| Pengarang |
Topik: Rain Probability Wrong??? |
47 balasan
|
|
|
|
#1 dikirim Apr 11 2012, 09:49:56
|
Sebut
|
Last Race the rain probability said....
Start - 0h30m
Rain Probability 60% - 70%
0h30m - 1h00m
Rain Probability 0%
So why the hell was it still raining round 50th min into the race?????
I know this might be a FOBY question but in my eyes it is a floor in the game if the rain probability is giving false information. Much help on this matter will be really appreciated :)
|
|
|
|
|
|
#2 dikirim Apr 11 2012, 09:51:05
|
Sebut
|
It's about how likely it is to START raining, not how likely it is to stop.
If it rains already when the 2nd quarter starts and it says 0%, you can be sure that it WILL end during that period, but you don't know when.
|
|
|
|
|
|
#3 dikirim Apr 11 2012, 09:52:53
|
Sebut
|
|
Quote ( Vincent Wilkins @ April 11th 2012,09:49:56 )
Probability
that is a word for your answer :)
|
|
|
|
|
|
#4 dikirim Apr 11 2012, 09:54:51
|
Sebut
|
|
As is the case with several other things in GPRO, you should not interpret the in game weather forecast like any sort of real world equivalent.
|
|
|
|
|
|
#5 dikirim Apr 11 2012, 09:55:06
|
Sebut
|
|
Santtu gave the perfect answer :)
|
|
|
|
|
|
#6 dikirim Apr 11 2012, 09:57:15
|
Sebut
|
Rain probability explained:
Rain Probability
0% rain probability means you can be sure that it won't start raining during the period. If it is already raining, you can be sure that it will stop raining during the period, but you don't know when it will happen (it can stop after one minute or after 29 minutes).
100% rain probability means that rain won't stop during the period. If it is dry at the start of the period, you can be sure it will start raining during the period - but again you don't know when that will happen.
http://wiki.gpro.net/index.php/Weather
|
|
|
|
|
|
#7 dikirim Apr 11 2012, 10:00:35
|
Sebut
|
|
Cheers Chinmay now i understand :)
|
|
|
|
|
|
#8 dikirim Apr 11 2012, 10:00:50
|
Sebut
|
|
The same question after each lottery rain race :D
|
|
|
|
|
|
#9 dikirim Apr 11 2012, 10:04:51
|
Sebut
|
|
Quote ( Artur Balcerzak @ April 11th 2012,10:00:50 )
The same question after each lottery rain race :D
Some might draw the conclusion that the current format actually is misleading...
|
|
|
|
|
Quote ( Artur Balcerzak @ April 11th 2012,10:00:50 )
The same question after each lottery rain race :D
Which is to be expected with new players arriving all the time. It is not an intuitive system that makes sense from a real-world perspective. It can be debated in endlessness whether it is good from a game-perspective or not and this has also been done, multiple times..
|
|
|
|
Again this has probably been discussed a million times but taking yesterdays race:
There was a 60-70% chance it would start raining in the first 30mins, it did so after 4 laps, we then knew it would stop raining at some point between 30-60mins, I believe it was around the 48th min how is this a prediction?
Basically for half the race the weather could have done anything, I dont believe there is a way to predict it (?) so for half the race it was a complete lottery. Is there no way of improving forecast or adjusting so the forecast is at 15mins intervals (not that this will help too much unless the forecast is either 0 or 100%). Any ideas or opinions?
|
|
|
|
|
|
#12 dikirim Apr 11 2012, 12:02:24 (terakhir disunting Apr 11 2012, 12:12:56 oleh Georg Petrov)
|
Sebut
|
I agree with you, Kevin.
The current system is flawed and based on luck, to a huge extent.
Comming into last race, I thought about my strategy for a long long time.
Needless to say, I failed miserably, but that's not the point of my post. The race before yesterday, we had a 20-30% rain probability and saw it rain for 12 laps.
In Monaco, same year, we had a probability of 25-30% with higher humidity and the race was a completely dry one.
Now explain to me, how on earth are you to cover both strategies, if it (lets say hypothetically) rains for 5-6 laps, stops, rains agein for 7-10 laps. In this case, it could have rained for half the race, like it did yesterday, but it could have also rained for a few laps and stopped, in which case NOT stopping for tyres would win you a huge amount of time.
Comparing GPROs weather to the real one is flawed, as during a real race, teams can and DO intervene during a weather change. We on the other hand have to sit infront of the screen and watch our races get screwed up by the very random rain mechanics of the game.
How does this random element fit in a (generally speaking) calculative game, where the best managers plan their races several seasons in advance?
A bit randomness is ok, and I salute it. But the way weather works on Gpro, it's a lose-lose situation. There is no way to cover all probabilities and that gets me really worked up. Here are the possibilities we had for last race:
1. Start on wets and hope for rain to come very fast (was the right strategy, no doubt)
2. Start on dries and change to wets as soon as weather changes (2nd best strategy in our case)
3. Start on dries and not pit for wets, due to the probability that rain might - if at all come for a brief period (absolute race killer yesterday)
Lets emphasise a bit more:
Strategy 1: What happens when it stays dry for more than 7 laps and my tyres are already dead when it starts raining?
Strategy 2: What happens if we have several weather changes and I find myself pitting every 3-4 laps?
Strategy 3: What happens if it rains early and stays wet for a very long time?
Now depending on how the weather gods feel like it, any and every single of these 3 strategies could have payed off big time or flopped big time. How does this fit in GPRO?
I am open for critisism, if you feel I am wrong about having 3 strategy options yesterday, but the way I see it, this was an absolute lottery race, hands down.
|
|
|
|
Well yes, the weather system could do with an improvement but last nights race was quite straightforward to plan for. The general idea with the current weather system is to cover all weather change possibilities. If course, you could still get caught out in certain cases but you should give yourself the best chance.
And before anyone points it, I didn't quite get it right last night either.
|
|
|
|
I admit yesterdays race is not the best of examples for a "complete" lottery race, but I am sure you will agree that we have had absolutely impossible races to make "the right call" from the get go.
Still, even yesterday, 2-3 laps less rain /later rain/ more rain could have changed the outcome of the race COMPLETELY.
|
|
|
|
|
Quote ( Shoaib Mohamed @ April 11th 2012,12:12:22 )
Well yes, the weather system could do with an improvement but last nights race was quite straightforward to plan for. The general idea with the current weather system is to cover all weather change possibilities. If course, you could still get caught out in certain cases but you should give yourself the best chance.
I agree with this. There are always ways to play on the safe side - you may not get any big advantage from your strategy but you also reduce the risk of ruining your race completely (like you would with a more risky strategy). I don't think the system needs major changes, i'd just like a little bit less randomness.
|
|
|
|
|
The simple answer is to change your strategy to suit most weather conditions, fuel up well and pick tyres which last longer, after all its a game of taking risks, your either being rewarded or penalized in the end but at least you cover yourself for fuel and tyres. This is the best way to deal with it instead of moaning about it. We all have to deal with the unknown, after all Fred the weather man did yesterdays rain forecast so what you expect?
|
|
|
|
I'm not complaining, Andrew. (Did enough of that yesterday :P )
I just feel, a new system could mix things up a bit and with a touch less randomness make the game a better experience for everyone.
I could bring at least 10 ideas for positively changing the way weather gets calculated. The thing is, I doubt it is worth the time. (mind you, I am not sure I actually know EXACTLY how the weather gets calculated)
|
|
|
|
|
Quote ( Georg Petrov @ April 11th 2012,12:26:03 )
I am not sure I actually know EXACTLY how the weather gets calculated
a little help Georg ;) this speaks in favour of admins imo.
Complex mathematical equations are used to predict the physics and movement of the atmosphere and to determine the rates of change in the atmosphere over time. The rates of change predict the state of the atmosphere a short time in the future. The equations are then applied to this next atmospheric state to find new rates of change, and these new rates of change predict the atmosphere at a yet further time into the future. This is called time stepping.
This time stepping is continually repeated until the desired forecast time in the future is reached. Computation time for a regional weather model is from a few seconds to a few minutes.
Forecasts cannot accurately predict the state of the atmosphere in two weeks time. Tiny errors in the initial data input, such as for temperature and wind, double for every 5 days of prediction within numerical models. This means that a forecast for the next few hours is more accurate than a forecast for next week.
|
|
|
|
Yesterday was pretty much as I expected it to be, maybe a few laps longer than expected but not unreasonably so. Shoaib is absolutely correct, you try to cover as many reasonable options as possible and there is actually some skill involved in doing this. I've been burnt a number of times on seeing the result that a small change to my fuel would have saved me loads of time while having small impact when not happening.
In addition, everyone is playing from the same forecast and strategies are usually not all that different. There are a few that plays against the odds on the hope that they will hit jackpot, this'll work once in a while but not consistently. My main issue here is that these races are relatively "rare" which means that streaks of weird races have a higher impact than warranted.
All that said, I think weather could be made more interesting while remaining a largely random factor. However, I've gotten past the stage where I found it really annoying though.
|
|
|
|
|
Quote ( Vincent Wilkins @ April 11th 2012,09:49:56 )
in my eyes it is a floor in the game
lol
|
|
|
|
|
|
#21 dikirim Apr 11 2012, 12:36:05 (terakhir disunting Apr 11 2012, 12:39:33 oleh Georg Petrov)
|
Sebut
|
|
Quote ( Fernando Fernandes @ April 11th 2012,12:28:53 )
Complex mathematical equations are used to predict the physics and movement of the atmosphere and to determine the rates of change in the atmosphere over time. The rates of change predict the state of the atmosphere a short time in the future. The equations are then applied to this next atmospheric state to find new rates of change, and these new rates of change predict the atmosphere at a yet further time into the future. This is called time stepping.
This time stepping is continually repeated until the desired forecast time in the future is reached. Computation time for a regional weather model is from a few seconds to a few minutes. thanks good guy greg!
Quote ( Fernando Fernandes @ April 11th 2012,12:28:53 )
This means that a forecast for the next few hours is more accurate than a forecast for next week.
Thanks captain obvious. ;)
Btw, I forgot to mention I was talking about GPRO weather. My bad :(
Quote ( Pål Göran Stensson @ April 11th 2012,12:32:01 )
Yesterday was pretty much as I expected it to be,...
Ok, lets talk about hungary's weather, then. Did that go "to plan" too? I feel it shouldn't have rained in hungary. Am I very far off?
|
|
|
|
|
Quote ( Georg Petrov @ April 11th 2012,12:36:05 )
Btw, I forgot to mention I was talking about GPRO weather. My bad :(
No problem at all :p But i guess that´s why the weather formula is GPRO´s most well kept secret :D
|
|
|
|
|
|
|
Quote ( Vincent Wilkins @ April 11th 2012,09:49:56 )
So why the hell was it still raining round 50th min into the race?????
And what's wrong with that? % rain means a chance of rain START, not STOP.
Read the rules, it's clearly stated there that rain DOES NOT stopped immediately.
The 0% chance in second quarter only means that after rain stops, it WONT rain again at that point. It does not absolutely mean the rain HAS to stop in the first lap of the second quarter, Elite time.
It's really disapointing that everytime something like this happens (which was totally normal by game rules and weather conditions), and someone has interpreted it wrong and messed its own strategy, it has to be a "bug" in the game.
It is not. People must learn how to read the forecast. Just that.
Or do you really think that double champ Leandro Sereno tossed a coin up there in Elite and was the only wet-starter in the top 20 and ended the race with a spetacular 1-stopper with 1+ min over the others by a lucky shot?
|
|
|
|
|
|
#24 dikirim Apr 11 2012, 12:52:45 (terakhir disunting Apr 11 2012, 12:53:12 oleh Georg Petrov)
|
Sebut
|
Tell me about how Sereno's race would have looked if it had stayed dry for only 2-3 more laps. ;)
(And care to tell me this was not possible at all, yesterday)
The OP already had his answer, the discussion has taken a new direction now. Surely not the first time something like this has been discussed, but who cares. I find people's views on the matter interesting.
|
|
|
|
|
An improved weather forecast or some more options for the managers to make their strategy more flexible would really be nice. There are just too many lottery races recently. In this season this one was the third IIRC. At this point the championship becomes less and less about strategy and more about pure luck, which is clearly not an ideal situation.
|
|
|
|
|
|
#26 dikirim Apr 11 2012, 14:03:52 (terakhir disunting Apr 11 2012, 14:04:29 oleh Jon Day)
|
Sebut
|
|
Quote ( Merse Kovácsházi @ April 11th 2012,13:58:48 )
At this point the championship becomes less and less about strategy and more about pure luck, which is clearly not an ideal situation. There are only 2 types of luck, good luck and bad luck so its 50/50. :D Leave the weather at it is, imo it makes the races all the more exciting :)
|
|
|
|
|
Quote ( Fernando Fernandes @ April 11th 2012,12:28:53 )
a little help Georg ;) this speaks in favour of admins imo.
Complex mathematical equations are used to predict the physics and movement of the atmosphere and to determine the rates of change in the atmosphere over time. The rates of change predict the state of the atmosphere a short time in the future. The equations are then applied to this next atmospheric state to find new rates of change, and these new rates of change predict the atmosphere at a yet further time into the future. This is called time stepping.
This time stepping is continually repeated until the desired forecast time in the future is reached. Computation time for a regional weather model is from a few seconds to a few minutes.
lol, no such thing is used in GPRO. GPRO's weather system is equivalent to casting a loaded dice every lap. And it doesn't speak in favour of admins...
|
|
|
|
|
Quote ( Georg Petrov @ April 11th 2012,12:02:24 )
1. Start on wets and hope for rain to come very fast (was the right strategy, no doubt)
2. Start on dries and change to wets as soon as weather changes (2nd best strategy in our case)
3. Start on dries and not pit for wets, due to the probability that rain might - if at all come for a brief period (absolute race killer yesterday)
Lets emphasise a bit more:
Strategy 1: What happens when it stays dry for more than 7 laps and my tyres are already dead when it starts raining?
Strategy 2: What happens if we have several weather changes and I find myself pitting every 3-4 laps?
Strategy 3: What happens if it rains early and stays wet for a very long time?
Now depending on how the weather gods feel like it, any and every single of these 3 strategies could have payed off big time or flopped big time. How does this fit in GPRO?
I think Georg sums it up very well here. Any one of those 3 strategies could have worked but with no idea as to when the rain would come and stop it was a impossible task guessing what the best strategy would be.
I went for 1 stopping on hards staying out when it rained but I guessed that the rain would come alot later then lap4 and would have stopped before 48mins in. Now I know everyone had to deal with the same choices and I got it wrong but its so frustrating that my race could have gone good or bad purely because I had to guess what the weather was going to do
|
|
|
|
|
Quote ( Vincent Wilkins @ April 11th 2012,09:49:56 )
Last Race the rain probability said....
Start - 0h30m
Rain Probability 60% - 70%
0h30m - 1h00m
Rain Probability 0%
So why the hell was it still raining round 50th min into the race?????
And yet you somehow managed to have 131 races of experience behind you..
|
|
|
|
I think what needs to be discussed is if the weather system adds to the game or makes it worse.
I do like the variation added by weather changes, but we have to admit it is almost completely random where the rest of the game favours precision.
You can collect data and make somewhat precise calculations for almost everything in this game but weather, that in my opinion is that this becomes quite a controversial subject, it almost goes against the "essence" of the game.
And I do understand the complains, I have seem races where it would start raining, rain for just a few laps and then stop.
With the limitations we have, orders entered before the race start, one can never optimally prepare for that, so a rain race becomes way more a lottery then the rest of the game is.
The problem lies in introducing a random component but not allowing one to react to it, just giving someone a best guess.
|
|
English
Deutsch
Español
Română
Italiano
Português
Português
Français
Polski
Русский
Български
Македонски
Српски
Nederlands
Vlaams
Suomi
Lietuvių
Čeština
Magyar
Eesti keel
Slovenčina
Türkçe
Shqip
Ελληνικά
Hrvatski
Dansk
Pirate
