Tracks

Below is a list of all of the tracks that I have developed, roughly grouped by style, and defined with some metrics. Stylistic groupings are something that I'm constantly tweaking but this represents my current thinking. For consistency, I've standardized the name of all tracks based on their locations (city/country). Tracks themselves and the FIA have been inconsistent over the years about what a track / grand prix is called.

Many tracks have changed layouts over the years. The year that a particular layout was first run is noted.

If you are interested in a detailed description of how I build tracks I wrote three posts on my blog describing my process:

1. Preparing to Build a New Track
2. Turning Research Into a New Track
3. Testing, Testing, Testing A New Track

The color-coded raw and adjusted scores are my current method for grading a track. Negative numbers indicate that a track seems to favor running from the front. Positive numbers indicate that a track may favor running from behind. I base the scores as much on actual results as possible, but where my sample size is lower than 10 races I add guesses based on the track layout. The raw score is more of an absolute score while the adjusted score relative to all other tracks. The colors are a visual way to group the tracks. MUCH more information below.

Location	1st Year	L	C	Raw Score	Adj Score	Results Score	Q: F | M | B	SS: F | M | S	Sample: Q | SS	Track Score	3wC	3wL	M Straights	L/SlowC
Monza, Italy	2000	81	6	-2.19	-1.66	-2.19	10.1 | 6.7 | 6.0	14.8 | 7.2 | 6.4	11 | 9	-1.64	0	56%	2	40.5
Nürburg, Germany	2002	74	8	-2.07	-1.54	-2.07	9.9 | 6.6 | 4.7	11.4 | 6.1 | 5.3	11 | 11	-0.13	2	31%	5	10.6
Suzuka, Japan	2003	83	8	-1.68	-1.35	-1.13	8.0 | 6.7 | 6.4	9.7 | 5.6 | 3.0	5 | 2	-1.99	0	18%	1	20.8
Francorchamps, Belgium	2007	97	9	-1.47	-1.14	-1.59	10.1 | 6.1 | 5.0	9.6 | 8.3 | 6.2	11 | 7	-0.40	1	61%	3	19.4
Monte Carlo	1986	51	7	-1.33	-0.80	-1.33	9.8 | 7.7 | 4.3	9.6 | 7.3 | 8.2	13 | 10	-0.66	0	45%	3	7.3
Silverstone, UK	1997	74	8	-1.32	-1.00	-2.04	10.3 | 6.3 | 5.2	13.0 | 5.2 | na	4 | 1	-1.09	1	35%	2	18.5
Barcelona, Spain	2007	60	8	-1.14	-0.82	-1.59	8.7 | 7.3 | 5.7	9.6 | 5.9 | 3.2	6 | 4	-0.67	1	40%	2	8.6
Buenos Aires, Argentina	1953	55	8	-1.11	-0.78	-1.76	10.6 | 5.9 | 4.7	9.6 | 6.6 | 5.8	7 | 7	0.42	3	51%	4	11.0
Location	1st Year	L	C	Raw Score	Adj Score	Results Score	Q: F | M | B	SS: F | M | S	Sample: Q | SS	Track Score	3wC	3wL	M Straights	L/SlowC
São Paulo, Brazil	1990	67	8	-1.02	-0.70	-1.12	8.1 | 6.3 | 5.2	10.2 | 5.6 | 6.8	7 | 5	-0.87	0	46%	2	9.6
Montreal, Canada	1996	65	7	-0.94	-0.41	-0.94	9.5 | 6.2 | 6.4	10.0 | 6.2 | 8.0	13 | 13	-0.09	2	65%	2	16.3
Valencia, Spain	2008	77	8	-0.85	-0.53	-0.16	8.0 | 8.6 | 7.1	5.7 | 10.1 | 10.9	7 | 3	-1.57	0	31%	0	11.0
Sakhir, Bahrain	2004	80	8	-0.90	-0.58	-2.67	10.6 | 6.1 | 3.1	11.9 | 6.4 | 4.7	4 | 4	0.28	3	51%	3	10.0
Mogyoród, Hungary	2003	63	9	-0.80	-0.48	-2.00	9.8 | 4.8 | 5.1	9.5 | 5.5 | 2.8	4 | 3	-0.20	2	38%	3	7.0
Spielberg, Austria	1997	63	7	-0.56	-0.03	-0.56	10.3 | 7.2 | 8.7	9.1 | 9.0 | 7.7	12 | 12	0.20	5	38%	1	12.6
Singapore	2008	74	11	-0.46	-0.13	-0.33	10.0 | 5.1 | 8.8	9.5 | 6.4 | 8.9	9 | 8	-1.18	0	21%	3	7.3
Istanbul, Turkey	2005	76	8	-0.45	-0.13	-0.56	9.2 | 6.7 | 5.0	3.7 | 7.7 | 6.0	9 | 6	-0.09	2	47%	4	15.2
Yeongam, Korea	2010	82	8	-0.36	-0.04	-0.64	8.4 | 7.2 | 8.6	11.1 | 7.9 | 6.8	5 | 4	-0.14	2	48%	3	11.7
Francorchamps, Belgium	1985	97	9	-0.35	-0.02	-2.36	12.0 | 7.0 | 7.0	14.7 | 6.0 | 6.7	1 | 1	-0.12	2	60%	3	19.4
Location	1st Year	L	C	Raw Score	Adj Score	Results Score	Q: F | M | B	SS: F | M | S	Sample: Q | SS	Track Score	3wC	3wL	M Straights	L/SlowC
Silverstone, UK	2011	84	9	-0.33	0.11	-0.33	8.5 | 7.1 | 7.8	9.2 | 7.0 | 8.2	12 | 12	0.32	4	55%	3	21.0
Melbourne, Australia	1996	76	8	-0.30	0.02	-0.36	8.9 | 8.1 | 8.0	8.7 | 8.2 | 7.5	9 | 7	-0.07	2	42%	4	10.9
Hockenheim, Germany	2002	68	7	-0.28	0.04	-0.44	8.0 | 5.5 | 6.7	7.4 | 6.7 | 6.4	10 | 8	0.98	4	75%	3	13.6
Shanghai, China	2004	80	7	-0.24	0.09	-0.41	7.8 | 6.9 | 6.2	7.6 | 7.2 | 6.5	5 | 5	-0.07	0	78%	4	16.0
Estoril, Portugal	1984	65	7	-0.22	0.11	-0.08	8.9 | 5.3 | 7.3	8.1 | 6.0 | 9.3	4 | 4	-0.32	1	55%	3	13.0
Sakhir Outer Circuit, Bahrain	2020	56	6	0.04	0.13	na	na | na | na	na | na | na	0 | 0	0.04	2	75%	1	14.0
Abu Dhabi, UAE	2009	80	8	-0.11	0.22	-0.50	8.4 | 6.9 | 6.4	9.7 | 6.3 | 9.3	6 | 4	0.28	2	65%	3	10.0
Indianapolis, USA	2000	61	9	-0.10	0.23	-1.88	9.6 | 6.2 | 3.9	8.0 | 6.6 | 3.4	4 | 4	1.09	4	67%	4	10.2
Francorchamps, Belgium	1983	97	9	0.01	0.34	0.56	10.0 | 11.3 | 6.0	4.0 | 11.1 | na	2 | 2	-0.12	2	60%	3	19.4
Oyama, Japan	1976	60	5	0.02	0.35	-0.09	10.0 | 5.4 | 11.0	9.6 | 9.0 | 8.1	8 | 8	0.48	3	75%	3	20.0
Location	1st Year	L	C	Raw Score	Adj Score	Results Score	Q: F | M | B	SS: F | M | S	Sample: Q | SS	Track Score	3wC	3wL	M Straights	L/SlowC
Imola, Italy	1995	74	9	0.07	0.39	0.33	8.3 | 5.9 | 7.2	5.3 | 8.3 | 8.1	6 | 6	-0.33	1	24%	6	8.2
Castellet, France	1971	83	7	0.10	0.43	0.40	6.2 | 8.6 | 6.3	6.0 | 6.8 | 8.1	4 | 4	-0.09	1	71%	3	16.6
Silverstone, UK	2010	84	9	0.32	0.40	na	na | na | na	na | na | na	0 | 0	0.32	4	55%	3	21.0
Rouen, France	1957	93	8	0.44	0.77	0.12	7.6 | 7.9 | 8.8	9.6 | 6.8 | 9.1	7 | 7	1.20	6	56%	5	23.3
Austin, US	2012	80	9	0.45	0.77	0.09	6.2 | 6.6 | 7.3	7.7 | 5.8 | 7.0	8 | 8	1.90	6	79%	4	11.4
Greater Noida, India	2011	73	9	0.62	0.95	0.20	5.5 | 16.0 | 7.3	6.7 | 6.4 | 6.0	4 | 4	0.91	4	62%	4	12.2
Mexico City	2015	63	8	0.62	0.95	0.62	6.3 | 7.0 | 6.5	5.5 | 6.3 | 8.6	3 | 3	0.62	3	71%	2	7.9
Sepang, Malaysia	1999	82	8	1.01	1.53	1.01	7.4 | 6.7 | 8.3	5.5 | 7.8 | 10.5	12 | 10	0.50	3	59%	4	13.7
Sepang, Malaysia	2016	82	8	na	na	na	8.3 | 9.0 | 3.5	na | na | na	1 | 0	0.51	3	59%	4	13.7
Baku, Azerbaijan	2016	86	10	1.02	1.35	1.02	8.8 | 9.9 | 9.5	7.0 | 8.6 | 11.9	7 | 7	1.03	3	70%	5	9.6
Sochi, Russia	2014	83	8	1.65	1.97	0.24	7.8 | 5.3 | 6.0	5.6 | 5.4 | 8.6	5 | 5	3.10	8	100%	5	11.9
Reims, France	1953	90	7	2.14	2.44	0.40	3.5 | 9.8 | 8.0	7.7 | 7.5 | 5.3	1 | 1	2.34	7	100%	4	22.5
Location	1st Year	L	C	Raw Score	Adj Score	Results Score	Q: F | M | B	SS: F | M | S	Sample: Q | SS	Track Score	3wC	3wL	M Straights	L/SlowC
Median Track	--	77	8	--	--	-0.53	8.8 | 6.7 | 6.4	9.5 | 6.6 | 6.8	6 | 5	-0.08	2	56%	3	12.6	--

Key and Notes

• General Information
• 1st year The first year that this track was run in this configuration.
• L Length of the track measured in total spaces the shortest way around the track. One space converts to about 0.07 km, 0.04 miles, or 76 yards.
• C The number of corners.
• Scores
• Raw Score If combined sample size is 20+ then this is equal to the Results Score. If I don't have any sample races, the score is equal to the Track Score. Otherwise it is an average of the Results Score and the Track Score -- weighting the Results Score score more and more the closer to 20 total samples I have.
• Adjusted Score Raw Score minus the median score. Note that I calculate the median differently if the score is a Results Score, Track Score, or average of the two.
• Results
• Results ScoreMore or less... pts scored by cars in the back and cars with 20 start speeds - pts scored by cars in the front and with 100+ start speed -- weighted if I have more samples for one of those data sets than the other. Then I divide by the standard deviation. So a Results score of 0 means that historically cars in the front and/or with 100+ start speeds score as many points on this track as cars who start in the back and/or with 20 start speeds. A -1 score means that the track has historically favored 100+ start speeds and/or starting in the front by a standard deviation... that's a significant amount.
• Q: F | M | B Qualifying data. Average points scored by a car that started in the (F)ront 2 rows, (M)iddle 2 rows, and (B)ack 2 rows.
• SS: F | M | S Start Speed data. Average points scored by a car that has a (F)ast Start Speed (100 or 120), (M)iddle Start Speed (60), and (S)low Start Speed (20).
• Sample: Q | SS Number of races where I have data related to (Q)ualifying position and (S)tart (S)peed. There are a few tracks where I do not have any Start Speed related data for one of the three groups and so that grouping is marked "na."
• Track
• Track Score An attempt to determine how a track will play out based on some objective measurments. The four measurements described below are combined to derive a score for how much I think the track will reward play from the front (negative scores) or play from the back. First I normalize the measures by subtracting the median and dividing by the standard deviation. Then I add all four together after a multiplier for the measures I have found are most predictive of actual results. The resulting raw track score is then divided by the standard deviation to normalize it to the previous 3 scores.
• 3wC 3-wide ending corners. This is the number of corners on a track whose last row is 3-wide. Of all the track atributes I have thought to count this is the most predictive of final result data so far and so get multiplied by 3 during the calculating of Track Score.
• 3wL Percentage of the track that is 3-wide. This is the second most predictive attribute I've found so far so it gets multiplied by 2.
• M Straights Number of "Medium" straights or straights longer than 3 spaces but shorter than 10 spaces. This attribute does not appear staistically relevant on its own but is more predictive than anything else. It is included in Track Score because it seems to help balance out some of the blind spots that the main two attributes have. It gets no multiplier.
• L/SlowC Track length divided by the number of "slow" corners or corners less than 120. This measure gets treated just like Medium straights BUT interestingly lower values here (which indicate fewer spaces between slow corners) tend to predict more points for people with lower start speeds and/or who start from the back so it is subtracted from the other measures to create Track Score.