January 2018 temperature forecast contest -- "Regional Rumble" begins

[Roger Smith]

Thanks for your entries, we seem to have every regional forum represented. 

Late penalties are going to kick in now, usually they start on the first day of each month at 06z, but we relaxed the deadline considerably to get this Rumble started. For any new entries, let's say 2% per hour from now to 10z 6th. (added 6th 00z -- only one late forecast received, below this post, going to cut them a break at 10%, would advise any really late but interested parties, wait to February now, you're going to get dinged for 80% plus 2% per hour to extinction at 10z).

Forecasts are arranged from warmest DCA forecast to coldest, with NYC then BOS etc used to separate DCA (or NYC) ties.

This year, your regional home will be identified beside your forecaster name.

For the rumble, highest regional scores in each set of three locations will be counted. For the individual contest, it's your total score in various groups. 

 

Table of Forecasts January 2018

 

FORECASTER ____ Forum ___ DCA _ NYC _ BOS ___ ORD _ ATL _ IAH ___ DEN _ PHX _ SEA

 

H2OTown_wx ____ NE _____ +1.1 _+0.4 _--0.1 __ +2.0 _+3.5 _+3.2 __ --1.6 _+1.3 _--0.8

SD _____________ PHL ______0.0 __0.0 __0.0 ____0.0 _+0.5 _+0.5 ___+1.0 _+1.0 _+1.0

___Normal ______ -- -- ______ 0.0 __0.0 __0.0 ___ 0.0 __ 0.0 __ 0.0 ____0.0 __ 0.0 __ 0.0

Mercurial ________ C W ____ --0.5 _--0.9 _--1.5 __ --2.4 _--2.0 _--1.2 __ +2.5 _+3.2 _+1.7

wxdude64 _______ M A ____ --1.0 _--1.2 _--1.4 __ --2.0 _--0.6 _--0.2 ____0.0 _+0.5 _+0.8

nrg jeff __________ TNV ____--1.0 _--2.0 _--3.0 __ --4.0 _--1.0 __0.0 __ +1.0 _+4.0 _+1.0 

jaxjagman _______ TNV ____ --1.5 _--3.0 _--2.0 __ --3.3 _--2.2 _--2.1 __ +1.5 _+2.5 _+1.5

tnweathernut * ___ TNV ____ --1.6 _--2.8 _--3.4 __ --2.6 _--1.6 _--0.8 __ +1.4 _+3.0 _+0.8

BKViking ________ NYC ____ --1.8 _--1.5 _--2.0 __ --3.4 _--0.8 _--2.0 __ --2.5 _+2.0 _+0.9

Roger Smith ______C W ____--2.0 _--3.5 _--5.0 __ --3.5 _--4.0 _--1.0 __ +2.2 _+3.7 _+1.8

DonSutherland.1 __ NYC ____--2.1 _--2.6 _--2.9 __ --3.8 _--2.1 _--1.7 __ +2.4 _+2.8 _+2.5  

H2O ____________M A _____--2.2 _--2.8 _--2.5 __ --3.2 _--3.1 _--1.8 __ +1.5 _+2.0 _--0.5 

hudsonvalley21 ___ NYC ____--2.2 _--2.8 _--3.1 __ --2.3 _--2.4 _--2.2 __ --0.2 _+1.5 _+1.2

___Consensus ____ - - _____ --2.2 _--2.8 _--3.0 __ --3.2 _--2.4 _--1.5 ___ 0.0 _+2.0 _+0.6

wxallannj ________ NYC ____--2.3 _--2.9 _--3.5 __ --2.7 _--3.2 _--2.4 __ --1.2 _+2.2 _+0.6

dmillz25 _________ NYC ____--2.5 _--3.3 _--3.7 __ --3.0 _--2.7 _--2.0 __ --2.5 _+1.5 _--0.5

buckeyefan1 ______ SE _____--2.8 _--2.6 _--3.0 __ --3.3 _--2.5 _--1.3 __ +1.0 _+1.5 _+0.3

so_whats_happening _PA ___ --3.0 _--2.6 _--2.8 __ --3.7 _--2.4 _--1.6 __ +1.1 _+1.7 _--0.8

RJay ____________ NYC ____--3.0 _--3.0 _--3.0 __ --4.0 _--2.0 _--2.0 __ --1.0 _+2.0 _+1.0

mappy ___________ M A ___ --3.0 _--4.1 _--3.2 __ --3.0 _--4.2 _--3.0 __ --1.8 _+1.5 _--2.5

Stebo ___________ GL/OV __--3.3 _--3.8 _--3.5 __ --2.8 _--4.0 _--3.2 __ --2.0 _+1.0 _--2.0

RodneyS _________ M A ___ --3.5 _--4.3 _--3.9 __ --5.7 _--3.8 _--1.5 __ --1.4 _+3.9 _--0.2  

Tom ____________ PHL ____--3.8 _--3.4 _--3.3 __ --4.1 _--3.8 _--0.9 __ --1.0 _+0.1 _--1.1

___________________________________________________________________________________

later this month, a post will appear in the thread showing anomalies to date, and projections based on 7-day NWS forecasts and days 8 to 16 from GFS model runs. First instalment will be after 7 days, on 8th.

 

* late penalty of 10% will apply

Consensus is 11th ranked of the 21 forecasts (not incl Normal) -- median value

Warmest forecast (not incl Normal) 

Coldest forecast (not incl Normal) 

Good luck everyone and welcome to the contest. 

UPDATE _ Jan 10th _ Have cross-checked all entries above and entered into excel file. In doing this, I discovered one minor mistake that I made in the table, DonS had -2.1 for DCA and I entered -2.3 ... this changes his location in the table slightly, but does not change consensus as 11th ranked is still -2.2.

[tnweathernut]

Hardly knew how to find the "main" board.... lol 

DCA -1.6

NYC -2.8

BOS -3.4

ORD -2.6

ATL -1.6

IAH -0.8

DEN +1.4

PHX +3.0

SEA +0.8

TN VALLEY 

[Roger Smith]

Snowfall 2017-18 contest -- Table of Forecasts

 

This post will only be updated in the February thread, this was the situation at the end of January 2018. 

 

FORECASTER _________ DCA _ NYC _ BOS __ ORD _ DTW _ BUF ___ DEN _ SEA _ BTV 

 

so_whats_happening___ 22.0 _ 38.0 _ 63.0 __ 31.0 _ 36.0 __ 85.0 __ 53.0 _ 14.0 _ 96.0

Tom ________________ 19.6 _ 43.8 _63.1 __ 59.4 _ 53.4 _ 103.8 __ 71.5 __ 9.4 _109.3

Roger Smith __________19.5 _ 40.0 _ 60.0 __ 33.0 _ 45.0 __ 80.0 __ 55.0 _ 17.5 _110.0

dmillz25 _____________ 17.0 _ 41.0 _ 57.0 __ 67.0_ 70.0_ 105.0 __ 60.0 _ 15.0 _ 96.0

wxdude64 ____________16.3 _ 27.7 _ 37.5 __ 54.8 _ 55.5 _ 102.2 __ 77.6 _ 11.2 _ 99.5

hudsonvalley21 _______ 15.5 _ 42.0 _ 57.0 __ 47.0 _ 51.0 __ 91.0 __ 64.0 _ 10.0 _ 89.0

DonSutherland.1 ______ 15.1 _ 34.3 _ 53.4 __ 48.9 _ 54.2 _ 112.6 __ 55.8 __ 7.3 _ 96.7

H2OTown__wx ________14.4 _ 23.2 _ 41.2 __ 44.4 _ 38.7 __ 93.4 __ 40.6 __ 5.4 _ 67.1

 

___ Consensus _____14.4 _33.7 _53.4 __48.9 _51.0 __95.8 __55.8 _10.0 _96.0

 

Stebo _______________ 12.5 _ 33.7 _ 50.0 __ 50.0 _ 57.3 __ 65.0 __ 50.5 __ 7.0 _ 72.5

wxallannj ____________ 12.3 _ 27.0 _ 33.0 __ 38.0 _ 43.0 __ 53.0 __ 58.0 _ 19.0 _ 57.0

BKViking _____________11.0 _ 33.0 _ 70.0 __ 40.0 _ 50.0 _ 110.0 __ 77.0 _ 10.0 _110.0

RJay ________________ 10.0 _ 40.0 _ 70.0 __ 50.0 _ 55.0 _120.0 __80.0 __ 6.0 _ 105.0

SD __________________10.0 _ 23.0 _ 45.0 __ 55.0 _ 67.0 _ 110.0 __ 85.0 __4.0 _ 87.0

RodneyS ______________8.1 _ 25.0 _ 35.8 __ 52.4 _ 45.7 __ 95.8 __ 52.1 _ 12.5 _ 57.0

SnoSki14 _____________ 2.0 __ 5.0 _ 35.0 __ 40.0 _ 45.0 __ 80.0 __ 55.0 _ 30.0_ 75.0

 

Amounts to date ____ 3.1 _18.9 _27.0 __10.0 _ 31.7 _ 65.1__13.2 __3.0 _ 41.4

... snowfalls through Jan 31st updated daily (when snow falls).

_____________________________________________________________________________

15 entrants, the consensus is the median value (8th ranked).

Heaviest snowfall forecasts for each location are shown in bold.

Least snowfall forecasts are shown in italics.

Forecasts passed by totals to date are in red. 

[Roger Smith]

Here are the anomalies for the first week, and they are extreme to say the least ... 

The p14d anomalies are based on NWS 7-day forecasts and reflect what the anomalies will be after 14 days, if those forecasts verify. Midnight lows and such quirks are factored in where obvious. 

The p24d anomalies then take days 8 to 16 of the GFS and estimate where the anomalies will sit after 24 days. 

This table is updated whenever I feel that there's enough new information to make an informed change, and towards the end of the month, on a daily basis as a countdown, with provisional scoring provided so you can track how you're doing (or if you think my estimates are off, you can adjust that) but it all means nothing until the real numbers come in at the end of each month.

I imagine the next update to this table will be around the weekend for 12 days perhaps, then maybe every 3 days to 24th.

Now that have corrected the table for the estimated values posted earlier, I have slightly adjusted the p24d estimates as the 12z GFS looks slightly warmer than the 06z GFS which I used to derive the first set posted. Changes are all +0.5 positive. 

 

_______________ DCA _ NYC _ BOS ____ ORD _ ATL _ IAH ____ DEN _ PHX _ SEA

______ (7d) ____-17.0 _-16.6 _-16.0 ___  -17.4 _-14.8_ -10.7  __ +3.5 _+7.5 _+1.1 

___ (p 14d) _____-6.3 __-6.0 __-6.2 ____ --8.0 __-7.0 __--4.4 __ +6.0 _+5.5 _ +2.7

___ (p 24d) _____-3.5 __-3.7 __-3.7 ____ --4.5 __-3.5 __--2.5 __ +3.5 _+4.0 _ +2.0

_____________________________________________________________________________

Meanwhile, I calculated score differentials from consensus for all 21 forecasters. This is your average (absolute) difference per forecast from our consensus values. The number in brackets beside your name is your average departure from consensus (not absolute, shows warm or cold bias). How to interpret the combination is discussed after this table. 

0.33 __ Buckeyefan1 (-0.02) and hudsonvalley21 (0.0)

0.46 __ H2O (-0.01) and RJay (-0.28)

0.48 __ wxallannj (-0.32)

0.51 __ so_whats_happening (-0.18)

0.63 __ tnweathernut (+0.54) and jaxjagman (+0.43)

0.73 __ DonSutherland1 (+0.56) and dmillz25 (-0.69)

0.87 __ BKViking (+0.16)

1.01 __ NRGjeff (+0.83)

1.11 __ Tom (-0.98)

1.16 __ wxdude64 (+0.82) and Roger Smith (+0.13)

1.24 __ mappy (-1.2)

1.27 __ mercurial (+1.27)

1.30 __ RodneyS (-0.88)

1.32 __ Stebo (-1.23)

1.97 __ Normal (+1.39)

2.06 __ SD (+1.83)

3.21 __ H2OTown_wx (+2.39)

The average absolute departure from consensus reveals how far from the consensus your forecasts are, on average, and the average departure shows in what direction you stray. Those with large average absolute values are likely to show a directional bias, unless they tend to go to extremes either at random or in sync with the consensus values. It appears that this month, the outliers have decided to go warmer than consensus. For those with small average absolute departures from consensus, they are not likely to show a warm or cold bias (that value has to be equal to or smaller than the first value) but we can see examples of those who consistently go a little warmer or a little colder than consensus. Then there are a few (including me) who have a moderate tendency to stray from consensus and average out on either side with a relatively small "bias" tendency. 

The scores for Normal tell us that on average consensus is 1.97 degrees removed from 0.0, and that most of those departures are negative (because Normal has a warm bias of +1.39). If all the consensus values were negative (PHX and SEA are positive) then these two statistics would be equal. 

This is not any kind of test of quality of forecasting, scores will tell that story later. 

 

[Roger Smith]

Anomaly tracker updated ...

_______________ DCA _ NYC _ BOS ____ ORD _ ATL _ IAH ____ DEN _ PHX _ SEA

______ (7d) ____-17.0 _-16.6 _-16.0 ____  -17.4 _-14.8_ -10.7  __ +3.5 _+7.5 _+1.1 

_____ (14d) ____ --6.6 _ --6.1 _ --5.0 ____ --7.0 _ --6.9 _ --7.6 __ +5.3 _+6.8 _+3.1

_____ (21d) ____ --4.3 _ --3.8 _ --2.9 ____ --3.8 _ --6.2 _ --7.2 __ +3.8 _+5.4 _+3.8

_____ (24d) ____ --1.8 _ --2.0 _ --1.4 ____ --1.8 _ --4.7 _ --6.2 __ +3.0 _+4.6 _+3.5

_____ (28d) ____ --0.4 _ --0.8 _ --0.4 ____ +0.5 _ --3.2 _ --4.9 __ +2.9 _+4.3 _+5.1

 

(end of month projections are retained from each calculation, first number is date of prediction)

(have added the all nine scores that these predictions would have received, not much skill until 25th)

15_ (p 31d) _____--4.0 _ --3.5 _ --2.5 ____ --3.2 _ --2.0 _ --3.0 __ +2.5 _+4.8 _+1.7 ___ 550

22_ (p 31d) _____--2.4 _ --2.2 _ --1.0 ____ --0.7 _ --2.8 _ --3.5 __ +3.5 _+4.0 _+3.0 ___ 744

25_ (p 31d) _____--1.4 _ --1.2 _ --1.0 ____ +0.3 _ --3.3 _ --4.0 __ +3.5 _+4.5 _+3.0 ___ 818

29_ (p 31d) _____--0.9 _ --1.2 _ --0.5 ____ +0.8 _ --3.3 _ --4.0 __ +3.5 _+4.5 _+3.0 ___ 848

 

actual 31d ______ --0.3 _ --0.9 _ --0.4 ____+0.8 _ --2.9 _--4.5 __+3.9 _+4.8 _+3.0 ____ 900

 

_our Consensus _ --2.2  _--2.8 _--3.0 _____ --3.2  _--2.4 _ --1.5 ___ 0.0 _+2.0 _+0.6 ___ 440

Scoring is now fully updated. 

[Roger Smith]

Final scoring for January 2018

(part 1, forecasters) (regional rumble scored in part 2)

___ scoring is based on confirmed end of month anomalies in previous post.  This year the east plus central totals will only generate a scoring table that will appear in the annual update starting after February's contest, and scoring of each month will be based on "all nine" totals. This change will result in one scoring table for each month. The annual update tables will look the same as last year though, and for January the e/c totals will appear after the central sub-totals. Ranks can be ascertained by checking that column although top score will be highlighted. Scoring is now rank ordered (by all nine totals).

 

FORECASTER __ Forum __ DCA_NYC_BOS ______ ORD_ATL_IAH ___________ DEN_PHX_SEA

Final January anomalies __--0.3_--0.9_--0.4 ____+0.8_--2.9_--4.5___________+3.9_+4.8_+3.0

_____________________________________east________________cent _(E+C) ____________ (west) ___ all nine 

 

Mercurial ________ C W __96 _100_ 78___274__ 36 _ 82 _ 34 ___152 __426___ 72_ 68 _ 74 __ 214 ___ 640

DonSutherland.1 __ NYC __64 _ 66 _ 50 ___180___08 _ 84 _ 44 ___136 __316___ 70 _ 60 _ 90 __ 220 ___ 536

jaxjagman _______ TNV __76 _ 58 _ 68 ___202___18 _ 86 _ 52 ___156 __358___ 52 _ 54 _ 70 __ 176 ___ 534

Scotty Lightning (SD)_PHL_94 _ 82_ 92 ___268___84 _ 32 _ 00 ___116 __384___ 42 _ 24 _ 60 __ 126 ___ 510

nrg jeff __________ TNV__86 _ 78 _ 48 ___212___04 _ 62 _ 10 ___076 __288___ 42 _ 84 _ 60 __ 186 ___ 474

H2O ____________M A __ 62 _ 62 _ 58 ___182___20 _ 96 _ 46 ___162 __344___ 52 _ 44 _ 30 __ 126 ___ 470

 

___Normal ______ -- -- ___94 _ 82 _ 92 ___268___84 _ 42 _ 10 ___136 __404___ 22 _ 04 _ 40 __ 066 ___ 470

 

hudsonvalley21 ___ NYC __62 _ 62 _ 46 ___170___38 _ 90 _ 54 ___182 __352___ 18 _ 34 _ 64 __ 116 ___ 468

wxdude64 _______ M A __ 86 _ 94 _ 80 ___260___44 _ 54 _ 14 ___112 __374___ 22 _ 14 _ 56 __ 092 ___ 464

Roger Smith ______C W __66 _ 48 _ 08 ___122___14 _ 78 _ 30 ___122 __244___ 66 _ 78 _ 76 __ 220 ___ 464

BKViking ________ NYC __ 70 _ 88 _ 68 ___226___16 _ 58 _ 50 ___124 __350___ 00 _ 44 _ 58 __ 102 ___ 452

wxallannj ________ NYC __ 60 _ 60 _ 38 ___158___30 _94 _ 58 ___182 __340___ 00 _ 48 _ 52 __ 100 ___ 440

 

___Consensus ____ - - ___ 62 _ 62 _ 48 ___172___20 _ 90 _ 40 ___150 __322___ 22 _ 44 _ 52 __ 118 ___ 440

 

buckeyefan1 ______ SE ___ 50 _ 66 _ 48 ___164___18 _ 92 _ 36 ___146 __310___ 42 _ 34 _ 46 __ 122 ___ 432

tnweathernut _____ TNV __74 _ 62 _ 40 ___176___32 _ 74 _ 26 ___132 __308___ 50 _ 64 _ 56 __ 170 _478

_________ (-10%) _______67 _ 56 _ 36 ___159___29 _ 67 _ 23 ___119 __278___ 45 _ 58 _ 50 __ 153 ___ 431

so_whats_happening _PA __ 46 _ 66 _ 52 ___164___10 _ 90 _ 42 ___142 __306___ 44 _ 38 _ 24 __ 106 ___ 412

RJay ____________ NYC ___46 _ 58 _ 48 ___152___04 _ 82 _ 50 ___136 __288___ 02 _ 44 _ 60 __ 106 ___ 394

dmillz25 _________ NYC ___56 _ 52 _ 34 ___142___24 _ 96 _ 50 ___168 __314___ 00 _ 34 _ 30 __ 064 ___ 376

H2OTown_wx ____ NE ____ 72 _ 74 _ 94 ___240___76 _ 00 _ 00 ___076 __316___ 00 _ 30 _ 24 __ 054 ___ 370

RodneyS _________ M A ___36 _ 32 _ 30 ___098___00 _ 82 _ 40 ___122 __220___ 00 _ 82 _ 36 __ 118 ___ 338  

mappy ___________ M A __ 46 _ 36 _ 44 ___126___24 _ 74 _ 70 ___168 __294___ 00 _ 34 _ 00 __ 034 ___ 328

Stebo ___________GL/OV__ 40 _ 42 _ 38 ___120___28 _ 78 _ 74 __180 __300___ 00 _ 24 _ 00 __ 024 ___ 324

Tom ___________ PHL ____ 30 _ 50 _ 42 ___122___02 _ 82 _ 28 ___112 __234___ 02 _ 06 _ 18 __ 026 ___ 260

______________________________________________________________________________

Regional Rumble -- provisional scoring for January 2018

Scoring is based on high scores in the three groups (east, central, west). If you have one entrant then all their scores are used. If you have more than one, highest scores for each group are used.

For those forums with more than one entrant, the number is shown after the forum initials.

 

Regional forum _______ # __ High score east __ High score central __ High score west ___ TOTAL

 

Central/western (C W) __ 2 _____ 274 ____________152 ____________ 220 ___________ 646

New York City (NYC) ___ 6 _____ 226 ____________182 ____________ 220 ___________ 628

Mid-Atlantic (M A) _____ 4 ______260 ____________168 ____________ 126 ___________ 554

TN valley (TNV) _______ 3 ______212 ___________ 156 ____________ 186 ___________ 554

Philadelphia (PHL) _____ 2 _____ 268 ____________116 ____________ 126 ___________ 510

___Normal __________________ 268 ____________136 _____________ 066 __________ 470

___Consensus ________________172 ____________150 _____________118 ___________ 440

Southeast (SE) _______________ 164 ____________146 ____________ 122 ___________ 432

PA / upstate NY (PA/NY) _______ 164 ____________ 142 _____________106 __________ 412

New England (NE) ____________ 240 ____________076 _____________ 054 __________ 370

Great Lakes Ohio valley (GLOV)__ 120 ____________180 _____________ 024 __________ 324

 

________________________________________________________________________________

 

Thank goodness I have Mercurial covering my back ... looks like we're all going to have our hands full with this NYC crew. 

Kind of tough sledding for the lone wolves in this contest, get yourselves some helpers in your regional banter ... I will be posting invitations for February forecasts on the 25th. 

 

[Roger Smith]

January scoring is final (previous post) and Mercurial is the champion, anchoring Central / Western to first place in the regional rumble, despite a strong performance by the deep NYC crew. I was able to boost Mercurial's individual score in the rumble by tying DonS for high score in the western portion. Third place in the rumble is a tie between Mid Atlantic and Tennessee Valley, with Philadelphia taking fifth place (on SD's score, SD now being Scotty Lightning). 

 

Extreme Forecasts

 

Here's a summary of extreme forecasts that succeeded in January.

(An extreme forecast must be either the most extreme or second most extreme capturing high score to qualify ... and in our scoring table, a loss is given only if the extreme is not high score but the second most extreme is ... so quite often, a loss in this table is not a bad outcome for the forecaster, depending on how wide of the mark they were. In 2015 to 2017, tracking extreme forecasts, we found that more than half to two-thirds of the months qualified, showing our generally conservative forecasting approach).

DCA just missed qualifying (for Scotty Lightning and Normal to share a win, H2OTown_wx a loss), as Mercurial had high score with third coldest forecast.

NYC had the same outcome although here Mercurial and several others had higher scores than the extreme entries, but there were only two plus Normal who had a warmer forecast than the outcome.

BOS (--0.4) did qualify with H2OTown_wx (-0.1) getting a win, Scotty Lightning and Normal (0.0) sharing a loss.

ORD (+0.8) saw Scotty Lightning and Normal (0.0) sharing a win, H2OTown_wx the loss at +2.0 (although a great recovery from where ORD stood at mid-month).

ATL (-2.9) failed to qualify, it was fairly close to our consensus and seven of 21 forecasts were colder.

IAH (-4.5) gave Stebo (-3.2) a win. 

DEN (+3.9) produced a win for Mercurial (+2.5).

PHX (+4.8) was a winner for NRG Jeff (+4.0).

SEA (+3.0) was also warmer than all forecasts and gave DonSutherland.1 (+_2.5) a win. 

 

Extreme Forecasts wins and losses

 

DonSutherland1 ________1-0

Mercurial _____________ 1-0

NRG Jeff ______________1-0

Stebo ________________ 1-0

H2OTown_wx _________ 1-1

Scotty Lightning (SD)____1-1

Normal _______________1-1