Quarterly cost: �0
 
March 29, 2024, 02:21:27 AM +0000 *
Welcome, Guest. Please login or register.
Did you miss your activation email?

Login with username, password and session length
Series
S4455GPP
S4455GPW
S4466CA
S4467F1P
S4467F1W
S4467F2A
S4467F2P
S44JSMT
Recent
S4455GPW Indianapolis …
S4455GPP Indianapolis …
S4466CA Road America (…
S4467F1W Imola (1953-7…
S4467F1P Imola (1953-7…
S4467F2A Ingliston
S4467F2P Ingliston
Forthcoming
S44JSMT Goodwood (Circ…
S44JSMT Boreham Airfie…
 
   Home   Help Search Calendar Login Register     LM2 Replays Rules Links Circuits Teams  
Poll
Question: Should we mandate the use of FairShift with H-shifters?  (Voting closed: January 02, 2015, 07:25:49 PM +0000)
Yes - 5 (41.7%)
No - 7 (58.3%)
Total Voters: 12

Pages: [1]
  Print  
Author Topic: H-Shifters  (Read 1989 times)
0 Members and 1 Guest are viewing this topic.
Phil Thornton
UKGPL Consigliere
Administrator
Hero Member
*****
Posts: 7963


View Profile
« on: December 16, 2014, 07:25:49 PM +0000 »

H-Shifters are very popular and give an added realism to GPL.  GPL Shift is needed to enable the use of H-Shifters with GPL.  However it has recently come to light (well it has been on SRMZ for a while see here) that quicker shifts are possible with GPL Shift which gives an unfair advantage.  GPL Replay Analyzer can be used to check shift times but the analysis of server replays appears to be suspect, client replays appear to give more accurate results.  The apparent advantage is open to some debate however there is a utility called FairShift which cancels out the apparent advantage.  

As with any utility there are potential compatibility issues and so it might not work with all installations.  What do we do if someone genuinely can't install it?  How do we check if someone has or has not installed it?
« Last Edit: December 16, 2014, 07:32:07 PM +0000 by Phil Thornton » Logged
FullMetalGasket
Director, AC
SimRacing.org.uk Staff
Hero Member
****
Posts: 4238



View Profile
« Reply #1 on: December 16, 2014, 08:42:31 PM +0000 »

If I'm honest I don't think we have a problem with that here?
I tend to race with the option (forget what it's called now) that stops gear changes being recognised if you don't press the clutch enough, although can't speak for others.

I believe that at Pro levels engine longevity might become and issue for people flat shifting while throwing the gearstick about!?

Additionally - does this program stop someone with a H-pattern shifter going from say 5th to 2nd?
I'll do this when the circuit suits it and wouldn't be too pleased if I was held in neutral for a second or so while waiting!  Wink
Logged
EvilClive
UKGPL Senior Consultant
UKGPL Moderators
Hero Member
****
Posts: 7700

I always play by the rules.... they are MY rules!


View Profile
« Reply #2 on: December 16, 2014, 09:23:24 PM +0000 »

IMHO
I am totally in favour of keeping things fair and equal in GPL, but I think we have to have a reality check here. I can understand why this issue has been raised, but I am not convinced that it would change any results on the track.

Personally I use a gearstick and GPLShift as a sequential box, pull back to upshift...push forward to downshift. I don't use the clutch when shifting but I do lift off the gas to preserve the engine, although the clutch pedal is calibrated and comes in useful in getting out of gravel traps occasionally. Roll Eyes

If we are talking about 2 uber aliens lapping perfectly at world record pace lap after lap, where one has an "H" shifter and the other is using flappy paddles, then a 0.05 second advantage on each gearchange with the "H" shift will give one an advantage that will probably decide the race.

However, if we check the lap times of the fastest guys who currently race in UKGPL I doubt that any of them can string 3 laps together at a pace that varies by less than 2 tenths, let alone 30 laps with no greater differential than 0.5 secs. One could also argue that the guy using the paddle gearchange has an advantage because he keeps both hands on the wheel and can therefore gear shift mid corner if he wants to, without releasing the steering wheel. Does that give him an advantage of 0.05 secs per corner?? very difficult to tell I think.

There are far bigger margins to be gained or lost elsewhere in the driving of GPL than just with gear shifting? I am sure that I could not tell if the car I was following in a race was using paddles or "H" shift just by the pace of the car on the track, and frankly, I would prefer to concentrate on using what I am comfortable with in terms of controls as this is far more likely to produce faster lap times for me.

As FMG points out, a far bigger advantage is being able to downshift from 5th ( or even 6th in some cars) directly into 2nd or 1st thereby eliminating at least 4 gearchanges at a single corner!!! and saving far more time in one hit than what is being discussed here.

We would be relying completely on each driver using the fairchange software correctly and we would have no way of checking, as the server replay seems inconclusive.
Logged

Evil Waza, now a completely reformed character!
          **NOW AVAILABLE ON TWITCH @ Evilclive67**
Rank   Only when I sweat
Ronniepeterson
Full Member
***
Posts: 1212


View Profile
« Reply #3 on: December 16, 2014, 10:32:36 PM +0000 »

The use of any shifting controller which allows down/upshift times below the minimum thresholds of GPL (0.222/0.383) offers that driver a significant advantage. Some might think "so what, some might think its unfair or even cheating. For sure there are plenty of factors which determine a drivers speed over the course of a single lap and a race. But the use of shifting which tricks GPL and allows times below those minimum thresholds is a major one. As for drivers not being able to string more than three fast laps together that applies to anyone whether they are using a GPL compliant gear change method or not. So it is most likely the advantage is maintained over the course of a single lap and race.  

As I said in my earlier post I admire the speed of these drivers but I am not a fan of the method. I therefore welcome the fact that the issue was originally raised by a moderator, a second moderator revealed that the moderation team have been concerned about this in private and a third moderator has now set up a poll on whether UKGPL needs to do something about it.

I just wonder how may drivers currently using shift controllers which allows down/upshift times below the minimum thresholds of GPL will vote for the compulsory use of the FairShift utility.

For those wavering on how to vote the clue is in the name of the utility  laugh


 
« Last Edit: December 16, 2014, 11:26:06 PM +0000 by Ronniepeterson » Logged
Phil Thornton
UKGPL Consigliere
Administrator
Hero Member
*****
Posts: 7963


View Profile
« Reply #4 on: December 16, 2014, 11:07:48 PM +0000 »

This issue has been on SRMZ for quite a while now but I only found out about it via the mods forum on 26 Nov 2014.  Since I use a H-Shifter I did a few tests of my own.

I checked my client reply of my fastest lap at Watkins Glen in the Privateers race on 23 Nov 2014.  Original without FairShift:


Had another go with Fairshift 2.0 enabled (in training mode and saved the client replay).  Put 3 gallons in so had same fuel load as my race fastest lap but without the benefit of a tow from JonnyO (I was right on his gearbox when I did my fastest lap in the race):So with FairShift:

Managed to better my time but that doesn't prove anything conclusively either way IMO.  The GPLRA report does show a sightly longer time in neutral with the Fairshift lap but does that just mean I was more aggressive in the race when the adrenaline was pumping when I was trying to catch JonnyO? 

So I decided to have a look at the server replay of my race fastest lap:

There is a difference, there are more .111 times on the server replay but the overall time in neutral is longer.  Not sure what conclusion we can draw from this except that we can't rely on analysing the server replay when it comes to gear shift times.  I assume the client is more accurate as there is no internet lag to take into consideration.

I'm not sure GPL Shift does give an advantage in practice but it would still be prudent to mandate the use of Fairshift 2.0 just in case.  Happy to ask the members what they think.  Any modification that does actually give an advantage is against the rules see here.
Logged
dave curtis
UKGPL Server Provider
UKGPL Moderators
Sr. Member
****
Posts: 1507



View Profile
« Reply #5 on: December 17, 2014, 12:08:58 AM +0000 »

I'm not suggesting in my position that I  have much to offer here in technique (definitely don't copy me!) or any insight,  but looks like this is a good place for a bit of a ramble...

I doubt I'd see any difference either way.   But without knowing full details,  I read this as being something that could not easily be tracked/enforced/actively encouraged;  so as a result have voted with a 'no' for now at least.     Although,  granted - IF it would work out to be the same for everybody,  then fair enough!   at which point, I could be talked round!


I guess that most H-shifters probably use GPLshift?

So yes - paddles vs. H-shifters.  Hand-on wheel vs. hand-off thing.   with H-shifter,  I've certainly hit the wrong gear/ended up in neutral too many times;   not sure you'd manage that with paddles/sequential Smiley
How about different control methods? (joypads/joysticks/keys?) which admittedly would probably be subject to the default shift times anyway,   but the mechanics of shifting is different of course.    And then Clive's hybrid sequential shift technique?

Maybe drivers (all/points finishers/top 6/podium?) should provide a sample  (well, client replay) following each event?  Of course,  disconnects & replays failing to be saved etc  probably invites further scrutiny/finger-pointing.  Not to mention the moderator overhead.  In addition to that, from what Phil mentions, it sounds as if the analysis may be not completely straightforward anyway?  


I guess if all drivers were made/encouraged to use shifters & in H-mode, then the starting point is at least consistent?
Does this require additional poll as to gathering current shift methods & details of shift utility?


Looks as though Phil managed to change from 4th->2nd without potential issue,  so maybe skipping gears with H-shifter is still possible/fair after all Smiley


Personally, I was happy using GPLshift  [maybe with a bit inadvertent cheating or external/artificial assistance] - but eventually realised that the enforced FOV limits were causing me an issue with my new wide monitor...
As a result,  currently using ALshift with the registry tweak for G25/G27/other instead.  Although, then also ended up using xpadder so the pit in/out/shift-r stuff would all work from the shifter buttons.   It all ended up being more tedious to get working than just firing up GPLshift.   Maybe moving to fairshift would be straightforward,   but I'm at least a bit wary.

In closing,  despite it being what surely would keep things in harmony if 100% across the board;   I have reasonable doubt despite the intentions.
And if the glove does not fit...


EDIT:  and of course,  not even mentioned clutch.  Is that relevent here at all if the discussions are around gearchanges?    Personally, use it for starts,  not for changing gear,  but it's there to be partially pushed in at other times.  Probably doesn't help, but sometimes just makes things sound a bit better with a slight dip now & again.



Cheers,
Dave.






« Last Edit: December 17, 2014, 12:16:02 AM +0000 by dave curtis » Logged
Ronniepeterson
Full Member
***
Posts: 1212


View Profile
« Reply #6 on: December 17, 2014, 10:10:04 AM +0000 »

From the data above on the FairShift utility I think the inclusion of the word fair is a little generous as the upshift/downshift times seem low compared to the target of 0.222/0.383 even when the utility is used Shocked  However at least the GPL community is trying to do something about this issue. Like wise I appreciate the UKGPL moderation team taking an active interest as well.

As for this belief that the advantage offered is debatable, I don't understand?

Monoco F2 lights race, top six, fastest laps in race, upshift/downshift times

Driver 1: 0.229/0.389
Driver 2: 0.083/0.097
Driver 3: 0.208/0.392
Driver 4: 0.211/0.394
Driver 5: 0.222/0.384
Driver 6: 0.216/0.400

This data seems to show a clear advantage for Driver 2, which helped to contribute to a lap 0.733s faster than the next best.
Logged
Cookie
UKGPL Divisional Moderator
UKGPL Moderators
Hero Member
****
Posts: 6189


Chris Amon fan


View Profile
« Reply #7 on: December 17, 2014, 11:32:30 AM +0000 »

The main advantage comes from the time spent in neutral per lap during upshift: 2nd - 0.583sec vs 1st - 1.833sec  Roll Eyes

I first was confronted with this phenomen after some laps in Monza following a driver using a H-shifter without clutch in the same car,
so I got an impression what a difference it makes accelerating out of Lesmo2 and Parabolica...

I admire the drivers using a H-shifter with clutch and grant them the little advantage they have with downshifts eg 5->2

But when it comes to shifting without clutch there seems to be a big grey zone!

Logged

Axel "Cookie"

poor, he who sees no stars without the punch in the face

an aphorism of  Stanislaus Jercy Lec
Tom van Ostade
Full Member
***
Posts: 397


"anything can happen, and it usually does"


View Profile
« Reply #8 on: December 18, 2014, 01:22:31 PM +0000 »

I guess the proof is in the puddin' that the league who invented FairShift two seasons ago is now on it's last legs and needs something major happening in order to survive...

I drive in the same way as Clive, so with a sequential stick shift. I tried H-shifting a couple of times but it's quite hard to do and would take some practice to get right. Yes, people using an H-shifter are faster over a lap, but because of the added difficulty they are also more prone to mistakes, especially when pressurized during a race. Also, they can miss a shift and loose valuable time that way. Obviously, the more powerful the car, the more dire the consequenses if you get a shift wrong. The same way for increased difficulty of track.

So, I would say that ONLY in the combo of easy track (i.e. Monza) and easy car (i.e. 65's) the H-shifters have a significant advantage, where you can get closer to the absolute limit without pushing. However, in these cases it is also easier for the trailing drivers to keep up due to slipstreaming, so much of the H-shift advantage is gone in that case.

So using FairShift or GPL Shift doesn't really make a difference, other than completely destroy the car count.
Logged

Robert Fleurke
Former UKGPL Moderators
Sr. Member
**
Posts: 1701


GPL Forever!


View Profile
« Reply #9 on: December 19, 2014, 12:16:55 AM +0000 »

It is a fact that less neutral upshift time will lead to slightly better accelerating, it's just simple physics.

However, after a lot of (re)thinking, even though Fairshift might help to make the advantage less, it is not that important to me. I would hate to see losing drivers because of enforcing Fairshift, and the community divided. As some pointed out, it might not work for all etc., and in the end the driver makes the difference (or the car).
Logged

Driving for Antipasti Racing

maddog
Sr. Member
****
Posts: 1709


View Profile
« Reply #10 on: December 19, 2014, 11:39:04 AM +0000 »

I think Ronnie has put forward a good argument to highlight why fairshift was developed.  There are drivers gaining an advantage from faster shifts.  There should be a minimum shift time requirement at Ukgpl.  The real F1's of our era required manual shifting, and a pause between gears.  The GPL auto shift option for learners, has a long pause as a reminder of this.  BUT, if detection of short shift times is not reliable, it's a non-starter.  I'd be surprised if that was the case, as we would not have been asked to vote.  I suspect a hairy plot! Grin
Logged
Cookie
UKGPL Divisional Moderator
UKGPL Moderators
Hero Member
****
Posts: 6189


Chris Amon fan


View Profile
« Reply #11 on: December 19, 2014, 03:04:03 PM +0000 »

I found this interesting answer in an SRMZ thread

Quote from: brr
Quote from: JohnnyAck
How can times below .111 be possible? Enlighten me.

The shifting in unmodified gpl.exe responds to "shift up" and "shift down" button presses, and imposes a delay (i.e. a few frames spent in neutral). Using an H-shifter works differently since there is no "shift up/down" but directly the engaged gear or neutral. The patches (gplshift and perhaps others?) which allow one to use an H-shifter must modify the shifting physics and probably just force the gear selected in the shifter. In principle there might not be any delay if the shifter mechanics allows it. I once made an experimental shifter using reed switches and a fridge magnet and it felt like it shifted without delay. I've never used it in races or setting rank laptimes, however.

Removing the time spent in neutral from shifting physics is an advantage as it improves acceleration. Its another question how big an advantage.

So I understand that those times can be made by special shifter hardware!
« Last Edit: December 19, 2014, 03:06:36 PM +0000 by Cookie » Logged

Axel "Cookie"

poor, he who sees no stars without the punch in the face

an aphorism of  Stanislaus Jercy Lec
dave curtis
UKGPL Server Provider
UKGPL Moderators
Sr. Member
****
Posts: 1507



View Profile
« Reply #12 on: December 19, 2014, 05:05:32 PM +0000 »

So I understand that those times can be made by special shifter hardware!

So now,  as well as potentially running same shift method (software implementation) if deemed acceptable,  there is the possibility to develop home-brew shifter hardware to try shave more time per shift? Wink

Every shift is sacred.
Every shift is great.
If a shift is cheated,
Mods gets quite irate.

Logged
Ronniepeterson
Full Member
***
Posts: 1212


View Profile
« Reply #13 on: December 19, 2014, 06:19:23 PM +0000 »

First up let me make it very clear that I have been, am and will be happy to race against anybody whatever they want to use as a controller. I love GPL offline and online.

I do think the claims of pole times and fastest lap times by some are a little disingenuous to those who use shift times within the intended GPL parameters but hey I just race to the rules agreed by all.

In any other "game" the use/abuse of GPLshift would be called a hack at best and a cheat at worst.

But I think its good that the admin/mod team are already on the case.  I am confident that whatever they decide will quell any rumblings of discontent from myself or anyone else if that is the case and I will  abide by the results of the poll. Enough said.
« Last Edit: December 19, 2014, 06:40:32 PM +0000 by Ronniepeterson » Logged
Pages: [1]
  Print  
 
Jump to:  

Hosted by DaveGymer.com
Powered by MySQL Powered by PHP Powered by SMF 1.1.21 | SMF © 2015, Simple Machines Valid XHTML 1.0! Valid CSS!
Page created in 1.913 seconds with 33 queries.