2014 NATL Tropical-Hurricane-Seasonal Forecast Contest

[Roger Smith]

(May 17th original thread starter)

 

Just a preliminary announcement that a 2014 North Atlantic tropical (hurricane) forecast contest will be available. If you'd like to organize it, let me know before May 25th. I don't mind at all handing this off to somebody else. But if nobody wishes to take it on, I will set it up in a similar way to last year. Hopefully we will have a better season than last year where Normal 1851-1880 was the unofficial winner (Normal 1993-2012 was the actual winner).

 

Also, it doesn't have to follow last year's template exactly, let me know if you think anything about that scoring system needs a tweak.

 

Watch for a further announcement of the contest rules and entry information.

 

 

 

Contest Rules

 

 

1. Post a forecast of seasonal totals, in the format ALL NAMED STORMS / HURRICANES / MAJOR H on or before June 1st (deadline June 2 06z). Late penalty will be 1% per day (06z to 06z) which means that if you enter the contest on July 1st you'll have a 30% penalty applied to this portion, etc. The seasonal forecast will be worth 50% of the total points.

 

2. Each month, in this thread, post your forecasts for the monthly totals using the same format. You can post all of these at the start of the season if you wish (perhaps a good idea if you're forecasting low seasonal totals) and starting with June, the deadline will be 06z of the first day of the month (except June will include a one-day extension so June 2 06z, July 1 06z, Aug 1 06z, Sep 1 06z, Oct 1 06z and Nov 1 06z will be the deadlines). Time penalties will be relaxed this year, 10% of your actual score per late day but calculated to nearest 1%. These will apply regardless of holidays or weekends intervening. Contestants can invoke an "assumed zero forecast" by claiming this right in any post at any time. It means that from that point on, the absence of any required monthly forecast is equivalent to a 0/0/0 forecast. The assumed zero is not an automatic condition, and in the absence of setting the condition, scores will be zero for non-entered months.

 

3. Months will be scored so that the totals of all forecasts add up to 50% of total points. June will have 4%, July 6%, August 12%, September 16%, October 10% and Nov-Dec 2%.

 

4. All scoring will be based on (Max Points) minus (Error deductions) where Error deductions will be calculated as the mean of absolute error plus square of absolute error. Example, seasonal, you predict 10/8/4 and actual is 12/7/3. Max points 50, minus (2+4)/2 or 3 for named storm error, [(1+1)/2] or 1 for hurricane error and same for major storm error, total error deductions 5, total score 45. For months, the same procedure is used. However, to protect scoring against anomalous outcomes, highest score must be at least 80% of max points, second highest 70%, third highest 60%, fourth highest 50% and from there down, intervals of 5%.

 

5. Post your thoughts about a maximum landfalling storm, but no bonus points will be awarded this year.

 

----------------------------

 

As stated in my post on May 21, contest entries made between this post and that clarification post can now be edited through to June 2 (06z) without the need to make any further posts; no forecasts will be gathered into tables or otherwise noted by the organizer until after June 2 so edit as often as you desire before then.

 

Good luck.

[PSUBlizzicane2007]

5/1/0

 

June:  1/0/0

 

Literally we could all probably just submit one set of numbers this year instead of monthly and get the same result in the end.

[Ed Lizard]

7/3/1

[Superstorm93]

10/3/1

 

Probably gonna go down...

[andyhb]

8/2/1.

 

Lol the Atlantic basin.

[Ground Scouring]

12 NS / 4 H / 2 MH

 

The season will overperform (in terms of total NS) relative to expectations and at least one major hurricane will hit the U.S. Otherwise the season will be a dud with no other U.S. landfalls except a TS.

 

Edit: I would like to specify that the major landfall in the U.S. will be at least 125 kt at LF.

[Stebo]

9/3/1

[Mike.Ventrice]

11/5/0

[IsentropicLift]

13/5/1

 

Going with 2 US landfalls, a category 2 into Florida and a tropical storm into North Carolina.

[metalicwx366]

12/3/1

[icebreaker5221]

9/3/1

[wxmx]

9/4/1

[SlantStickers Anonymous]

10/4/1

[Eyewall2005]

_10/5/2

[Srain]

10/4/2

[Harry]

11/5/2

[Roger Smith]

Okay then, a contest will go forward. I will edit the original post to include the rules, and those who have already posted take note, you can edit your submissions to meet contest rules without an editing note, as I will not be gathering the forecasts until the first few days of June.

 

Will go with the same basic format as previous years despite some consensus already evident that it will be a weak season, but will leave the landfall bonus question out of the points calculation entirely, just post your thoughts for the interest value alone.

 

Readers starting here who have not seen the thread before today, go to post #1 and then post your forecast as per the rules given there.

[biodhokie]

6/3/1

[WxFreak11]

8/4/1

[Amped]

9/1/1

[TropicalAnalystwx13]

10/2/1

[uncle W]

June...1/0/0

July....3/1/0

Aug....5/3/0

Sept...4/2/1

Oct.....1/1/0

Total. 14/7/1

[Jim Marusak]

total: 13 storms, 6 hurricanes, 1 major

month by month:

June: 1/0/0

july: 1/0/0

august: 3/1/0

September: 5/3/1

October: 2/2/0

November: 1/0/0

December: 0/0/0

 

maximum landfalling storm: between Tampa and Homosassa Springs, probably cat 2, after crossing florida and exiting between jacksonville and daytona re-strengthens in the gulf stream and zips up the coast, making a 2nd landfall as a transitioning cat-1 or cat-2 near Cape Breton, NS east of YQY. 

[isohume]

6-12/1-5/0-3

[hurricaneman]

9\3\1 seasonal

 

1\0\0 for June

1\1\0 for July

2\1\1 for August

3\1\0 for September

1\0\0 for October

0\0\0 for November

1\0\0 for December

 

Max Landfall: August 22nd 135mph Ft. Lauderdale, FL

[Ed Lizard]

6-12/1-5/0-3

 

You work for NOAA...

[isohume]

You work for NOAA...

 

Yeah, but that's CSU's outlook. 

[PSUBlizzicane2007]

I'm amazed at how high people are going for this year.

[Roger Smith]

Okay, just so that nobody's confused about the rules, you can enter monthly forecasts up to the first day of each month without penalty. If you've entered all months, you can repost updated forecasts later, or edit the original post with a note. I will make a habit of checking back to see if posts are edited.

 

Just realized that there's an implied shift in the rules from previous years, in that the rules did not state any opportunities to update seasonal forecasts after your first entry (which quite a few have already made). Thus, to amend that situation, you can update any time you want under the same rules as first entries, namely, a 1% penalty per day starting June 2nd. So if you change your mind let's say on July 21 and figure to score better at 50% of your forecast score than 100% of your original, make the amended forecast.

 

I will enter on the 31st, still looking at the variables.

[Stebo]

June 1/0/0