Xtreme Gaming Network

Japanese pack server

Author Topic: Japanese pack server  (Read 28647 times)

Offline slideways

  • Sr. Member
  • ****
  • Posts: 371
  • Karma: 3
    • View Profile
Re: Japanese pack server
« Reply #30 on: July 12, 2016, 07:50:31 AM »
Will get back on later for some more laps. Did 1 session and had to call it quits last night.

Offline Wally

  • Administrator
  • Hero Member
  • *****
  • Posts: 11033
  • Karma: 152
  • AC Admin
    • View Profile
Re: Japanese pack server
« Reply #31 on: July 12, 2016, 08:07:47 AM »
We've discussed the methodology many times before. You can't just look at the fastest laps, because what if YoRap is an alien who sets times in all cars but the Supra? You'd conclude,  wrongly, that the Supra was the much slower car.

You can't just look at one person's laps in all 4 cars either.  That one person might have a knack of being able to drive the Skyline the fastest, and you'd conclude, wrongly, that the Skyline was the fastest car.

You have to average the top percentage of times (to exclude all the unrealistic half arsed laps at the tail end). That also gives you a better indication that the spread of times in a league will be close, not just the most skilled guys.

You can only go by the data you've got. If we only have 9 Supra times, then that's what we've got.  9 laps is still better than 1 lap.
“You can please some of the people all of the time, you can please all of the people some of the time, but you can’t please all of the people all of the time”

Offline slideways

  • Sr. Member
  • ****
  • Posts: 371
  • Karma: 3
    • View Profile
Re: Japanese pack server
« Reply #32 on: July 12, 2016, 08:50:44 AM »
I'll get a heap of laps in today in all the cars (crappy weather and can't do much else). BTW all my times on RSR have been with AI out on track. Trying to keep it realistic..... ;)

Offline marty

  • Hero Member
  • *****
  • Posts: 3747
  • Karma: -63
    • View Profile
Re: Japanese pack server
« Reply #33 on: July 12, 2016, 09:23:23 AM »
I had the wrong link for the supra at imola
http://52.62.170.191:50041/lapstat?track=imola&cars=ks_toyota_supra_mkiv&valid=1,2&date_from=&date_to=&currservers=acserver3Ballast

Using the top 5% times just doesnt work in that case, there is no way the car is 2 seconds faster.

Here is the data showing only the best lap per driver so it shows what car people were quickest in.

http://52.62.170.191:50041/lapstat?track=imola&cars=ks_mazda_rx7_spirit_r,ks_nissan_370z,ks_nissan_skyline_r34,ks_toyota_supra_mkiv&valid=1,2&date_from=&date_to=&currservers=acserver3Ballast&ranking=1

This isnt showing that the supra is clearly quickest for most people with only 2 of the top 15 being in the Supra showing that method isnt giving a true indication if it was 2 seconds faster then anyone would easily set their quickest time in the car I think.



Offline marty

  • Hero Member
  • *****
  • Posts: 3747
  • Karma: -63
    • View Profile
Re: Japanese pack server
« Reply #34 on: July 12, 2016, 09:29:37 AM »
I'll get a heap of laps in today in all the cars (crappy weather and can't do much else). BTW all my times on RSR have been with AI out on track. Trying to keep it realistic..... ;)

All the server laps of mine were with other people on track with many laps limited by traffic so fairly realistic. Get more then 10 people on that short track no tyre warmers and it takes 4 to 6 laps to get the tyres up to speed. Then the odds of getting no traffic on the lap is fairly tricky, the only car I ran on RSR was in hotlap mode with tyre warmers and this would be .3 or so quicker at least as you can set pressures for first laps and not deal with traffic get a lap in with fresher tyres.

If you do it during the day it will likely be quiet and I can extend the qualy session time if you want on that server.

Offline slideways

  • Sr. Member
  • ****
  • Posts: 371
  • Karma: 3
    • View Profile
Re: Japanese pack server
« Reply #35 on: July 12, 2016, 09:34:28 AM »
That would be great. And you're right, less peeps who want to go straight to race during the day. Was only on for a few seconds last night when they started rolling in.. :)

Offline Wally

  • Administrator
  • Hero Member
  • *****
  • Posts: 11033
  • Karma: 152
  • AC Admin
    • View Profile
Re: Japanese pack server
« Reply #36 on: July 12, 2016, 09:46:55 AM »
I had the wrong link for the supra at imola
http://52.62.170.191:50041/lapstat?track=imola&cars=ks_toyota_supra_mkiv&valid=1,2&date_from=&date_to=&currservers=acserver3Ballast

Using the top 5% times just doesnt work in that case, there is no way the car is 2 seconds faster.

Here is the data showing only the best lap per driver so it shows what car people were quickest in.

http://52.62.170.191:50041/lapstat?track=imola&cars=ks_mazda_rx7_spirit_r,ks_nissan_370z,ks_nissan_skyline_r34,ks_toyota_supra_mkiv&valid=1,2&date_from=&date_to=&currservers=acserver3Ballast&ranking=1

This isnt showing that the supra is clearly quickest for most people with only 2 of the top 15 being in the Supra showing that method isnt giving a true indication if it was 2 seconds faster then anyone would easily set their quickest time in the car I think.
How many of those people put in a decent Supra lap? If they didn't do a hot lap in the Supra at all, then it's a meaningless comparison.
“You can please some of the people all of the time, you can please all of the people some of the time, but you can’t please all of the people all of the time”

Offline marty

  • Hero Member
  • *****
  • Posts: 3747
  • Karma: -63
    • View Profile
Re: Japanese pack server
« Reply #37 on: July 12, 2016, 10:20:12 AM »
Its the same comparison of the same 9 people that did a lap in the Supra from the total data in all the cars  ;) The same data that comes up with the supra 2 seconds faster then the other cars.

Looking at the actual data Yorap did 2:06.116 in Supra and 2:06.213 in the 370 so he is just .097 quicker in the Supra and this is also the overall best time. RX7 I am .296 off that Supra time of Yoraps with 2:06.412 and I am .171 slower in the Supra 2:06.583 and another .153 2:06.736 in the 370.

Howzat who is 3rd fastest in the Supra with 2:08.052 has done 2:07.718 in the rx7 so his gap is .334 to the Rx7

Wally is 2:09.621 in the Supra which is your fastest, then 2:10.477 in the r34 so .856 slower in it with your times in the other cars within .3

This shows that only 2 people are actually quicker in the Supra who have driven more then 1 car you and Yorap with the biggest gap of .856 and this can just simply be a better lap or your more comfortable with the car. The 5% method showed it as 2 seconds faster on average yet it is only .097 quicker in the fastest times and the largest gap anyone has in its favor is yours showing I would discount that calculation.

The sample size is a bit bigger then RSR data but still not quite enough to look at overall averages too much I think. Best to look at relative gaps in individual drivers and overall best times, the server has 4 cars to choose so people can run whichever they like most. The fact that not many are actually quickest in the Supra just makes that calculation seem bugged to me. The other thing is Imola is quite a tricky track to get a solid lap in, very easy to loose time and so unless someone has nailed a lap or done a considerable amount of them its hard to say if they are close to their actual potential yet.

I do like to see how all people stack up in the cars but also the top times are quite telling If my best time in the slowest car is only .766 off the best time in the server then I think they arent too far off. Also the car that should be showing the largest gap is the r34 as its had the most laps in the server yet is the slowest overall. Im taking into account my actual PB in the same conditions but not one in the server to even bring the r34 back into the ballpark there. It is actually 1.2 slower on best overall yet your 5% figure shows the Supra 2 seconds faster and the other 3 within a tenth or so.

Here is the best lap in each car and Laps column shows total laps done in each car.
http://52.62.170.191:50041/lapstat?track=imola&cars=ks_mazda_rx7_spirit_r,ks_nissan_370z,ks_nissan_skyline_r34,ks_toyota_supra_mkiv&valid=1,2&date_from=&date_to=&currservers=acserver3Ballast&ranking=2
« Last Edit: July 12, 2016, 11:03:25 AM by marty »

Offline Wally

  • Administrator
  • Hero Member
  • *****
  • Posts: 11033
  • Karma: 152
  • AC Admin
    • View Profile
Re: Japanese pack server
« Reply #38 on: July 12, 2016, 10:46:45 AM »
Its the same comparison of the same 9 people that did a lap in the Supra from the total data in all the cars  ;) The same data that comes up with the supra 2 seconds faster then the other cars.
I  shall ponder when I can look at this off my phone :)
“You can please some of the people all of the time, you can please all of the people some of the time, but you can’t please all of the people all of the time”

Offline marty

  • Hero Member
  • *****
  • Posts: 3747
  • Karma: -63
    • View Profile
Re: Japanese pack server
« Reply #39 on: July 12, 2016, 11:26:55 AM »
I understand why you do the top 5% but it really needs a larger sample size to really work, I did run quite a few laps in the servers as that way more join and it gets a larger data set but it shows that we dont quite have enough for all the cars. The cool thing with stracker once you start digging in to the + section you can gather lots of info. You can manually compare any 2 laps and see where the cars differ.

Here is Yoraps lap in the Supra vs 370 with ballast, this also shows why the Supra has the slight edge as it has more top speed.



Here is me in the RX7 vs Supra


Wally Supra vs GTR


Wally RX7 vs Supra (this shows you made an error into t1 and so probably cost that 1 second just there.


Me in RX7 vs Yorap in Supra



Offline marty

  • Hero Member
  • *****
  • Posts: 3747
  • Karma: -63
    • View Profile
Re: Japanese pack server
« Reply #40 on: July 12, 2016, 11:39:16 AM »
Here is no ballast with me in the RX7, Supra and 370z


Offline Wally

  • Administrator
  • Hero Member
  • *****
  • Posts: 11033
  • Karma: 152
  • AC Admin
    • View Profile
Re: Japanese pack server
« Reply #41 on: July 12, 2016, 12:50:29 PM »
It's all very, very close and looking like it will be fine. All I'm trying to avoid is that any car is under or over-ballasted and turns into a dog or an unfair advantage.

I didn't get my best RX7 lap in.  I was quite a bit up on my best time when I ran out of fuel, and just left it at that.
“You can please some of the people all of the time, you can please all of the people some of the time, but you can’t please all of the people all of the time”

Offline marty

  • Hero Member
  • *****
  • Posts: 3747
  • Karma: -63
    • View Profile
Re: Japanese pack server
« Reply #42 on: July 12, 2016, 01:19:54 PM »
The main reason I am doing the testing is to hope each car has the same chance of being competitive, still some will find one easier to drive then others and the extra ballast does make the heavier cars a bit more awkward but it brings them back to the pack. If the ballast is done correctly and atm it looks fairly close then people should be able to pick any car and if they get the most out of it should be able to post similar lap times. These cars also arent too far apart in how they get their times but in order to take 3 seconds off the rx7 it does need quite a bit of weight.

When Redbull ring comes out I will move this server over there as it should be popular and its also a track that does favour horsepower so it will be interesting to see if the GTR is still competitive there at these levels. There are still a few weeks left before the next season so this test can be run at a few locations and we should get a reasonable idea. From my tests at Brands they are really close but I think the rx7 and 370 will still be slightly the quickest there, the GTR is easiest to get pace but also eats front tyres quickest. The Supra is quick but not the greatest handling so getting a laptime out of it is probably the most challenging.

Offline Wally

  • Administrator
  • Hero Member
  • *****
  • Posts: 11033
  • Karma: 152
  • AC Admin
    • View Profile
Re: Japanese pack server
« Reply #43 on: July 12, 2016, 02:06:03 PM »
Yep, we're both trying to achieve the same thing.
“You can please some of the people all of the time, you can please all of the people some of the time, but you can’t please all of the people all of the time”

Offline marty

  • Hero Member
  • *****
  • Posts: 3747
  • Karma: -63
    • View Profile
Re: Japanese pack server
« Reply #44 on: July 13, 2016, 03:55:46 PM »
Good to see the Jap pack server getting a fair bit of use with around 2000 valid laps just in the brands indy combo with ballast.

Here are the total amount of laps driven in all cars by everyone in the server and the fastest in each with ballast.



I also ran each car in SP hotlap mode to set a time with no ballast.



The 370z is just not quite geared well for a few critical corners so it doesnt quite get as good a lap as it would with a taller 2nd or shorter 3rd gear.

The ballast looks rather close here but I will see how they stack up at redbull ring, with one more week to go I think we can run both the short and long versions there and get an idea. The long version will likely hurt the GTR the most but the short version they will hopefully be fairly close there too.

Being a new DLC track it may get some running but possibly it will be more popular with the new cars rather then these. I will keep it ABC for these and make the new cars all AB there to encourage more to run laps in these still.

This is quite a short track and the cars are fairly close even with ballast, not sure if the rx7 is actually much quicker with ballast maybe it was just a better lap there as in a race the pace in all 4 I am quite close.

Most people are fairly close in all the cars except the Supra is still in general the slowest here but also has had the least amount of laps.

 

SimplePortal 2.3.5 © 2008-2012, SimplePortal