January 2022 temperature forecast contest

[Roger Smith]

Merry Christmas and Happy New Year to all. Will start this thread and see if we have a quorum for a contest in 2022. No late penalties until perhaps the 3rd of January, encourage your friends and even perfect strangers to join, I need people I can beat. Besides Normal. 

As always, predict the temperature anomalies relative to 1991-2020 in F deg, for these nine locations:

DCA _ NYC _ BOS __ ORD _ ATL _ IAH ___ DEN _ PHX _ SEA

and let me know if you want to see any changes in the contest, either on the thread or by PM. 

[Scotty Lightning]

DCA _ NYC _ BOS __ ORD _ ATL _ IAH ___ DEN _ PHX _ SEA

1.0       1.0       0.5       0.5      1.5     1.5        0.5      0.5     -0.5 

[StormchaserChuck!]

DCA _ NYC _ BOS __ ORD _ ATL _ IAH ___ DEN _ PHX _ SEA

-2.5       -3.0       -3.5       -2.5      -1.0     0        -0.5      2.5     2.0

[so_whats_happening]

DCA -0.5

NYC -0.6

BOS-0.9

ORD -3.1

ATL 0.6

IAH 1.2

DEN -1.4

PHX 1.4

SEA -1.8

 

[Rjay]

DCA +1.0

NYC +1.0

BOS +1.0

ORD -1.5

ATL +1.0

IAH  +3.0

DEN  0.0

PHX  +2.5

SEA  0.0

[Tom]

DCA +0.1

NYC +0.3

BOS +0.4

ORD -1.9

ATL -1.1

IAH -0.9

DEN -1.1

PHX +1.1

SEA +0.9

[donsutherland1]

DCA _ NYC _ BOS __ ORD _ ATL _ IAH ___ DEN _ PHX _ SEA

0.6       0.4     -0.1       -2.6   3.0     2.5        -0.5    -0.3    -5.0

[Roger Smith]

+0.3 _ +0.1 _ +0.5 __ --1.0 _ +1.2 _ +2.5 ___ +1.5 _ +2.4 _ -0.7

[wxdude64]

DCA:  -0.8

NYC:  -1.3

BOS:  -2.1

ORD:  -2.7

ATL  -0.3

IAH:  +0.8

DEN:  +0.4

PHX:  +1.7

SEA:  -1.2

[bkviking]

DCA +1.1

NYC +0.9

BOS +0.8

ORD -1.6

ATL +1.2

IAH  +2.8

DEN  -0.7

PHX  +1.5

SEA  -0.3

[hudsonvalley21]

DCA +0.3

NYC +0.1

BOS +0.0

ORD -2.8

ATL +1.5

IAH  +2.1

DEN  -0.9

PHX  +0.5

SEA  -2.4

[wxallannj]

DCA -0.6

NYC -0.7

BOS -0.8

ORD -2.2

ATL -1.0

IAH  +0.6

DEN  -0.9

PHX  +1.5

SEA  -2.2

[RodneyS]

DCA _ NYC _ BOS __ ORD _ ATL _ IAH ___ DEN _ PHX _ SEA

 1.0      1.5       1.4      -0.5     1.5      2.0       -0.1     0.1      -2.1

[Roger Smith]

Thanks for these entries, I will check my e-mail for one more possibly ... and please note, as is our tradition in January, relaxed late penalty situation, anyone may adjust their numbers up until end of the 3rd if they wish, the table of entries will appear early on the 4th. No late penalties will be assessed so I wanted each on-time entrant to have the same opportunity to see and possibly utilize any last minute guidance. Currently nearly 00z Jan 3 so you have all day in z time on Monday to make adjustments.

And if you haven't been in to check, all the scoring is finished over in the Dec 2021 thread. Don Sutherland1 held on to his lead despite a valiant late charge up the tables from RodneyS which if memory serves began around May or June from a position well back of the leaders at that point. 

Just a reminder of how the forecasts are scored ...

Most months will be scored from 100 minus 2 points per 0.1 F deg error. 

Months that exceed a 5F anomaly (warm or cold) are scored with the excess portion counting only 1 point per 0.1 F deg error. That reduction applies both at the top end (5.0 to the value) and at the lower end (0.0 to corresponding equal interval). An example from last month, Dec 5.9 at DCA, your forecast would have scored 9 points to 0.9 then an additional 2 points per 0.1 F above 0.9, so if you predicted 2.5 then your raw score was 41. (see down for raw score adjustments)

Months that exceed a 10F anomaly (warm or cold) are scored by taking your error and dividing it by the anomaly, times 10, so for example last month IAH was +12.4, a forecast of +3.0 would score (30/12.4) x 10 which is 24. 

If no raw score is 60 or higher, then forecasts are scored from rank order in equal increments depending on number of forecasts. With 12 forecasts I have been using 0.5 as the increment which means even the worst forecast got five points. Ties are given highest scores in the sequence but those behind tied forecasts drop down all the intervals used up. Anyone whose raw score would exceed the awarded adjusted score retains their raw score. However, I plan to adjust this in 2022 to the following system. Instead of going with intervals, I will score the forecasts from a statistical basis from 60 down to the lowest required interval. Then for example, if I had to boost scores from the 35 level to 60 and those from 30 down were okay as raw scores, the adjusted scores would be on a differential basis in the boosted interval, and could be either closer together or further apart than we've seen in the direct application of equal intervals. As an example, if I had to boost 3.0, 2.9, 2.8, 2.6, 2.3 and 1.9, with 1.7 the next highest forecast having a raw score of 32 for an outcome of +5.2, above the progression value of 30, then I would take (60-32) = 28 as the total range of adjusted scores, then 60 to xx the specific adjusted scores so that xx was at a value derived from the differential of 3.0 and 1.7 but using 30 for the base score (so it would be 0.2 of 1.3 of the 30 point interval which is still five rounded off). The adjusted scores for those that needed boosting would be 60, 57, 55, 50, 45, 35 which seems like a fairer outcome than the previous 60, 55, 50, 45, 40, 35 for them. It may not make a lot of difference in many cases but it will eliminate the odd scoring jogs between similar forecasts when raw scores don't work.

I suppose the other approach would be to eliminate boosted scoring and just go with raw scores, if they are all low, too bad. But this makes each month more equally weighted and using golf as an example, it's worth the same to win a tournament at four over as it is at eighteen under par. 

A significant advantage can be gained by a forecast that eliminates the use of boosted scoring. If anyone has a raw score over 60, then all scores are raw scores. RodneyS had that going for DEN last month where his forecast was quite a bit closer to the large anomaly of +7.2 (at +5.6) than most of ours were, and he gained the full advantage as a result. 

[Roger Smith]

Table of forecasts for January 2022

 

FORECASTER ____________ DCA _NYC _BOS __ ORD _ATL _IAH ___DEN _PHX _SEA

BKViking _________________+1.1 _+0.9 _+0.8 __ -1.6 _+1.2 _+2.8 __ -0.7 _ +1.5 _ -0.3

RodneyS _________________+1.0 _+1.5 _+1.4 __ -0.5 _+1.5 _+2.0 __ -0.1 _ +0.1 _ -2.1

RJay _____________________+1.0 _+1.0 _+1.0 __ -1.5 _+1.0 _+3.0 ___ 0.0 _+2.5 __ 0.0

Scotty Lightning _________ +1.0 _+1.0 _+0.5 __ +0.5 _+1.5 _+1.5 __ +0.5 _+0.5 _-0.5

DonSutherland1 __________ +0.6 _+0.4 _-0.1 __ -2.6 _+3.0 _+2.5 __ -0.5 _-0.3 _-5.0

Roger Smith _____________ +0.3 _+0.1 _+0.5 __ -1.0 _ +1.2 _+2.5 __ +1.5 _+2.4 _-0.7

____ Consensus __________ +0.3 _+0.2 _+0.2 __ -2.2 _ +1.1 _+1.8 __ -0.5 _+1.5 _-0.9

hudsonvalley21 __________ +0.3 _+0.1 __ 0.0 __ -2.8 _ +1.5 _+2.1 __ -0.9 _+0.5 _-2.4

Tom ______________________+0.1 _+0.3 _+0.4 __ -1.9 _ -1.1 _ -0.9 ___ -1.1 _ +1.1 _ +0.9

_____ Normal ______________ 0.0 __ 0.0 __ 0.0 ___0.0 __ 0.0 __ 0.0 ___0.0 __ 0.0 __ 0.0

so_whats_happening _____ -0.5 _-0.6 _-0.9 __ -3.1 _ +0.6 _ +1.2 __ -1.4 _+1.4 _ -1.8

wxallannj _________________ -0.6 _-0.7 _-0.8 __ -2.2 _ -1.0 _ +0.6 __ -0.9 _+1.5 _-2.2

wxdude64 ________________ -0.8 _-1.3 _-2.1 ___-2.7 _ -0.3 _+0.8 __ +0.4 _+1.7 _ -1.2

Stormchaser Chuck _______-2.5 _-3.0 _-3.5 __ -2.5 _ -1.0 __ 0.0 ___ -0.5 _+2.5 _+2.0

_____________________________________________

12 forecasts, consensus is the median or average of 6th and 7th ranked forecasts.

DZ who joined us by e-mail in 2021 not entering this month. 

[Roger Smith]

Updated winter snowfall totals migrating to the Feb thread (Feb 11th) ( as of Feb 5th)...

DCA _12.3"

NYC _15.5"

BOS _ 37.4"

ORD _ 21.9"

DTW _ 28.6"

BUF _ 73.5"

DEN _ 21.7"

SEA _ 9.2"

BTV _ 46.3"

____________________________

Will be updating this list whenever new snow occurs, no new posts will appear. 

[Roger Smith]

Anomalies and projections ...

________________________ DCA _ NYC _ BOS __ ORD _ ATL _ IAH ___ DEN _ PHX _ SEA

____ (10d anom) _______+0.5 _ +1.0 _ +1.9 ___ -7.1 __+4.3 _ +3.0 __ -5.5 _ -1.1 __ -3.6

____ (20d anom) _______-0.6 _ -0.8 _ -0.4 ___ -3.1 __+1.6 _ +1.4 __ +0.2 _+1.0 __ -0.7

 

11th__ (p20d anom) ___-3.0 _ -2.5 _ -2.0 ___ -8.0 __ 0.0 _ +2.5 __ +1.5 _ +2.0 __ 0.0

11th__ (p27d anom) ___-4.0 _ -4.0 _ -4.0 ____ -9.0 _ -2.0 _ 0.0 ___ +2.0 _ +2.5 _ +1.0

21st __ (p31d anom) __ -2.0 _ -2.0 _ -2.0 ____ -4.0 __ 0.0 _ 0.0 ___ -1.0 _ 0.0 __ +0.5

30th __ (p31d anom) __ -2.5 _ -3.0 _ -2.5 ____-5.0 _ -0.5 _-1.0 ___ -1.0 _ +1.0 _ -1.5

 

end of month anom ___ -2.9 _ -3.4 _ -2.5 ____ -4.6 _ -0.3 _ -1.0 ___ -0.8 _ +0.9 _ -1.8

 

(see previous post for updated snowfall totals)

11th _ The month started very mild in the east which is why the anomalies are still holding on to small positive values after ten days but basically the pattern changed around the 3rd and has remained rather cold since then, and is about to get even colder, with the last third of the month looking even colder. That cold air may eventually spill down to the Gulf coast and as we learned last February, a few days in the deep freeze at Houston can wipe out a significant positive anomaly very quickly, so I have tentatively dropped the projection to zero by the 27th but if that cold air fails to reach IAH then it could be closer to +3, if the outbreak is severe for 4-6 days than it could be -3.0 very quickly. ORD never gets much of a break from the cold but SEA is quickly coming out of the deep freeze and should eventually get to positive numbers. DEN had some deep cold for a few days but has flipped to very mild and could push past the conservative +2.0 that I show here. Nothing very simple about this pattern, but historic cold appears to be heading for the east after the end of this week which will produce a couple of near normal days before the cold arrives. 

21st _ The colder trend was weaker than expected for the east but the milder trend was about as predicted in the west. This has left most locations with small negative anomalies to the 20th. Another cold spell looms for the east so the projections there go back below normal. DEN looks to be in a variable regime with some quite cold days keeping the average close to normal there. SEA will continue a slow rise towards an outcome just above normal, possibly. 

Snowfall amounts to date were updated in the previous post. Only BUF has had a heavy snowfall so far and SEA has not added to its impressive earlier totals. 

30th _ Snowfall totals updated after the big storm yesterday (in the northeast). DEN and ORD have been adding some significant amounts too. Also, a new set of provisionals is available, so that scoring below can be updated. These new provisionals will be converted into the final values by mid-day Tuesday 1st. 

31st _ Preliminary final anomalies have been posted from known data including the 31st, these will be adjusted when finalized, and then scoring will be adjusted (so I don't do the same thing twice, the scoring you see as of now is based on the earlier provisionals -- these new ones may change the scoring order a bit here and there). 

1st __ Final anomalies posted, scoring is final. 

[StormchaserChuck!]

Wish I would have went colder in Chicago, NAO was the wild card.. it had been negative against trend-analogs going into this 4/6 months 

-EPO really hit it too. In October I said +EPO Winter's averaged 7 named storms that point on, and we had 0 then after 19 early.

[Roger Smith]

The cold looks less extreme on more recent model runs than when I posted, so maybe the values shown will be too cold now. Reality could be closer to -2 in the east and -5 for Chicago if the cold is just mundane. We have to get the eastern locations below normal first, then see how sustained the cold trend really proves to be. Your numbers look pretty good in general, especially in terms of relative scoring. (that is by the way against the law in 37 of the 50 states and most provinces of Canada)

[Roger Smith]

Final scoring for January 2022

Scoring in this table is based on final anomalies as shown.  

 

FORECASTER ____________DCA_NYC_BOS__east__ORD_ATL_IAH__cent_ c/e__DEN_PHX_SEA__west___ TOTAL

 

Confirmed anomalies ___ -2.9 _-3.4 _-2.5 _______-4.6 _-0.3 _-1.0 ___________ -0.8 _+0.9 _-1.8

 

wxdude64 _______________ 58 _ 58 _ 92 __ 208 __ 62 _100_ 64__ 226 __ 434 __ 76 _ 84 _ 88 __ 248 ____ 682

Stormchaser Chuck ______ 92 _ 92 _80 __264 __ 58 _ 86 _ 80 __ 224 __ 488 __ 94 _ 68 _ 24 __ 186 ____ 674

wxallannj _________________54 _ 46 _ 66 __ 166 __ 52 _ 86 _ 68 __ 206 __ 372 __ 98 _ 88 _ 92 __ 278 ____ 650

so_whats_happening _____52 _ 44 _ 68 __ 164 __ 70 _ 82 _ 56 __ 208 __ 372 __ 88 _ 90 _100 __ 278 ____ 650

Tom ______________________40 _ 26 _ 42 __ 108 __ 46 _ 84 _ 98 __ 228 __ 336 __ 94 _ 96 _ 46 __ 236 ____ 572

hudsonvalley21 __________ 36 _ 30 _ 50 __ 116 ___64 _ 64 _ 38 __ 166 ___282 __ 98 _ 92 _ 88 __ 278 ____ 560

 

____ Consensus __________ 36 _ 28 _ 46 __ 110 ___52 _ 72 _ 44 __ 168 __ 278 __ 94 _ 88 _ 82 __ 264 ____ 542

_____ Normal _____________ 42 _ 32 _ 50 __ 124 __ 08 _ 94 _  80 __ 182 __ 306 __ 84 _ 82 _ 64 __ 230 ____ 536

 

BKViking _________________ 20 _ 14 _ 34 __ 068 __ 40 _ 70 _ 24 __ 134 __202 __ 98 _ 88 _ 70 __ 256 ____ 458

Roger Smith ______________36 _ 30 _ 40 __ 106 __ 28 _ 70 _ 30 __ 128 __ 234 __ 54 _ 70 _ 78 __ 202 ____ 436

DonSutherland1 __________ 30 _ 24 _ 52 __ 106 __ 60 _ 34 _ 30 __ 124 __ 230 __ 94 _ 76 _ 36 __ 206 ____ 436

RodneyS _________________ 22 _ 02 _ 22 __ 046 __ 18 _ 64 _ 40 __ 122 ___168 __ 86 _ 84 _ 94 __ 264 ____ 432

Scotty Lightning _________ 22 _ 12 _ 40 __ 074 __ 00 _ 64 _ 50 __ 114 ___ 188 __ 74 _ 92 _ 74 ___240 ____ 428

RJay _____________________ 22 _ 12 _ 30 __ 064 __ 38 _ 74 _ 20 __ 132 ___ 196 __ 84 _ 68 _ 64 __ 216 ____ 412

===========================================

EXTREME FORECAST REPORT

_ DCA (-2.9) and NYC (-3.4)  are wins for Stormchaser Chuck with lowest forecasts (-2.5, -3.0). 

_ BOS (-2.5) has ended up with a win for wxdude64 (-2.1) and a loss for Stormchaser Chuck (-3.5).

_ ORD (-4.6) is confirmed to be a win for so_whats_happening (-3.1) with low forecast. 

_ ATL (-0.3) does not qualify as fourth coldest forecast has highest score. 

_ IAH (-1.0) is a win for Tom with coldest forecast (-0.9). 

_ DEN (-0.8) did not qualify with third and fourth coldest forecasts tied for highest scores. 

_ PHX (+0.9) did not qualify with fifth coldest forecast the highest score. 

_ SEA (-1.8) did not qualify for an extreme forecast as fourth coldest forecast is the highest score.  

Scoring summary after one month __

Stormchaser Chuck___ 2-1 

so_whats_happening__ 1-0

Tom __________________ 1-0

wxdude64 ____________ 1-0

Five locations qualified for coldest and none for warmest. 

 

Best forecasts will appear in tabulated format next month

You can check them in the scoring table (bold highlighted). Congrats to wxdude64 with top score for January. 

 

[StormchaserChuck!]

Nice!