Log in

I forgot my password

Facebook
Instagram
Sponsors & Affiliates
TORA ShirtRLM_10_main_logoMSA_full
Donations
Donations options

Forza 7 Discussion

Page 4 of 4 Previous  1, 2, 3, 4

Go down

Re: Forza 7 Discussion

Post by KilroyK95 on Fri 1 Jun 2018 - 21:45

I miss Forza Sad
avatar
KilroyK95

Playstation 4 Xbox One

Number of posts : 3117
Location : Leeds/Newcastle
Registration date : 2011-05-30
Reputation : 107

View user profile

Back to top Go down

Computer Rating The ForzaRC

Post by GTDon2 on Tue 26 Jun 2018 - 18:00

For those who are interested in the ForzaRC and the very fastest drivers of us all, the IFCA has taken it upon themselves to computer rate each round and get a handle on who has been the strongest driver over time.

The formula used is an Elo formula not dissimilar to that used by iRacing but customized for Forza. It gives an accurate performance-based snapshot of each driver within a particular race. It takes about 12 rated races for a rating to become relevant.

The computer logs a pre-race rating, gives a predicted finishing order, an actual finishing order, what each driver performed like, a post-race rating, and, how many ratings points he won or lost to the other drivers in the race.

So far the IFCA computer has identified that the Euro group is significantly stronger than the American group of racers. The numerical ratings are divided into classifications to help understand the grade or different skill levels. Borrowing from professional chess, the highest rated drivers are classified as GM's or "Grand Masters".

A closer look at the ratings also shows that there is a very small difference between the top 20 drivers, but a large difference between those outside of the top 50, and those top drivers. This is because the top drivers are truly the professional console racers, and the rest of us, well, not so much. lol

Ratings by each round here:
http://ifca.highforum.net/t6524-forzarc


World Rankings here:
http://ifca.highforum.net/t6516-ifca-forza-7-ratings-and-world-rankings

ForzaRC Season 1 results included:

http://ifca.highforum.net/t6516-ifca-forza-7-ratings-and-world-rankings

Due to the short sprints and reverse order grid format in the ForzaRC, and the extremely close skill level of all the drivers, the final race threw everyone out of order. The eventual Champion of Season 1 should have been #1 or #2, either b0x, or defending 2017 Champion RoadRunner. But due to starting at the back and then crashing, they both lost a lot of points as a result from this poor last race. b0x fell to #5 in the World and RoadRunner to #6 in the world, even though up to the last race they were both fighting to be #1.

It is the opinion of the IFCA that the format used in the ForzaRC is not representative of the kinds of performance that can be achieved by these incredibly fast drivers. It is a format that entices luck to play too big of a role.

Sprint races promote lucky wins and don't allow enough time to recover from bad luck. A bit too much like public lobby racing. They are designed to be high-pressure thrill competitions and not mimic real racing in any way. For example, there are no pit stops. T10 tries to compensate by forcing a reverse order grid to mix things up, but when the drivers are this good and this close it just ensures the guy who was at the back will almost certainly win now that he is at the front.

The real cure is not a complete overhaul of the format but simply a lengthening of the races, with a pit stop. Give the drivers a chance to recover from no fault contact and so on, and have an opportunity for some pit lane passing strategies. Then with as much luck taken out as you can, only then will you see who really is truly the fastest Forza 7 driver in the World.



==========================================================================


Re: Computer Rating The ForzaRC
Post by LMR Harmonic on Sun 24 Jun 2018, 12:01


This is irrelevant. FRC determines the ranking for their own competition. You can come up with whatever numbers you want, but they don't mean anything.


==========================================================================




Hi Harmonic! Thanks for replying.

Yeah, I get this from time to time even though I've been doing it for about 5 years now. I would tend to agree with you if it were just my opinion or some other subjective evaluation. But it's not. All it is, is, numbers, period. It's just the objective performance results of wins and losses in a mathematical formula. Your argument isn't really with me as I did not invent the formula, but with Dr. Arpad Elo as it were. Here, allow me to quote:

"The Elo[a] rating system is a method for calculating the relative skill levels of players in zero-sum games such as chess. It is named after its creator Arpad Elo, a Hungarian-American physics professor.

The Elo system was originally invented as an improved chess rating system over the previously used Harkness system, but is also used as a rating system for multiplayer competition in a number of video games,[1] association football, American football, basketball,[2] Major League Baseball, Scrabble, board games such as Diplomacy and other games.

The difference in the ratings between two players serves as a predictor of the outcome of a match. Two players with equal ratings who play against each other are expected to score an equal number of wins. A player whose rating is 100 points greater than their opponent's is expected to score 64%; if the difference is 200 points, then the expected score for the stronger player is 76%.

A player's Elo rating is represented by a number which increases or decreases depending on the outcome of games between rated players. After every game, the winning player takes points from the losing one. The difference between the ratings of the winner and loser determines the total number of points gained or lost after a game. In a series of games between a high-rated player and a low-rated player, the high-rated player is expected to score more wins. If the high-rated player wins, then only a few rating points will be taken from the low-rated player. However, if the lower rated player scores an upset win, many rating points will be transferred. The lower rated player will also gain a few points from the higher rated player in the event of a draw. This means that this rating system is self-correcting. A player whose rating is too low should, in the long run, do better than the rating system predicts, and thus gain rating points until the rating reflects their true playing strength."

In fact, this is the same basic system that iRacing uses truth be told. So while I understand your view in that the FRC has something that they feel represents a way to rank the drivers, it is not based purely on performance as it is not based on the Elo formula. Rather, it is based on accumulated series points which is far different and doesn't really get at the meat of the matter in terms of true driving strength. Why? Because a slower driver can do fairly well if the measurement is just about series points. Doing so brings into play certain strategies apart from raw driver skill. And maybe this is just fine for the FRC, but you ought not to be fooled into thinking it is the same as what a mathematical performance rating is. Just sayin.

The IFCA computer rankings formula and list have already proven itself to be generally highly reliable over the years. No one should be surprised that Laige is listed as the number 1 driver for example. He's been #1 before and has always been by consensus among the top 5 in FM7 anyway. No hard prediction there.

What's interesting is when you look at the results of each regular race. Those expected to finish mostly towards the top generally do, and those expected to finish mostly towards the bottom generally do. What this means is that the formula is working correctly. The faster drivers consistently finish higher than the slower drivers as a rule.

Just as an example let's pick on the very fast Daveyskills for a moment. He was somewhat overrated as most provisional drivers are in that he hadn't yet had enough races to establish his actual rating yet. He entered the finals with only 7 rated races to his credit. He needs at least 12 to be fair. Because he was highly rated the computer said he should finish 4th in race 1. But because his true rating is not firm yet and thought to be inflated, he finished last, or 12th.

In the second race with his rating adjusted downward from a high of 2593.036 to 2521.093 a loss of -72 points, the IFCA computer predicted he would likely finish 9th in race 2. And guess what? He finished 9th.

In race 3 Daveyskill in this group was expected to finish 11th, but again finished last in 12th. But only lost -4 points because he did what was expected for the most part.

In the final 4th race with inverted grid order, the computer predicted he would finish dead last again not because of the format, but because he was the lowest rated driver of the group. Since he started at the front and the margin between all the drivers is fairly small he was able to finish in 3rd. This bumped his rating up to 2552.803 with a total of just 10 rated races, or 2 shy of the 12 usually needed to establish a solid rating. Chances are he's still slightly overrated at this point and will come back down into the low 2500's after two more races.

This is no dig on Daveyskills or shame understand, he is faster than nearly everyone who might read this. This is just the objective quantifiable truth of his performance with this group of drivers and nothing more.

Thanks for the opportunity to explain all of this to everyone here at TORA. Hopefully, some will have a better understanding and greater appreciation of how performance ratings work.

And good luck next season Harmonic, I am rooting for you!

GTDon
avatar
GTDon2

TORA Race Number : 126
Xbox 360 Xbox One

Number of posts : 190
Location : Minnesota
Registration date : 2010-02-02
Reputation : 9

View user profile http://theforzaunderground.forumotion.com/

Back to top Go down

Page 4 of 4 Previous  1, 2, 3, 4

Back to top


 
Permissions in this forum:
You cannot reply to topics in this forum