FRS Season 1 Review & Discussion

[95GTIVR6]

As most of us prepare out final Hot Lap of the season, I would like to take this moment to give a simple overview of Season 1 and open it to discussion on all fronts.

The below data reflects just some info I compiled throughout the season (CURRENT THROUGH WEEK 15):

Class Races:
F=0
E=0
D=0
C=2
B=1
A=5
S=6
R3=1
R2=2

Stock/Tuned Options:
Stock=11
Tuned=6

Perfect Attendance Awards: Participated Every Week
Farming The Fog & I2valve

Most Golds: (Per Division)
D1: lukas2121 (8)
D2: fastulus (6)
D3: bergj42 (11)

Racers Per Week:
Week 1: 25
Week 2: 21
Week 3: 15
Week 4: 17
Week 5: 16
Week 6: 18
Week 7: 16
Week 8: 18
Week 9: 14
Week 10: 13
Week 11: 16
Week 12: 14
Week 13: 10
Week 14: 11
Week 15: 11
Week 16: ??

Based off my observations throughout the year, it would appear that the divisions are VERY close to being accurate. I know we will probably run another qualifier for season 2 so the only change I would have made this year was to be moving hennessey861 to D1 based off performance alone.

 

[Jump_Ace]

We won't be re-running a new qualifier, just so you know. But the end of season 1 will be a good time to look over promotions and your information here will help us out. Thanks for your hard work on this!


Jerome

 

[95GTIVR6]

I have been wondering how to incorporate classes F-D and was wondering if we would try to have 2 races each week to choose from. Basically, my idea is to have the main FRS week already in place with a car and track option but also have a "partner" series that uses a selected F-D class car at the club/small version (or same version) of the track being raced.

An example of this might be week 16 for example:

Racing a stock B Class SRT Charger at Sebring Full

Racing a stock D Class 69 Charger at Sebring Short

 

[Jump_Ace]

That would double our work load and isn't something we want to take on unfortunately.

But we'll have some 'slower' combos in the somewhat near future 👍


Jerome

 

[95GTIVR6]

I wasn't asking for anyone else to take it on, I just thought I might host my own FRS Lite series in conjunction with the weekly events.

I have been fooling with this idea for a while. Whether it is bi weekly or a monthly challenge. Basically, just another option to where if a driver gets tired of doing Hot Laps a certain week, they can take a break by trying to set another fast time.

Who knows, I just want more people to race in these events. It was off to a nice start but many variables are playing into why we aren't getting a lot of race attendance, the official Forza site being one of them.

I totally understand If I cant put the threads in this sub-topic (if I make them at all) but I wouldn't ask anyone else to control this. I still plan on FRS participation but I do want MORE participation.

 

[Im_Lukas]

Wow it's interesting how the participation has dropped so suddenly, I guess with it being a lot nicer weather in many places people are taking advantage of it which gives less time for Forza. I would expect the participation to rise again as we near the release of Forza Horizon as many people start looking for information and discussion boards, I know I originally found GTPlanet through looking for information on GT PSP. Speaking of Forza Horizon would there be any chance of a competition like the FRS for that game when it's released?

@ Mr. TurboWagon I like the idea and could probably find time for a few laps every now and then but my main focus would still be on the FRS. Well done to you and Farming the Fog for 100% attendance throughout the season.

 

[shawtyoner]

Do most people not like the lower classes? I for one don't mind so long as its a car with some history or cult following. C and B classes are actually my favorite classes in game.

 

[95GTIVR6]

For every Forza, I found that I have almost NEVER used the race cars. I love them and will drive them if I have to but the lower classes are always what makes me smile.

They're not as stressful to drive and in many cases, the races tend to lend themselves to being closer matches.

When referring to online play, I almost exclusively race E-B classes as I have found that leaderboard cars tend to dominate some of the other classes and the races are just about who has the best leaderboard car.

This led me to NOT tune cars to their full class potential. I figured If I wasn't placing first with my non-leaderboard car, then why even tune a car to the class maximum? So I would begin racing a tuned F-class to maybe a mid-range D class (for example) and still finish at least in the top half of then pack.

I am trying to buy and tune every car in the game so I started with F class cars. So far I have purchased and tuned every F and E class car in the game (not including any unicorns) and I am starting now with D class.

Im not sure what this post was meant to say, I just began to ramble.

 

[shawtyoner]

^ So every car that you have purchased and tuned is no longer in its original class is what you are saying? Or is it? Thats dedication btw.

And I am 100% in agreement with what you said, lower class races are less stressful, however they can be tough as hell being that matches are generally a lot closer.

 

[95GTIVR6]

I should have been more clear.

I have a strange way of modifying my car, I call this "Stage 1". Basically, there are always 3 modifications for each part (Air Intake has a Street, Sport and Race Option). I proceed to put on the Street mod for every part, if the street modification is already installed "stock" (much like the throttle body on most cars) I leave it be and continue. I do this for all engine and suspension modifications, when I get to the transmission, I get the second, Sport options for all and a race differential. Tires are the same, I get the first street performance tire and also some wheels.

This means that all that is adjustable is usually just the tire pressure, differential and final drive ratio. After this I usually try and knock out a few Events with whatever class the car is in.

So far I have done this to every F and E class car. Then I decide what class would be most appropriate, whether it moves up or down. I have also done this with other cars but I figure I would start from the bottom.

I want to find a way to record my findings on this but it would take a while. But to be short, yes, every car I have at LEAST has a stage 1 tune. Because some cars are tuned more from the factory than others, some of the classes where the cars end up are interesting.

 

[YakieS1]

I call this "Stage 1". Basically, there are always 3 modifications for each part (Air Intake has a Street, Sport and Race Option). I proceed to put on the Street mod for every part, if the street modification is already installed "stock" (much like the throttle body on most cars) I leave it be and continue. I do this for all engine and suspension modifications, when I get to the transmission, I get the second, Sport options for all and a race differential. Tires are the same, I get the first street performance tire and also some wheels.

This means that all that is adjustable is usually just the tire pressure, differential and final drive ratio. After this I usually try and knock out a few Events with whatever class the car is in.

Really like that idea, Turbo! I don't think I'll use it for my event list, but I think it certainly has some application in the FRS...

It sure would be a neat way to get some "semi-stock" lower class cars up in PI enough to be interesting, without leading to a whole lot of tuning, as people were complaining (*a little*) about a few weeks ago in this season. Maybe post some of your favorite stage-1 or stage-2 cars in the suggestion thread?

 

[shawtyoner]

^ I'm with you on that. I like lightly tuned cars, anything fully tuned then you might as well just use a race car. Maybe its just because I suck at tuning that I say this, but all in all I like street cars more.

 

[YakieS1]

Did a little work with excel this weekend in search of the "most improved driver" for season 1.

Method: Took a spreadsheet of all the times for each week, and assigned a "percent behind the leader", such that if the fastest time for the week was a 1:40.00 (100 seconds) and I did the course in 1:50.000 (110 seconds), I would be 10% behind the leader. The goal was to normalize each week for course length, and not give extra weight to the long or short tracks.

Next, I did a line fit for each person's "percent behind the leader" to see if any of us consistently closed (checked the slope of the fit) on the leader's time as the season progressed. I used the D1 winner each week as the benchmark, with the idea that the D1 driver was as close as possible to the "ideal time" for the car and track combination.

What I found: VERY SURPRISED with the results. Only ONE driver had a "positive slope" indicating that they had consistently improved relative to the winner of each week! Some of us (like myself) appeared to get slower (relative to the leaders) as the season progressed!! Maybe we all did improve, but the D1 winners improved faster.

I will reveal the identity of that talented driver after the end of week 16.

 

[Im_Lukas]

That sounds confusing but I'd put money on who I think has improved the most. My slope would be going down if compared to the ideal time because every week I find less time to put in to the FRS and I'm not improving enough to out-balance my decreased time.

 

[95GTIVR6]

Yakie:

I would be interested to compare your spreadsheet and mine. I for one want to see if there is a difference in which classes people seem to perform better in.

 

[YakieS1]

OK, it's taken me a long time to get this together, but it's finally here: A few graphical products that show some of the results from Season 1 of the FRS.

Because we ran a different track and car each week, just comparing peoples times week to week wasn't going to give any useful information, so I had to find a way to normalize the race results for track and car.

As described in my previous post, I used "percent behind the leader" as the parameter to normalize for track and car. As an example calculation: If the fastest time for the week was a 1:40.00 (100 seconds) and I did the course in 1:50.000 (110 seconds), I would be 10% behind the leader (or -10.00%), as shown in the charts below.

The first chart is a histogram that includes all the time entries of the season, and shows where their times fell out relative to the leaders. As you can see, the Division 1 times are always really close to the leader, and in general the Division 2 drivers were massed a little bit slower than the leaders. There wasn't a very significant difference between the Division 2 and Division 3 drivers, though.

The second chart compares my times to Hennessey861's times. As you can see, he was consistently faster than me. The positive slope to Hennessey's trendline shows that he got consistently faster throughout the season, while the negative slope on my trendline shows that I did NOT get consistently faster!

Only three racers had a positive trend through the whole season, but Hennessy861 had the most positive slope of all the competetors. As such, I nominate him as the most improved driver for the season!

I can make more charts like this, on request, and I will be tracking everyone's progress through Season 2!

Great season everybody,
'Mater

 

[bergj42]

The second chart compares my times to Hennessey861's times. As you can see, he was consistently faster than me. The positive slope to Hennessey's trendline shows that he got consistently faster throughout the season, while the negative slope on my trendline shows that I did NOT get consistently faster!

I like how this is presented but I'm not quite sure on your line showing that you got slower. If you take out the one point way down later in the season (week 14) and and start your line at week 5, it would appear that you stayed even and or improved.
The one week where you were way off the pace looks to be more of an anomaly in relation to the rest of your times. It also looks like there are no data points for you before week 5?
Overall, nice job with this. I wish I had the time to put into charting it like that.

I would like to see what mine look like if its not too much of a problem.

 

[Brian 7h3 King]

.

This is really cool, I must say 👍

 

[Jump_Ace]

Thanks for your hard work on that 'Mater. It doesn't go unnoticed

As for Forza Horizon, I honestly haven't gotten anything new from T-12 (who is always busy, haha) but I'm sure once the game gets closer to release, we'll know what's what. But for now I can't promise anything.

As for online racing, I personally would only want the FRS to take it on if participation increased a lot. It'd have to be a lot because there are weekly races with the Forza Planet Motorsport club, found here. I've raced with them and they are a blast, very courteous, clean and most important...fast club of drivers 👍

Jerome

 

[YakieS1]

I would like to see what mine look like if its not too much of a problem.

It's no problem at all! Here you go:

I like how this is presented but I'm not quite sure on your line showing that you got slower. If you take out the one point way down later in the season (week 14) and and start your line at week 5, it would appear that you stayed even and or improved.
The one week where you were way off the pace looks to be more of an anomaly in relation to the rest of your times. It also looks like there are no data points for you before week 5?

You're absolutely right, and you've hit on a big limitation of linear regression: It is heavily influenced by outliers, particularly at the beginning or end of the (short) data set.

Your season is a good example: Week one and two you were were right up by the leaders, but as the divisions got set and D1 really started getting faster, your times started to stabilize at 3-5 percent behind the leaders. If I ignored week one and two, the slope of your trend turned more positive, and is probably indicative of your true performance.

I was searching for a totally objective and numerical way of picking a "most improved" player, but the system I used certainly isn't perfect... I could have looked for outliers and trends and "edited" the data to make the line fits look better, but I thought that would be a very subjective process...

As for my times: I didn't start participating in the series until week 5 (and the trendline isn't changed by the blank data area) and for week 14? I was just flat slow that week. Couldn't find a tune that I liked, and if I recall I wasn't the only one who had trouble that week.

 

[VICGT1970]

I would like to see how much I improved over the season

It must have taken alot of time to do all time and math to figuer all this out

 

[YakieS1]

I would like to see how much I improved over the season

Well, VIC, it's hard to improve much when you win 6 out of 16 of the races... Since the parameter I used was "percent behind the leader" in 6 cases you show 0.00%. Either you or Lukas2121 won all but 2 of the race weeks for D1, and you domiated the end of the season.

The three people on this chart are the three with a positive slope to their season trend.

 

[VICGT1970]

thanks alot YakieS1 I knew that I finished strong just wanted to see if I was 1 of the three to improve

And with attendents down maybe this season all the divisions should be put into 1 larger division just a thought

 

[YakieS1]

When I get some more time, I'll try to nail down a statistical method for recommending someone for promotion... If you look at Hennessey's record, he's only been consistently running D1 times for about 5 weeks... I will work on it!