GTR2 Weather

[H@L9000]

Some of the comments about weather that started in the U11 Spa race thread have got me thinking:
https://www.simracing.org.uk/smf/index.php?topic=7272#event1964

What level of accurate forecast could we expect at a modern race track in the real world?

My knowledge only comes from watching F1 and as far as I know, there is a track forecast done by local weather groups and some teams also send a chopper up to take a look. I know that some jets have a Doppler Radar on board and I assume that some choppers do as well and that is the significance of sending a chopper up. The radar can tell how dense clouds are within a limited range.

This seems to give reports such as 'no rain expected for the next 35 mins'. Probably 6 out 10 times they will be correct, the other times they could be a little bit wrong or they could be completely wrong.

I think I need to speak to a Meteorologist.

What would be the best possible way that a Weather generator for GTR2 could be used to realistically simulate the Weather?

From the discussion in the above link, it would seem that any notification of a big change in the weather should be passed to the drivers during a race and before that big change happens. The problem with this is that we can not get any extra warning once the race has started. Not unless the makers of XD or some one with similar skill wants to get more serious about weather.

I think I have seen F1 drivers get as little as 10 mins warring about rain that was not expected before the race. Also what about the situation of clouds that look like rain but dont rain for like 50 mins then just poor it out unexpectedly?

A Weather generator could potentially report what it will do for a entire race before that race has started but, at what point does it take the unpredictability away?

Feel free to comment so I dont look silly.

[Legzy]

Well I for one am chuffed that it's got someone thinking in a constructive way about how we could improve what we 'mostly' enjoy currently. Nice one H@L.

What level of accurate forecast could we expect at a modern race track in the real world?

This seems to give reports such as 'no rain expected for the next 35 mins'. Probably 6 out 10 times they will be correct, the other times they could be a little bit wrong or they could be completely wrong.

What would be the best possible way that a Weather generator for GTR2 could be used to realistically simulate the Weather?

A Weather generator could potentially report what it will do for a entire race before that race has started but, at what point does it take the unpredictability away?

I've taken the highlights from what you said above.
I'm fairly sure we only have a couple of people here capable of producing an updated weather generator (& I'm not one of them ), but I guess that if a new generator was told to do "dry start with showers later in the race" for example, it could be done at different levels & with a 'slight' randomness factor, to already have a forecast for everyone, that maybe 7 times out of 10 would be totally accurate & it's more down to when the 'probable' changes (if any) will take place . The other times it would work out fairly close but not quite right. Just like real forecasting.

But as I said, I don't know how to create these things, so I could be speaking out my Bass.

[Joss]

The only thing that seems to be dodgey is how large the weather changes can be; and how quickly it changes.

Saying that, changable weather races are usually the most fun. It's one of the main reasons I like GTR2 over rFactor.

[Truetom]

I'd like to keep dynamic weather if possible.
H@l, I remember you sent me some clever weather file where you could specify when in the race (xxx minute) conditions would shift and into what direction. As far as I can see the main problem weather not-haters have is that it's too much randomness in the changes. I guess the conditions could be made according to how the weather actualy was during the day of the race, not at the conditions at 8 pm. The race time is set to 2 pm, right? I guess this would take away random factor for the guy who actualy sets the weather as he would know what's gonna happen.   Though we also could set margins of weather changes a bit closer to keep some randomness.

The internal weather forecast could be made in announcement that the "starting race conditions would be wet, with drying out in the second half of the race". Or whatever. So the minute by minute changes would be known to one guy only and others would have the direction of weather pattern throughout the race - with an occasional surprise. The weather creator would have to be a person above any conspiratory suspicion, of course.  

What was it? Oh - feel free to comment so I don't look silly.  

[Simon Gymer]

For me I think it's two fold:
1) The fact that the forecast at the start of the race is often useless which makes the initial tyre gamble a lottery.
2) As much as I enjoy the odd wet race, having one every week spoils the season for me because I want to actually drive the car "properly" more than I want to have fun staying on.

Once you're in the race you're all in the same boat (no pun intended) and if weather changes, weather changes, just like it does in the real world. I don't mind big changes in weather during the race, that happens, it's annoying, but it happens.

I just find it "not very helpful" that when it starts off and says damp and when you get out on the pace lap and find it's chucking it down, your tyre selection has been a lottery. If damp was correct most people would choose a more appropriate tyre. We have to be in the business of helping people finish races and the current weather forecasts don't help. Is this really a game "problem", well I assume so, I believe we just set the start weather for the session and a changeability factor and then the game does the rest.

So what can you do about it?
1) Not use changeable weather? Is that better?
2) Reduce the changeability factor used when generating the weather?
3) Have only 4 out of 8 races run with changeable weather, guaranteeing dry weather for 50% of the races? That might work well in winter, but in summer we might get 100% dry races, but then again is that a problem? Then maybe also guarantee 1 wet race a season with fixed weather?
4) Not base the weather on current real world weather? Seems like a good way of seeding the weather and gives people a better idea what the weather may do than if we just make it up at random each week, but in winter it means and has meant in the past that we get a lot of wet races in the winter seasons.

[Clive Loynes]

Do they actually race at Spa in January?

If real weather is used to seed the conditions, shouldn't a winter series be made up of Australia, Brazil and South Africa etc.?

It might still rain but not as often.

[Dave 'Gizmo' Gymer]

I just don't see the point in an 'accurate' forecast. If it's generated right before the race, people will still complain that they haven't tested for the conditions, and if it's generated a few days in advance, all that will happen is that the people with lots of time to test tyre and setup combinations will have a huge advantage over those who don't - even bigger than they do now.

I'd rather just fix the weather a few days beforehand, at least then we all get a chance to do some simple testing in those conditions without spending hours at it.

[Busi]

For me I think it's two fold:
1) The fact that the forecast at the start of the race is often useless which makes the initial tyre gamble a lottery.

There is an easy solution to this one which I think must should be used in races with this weather system: add a 2 minute warm up between quali and race

This way people can give tyres a quick try. Make it a 5 minute warmup if necessary. With a couple of quick laps you can see if the tyres you're pretending to use for the race start reach a healthy temperature. As I understand that the warmup always has the same weather as the race start. Or at least this is what I've learned from Gamers Crib AND Race2Play. I think it's a pretty sensible (and logical) thing to do.

[Simon Gymer]

If it's generated right before the race, people will still complain that they haven't tested for the conditions

I'm not worried about those people. That's just tough as it is in real life. You don't need to practice in all conditions to be ready to race. I VERY rarely do any practice in any car in any race anymore, I simply don't have the time or enthusiasm to do it. So what? I don't use that as an excuse anymore, I just accept that I'm not going to do it and that means that I'm not as fast as I would be if I did practice. Big deal.

You have to remember Dave that just cause people complain isn't a reflection on the league or format or anything outside of themselves necessarily. I can complain I had a bad race and that I can blame it on x, y and z, but so what, just ignore complaints that are cause people only have themselves to blame or are trying to achieve things they simply are never going to acheive. Don't take it all to heart.

This is just trying to find improvement where and IF they can be made not because people complain but because we want to a) have as much enjoyment for as many people as possible b) keep people driving to the 80th minute and not crash out.

[Dave 'Gizmo' Gymer]

There is an easy solution to this one which I think must should be used in races with this weather system: add a 2 minute warm up between quali and race

This way people can give tyres a quick try. Make it a 5 minute warmup if necessary. With a couple of quick laps you can see if the tyres you're pretending to use for the race start reach a healthy temperature. As I understand that the warmup always has the same weather as the race start. Or at least this is what I've learned from Gamers Crib AND Race2Play. I think it's a pretty sensible (and logical) thing to do.

The weather at the start of prac/qually is the same as at the start of the race. That's how the files are constructed now. Warmup won't add anything but another yawnworthy 2 minute delay on top of the existing 2 minute gridding period.

[spanner]

I'm not worried about those people. That's just tough as it is in real life. You don't need to practice in all conditions to be ready to race. I VERY rarely do any practice in any car in any race anymore, I simply don't have the time or enthusiasm to do it. So what? I don't use that as an excuse anymore, I just accept that I'm not going to do it and that means that I'm not as fast as I would be if I did practice. Big deal.

You maybe able to jump in a car with no practise or setup and quallify at the head of the field but the rest of us cant. Its more about be safe, if your not given the time to adapt the car and yourself to the conditions, and everyones different here, then your not going to enjoy it, knowing it ends up being a test of mental endurance runing on your own with no one to even think about pretending to race with. It probably means your more likely to crash out anyway.

[Truetom]

How I prepare for dynamic weather?
- make a dry setup
- have the same setup tried in 50% wet
- adjust changes if necessary (higher wing, softer suspension overall, diff)
- check how modified setup acts in dry

I find I need less and less difference between wet and dry setup. The car is either ok or it isn't - for either. All in all - 30 minutes. And I don't need UKGTR server running for it. I come on for company and spying mostly.

[goldtop]

I created dry, damp and wet settings for Spa and chose the damp (inters) for the start of the race because the game described the conditions as "damp". When the race started I had little grip and this soon got worse as the rain became heavier. Yesterday I checked my damp setting again on a damp track (20%) offline and had lots of grip in conditions which appeared to be very similar. Even at 30% I had more grip than I had at the start of the race.

I do need to prepare and practice a lot for all my races to be anywhere near competetive, but at Spa I found this to be a complete waste of time. Surely the whole idea of this league is to make racing an enjoyable experience for everyone. I welcome any improvements that can be made to this end.

[Dave 'Gizmo' Gymer]

Surely the whole idea of this league is to make racing an enjoyable experience for everyone.

Unfortunately, not "everyone" has the same criteria for an enjoyable experience.

[Legzy]

This is just trying to find improvement where and IF they can be made not because people complain but because we want to a) have as much enjoyment for as many people as possible b) keep people driving to the 80th minute and not crash out.

Well said Simon.

I created dry, damp and wet settings for Spa and chose the damp (inters) for the start of the race because the game described the conditions as "damp". When the race started I had little grip and this soon got worse as the rain became heavier. Yesterday I checked my damp setting again on a damp track (20%) offline and had lots of grip in conditions which appeared to be very similar. Even at 30% I had more grip than I had at the start of the race.
I do need to prepare and practice a lot for all my races to be anywhere near competitive, but at Spa I found this to be a complete waste of time. Surely the whole idea of this league is to make racing an enjoyable experience for everyone. I welcome any improvements that can be made to this end.

If you'd had a forecast on the forum just 10 mins before Spa saying "Damp start with heavier rain likely to follow", you may well have been able to make an educated decision at HW's instead of Inters, especially if it was a close call anyway. As it was, you joined the server, saw it was damp & effectively had no indication what weather would do, so picked Inters (like I did)... Which it turned out were soon not man enough for the job, your "race" was effectively over at that point due to an unlucky underinformed choice. Funnily enough, if we'd both been in different cars, we may well have been on HW's because the Inters on some cars would not have been suitable (i.e. the performance difference between Pirelli/Michelin in wet conditions)