Azərbaycan dili Bahasa Indonesia Bosanski Català Čeština Dansk Deutsch Eesti English Español Français Galego Hrvatski Italiano Latviešu Lietuvių Magyar Malti Mакедонски Nederlands Norsk Polski Português Português BR Românã Slovenčina Srpski Suomi Svenska Tiếng Việt Türkçe Ελληνικά Български Русский Українська Հայերեն ქართული ენა 中文
Subpage under development, new version coming soon!

Subject: Weather...

2006-07-13 16:17:46
My matches, from most recent to oldest, inc friendlies

moderate clouds, cool
part cloudy, drizzle
moderate clouds, warm
part cloudy, drizzle
moderate clouds, warm
moderate clouds, cool
partly cloudy, drizzle
partly cloudy, drizzle
no clouds, warm
moderate clouds, cool
partly cloudy, drizzle
sparse clouds, warm
2006-07-13 16:39:20
Ok, Ok I've got enough now guys. Thanks very very much!
2006-07-13 16:39:37
Analysis will come over the next week or so. Lots on atm....
2006-07-17 21:58:44
OK, Here are some preliminary results on my weather Analysis.

1. Probability of types of Weather

Weather has two settings what I will call "cloud setting" and "weather" settings.

Each has the following settings (from best weather to worst)

Cloud Settings

No Clouds
Sparse Clouds
Moderate Clouds
Partly Cloudy
Heavy Clouds

Weather Settings

Hot
Warm
Cool
Drizzle
Shower
Heavy Shower.

Now here is the interesting bit - there is a relationship between the two. The cloud type seems to determine the weather setting. For example, you will only get a hot day when there are no clouds.

By doing a load of analysis of over 200 different weather events, I can give a pretty accurate basis for the probability of each type of weather. When I have some spare time I will increase those numbers, but I wouldn't expect the numbers to change substantially.

Cloud Settings

No Clouds - 20%
Sparse Clouds - 15%
Moderate Clouds - 35%
Partly Cloudy - 15%
Heavy Clouds - 15%

i.e. 20% of days appear to have No Clouds as the "cloud weather"

Weather Settings

Each "cloud setting" impacts the type of weather - as detailed below....

Days with No Clouds - 25% Hot, 75% Warm
Sparse Clouds - 99% Warm, 1% Cool (maybe 100% warm?)
Moderate Clouds - 50% Warm, 50% Cool
Partly Cloudy - 100% Drizzle
Heavy Clouds - 60% Shower, 40% Heavy Shower

Therefore, you've guess it - multiply the two together and here is the probability of each weather type.

Hot ~5%
Warm ~47%
Cool ~18%
Drizzle ~15%
Shower ~9%
Heavy Shower ~6%

2. Impact on attendances

This is the tougher part to analyse. There are many factors that impact the attendance at games - these include the mood of your fans, your opponents fans, the size of your fanclub and the opponents fan club, the presence of national team players, the type of game and the weather. Therefore to try and keep the analysis accurate I looked at games played by mcoi or forest reserves in the Premiership. Both teams have a similar number of fans, similar results and therefore fan mood. Both tend to play a NT player. Also, it can be argued that the Premiership has a pretty equal "quality" of teams making that area less. Games without full sell-outs were used (except for the one hot datapoint we had..)

This was the result...

There appears to be a substantial impact of the weather type. The correlation was stronger than cloud type & makes me pretty sure that cloud type impacts weather type which impacts the crowds:

Here were the average crowds at each type and the number of datapoints:

Hot: 28100 (1 - and sold out but 1st vs 2nd..)
Warm: 21918 (5)
Cool: 19241 (3)
Drizzle:14056 (4)
Shower: 16789 (2)
Heavy Shower: 12033 (2)

More information is needed here, but this suggests the following approximate, but substantial impact of the weather.

Hot: 100%
Warm: 90%
Cool:80%
Drizzle:70%
Showers:60%
Heavy Showers:50%

i.e. being unlucky with the weather can cost you substantially....

3. Stadium Design - mitigation

The final part is unanswered...

a) How much does a roof help avoid the drop in attendance?

b) Is it financially worth it.

This is work in progress. Any help in ascertaining this is much appreciated.


I hope that it is useful. It's been a while since I did any analysis (which oddly is what I was "famous" for back in 2004/05 for the oldies that remembered when we knew nothing!

forestman
(edited)
2006-07-17 22:02:35
4. Impact on drain on players' stamina

Where is this section, fm?



Also, do not forget that "data" is a plural word. More data ARE needed.
2006-07-17 22:05:53
Have you swallowed a dictionary this week?

Changed to information to annoy you.

Yes, it would be interesting to know how the different weather conditions impacts the game engine
2006-07-17 22:07:59
And it should be "knew" nothing, not "new" nothing. :P
2006-07-17 22:08:21
Nice work there FM.
2006-07-17 22:10:16
that was a fliddy typing whilst watching a TV program about russian two headed dogs on five...
2006-07-17 22:11:04
I assumed that...
2006-07-17 22:13:22
that...I was watching channel 5?
2006-07-17 22:15:01
Does it hang out with the 2 faced cat?
2006-07-17 22:26:44
nice piece of work done there.

hopefully ill have a bit of luck with weather next season
2006-07-17 23:39:36
God job mate. I'll give you a hand with the stadium calcs at some point if I have time.
2006-07-20 02:25:57
if anyones doing any calcs on attendance i got 2 exact same weather conds

in my league game 875915 i got 4464 with furios fans and in friendly 1200090 i got 3428 with believe in the club and supporters are only dif of 4.
(edited)
2006-07-20 09:13:17
When you hover over the links it has the id in the status bar.
When you open in a new window/tab it has the id in the address bar.