SRAting is a weighted combination of three separate metrics: pace, performance, and safety. It is a data-driven approach to quantifying the overall strength of each member as a racing driver.
It is a tool to quantify each driver's racing potential to help delineate divisions for the SRA's GT3 Team Series.
You can view the current driver SRAting list at https://www.simracingalliance.com/about/srating.
Evaluates each driver's raw speed. It takes into account the best laps from both qualifying and race, as well as average race pace.
Evaluates each driver's ability to put a race together based on overall race time, pace of nearby drivers, and strength of grid.
Evaluates each driver's overall safety on track by evaluating each driver's consistency. (Collision and Penalty Points have been temporarily removed. Revamp coming up for future versions.)
Indicates the number of races each driver has logged within our system.
The overall rating is calculated based each driver's most recent 8 races. Only the most recent 14 events are considered. Drivers who haven't raced in the most recent 4 events are excluded from the list.
We provide a provisional rating once a driver has at least 3 races.
Currently, only the qualifying and race results from official SRA GT3 Team Series events are considered towards SRAting calculation (MCM and Endurance races are not included).
Why don't I see myself on the SRAting leaderboard?
- You must have 3 races in the GT3 Team Series to be rated. If you have less than 2 valid laps in a race, it won't be considered.
Why are there so many outliers? (e.g., D4 driver mixed in with D2 drivers)
- The divisions are very tight especially in the middle divisions, so accurate intra-division evaluation is challenging.
- Running at the front of the grid is “easier” than running mid pack, which can lead so slightly exaggerated ratings (even though the system tries to account for it).
- Initial placements with just the hotstint data is oftentimes inadequate.
- Many people improve drastically after joining the league (especially common in lower divisions).
Why do some people with poor race results have high SRAting?
- The system is designed to predict “optimal performance” of each driver for the purposes of generating competitive divisions. So it won't necessarily reflect each season's standings.
- Each data point is weighted by its total race time. If someone crashed out early, that race won't count as much as others.
- Repeated poor results will still be reflected by the rating.
How do you evaluate PACE and PERF in wet/mixed condition races?
- We utilize the meta data we have on drivers racing on track to calculate a “track condition coefficient”, which reflects the difficulty of a given race session.
- The “track condition coefficient” is evaluated separately for each Q/R session in each division. The coefficient is used to adjust the reference time before quantifying relevant metrics.
- This is so that varying conditions in different divisions held at the week can be evaluated together (as best as possible, but not perfect).
- This method is more accurate if we have confidence in our data from dry races. That's why we are currently limiting the wet races per season to 2 races.
Why did my rating change even though there hasn't been a new race?
- We're constantly trying to improve the system to get the most out of the limited data that's exported by the game. If I make some tweaks to the system, the ratings will change.
- All changes are fully retroactive and applied equally to all drivers.
- SAFE metric has been simplified to exclude contacts and penalty points due to inconsistent data collection. Mainly reflects drivers' consistency now. Will be revamped in future versions.
- SAFE metric weight reduced drastically
- PACE/PERF metric weight adjusted
- Peer PACE evaluation window reduced to 0.5s from 1.0s
- Confidence parameter added. Weighted averages are more favorable if you have more races (caps at 8 races).
- PERF metric weight increased
- Performance metric is now decoupled from finishing positions and overall grid size, instead overall race time and strength of grid metric replaces them in the formula. Removed arbitrary scaling parameters between divisions.
- Incorporates weighted average in place of standard arithmetic mean in PACE, PERF, and SAFE score calculations. Parameter tuning focuses on reducing the impact of outliers.
- Incorporates weighted average for track condition quantification for non-dry race weeks.
- Simplified the evaluation scope for SRAting. Now for each driver, 8 most recent races are evaluated, races older than 14 most recent events are not considered, and drivers without any races in the 4 most recent events are excluded.
- Overall parameter tuning across the board
- Reference lap time adjustments
- Parameter adjustments: PERF rating weight reduced slightly
- Race results older than 120 days are now ignored
- Weight reduced for consistency when calculating SAFE rating for wet races
- Rain PACE calculation now also considers average race pace
- Track condition evaluation algorithm (for wet races) improved to consider more data points
- PACE formula updated to include average race pace
- Lowest score per category dropped when driver reaches 9+ races
- Formalization of various metrics and formulas from version 1