at

...

read more »

Marc Collin

NR2003: AI still inconsistent from one track to the next

by Marc Collin » Mon, 17 Feb 2003 01:12:43

See my reply above.  I am taking about Bob Stanley's set-ups, not my skill
level--which is more than adequate and exceeds that of about 95% of the
purchasers of this title.  Perhaps you should read the whole thread.  This
is a discussion of the realism levels of the speed of the AI, not a plea for
help with my set-ups or driving/racing skills--I don't need that (at least
any more than we all need to improve constantly over time).

Marc

"Bill Bollinger" <teambricky...@bluemarble.net> wrote in message

news:v4shop9qr3vfce@corp.supernews.com...
> Marc, 90% of oval racing is about setups.  90% of road racing is the
driver.
> If you can't beat the AI at 100% at Michigan, it should tell you that you
> need to spend more time making/learning setups.  You need to get in there
> and learn how to do it yourself.

> "Marc Collins" <marc_coll...@sympatico.ca> wrote in message
> news:Lpf2a.30455$ns3.436218@news20.bellglobal.com...
> > What I mean by Voodoo is things like ai_accel_modifier...the layman
would
> > ask why would that vary from on track to the next?  Same goes for
> > ai_grip_modifier...if the track surface has a certain grip, varying from
> > track to track of course, why would you need a modifier for this?

> > I have no doubt that many, many hours, sweat and tears were poured into
> the
> > effort.  But I think I explained why the result is less than
satisfactory.
> > And I will eat my hat and shut up about this forever as soon as you or
> > someone else provides two "fast" set-ups, one for Michigan and one for
> > California, that permit an experienced user/driver/racer to hit the top
10
> > in each with approximately the same effort...since that would be
realistic
> > and also demonstrate balanced gameplay as a bonus.

> > If I can't do a single top-10 lap at Michigan in practice mode with the
AI
> > at 100% when I can beat the AI at 100% in GPL at every track (except the
> > 'Ring) and easily do top 5's at the sister track of California, then
> > something is wrong...and it ain't me.  It's either the game or the
set-up
> > and I am more than willing to believe you that it is the set-up since
you
> > are intimately involved with the title and I am a lowly purchaser/user.

> > As I asked in the other part of this thread, please attach your two
> set-ups
> > and we'll see what results.

> > Thanks,

> > Marc

> > "Jay Taylor" <jma...@prodigy.net> wrote in message
> > news:wKa2a.3765$Ly6.1187@newssvr16.news.prodigy.com...
> > > Over looked?
> > > If you had any idea the amount of time that went into play balancing
> > NR2003
> > > I doubt you'd say it was over looked.  Perhaps the settings don't work
> > > perfectly for everyone, I'll agree thats possible.  However they are
> > > certainly close, and should provide enjoyable racing.  I'm sorry if
you
> > feel
> > > otherwise.
> > > The effort was certainly put into it, and the beta team felt the
tracks
> > were
> > > balanced pretty well.  I'm not sure what you mean by "Voodoo settings"
> but
> > I
> > > dont remember the use of shaman totems, shrunken heads, or bat wings
> being
> > > involved in the balancing of NR2003 <G>

> > > Jay Taylor

> > > "Marc Collins" <marc_coll...@sympatico.ca> wrote in message
> > > news:Gsa2a.358$3g1.52683@news20.bellglobal.com...
> > > > These are the "voodoo" settings that Papy tweaks to come up with the
> > > > "balanced" playability at each track, which unfortunately has been
> > > seriously
> > > > imbalanced in N3, N4, NR2002 and now, finally, NR2003.

> > > > The rest of the sim is so damned good that it annoys me that
something
> > > like
> > > > this is overlooked year after year.  Oval racing isn't my cup of tea
> to
> > > > start with, but if I have to go to a track and practise and twist my
> > brain
> > > > to drive a car that has a twisted to the side unnatural set-up, I
get
> > > really
> > > > annoyed when my sense of accomplishment or competence (as measured
by
> > how
> > > > well I do against the only benchmark available--the AI at that
track)
> is
> > > as
> > > > much of a crap shoot as surviving an online pick-up race.

> > > > Marc

> > > > "John DiFool" <jdif...@earthlink.net> wrote in message
> > > > news:3E491723.9B016299@earthlink.net...
> > > > > Marc Collins wrote:

> > > > > > So I rest my case...each and every user needing to do a detailed
> > > > research
> > > > > > project just to enjoy the main premise of the game--a
championship
> > > > > > season--is ridiculous.  No wonder so many don't bother and just
> play
> > > > online.
> > > > > > Unfortunately, 90% of the people playing online drive like they
> > should
> > > > have
> > > > > > been forced to complete a few races offline and actually finish
> > higher
> > > > than
> > > > > > 43rd before being allowed online.

> > > > > > When the average experienced user can take the fast set-up at
> > > California
> > > > and
> > > > > > Michigan and achieve roughly the same results (since the skill
> level
> > > is
> > > > the
> > > > > > same), the game will have been designed properly (at least for
> those
> > > two
> > > > > > tracks).  When one track is easy to place in the top 3 and the
> other
> > > you
> > > > > > have to struggle to make it to the bottom 3, something is
> seriously
> > > > wrong.

> > > > > > Marc

> > > > >      [serious snippage]

> > > > > California.ini
> > > > > -----------------
> > > > > [ai track]
> > > > > ai_accel_modifier = 1.00                        ; acceleration
grip
> > > > efficiency
> > > > > ai_decel_modifier = 0.90                        ; braking grip
> > > efficiency
> > > > > ai_fuel_use = 0.98                              ; > 1.0 = more
fuel
> > > > consumed
> > > > > ai_grip_modifier = 1.05                         ; > 1.0 = more
grip
> > > > > ai_drag_modifier = 1.08                         ; > 1.0 = more
drag,
> > > which
> > > > is
> > > > > slower

> > > > > Michigan.ini
> > > > > -----------------
> > > > > [ai track]
> > > > > ai_accel_modifier = 1.20                        ; acceleration
grip
> > > > efficiency
> > > > > ai_decel_modifier = 0.88                        ; braking grip
> > > efficiency
> > > > > ai_fuel_use = 0.96                              ; > 1.0 = more
fuel
> > > > consumed
> > > > > ai_grip_modifier = 1.06                         ; > 1.0 = more
grip
> > > > > ai_drag_modifier = 1.04                         ; > 1.0 = more
drag,
> > > which
> > > > is
> > > > > slower

> > > > > The biggest difference is in the accel modifier (which I suppose
> > > > > means less slippage when accelerating out of the corners).  For
the
> > > > > others there are small advantages for the Michigan AI (and why the
> > > > > hey should fuel use be a variable from track to track?  From trial
> > > > > and error I found that .93 will ensure the AI pits when you do-
> > > > > and I will note that fuel mileage for _me_ seems better than
> before).

> > > > >     John DiFool

> > > > > --
> > > > > ============================================
> > > > > Reach heaven far too high
> > > > > ============================================


rec.autos.simulators is a usenet newsgroup formed in December, 1993. As this group was always unmoderated there may be some spam or off topic articles included. Some links do point back to racesimcentral.net as we could not validate the original address. Please report any pages that you believe warrant deletion from this archive (include the link in your email). RaceSimCentral.net is in no way responsible and does not endorse any of the content herein.