Hi Chris
It seems as though the COTA lap times are available through the API but not yet updated on the disk image (which says last updated yesterday). Is this as expected?
Cheers
Toby
there are results for 111 drivers that are not in the drivers table. For example drivers 800 to 816 are some of the driverIDs in the results table that are not in the drivers table. Also, team 201 is missing. Also, there are some drivers that have 2 results for a race. For example driver 624 for race 788 has two results. What is the meaning of the two results? Is this an error did he take over a team mates car after a wreck?
Just noticed Jacques’s laptimes for the 2004 Chinese GP are linked to Trulli instead.
2004’s the season where Trulli drove for Renault for rounds 1 to 15, skipped 16 (China), and jumped over to Toyota for rounds 17 and 18. Villeneuve took over Trulli’s seat at Renault for rounds 16, 17, and 18.
# Jacques’s and Trulli’s records in drivers
select *
from drivers
where driverId in (15,35); # 15 is trulli, 35 is Jacques
# raceId for 2004 China
select *
from races
where raceId = 105;
# query showing Jacques’s laptimes for 2004 China are all associated with Trulli
select *
from laptimes as l
join drivers as d on d.driverId = l.driverId
where l.raceId = 105 AND d.driverId in (15,35); # 15 is trulli, 35 is Jacques
AFAICT, the records in the results table are fine. Villeneuve has results for rounds 16 to 18. Trulli has results for rounds 1 to 15, 17, and 18.
select driverRef, c.name, races.*
from races
join results on races.raceId = results.raceId
join drivers d on d.driverId = results.driverId
join constructors c on c.constructorId = results.constructorId
where year = 2004
AND d.driverId in (15,35)
order by round;
Thanks for maintaining a fantastic project!!!
TK
# temp fix?
update laptimes
set driverId = 35 # Jacques
where raceId = 105 # China
AND driverId = 15; # Trulli
The first one appears to just require relinking the laptimes to the correct driver.
– 2014 Brazilian GP (raceId=917)
— Kevin Magnussen’s times (driverId=825) are linked to the “original” Magnussen (driverId=76). The laptimes and positions match up with other published records I found.
I wasn’t able to find laptimes and positions for the next ones, but they might just be linked to the wrong driver.
– 1997 European GP (raceId=223)
— Fontana’s times (driverId=82) are missing and might be linked to Morbidelli (driverId=81). Morbidelli had an injury (or injuries) and retired before the season ended. Fontana subbed for Morbidelli for this race.
– 1999 San Marino GP (raceId=177) and Spanish GP (raceId=179)
— Salo’s times (driverId=63) are missing and might be linked to Ricardo Zonta (driverId=41). Zonta was injured during practice in Brazil (round 2) and was subbed by Salo for the next three races (these two, plus Monaco).
– 2001 Italian GP (raceId=155)
— Alesi’s times (driverId=55) are missing and might be linked to Trulli (driverId=15). Trulli retired in the first lap, but has laptimes linked for this race.
The next two might be missing records and just need to have them re-ingested.
– 1999 Monaco (raceId=178)
— Salo (driverId=63) doesn’t have laptimes, but the results table and wiki say he completed 36 laps. However, there don’t appear to be any incorrectly linked laptime records for this race.
– 1999 Brazilian GP (raceId=176)
— Sarrazin (driverId=72) doesn’t have laptimes, but the results table and wiki say he completed 31 laps. However, like the instance above, there don’t appear to be any incorrectly linked laptime records for this race.
This last one might be tricky (or maybe not).
– 2009 Italian GP (raceId=13)
— Alguersuari (driverId=153) completed 19 laps in this race, but doesn’t have any laptime records. Webber (driverId=17) actually retired in the first lap, but has 52 laptime records associated to him. But that’s too many records to match Alguersuari’s lap count for this race (19 vs 52). However, Buemi (raceId=67) completed 52 laps, but only has 19 records; which matches Alguersuari’s number. So, maybe Webber has Buemi’s records, while Buemi has Alguersari’s. Might be easier to just drop all the laptime records for that race and just reimport the data.
1. Is the laptime record for the driver’s 1st lap missing?
2. Is the laptime record for the driver’s total laps missing?
The last one’s a bit heavier.
3. Do the number of laptime records match the driver’s results.laps value for the race?
I don’t know if all the scenarios identified with #2 and #3 are problems. #3 identified 2014 China, the race where Hamilton was shown the chequered flag a lap early. So, the laptime records for that race are probably okay. But #3 also identified Perez’s results in 2014 Hungary. His results.laps value is 23. But, the laptime records say he completed 22 laps, which matches the FIA and Wiki’s listed completed laps.
Anyway, there were 8 records returned for #1, 61 for #2, and 97 for #3. Given the overlap with what the queries identify, I’d guess there’s a max of 97 driver results that >might< have misplaced laptimes. So basically a small percentage, given that there are 8,497 total driver results during this period.
I stumbled over a few errors but in reading through the recent feedback above I see TK has already noted them. Presumably still on the list to be corrected (2014/MAG in particular was my issue).
As always, thanks for your awesome work and great work TK on the validation code.
Elio De Angelis’ time in the 1980 Austrian GP (resultId=12244) is incorrect – it’s a duplicate of Piquet’s time.
The correct values are: `time` = ‘+1:14.97’, milliseconds = 5250700.
There seems to be something wrong in the lap times for the 2018/1 race, most of them are way too low.
The lap times for Vettel seems to be right. For Hamilton and Raikkonen only the safety car laps are off. For the other drivers almost all laps are off.
i found mistake in laptimes from Bahrain 18 GP. Here Gasly’s lap 20 time is 1:56.448, but in FIA’s pdf this time was achieved in lap 16 after his pitstop. It looks like his laptimes are in wrong order.
Thanks to work by Miguel Bento the lap times for the Belgian and Italian Grand Prix are now available.
Hi Chris
In case you’re not already aware but there seems to be an issue with the lap times data again after Singapore.
Cheers
Toby
Hi Toby,
Yes, there’s an ongoing problem with the import of the lap times so they’re likely to be late. Try now.
Chris
Hi Chris
It seems as though the COTA lap times are available through the API but not yet updated on the disk image (which says last updated yesterday). Is this as expected?
Cheers
Toby
Hi Toby,
That’s odd. As it’s easier than checking, I’ve created a new dump and updated the date. Let me know if the problem remains.
Thanks,
Chris
there are results for 111 drivers that are not in the drivers table. For example drivers 800 to 816 are some of the driverIDs in the results table that are not in the drivers table. Also, team 201 is missing. Also, there are some drivers that have 2 results for a race. For example driver 624 for race 788 has two results. What is the meaning of the two results? Is this an error did he take over a team mates car after a wreck?
Sorry, missing drivers is error on my part. Still remains the question about why some drivers have two results for the same race. Thanks.
Hi Bob,
Drivers sharing or swapping cars was an interesting feature of early races – difficult to imagine nowadays!
Chris
Hi Awesome Ergast Admin(s),
Just noticed Jacques’s laptimes for the 2004 Chinese GP are linked to Trulli instead.
2004’s the season where Trulli drove for Renault for rounds 1 to 15, skipped 16 (China), and jumped over to Toyota for rounds 17 and 18. Villeneuve took over Trulli’s seat at Renault for rounds 16, 17, and 18.
https://en.wikipedia.org/wiki/Renault_in_Formula_One#2004_season
https://en.wikipedia.org/wiki/2004_FIA_Formula_One_World_Championship#Drivers.27_Championship
Anyway, here are the IDs and some verification queries.
# 2004 China :: raceId = 105
# Trulli :: driverId = 15
# Villineuve :: driverId = 35
# Jacques’s and Trulli’s records in drivers
select *
from drivers
where driverId in (15,35); # 15 is trulli, 35 is Jacques
# raceId for 2004 China
select *
from races
where raceId = 105;
# query showing Jacques’s laptimes for 2004 China are all associated with Trulli
select *
from laptimes as l
join drivers as d on d.driverId = l.driverId
where l.raceId = 105 AND d.driverId in (15,35); # 15 is trulli, 35 is Jacques
AFAICT, the records in the results table are fine. Villeneuve has results for rounds 16 to 18. Trulli has results for rounds 1 to 15, 17, and 18.
select driverRef, c.name, races.*
from races
join results on races.raceId = results.raceId
join drivers d on d.driverId = results.driverId
join constructors c on c.constructorId = results.constructorId
where year = 2004
AND d.driverId in (15,35)
order by round;
Thanks for maintaining a fantastic project!!!
TK
# temp fix?
update laptimes
set driverId = 35 # Jacques
where raceId = 105 # China
AND driverId = 15; # Trulli
Thanks TK,
I’ve reviewed and applied your fix.
Chris
Hi Chris,
Found a few other items for laptimes.
The first one appears to just require relinking the laptimes to the correct driver.
– 2014 Brazilian GP (raceId=917)
— Kevin Magnussen’s times (driverId=825) are linked to the “original” Magnussen (driverId=76). The laptimes and positions match up with other published records I found.
I wasn’t able to find laptimes and positions for the next ones, but they might just be linked to the wrong driver.
– 1997 European GP (raceId=223)
— Fontana’s times (driverId=82) are missing and might be linked to Morbidelli (driverId=81). Morbidelli had an injury (or injuries) and retired before the season ended. Fontana subbed for Morbidelli for this race.
– 1999 San Marino GP (raceId=177) and Spanish GP (raceId=179)
— Salo’s times (driverId=63) are missing and might be linked to Ricardo Zonta (driverId=41). Zonta was injured during practice in Brazil (round 2) and was subbed by Salo for the next three races (these two, plus Monaco).
– 2001 Italian GP (raceId=155)
— Alesi’s times (driverId=55) are missing and might be linked to Trulli (driverId=15). Trulli retired in the first lap, but has laptimes linked for this race.
The next two might be missing records and just need to have them re-ingested.
– 1999 Monaco (raceId=178)
— Salo (driverId=63) doesn’t have laptimes, but the results table and wiki say he completed 36 laps. However, there don’t appear to be any incorrectly linked laptime records for this race.
– 1999 Brazilian GP (raceId=176)
— Sarrazin (driverId=72) doesn’t have laptimes, but the results table and wiki say he completed 31 laps. However, like the instance above, there don’t appear to be any incorrectly linked laptime records for this race.
This last one might be tricky (or maybe not).
– 2009 Italian GP (raceId=13)
— Alguersuari (driverId=153) completed 19 laps in this race, but doesn’t have any laptime records. Webber (driverId=17) actually retired in the first lap, but has 52 laptime records associated to him. But that’s too many records to match Alguersuari’s lap count for this race (19 vs 52). However, Buemi (raceId=67) completed 52 laps, but only has 19 records; which matches Alguersuari’s number. So, maybe Webber has Buemi’s records, while Buemi has Alguersari’s. Might be easier to just drop all the laptime records for that race and just reimport the data.
Later,
TK
Thanks TK – I’ll work on these when the season closes.
Chris
Hey Chris,
I figured I’d make this a bit more automated and wrote some validation queries to identify potential problems like these.
https://git.io/vFANR
The first two are pretty quick.
1. Is the laptime record for the driver’s 1st lap missing?
2. Is the laptime record for the driver’s total laps missing?
The last one’s a bit heavier.
3. Do the number of laptime records match the driver’s results.laps value for the race?
I don’t know if all the scenarios identified with #2 and #3 are problems. #3 identified 2014 China, the race where Hamilton was shown the chequered flag a lap early. So, the laptime records for that race are probably okay. But #3 also identified Perez’s results in 2014 Hungary. His results.laps value is 23. But, the laptime records say he completed 22 laps, which matches the FIA and Wiki’s listed completed laps.
Anyway, there were 8 records returned for #1, 61 for #2, and 97 for #3. Given the overlap with what the queries identify, I’d guess there’s a max of 97 driver results that >might< have misplaced laptimes. So basically a small percentage, given that there are 8,497 total driver results during this period.
TK
Hi, I have been playing with your database and discovered the following discrepancies:
– https://en.wikipedia.org/wiki/1950_Italian_Grand_Prix you have listed Serafini and Ascari racing for different team, but acc to wiki they were both in ferrari
– https://en.wikipedia.org/wiki/1951_French_Grand_Prix you put Fangio and Fagioli in car number 24 (should be 4) and different teams (should be alfa romeo)
– https://en.wikipedia.org/wiki/1953_German_Grand_Prix you put Farina with car number 12 (it was 2)
– https://en.wikipedia.org/wiki/1953_Italian_Grand_Prix you put Landi in car number 2 (it was 42)
Hi! I just noticed that this query
https://ergast.com/api/f1/1978/constructors/brabham/races
feeds back the results twice per row.
race 1
race 1
race 2
race 2 and so on
Hi Giannis,
Thanks for the heads-up – now fixed.
Chris
Hi,
The 2018 calendar list Round 3 Chinese GP as 15 May 2018, it should be 15 April 2018.
Hi, The times for the 2018 season need to be updated to match the new start times as announced here: https://www.formula1.com/en/latest/headlines/2018/2/new-race-weekend-schedule-announced-for-2018.html
Also as mentioned before the Chinese Grand Prix is listed as the 15th of May, instead of the 15th of April.
Thanks Scott, Thanks Aaron – schedule updated. Does anyone know of a list somewhere with race start times in UTC?
Hi Chris
I stumbled over a few errors but in reading through the recent feedback above I see TK has already noted them. Presumably still on the list to be corrected (2014/MAG in particular was my issue).
As always, thanks for your awesome work and great work TK on the validation code.
Cheers
Toby
Hi Toby,
Thanks for the heads-up. I forgot about the second list. I’ve done 2014/MAG but the others need further investigation.
Chris
Hi,
Elio De Angelis’ time in the 1980 Austrian GP (resultId=12244) is incorrect – it’s a duplicate of Piquet’s time.
The correct values are: `time` = ‘+1:14.97’, milliseconds = 5250700.
Thanks Emkael – now corrected.
Chris
Hello,
There seems to be an issue with the FastestLap data on at least one result:
https://ergast.com/api/f1/2017/20/results
31
Esteban
Ocon
1996-09-17
French
Force India
Indian
9
55
Finished
+1:38.911
1:42.609
1:42.609
As you can see, the AverageSpeed field is the same as the Time field. Just wanted to make you ware. Thanks for the wonderful API!
Caleb
Thanks Caleb – now corrected.
Chris
Hiya –
Great API as always 🙂
Just noticed the fastest lap for the Australian GP (Ricciardo) is missing (link below), but all other data is available:
http://ergast.com/api/f1/2018/1/fastest/1/results
Thanks Mike – I’ve had a lot of issues with the data at the start of the season, due to unforeseen changes in formats.
Hi Chris
Lots of niggling changes to the data publication format, eh?!
Just spotted an issue in 2018/1 quali results – red_bull – HAR and GAS have constructor_id red_bull rather than toro_rosso
–tony
Thanks Tony – now corrected.
Chris
Hi Chris,
There seems to be something wrong in the lap times for the 2018/1 race, most of them are way too low.
The lap times for Vettel seems to be right. For Hamilton and Raikkonen only the safety car laps are off. For the other drivers almost all laps are off.
Jeffrey
Hi Jeffrey.
Thanks for warning – now corrected.
Chris
Hi Jeffrey,
Check the lastest image. They should be there.
Chris
Hi Chris,
Amazingly fast response! I can confirm the lap times are now correct.
Thanks!
Jeffrey
Hi,
i found mistake in laptimes from Bahrain 18 GP. Here Gasly’s lap 20 time is 1:56.448, but in FIA’s pdf this time was achieved in lap 16 after his pitstop. It looks like his laptimes are in wrong order.
Lukas
Hi Lukas,
Thanks for the warning. Now corrected.
Chris