Jump to content
  • Member Statistics

    17,508
    Total Members
    7,904
    Most Online
    Toothache
    Newest Member
    Toothache
    Joined

January 2013 Forecast Contest, Temperatures


Recommended Posts

A new year, with a fresh start for all participants. Pushed the deadline to the end of the 3rd due to the late thread start and the holiday weekend.

So who wants to take charge of it this year? I wouldn't mind doing it if I can get the files and calculations needed. Roger's kept up with things well, so maybe he can try it out. I love ya Mallow, but multiple months of backlogged verifications make me sad.

So let's throw in those guesses - I mean forecasts - for the new year!

January

DCA:

NYC:

BOS:

ALSO, add in forecasts for the following in case it's decided to add them into the contest (Chicago O'Hare, Hartsfield-Jackson Atlanta International and George Bush Intercontinental/Houston)

ORD:

ATL:

IAH:

Link to comment
Share on other sites

  • Replies 73
  • Created
  • Last Reply

That's a pretty interesting thought. Perhaps ORD for the Midwest, IAH for the Plains and ATL for the Southeast?

I believe these locations would attract more folks from those Regions to participate. We are American Weather after all...;)

Link to comment
Share on other sites

I will put in some entries on the 31st because I have been away from the weather for the past week.

On the subject of scoring and contest scope, I am willing to score the contests and that would more or less guarantee a real-time report as I have enough time to handle this. I had a look back to January as one example and my somewhat simpler scoring method gives almost the same numbers as the s.d. method as well as generating the same top ten with roughly the same differences from person to person. Therefore I would suggest using this simpler method because it takes less time and gives more or less the same results. FYI that method is simply to deduct 2 pts for every 0.1 F error and from the same base of 100 per site. The one variation that I tried was in March when the anomalies were so far from normal and scores seemed too low using the method, so I did something like deducting a smaller amount per error-degree. The s.d. method then generated roughly similar scores to what I had considered too low. Since the contest is really based on scoring differences anyway, it probably doesn't matter if we encounter one month with a lower range.

On the other topic, expanding the scope, that would be good and I like the three stations suggested. I would suggest keeping two sets of books, one to be called "Classic" reporting on the three traditional sites, and one to be called "Expanded" that reports on all six. I would not bother to keep a set of scores on just the new stations, but I figure some contestants from previous years might not feel motivated to enter the larger contest, so this would suit them too.

If we feel the situation is ambiguous about scoring, I would undertake to produce these simpler scores anyway and then if we have a set based on the Mallow method no problem, I would bet that the order of finish would be the same anyway. I think I understand the method we're using now and could produce scores that way, just needing to check one detail, the s.d. is derived from the spread of forecasts and not from the spread of actuals?

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.

×
×
  • Create New...