2017/18 Lakes/OV winter snowfall contest

[Roger Smith]

Do you mean the forecasters' relative to actual snow? 

In most cases, snowfall has not yet reached anyone's forecast, even the minimum.

The only real exception is PAH where it has passed most (all but four) of the forecasts quite rapidly. At both CMH and PIA snowfall has passed one forecaster (DMC76) and CMH has also passed cmillzz (code CM).

If you want to see a summary of all forecasts, I printed out the excel file and attached that to the December version of the table (at the moment we are updating the January table). 

But I have posted something below that will show the total departure from actual snowfall at this point, summed for each location. The table is almost the same as the order of total snowfall at this point.

[Roger Smith]

Forecaster errors (departures) to date

This table has been expanded and moved to the end of the thread (as of March 7th).

It will appear after the March (and April-May) update thread and was posted on March 7th. 

[Roger Smith]

A previous post shows the forecasts and snowfall amounts through Dec 31, 2017.

This post was moved on Jan 19th to end of the thread.

Added Jan 31, 2018 ... as this table is updated Feb 1st and 2nd to reflect end of January totals, a new table will be added to track February snowfalls. The red numbers show the seasonal totals. The orange numbers near the left end of each row show the January snowfalls and have no other contest-relevant significance.

This new post will remain active for January 2018. Snowfalls to date in red, 1986-2015 averages in green, Contest averages in blue, and the various forecasts showing our minimum and our maximum, with the others roughly to scale. 

Newly added, snowfall forecasts to go with the graphical presentation below. Many are in whole inches, have left out the .0 to save space, but decimal forecasts are shown in full. 

Just for orientation, note, green column of numbers in center of graph show 1986-2015 contest "normal" values. The contest average forecast is shown in blue. It could be on either side of the normal column but is usually close to it. The red numbers show snowfalls to date, and these change (numerically and in graph position) as they increase. Then all the fifteen forecasts are shown as close to a proportionate distance below to above normal, as space allows. I have taken out the notations of MIN (-50%) and MAX (+50%) that were in the header previously as they are not very helpful. But now you have not only a graphical overview of which forecasts are close or not close to actuals or expected final amounts, but the actual numbers too. The forecast numbers are below the forecaster codes, except for minimum and maximum forecasts which show up in the same row as the codes and the contest values. Forecasts that are equal in value are found whenever two or three forecaster codes have commas to separate them, for example, SP, CI and VP all have 39" for ORD.

 

LOC _ Jan totals __ season to date_MIN______________ 8615 avg  __ ConA  _______________________MAX_______________

 

APN  __14.9 _ 29.1______ 72 RS_DM _VP _CI _CM _ST SP 80.4 TH MA 85.3 _HA_ MI__ DA___ IT__  IW,CY 105

_________________________72.6_73_74_75_77_80___ 81.1_85 ___ 90_92.1_97.6_100 

ORD _ 4.6 ___10.0________ 27.6 DM_ CM,RS_TH__HA ST 38.1 SP,CI,VP 40.0 CY _ MA_ MI__ DA ___ IW __  IT 65

____________________________ 30 __30.5_35_37 ___ 39 _______ 41 _44 _44.6_44.9__54    

CLE 

_12.0_ 30.0_45.3 DM _CI ST CM TH_IT VP RS59.2HA MI 67.2 CY SP,IW MA ___  DA 89.7

_________________46 47 48 48.5 50 53 55__60 64.4___69_70 _ 72 

CMH  10.5_14.7 DM_CM ST18.6 RS,MA_TH_25.1 CI MI VP28.4SP,IT,HA DA CY IW 34

________________15_18 ___ 20 __ 22 ____26 27 28 ____ 30 _ 30.1_31 

DTW _ 9.2 ______________30.5 DM_31.7 VP RS,SP CM 44.8 CI,HA MA CY TH 49.3 _IW _ST MI __ DA ______ IT 83

_______________________________38 _ 40 _43____ 45 __46 48 49.1___52 _55_55.4_69.2   

FWA __ 7.2______ 18.6 ____ 22.2 TH __CM RS DM MA 34.0 HA SP 36.9 DA MI VP ST,CI,IW CY_ IT 47

______________________________ 28_30_32.2_33___35_36___ 38_39_41__42__ 45  

GRR _ 12.6 _________46.1 __57.8 DM__ST _TH_ CM CI  77.1 SP RS,HA 80.3 MA DA__ MI___ CY, VP _IW __ IT 106

______________________________68_69.6_70_71 __ 78__80_____83_83.1_86.8__90___91  

GRB _5.1_17.7_37 VP TH DM CM,RS CY DA CI SP 52.1ST MI54.2  HA __ IW _ MA _____________________  IT 84

______________42.4_43.3_45_47 50.3 51 52__53 53.3___56 __ 60 _62        

IND _ 3.7_ 7.6 _11 RS DM _ CM,CY HA MA,ST DA 23.5 MI 25.9 SP TH CI, VP IW  IT 35

_________________ 11.6 _ 19 _20 _ 22 _22.1___24.4___26 27__ 31_ 32

LSE 

10.8__15.2__31.7 DM RS VP,CI,CY,ST DA TH43.7SP,HA,CM46.1_ MI _ MA _ IW ______________________   IT 81

_________________ 34 __35 ___41.2 42.1 ___ 45 ______  47.7_51_52         

YXU _ 17.8 ___40.1____ 55 CM VP _ CY,CI MA ST,TH DM 75.7 SP 77.9  RS,HA_ IW _DA __ MI ______________ IT 145

_______________________ 59 __ 64_ 68 __70_73 ____76 _____ 80 __81_87.3_95.9  

SDF _5.0__5.2_7 CY_MA_ST,CI_DM TH _RS _MI_CM 13.1_13.9 IW,SP,HA__ DA ____ VP ____________________ IT 29

_______________7.5 _8.5_9.7_9.9_10_11.8_12__________ 15____ 17.2 ___21  

MQT 

_39.8_99.8_142 CI_ST DM__CM_ RS_TH__VP MA192.4_199.8 SP _HA __IT __IW __DA __ MI___  CY 233

_____________157 167.5 168 176 182.5 186 188______ 200 203.4 205 216 222.3 223.4    

MKE

11.4__16.2__35 DM,RS_VP CM CY TH_ MA ST,CI 45.9DA SP49.2 HA MI ___ IW __________  IT 71

__________________ 38_40_41_42.7_44_ 45 __47.2 48 __ 50_51.1_ 55      

MSP_20.4_27.5_29 CI_ST VP RS DM CY_HA,CM 48.1 DA SP 50.6 MA MI ___ IW _TH _________________ IT 87

________________ 34 37 38 42.5 43_ 45 ___ 49.9 50 ____52 52.6__57_59.8    

MLI 

2.2___11.8_14 DM__RS CY,HA,CM _VP TH DA30.9 CI ST SP33.5 MI __ MA IW _____________________  IT 58

________________ 22 ___ 25 __28 29 30.2___31 32 33__34.5__38 39     

PAH _ 12.4 __4 CI ___ ST _MA ______ TH _ HA ___  MI _ CM 9.2 _ 10.9 RS,CY SP,IW 12.4 DM __ DA __ VP ____ IT 23

_________________ 5 _ 5.5 _____7.7 _ 8 ___ 8.8 __ 9 ________ 11 ___ 12 ____ 12.6 _ 14.3 _ 20     

PIA _ 3.5__8.9 DM_10.9_RS _ST,HA CM _MA VP 24.0=SP 24.9 MI _TH __CY _DA __ IW ___ CI __  IT 40

__________________ 12 __ 18 _ 19__21_23___24 ___ 25.7_27.1_28_28.3_ 30 __ 37  

STL

_ 1.2_4.7_10 HA,CM_MA,DM ST TH RS MI IW,VP DA SP18.118.4__ CY CI _____________________________>> IT 52

_______________ 12.5_14_14.5_15_16.1_17_17.2_18 ____ 22 23   

YYZ _ 8.3____ 24.1____ 27 IW ____ RS __ TH __ CI,CM,SP 42.5 HA DM ST,MA 46.8 _CY_DA_ VP _ MI ______>> IT 104

____________________________ 35__37.5__ 40 _______ 43_44.6_45 ___ 48_48.2_49_55.5   

Total snow to date __ 477.3" (48%) __ PAH first snow Jan 12th, now every station in the game.

Feb 2nd_Amounts are now all updated through Jan 31 ...

... (totals to end of January are 105.0 cm for YXU and 61.3 cm for YYZ). 

We are using an alternate reporting site for YXU located in a similar zone relative to the Lake Huron snowbelt, namely Tillsonburg which is 40 km east of London (YXU not measuring snow this winter). The contest will be scored both with and without this YXU alternative data.

 

... percentages of 1986-2015 normal so far at ...

PAH (113)

DTW (71), CMH (65), GRR (60),

YYZ (57), MSP (54), FWA (53), YXU (53), MQT (50)

PIA (44), CLE (44),

SDF (36), APN (36), MLI (35), LSE (33),  GRB (33), MKE (31)

IND (29), STL (26), ORD (26)

(about 55% would be "normal" at this stage of winter assuming steady increase to end of season so lower MI and s ON appear to be doing better than an average winter, most other places perhaps less.)

* APN have now decided to go with posted totals and ignore missing snow Dec 24th (.04 LE and temps cold enough for snow, GRB 10:1 ratio on similar amount 0.3") since it appears to be added into next day's snowfall report. Will check this at end of contest. Otherwise going with APN posted seasonal totals.

 

Forecasters and snow totals, tie-breakers

... bold is shared win for Dec (ORD) and lone winner for Jan (IND).

 

code _ FORECASTER __ SNOW TOTAL ______T-1 __ T-2 __ T-3  ___  (actual T-1, T-2 to date)

DM ____ dmc76 ________ 787.6 __________ 2.3 __ 3.8 __ 6.8

CM ____ cmillzz ________ 841.0 __________ 4.2 __ 8.2 _ 15.0

RS ____ Roger Smith ____ 851.0 __________ 6.5 __ 3.5 _ 17.2

ST ____ Stebo __________882.5 __________ 7.0 __ 8.0 _ 13.5

CI ____ Central Illinois ___ 883.5 __________ 9.0 _ 11.0 __ 9.5

TH ____ Thomp2mp _____ 915.2 __________13.1 _ 4.5 _ 18.0

VP ____ VPBob21 _______ 943.0 ___________ 6.5 _13.6 _ 9.9

HA ____ Hawkeye_wx ___ 995.0 ___________ 7.0 __ 9.0 _13.0

MA ____ Madwx ________ 999.5 __________ 10.2 __ 6.3 _ 9.2

 

c avg __ Contest average _1003.5 __________ 8.2 __ 7.7 _13.3 ______ 5.3 (final) __ 3.7

 

SP ____ Slow poke ______1012.0 __________ 8.0 __ 8.0 __ 8.0

 

8615 _1986-2015 avg _1013.9 ________ 8.5 __8.6_10.2

 

CY ____ Cyclone77 ______ 1052.0 _________ 6.4 __ 5.0 _16.8

MI ____ michsnowfreak __ 1111.0 _________ 9.8 __ 9.0 _12.1

DA ____ DAFF __________ 1128.3 ________ 10.2 __ 8.1 _13.3

IW ____ IWXwx _________1145.0 _________ 8.5 _ 11.0 _16.5

IT ____ IndTenn ________ 1506.0 _________15.0 __ 6.0 _20.0

First tie-breaker was a tie with cmilzz 1.1" below ORD (5.3") in Dec and cyclone77 1.1" above.

The second tie-breaker was won by DMC76 (3.8" forecast, 3.7" actual Jan snow at IND).

+ ++ +++ ++++ +++++ ++++++ +++++++ +++++++ ++++++ +++++ ++++ +++ ++ +

(Forecaster departures from actual snow reported in previous post, this will gradually transition to being a contest results thread although at this early stage it basically reports those few cases where forecasts have been passed by the actuals (at PAH mostly, and also CMH (2) and PIA (1). ... we seem to be running just slightly below an average pace but reasonable projections still overlap the entire field of forecasts so anything goes from here on. 

This table is now retired (end of January) and further snowfalls are tabulated in the next post.

[Roger Smith]

February updates to snowfall amounts and contest 

The following chart continues on from January which was tracked in the previous post. That post is now retired. 

The contest results are beginning to take shape in the post before that one (two posts back), forecaster departures from actual snowfall. But so far, most of the "errors" are just snow in the future catching up to our forecasts. There are a few cases where actual amounts have passed a few predictions. Those errors can continue to increase.

So here is the February update, the orange numbers on the far left will track February snowfall, and other colored numbers are explained in the first line. Your forecasts appear across each line in the order they hold from lowest (left) to highest (right).

Actual seasonal snowfall (which continues to increase) are the red numbers moving slowly across the chart and sometimes passing your forecast. This has happened to most of us at PAH, a few at CMH, and just one forecaster as of Feb 2nd, at PIA and DTW. 

(update 11th _ Red numbers are moving faster across the charts and now almost a third of the forecasts at half the stations are lower than actual values. Some stations like APN, MQT and CLE continue to have some room left before anyone gets passed by their totals, if it even happens at all (APN is lagging the most). 

 

LOC _ Feb totals __ season to date_MIN______________ 8615 avg  __ ConA  _______________________MAX_______________

 

APN  __ 8.0 _____ 37.1 ____ 72 RS_DM _VP _CI _CM _ST SP 80.4 TH MA 85.3 _HA_ MI__ DA___ IT__  IW,CY 105

___________________________72.6_73_74_75_77_80___ 81.1_85 ___ 90_92.1_97.6_100 

ORD _20.3 _____________27.6 DM_ CM,RS 30.3_TH__HA ST 38.1 SP,CI,VP 40.0 CY _ MA_ MI__ DA ___ IW __  IT 65

____________________________ 30 _____30.5_35_37 ____ 39 ______ 41 _44 _44.6_44.9__54    

CLE _6.7___36.7_45.3 DM _CI ST CM TH_IT VP RS59.2HA MI 67.2 CY SP,IW MA ___  DA 89.7

___________________46 47 48 48.5 50 53 55__60 64.4___69_70 _ 72 

CMH  6.0____14.7 DM_CM ST RS,MA_TH 24.6_25.1 CI MI VP 28.4 SP,IT,HA DA CY IW 34

_________________15_18  20 __ 22 _______ 26 27 28 _____ 30 _ 30.1_31 

DTW _ 21.0 __________________30.5 DM___VP RS,SP _ CM _ 44.8 CI,HA MA CY TH 49.3__IW52.7ST MI _ DA __ IT 83

__________________________________38 _40 __ 43______ 45 _46 48  49.1____52___55_55.4_69.2   

FWA __ 9.5 ______________ 22.2 TH  __CM28.2 RS DM MA 34.0 HA SP 36.9 DA MI VP ST,CI,IW CY_ IT 47

______________________________ 28___30 32.2 33 ___ 35_36___38_39_41__42__ 45  

GRR _20.7_______________57.8 DM_66.8_ST _TH_ CM CI  77.1 SP RS,HA 80.3 MA DA__ MI___ CY, VP _IW __ IT 106

_______________________________68_69.6_70_71 ___ 78__80_____83_83.1_86.8__90___91  

GRB _6.7__24.4_ 37 VP TH DM CM,RS CY DA CI SP 52.1ST MI54.2  HA __ IW _ MA ____________________  IT 84

_________________42.4_43.3_45_47 50.3 51 52__53 53.3___56 __ 60 _62        

IND _ 1.4____ 9.0 _11 RS DM _ CM,CY HA MA,ST DA 23.5 MI 25.9 SP TH CI, VP IW  IT 35

____________________ 11.6 _ 19 _20 _ 22 _22.1___24.4___26 27__ 31_ 32

LSE _9.5__24.7_31.7 DM RS VP,CI,CY,ST DA TH43.7SP,HA,CM46.1_ MI _ MA _ IW ____________________   IT 81

___________________ 34 __35 ___41.2 42.1 ___ 45 ______  47.7_51_52         

YXU _14.5 __________ 55 CM 55.8 VP _ CY,CI MA ST,TH DM 75.7 SP 77.9  RS,HA_ IW _DA __ MI ____________ IT 145

__________________________ 59 __ 64_ 68 __70_73 ____76 _____ 80 __81_87.3_95.9  

SDF _0.8_____6.0_7 CY_MA_ST,CI_DM TH _RS _MI_CM 13.1_13.9 IW,SP,HA__ DA ____ VP ________________ IT 29

__________________7.5 _8.5_9.7_9.9_10_11.8_12__________ 15___ 17.2 __ 21  

MQT _16.2_116.0_142 CI_ST DM__CM_ RS_TH_VP MA192.4_199.8 SP _HA __IT __IW __DA __ MI___  CY 233

________________157 167.5 168 176 182.5 186 188______ 200 203.4 205 216 222.3 223.4    

MKE_16.5__32.7 35 DM,RS_VP CM CY TH_ MA ST,CI 45.9DA SP49.2 HA MI ___ IW __________  IT 71

_____________________ 38 40 41 42.7 44_ 45 ___47.2 48 __ 50_51.1 _ 55      

MSP__16.9___29 CI_ST VP RS DM CY 43.4_HA,CM 48.1 DA SP 50.6 MA MI ___ IW _TH _________________ IT 87

_______________34 37 38 42.5 43_____ 45 ____ 49.9 50 __52 52.6___57_59.8    

MLI  16.3_14 DM___RS_CY,HA,CM  VP28.1 TH DA30.9 CI ST SP33.5 MI __ MA IW ___________________  IT 58

_______________ 22 _ 25____28 ____29 30.2___31 32 33__34.5__38 39     

PAH  _ 0.1_____ 4 CI ____ ST _MA ______ TH _ HA ___  MI _ CM 9.2 _ 10.9 RS,CY SP,IW 12.5 DM __ DA __ VP ____ IT 23

____________________ 5 _ 5.5 _____7.7 _ 8 ___ 8.8 __ 9 ________ 11 ___ 12 ____ 12.6 _ 14.3 _ 20     

PIA _ 9.4_____8.9 DM__RS__ST,HA__CM_20.3MA VP 24.0=SP 24.9 MI _TH __CY _DA __ IW ___ CI __  IT 40

__________________12 __ 18 __ 19____21_23___24 _____ 25.7_27.1_28_28.3_ 30 __ 37  

STL_0.8_5.5_10 HA,CM_MA,DM ST TH RS MI IW,VP DA SP 18.1 18.4__ CY CI _____________________________>> IT 52

_________________ 12.5_14_14.5_15_16.1_17_17.2_18 ____ 22 23   

YYZ _ 11.5 _________ 27 IW _____RS  35.4__ TH __ CI,CM,SP 42.5 HA DM ST,MA 46.8 _CY_DA_ VP _ MI ______>> IT 104

__________________________ 35 _____ 37.5__ 40 _______ 43_44.6_45 ___ 48_48.2_49_55.5   

 

Total snow to date __ 690.2" (67.7%) 

Mar 1 _ Amounts are now updated through Feb 28 ... new snow on 28th, 2.0" MQT

... Total seasonal snowfalls at the two Canadian locations are 141.0 cm for YXU and 90.1 cm for YYZ.  

We are using an alternate reporting site for YXU located in a similar zone relative to the Lake Huron snowbelt, namely Tillsonburg WWTP which is 40 km east of London (YXU not measuring snow this winter). The contest will be scored both with and without this YXU alternative data, although I am satisfied that Tillsonburg should have a similar total to YXU especially given the rather low ratio of lake effect to synoptic scale snow this winter.

 

... percentages of 1986-2015 normal so far at ...

PAH (136) DTW (118),

GRR (87), CMH (87), MSP (86), YYZ (84), MLI (84), FWA (83), PIA (82), ORD (80),

YXU (74), MKE (66), MQT (58),  CLE (54), LSE (54),

APN (46), GRB (45), SDF (43), IND (35), STL (30)

(about 80% would be "normal" at this stage of winter assuming steady increase to end of season so lower MI and s ON appear to be doing better than an average winter, also recently central IL-IN-OH, most other places perhaps less than expected at this point.)

 

Forecasters and snow totals, tie-breakers

... bold is shared win for Dec (ORD) and lone winner for Jan (IND).

 

code _ FORECASTER __ SNOW TOTAL ______T-1 __ T-2 __ T-3 

_____________________________Actual___ 5.3  __ 3.7 _21.0 (to 9th)

DM ____ dmc76 ________ 787.6 __________ 2.3 __ 3.8 __ 6.8

CM ____ cmillzz ________ 841.0 __________ 4.2 __ 8.2 _ 15.0

RS ____ Roger Smith ____ 851.0 __________ 6.5 __ 3.5 _ 17.2

ST ____ Stebo __________882.5 __________ 7.0 __ 8.0 _ 13.5

CI ____ Central Illinois ___ 883.5 __________ 9.0 _ 11.0 __9.5

TH ____ Thomp2mp _____ 915.2 __________13.1 _ 4.5 _ 18.0

VP ____ VPBob21 _______ 943.0 ___________ 6.5 _13.6 _ 9.9

HA ____ Hawkeye_wx ___ 995.0 ___________ 7.0 __ 9.0 _13.0

MA ____ Madwx ________ 999.5 __________ 10.2 __ 6.3 _ 9.2

 

c avg __ Contest average _1003.5 __________ 8.2 __ 7.7 _13.3 

 

SP ____ Slow poke ______1012.0 __________ 8.0 __ 8.0 __ 8.0

 

8615 _1986-2015 avg _1013.9 ________ 8.5 __8.6_10.2

 

CY ____ Cyclone77 ______ 1052.0 _________ 6.4 __5.0 _16.8

MI ____ michsnowfreak __ 1111.0 _________ 9.8 __ 9.0 _12.1

DA ____ DAFF __________ 1128.3 ________ 10.2 __ 8.1 _13.3

IW ____ IWXwx _________1145.0 _________ 8.5 _ 11.0 _16.5

IT ____ IndTenn ________ 1506.0 _________15.0 __6.0 _20.0

First tie-breaker was a tie with cmilzz 1.1" below ORD (5.3") in Dec and cyclone77 1.1" above.

The second tie-breaker was won by DMC76 with 3.8" (actual 3.7") for Jan snow at IND.

The third tie-breaker is Feb snow at DTW. -- IndTenn had already won this one on the 10th with highest forecast of 20.0" (when 19.6" had fallen, the rest won't change this result which is already up to 21.0"). 

This table is now complete through February and will be retired with a new post starting off March. 

[Roger Smith]

March to May updates to snowfall amounts and contest 

The following chart continues on from February which was tracked in the previous post. That post is now retired. 

The contest results are beginning to take shape in the post before that one (three posts back but subject to being moved to end of the thread if I forget to edit this and it's gone, a notice will replace it and it may be after this post), forecaster departures from actual snowfall. 

So here is the March update, the orange numbers on the far left will track March snowfall, and other colored numbers are explained in the first line. Your forecasts appear across each line in the order they hold from lowest (left) to highest (right).

When March is finished, April and May additional snowfall will be shown where necessary, second and possibly third orange numbers will indicate April and May snowfall totals. 

Actual seasonal snowfall (which continues to increase) are the red numbers moving slowly across the chart and sometimes passing your forecast. This has happened to most of us at quite a few stations, notably PAH and DTW, but there are growing numbers at nine other locations as of March 1st (19 as of April 14th) and that number could continue to grow. We were nowhere near the minimum forecast at APN, MQT, or GRB as March began but in early April we passed LSE, GRB and MQT, and have made considerable progress at APN, passing the minimum on April 14th, with STL now the only location not to pass minimum forecast (needs 3").

In the table below, the forecasts that are MIN and MAX appear in the first line (e.g., APN min fcst is 72" by RS and max is 105" by IW and CY). All the other forecasts have just the two-letter forecaster codes in that same line, in order and their numerical forecasts are shown below their names (as close as possible, sometimes on a bit of an angle). The 86-15 avg and Contest Avg are shown in their appropriate positions relative to all those forecasts, and the red seasonal total advances left to right ... for APN it has recently passed five forecasts and is currently just below Stebo (ST) at 76.6", for ORD it has passed five forecasts. That part of the table is updated each day that snowfall occurs.

Note: April snowfalls are found in line two of each station's data under the March totals. (nothing has yet fallen in May as of 8th, but if any does, that will appear beside the April totals).

 

LOC _ Mar totals ____MIN __  season to date _________ 8615 avg  __ Contest Avg  ________________MAX_______________

 

APN  __14.7_________ _72 RS_DM _VP _  CI _CM_ ST SP 80.4 TH82.5 MA 85.3 _HA_ MI__ DA__ IT__ IW,CY 105

____30.7Apr ____________72.6_73__74_75  77 80 ____81.1___85 ____90_92.1_97.6_100 

ORD _2.6 _____________27.6 DM_ CM,RS _TH __ HA_36.1ST 38.1 SP,CI,VP 40.0 CY _ MA_ MI _ DA ___ IW __  IT 65

_____3.2Apr _________________ 30_ 30.5__ 35___ 37 ____ 39 _______ 41 _ 44_44.6_44.9__54    

CLE _12.7__45.3 DM_ CI ST CM TH_IT VP 53.1 RS59.2HA MI 67.2 CY SP,IW MA ___  DA 89.7

____ 3.3Apr  ______46 47 48 48.5 50 53 __ 55__60 64.4 ___ 69_70 _ 72 

CMH  3.2__14.7 DM_CM _ST RS,MA_TH__25.1 __CI _MI_VP_28.4  _ SP,IT,HA DA 30.7 CY__ IW 34

____ 2.6Apr _____ 15_18  20 __ 22 ________26 27_28_________30_30.1____31

DTW _ 5.9 ___________________30.5 DM___VP RS,SP _ CM_ 44.8 CI,HA MA CY TH 49.3_IW ST MI __ 61.0 _ DA __ IT 83

_____ 2.4Apr ______________________ 38 _40 __ 43______ 45 _46 48  49.1__ 52 55 55.4_____69.2   

FWA __ 2.1 ___________ 22.2 TH  __CM __RS_  DM 32.7 MA_ 34.0 HA SP 36.9 DA MI VP ST,CI,IW CY_ IT 47

_____2.4Apr_________________ 28__ 30_32.2 ___ 33 ____ 35_36___ 38_39_41__42__ 45  

GRR _ 4.9 _______________57.8 DM_____ST _TH_ CM _ CI_ 77.1 77.7  SP RS,HA 80.3 MA DA_ MI__ CY,VP_IW_ IT 106

_____ 6.0Apr ____________________68_69.6_70_71 ________ 78__80_____83_83.1 86.8__90_91  

GRB _7.0___37 VP _TH DM CM,RS  CY _DA CI SP_52.1_ST MI 54.2  HA __ IW _ MA ____ 68.1 ____________  IT 84

____36.7Apr ___42.4 43.3_45_47 50.3 51 52____53 53.3___ 56 __ 60 _ 62        

IND _11.6___11 RS DM___CM,CY HA MA,ST DA 23.2 23.5 MI 25.9 SP_ TH__ CI, VP _IW__  IT 35

___ 2.6Apr _____ 11.6 ___ 19 _20 _22_22.1______24.4____ 26 _27___ 31 _ 32

LSE _6.1___31.7 DM RS VP,CI,CY,ST DA TH  43.7  SP,HA,CM46.1_ MI _49.8 MA_ IW _______________   IT 81

___19.0Apr _____ 34 __35 ____41.2 42.1 _____ 45 ______  47.7____51 _ 52         

YXU _ 7.5 __________ 55 CM  VP_ CY,CI 65.7  _MA ST,TH _DM 75.7 SP 77.9  RS,HA_ IW _DA __ MI __________ IT 145

____ 2.2Apr____________ 59 __64 ______ 68 __70 _73 ____ 76 _____80 __ 81_87.3_95.9  

SDF _11.9____7 CY_MA_ST,CI _DM TH _RS _MI_CM___ 13.1 _ 13.9__ IW,SP,HA__ DA_ 19.1 _ VP _________ IT 29

____ 1.2Apr_____7.5 _8.5_9.7_9.9_10_11.8_12_____________ 15 ____ 17.2 _______21  

MQT _24.1  142 CI _ST  DM  CM_175.3 RS_TH_VP_MA 192.4_199.8 SP _HA __IT __IW __DA __ MI___  CY 233

____35.2Apr____157 167.5 168__176 182.5 186 188______ 200 203.4 205 216 222.3 223.4    

MKE_ 4.0__35 DM,RS_VP CM CY_TH  MA ST,CI 45.9 46.7 DA SP 49.2 HA MI ___ IW __________  IT 71

___ 10.0Apr ______ 38_40 41 42.7 44_45________ 47.2 48 __ 50_51.1 _ 55      

MSP__ 8.8 __29 CI_ST VP RS DM CY _HA,CM  48.1 ___ DA _SP 50.6  MA MI  ___ IW _TH ______ 78.3 __ IT 87

___26.1Apr ____ 34 37 38 42.5 43__45 _______ 49.9 50 ____52 52.6___57_59.8    

MLI   10.4_14 DM__ RS_CY,HA,CM  VP TH DA 30.9 CI _ST _ SP 33.5 MI __ MA _ IW_ 41.7__________  IT 58

____ 3.2Apr _____ 22 _ 25 ___ 28 29 30.2___31 _32  _33 __ 34.5__38 _ 39     

PAH  _ 0.2_____ 4 CI ____ ST _MA ______ TH _ HA ___  MI _ CM 9.2 _ 10.9 RS,CY SP,IW  DM__ 13.7 DA __ VP __ IT 23

____1.0Apr __________ 5 _ 5.5 _____7.7 _ 8 ___ 8.8 __ 9 ________ 11 ___ 12_12.6 ____ 14.3 _ 20     

PIA _ 9.3_____8.9 DM____RS___ST,HA__CM_ MA VP 24.0=SP 24.9 MI _ TH__CY _DA _ IW 32.1 __ CI __  IT 40

____ 1.4Apr __________12 ___ 18 __ 19_21_23___ 24 ____25.7_27.1_28_28.3_30 _____ 37  

STL_1.0_ 7.1 10 HA,CM_MA,DM ST TH RS MI IW,VP DA SP 18.1 18.4__ CY CI _____________________________>> IT 52

____ 0.6Apr _______ 12.5_14_14.5_15_16.1_17_17.2_18 ____ 22 23   

YYZ _ 3.5 __________ 27 IW __________RS_ TH ___ CI,CM,SP_ 42.5 HA43.1 DM ST,MA 46.8 _CY_DA_ VP _ MI ___>> IT 104

____ 3.8Apr _____________________35  37.5_____ 40 ______ 43___44.6_45 ___ 48_48.2_49_55.5   

 

Total snow to date __ 1037.7" (102.3% of 1986-2015 avg)

Apr 19 _ CLE 1.1" ... FWA 0.8" ... CMH 0.3" ... ORD 0.1"

Apr 18 _ LSE 6.0" ... MKE 3.2" ... ORD 0.4" ... GRR 0.3"

Apr 17 _ MQT 3.1" .. GRR 1.4" .. APN 1.0" ... DTW 0.8" .. CLE, CMH 0.3" .. YYZ 0.1"

Apr 16 _ MQT 7.8" .. APN 4.9" .. MKE 0.9" .. GRB, GRR 0.5" .. MLI 0.4" .. ORD, PIA, IND 0.3" .. FWA, DTW, SDF 0.2" .. MSP 0.1" (LSE revised to tr dropped 0.1) .. YXU running est 1.0" moved from 15th

Apr 15 _ MQT 15.8" GRB 10.2", MSP 3.5", MKE 3.2", APN 2.7", LSE 2.1" YYZ 1.5" GRR 0.7" MLI 0.1", YXU only tr reported

Apr 14 _ MSP 11.1", GRB 11.0", APN 7.3", LSE 3.3", YYZ 1.4" (sleet and snow), MKE 0.5" and GRR 0.1"

Apr 13 _ 3.4" APN, 2.5" GRB (tr LSE, 0.0" MQT) MSP 1.1" 

Apr 12 _ 4.8" MQT

Total seasonal snowfalls at the two Canadian locations are 163.5 cm for YXU (includes 2.5 cm est for April 16th-17th, plus 3.5 for March 14th missing data at Tillsonburg, hourly reports at YXU used to estimate, contest total 166.0 cm), and 109.3 cm for YYZ ... amounts confirmed through April 16 for YYZ and 15th for YXU.

We are using an alternate reporting site for YXU located in a similar zone relative to the Lake Huron snowbelt, namely Tillsonburg WWTP which is 40 km east of London (YXU not measuring snow this winter). The contest will be scored both with and without this YXU alternative data, although I am satisfied that Tillsonburg should have a similar total to YXU especially given the rather low ratio of lake effect to synoptic scale snow this winter. Radar on March 14th looked as though YXU was in a heavy Lake Huron squall band and Tillsonburg may report less but assuming there have been offsetting cases earlier this winter. Tillsonburg now reports missing on 14th, YXU hourly obs indicate several hours of snow and blowing snow, assigning the previously estimated 3.5 cm to this missing data (observer may have been unable to visit site due to weather). 

A new scoring table has been posted below this post. The old one a few posts back is not required now and has been deleted (all of the same information and more is in the new post). 

... percentages of 1986-2015 normal so far at ...

MSP (155), PAH (149), DTW (137), SDF (137), PIA (129), GRB (127), MLI (124), CMH (108), LSE (108), APN (103), GRR (101), YYZ (100)

 ... SDF became third station to achieve above normal status on March 21st, PIA was fourth on March 24th, MLI later 24th became fifth to reach above normal. MSP was sixth as of March 30th. ... CMH was seventh on April 1st. ... GRB was eighth as of April 14th. YYZ was 9th on 15th. APN passed normal late on April 16th to become the 10th location above normal. The total snowfall also passed total normal snowfall on April 16th. GRR joined the above normal club on April 17th (11th location). LSE became the 12th location above normal on April 18th. 

FWA (96), ORD (95), MKE (95), IND (90), MQT (88), YXU (87), CLE (79), STL (39)

 

Forecasters and snow totals, for tie-breakers ... this part is now complete

... bold is shared win for Dec (ORD) and lone winner for Jan (IND).

 

code _ FORECASTER __ SNOW TOTAL ______T-1 __ T-2 __ T-3 

_____________________________Actual___ 5.3  __ 3.7 _ 21.0

DM ____ dmc76 ________ 787.6 __________ 2.3 __ 3.8 __ 6.8

CM ____ cmillzz ________ 841.0 __________ 4.2 __ 8.2 _ 15.0

RS ____ Roger Smith ____ 851.0 __________ 6.5 __ 3.5 _ 17.2

ST ____ Stebo __________882.5 __________ 7.0 __ 8.0 _ 13.5

CI ____ Central Illinois ___ 883.5 __________ 9.0 _ 11.0 __9.5

TH ____ Thomp2mp _____ 915.2 __________13.1 _ 4.5 _ 18.0

VP ____ VPBob21 _______ 943.0 ___________ 6.5 _13.6 _ 9.9

HA ____ Hawkeye_wx ___ 995.0 ___________ 7.0 __ 9.0 _13.0

MA ____ Madwx ________ 999.5 __________ 10.2 __ 6.3 _ 9.2

 

c avg __ Contest average _1003.5 __________ 8.2 __ 7.7 _13.3 

 

SP ____ Slow poke ______1012.0 __________ 8.0 __ 8.0 __ 8.0

 

8615 _1986-2015 avg _1013.9 ________ 8.5 __8.6_10.2

 

CY ____ Cyclone77 ______ 1052.0 _________ 6.4 __5.0 _16.8

MI ____ michsnowfreak __ 1111.0 _________ 9.8 __ 9.0 _12.1

DA ____ DAFF __________ 1128.3 ________ 10.2 __ 8.1 _13.3

IW ____ IWXwx _________1145.0 _________ 8.5 _ 11.0 _16.5

IT ____ IndTenn ________ 1506.0 _________15.0 __6.0 _20.0

First tie-breaker was a tie with cmilzz 1.1" below ORD (5.3") in Dec and cyclone77 1.1" above. This tie went to second tie-breaker where cyclone77 was closer.

The second tie-breaker was won by DMC76 with 3.8" (actual 3.7") for Jan snow at IND.

The third tie-breaker is Feb snow at DTW. -- IndTenn had already won this one on the 10th with highest forecast of 20.0" (when 19.6" had fallen, the rest won't change this result which is already up to 21.0"). 

[Roger Smith]

Contest scoring -- total departures from actual snowfall

 

This table will be edited each time snow falls and the leaderboard may change before the contest is done.

The black numbers in the table are the departures where actual snowfall has already surpassed your prediction.

The red numbers are departures that could still improve, your forecast is still above the actual snowfall.

Stations in black (19) have at least one forecast below actual, stations in red (1, STL) have no forecasts below actual snowfall (yet).

The order of stations tells you in which order these locations passed minimum forecast (more or less this order anyway, I added four to a previous table that was showing all forecasts within five inches of actual, so the order that they go may not be the order shown here, confirmed Apr 4th as MQT beat STL and APN to this mark). APN joined on April 14th. 

The sum of all these numbers is your total departure. This will amount to total error at end of contest (red numbers are not really "errors" yet) although added April 4th some of the larger ones will be difficult to hack down now. Some red zone flexibility appears good for forthcoming events in the central zone.

I am now showing a table of average percentages also, to see if that looks any different (it will assign equal values to the heavy snowfall locations like MQT and the less snowy places in the south.

(note -- this table does not wrap on my screen, if it wraps on yours, try a smaller zoom % and conversely if it looks small but you have room, try a larger zoom %)

 

__ Forecast _// ___ red numbers = forecast can improve ___ black numbers = forecast already below actual __

___ TOTAL_ //___________________ Departures (at last report) _____________________________________

code_SNOW_PAH CMH PIA DTW MSP__MLI_FWA YYZ GRR YXU__ORD_MKE SDF IND CLE ___ LSE GRB _STL APN MQT___TOTAL

 

MA _ 999.5 _ 8.2_10.7_11.1 15.0 26.3___3.7_ 0.3_ 1.9 _5.6 _2.3 __7.9_ 2.7_11.6 _1.2_18.9___ 1.2 _ 6.1_ 5.4 _2.5 12.7__159.0 

 

8615 avg __ PAH CMH PIA DTW MSP___MLI FWA YYZ GRR YXU___ORD MKE SDF IND_ CLE___LSE GRB_STL APN MQT

__ 1013.9 _ 4.5_2.3_7.2_16.2_27.7___ 8.2_ 1.3_ 0.6 _0.6 10.0___2.0_ 2.5_5.2_ 2.7 14.1____ 3.7 13.9 11.3 _ 2.1_24.5__164.6

c av_1003.5_ 2.8_5.6_8.1_11.7_30.2__10.8_ 4.2_ 3.7_ 2.6_12.2___3.9 _0.8_ 6.0 _0.3 _6.1___ 6.1_16.0_11.0 _2.8_17.1__166.0

___________ PAH CMH PIA DTW MSP___MLI FWA YYZ GRR YXU___ORD MKE SDF IND_ CLE___LSE GRB_STL APN MQT

TH _ 915.2 _ 6.0_ 8.7_5.0_11.9_18.5__12.7_10.5_ 5.6_ 8.1_ 4.3___ 5.6_ 4.0_9.2_ 3.8_ 4.6___ 7.7_25.7_ 7.4 _ 1.4_ 7.2__171.9

SP _ 1012.0_ 1.7_ 0.7_8.1_21.0_28.3___ 8.7_3.3_ 3.1_ 0.3_10.3___ 2.9_ 1.3_ 4.1 _2.8 16.9___ 4.8_16.1 10.9 _2.5 24.7__176.5 

HA _ 995.0 _ 5.7_0.7_14.1 16.0 33.3___16.7_2.3_ 0.1 _2.3_14.3___ 1.1 _3.3_ 4.1 _3.2_ 6.9___ 4.8 _12.1_2.9_ 7.5 29.7__ 185.1

CM _ 841.0 _ 4.7 15.7 13.1 18.0 33.3___16.7_4.7_ 3.1_ 7.7_10.7___6.1_ 6.7 _ 7.1_ 4.2 _ 5.1__ 4.8_23.1_ 2.9 _ 7.5_ 7.3__198.5

ST _ 882.5 _  8.7 12.7 14.1_6.0 44.3___ 9.7 _9.3_ 1.9_ 9.7_ 4.3___ 0.9_ 1.7 10.6 _1.2_ 6.1___14.8 _15.1_ 6.9_ 5.5_18.3__207.4 

VP _ 943.0 __ 6.3_2.7_9.1 23.0 41.3 __13.7 _8.3_ 5.9 12.3_ 6.7___ 2.9_ 8.7_ 1.9 _ 7.8 _0.1___14.8_31.1_ 9.9_ 9.5_10.7__222.7 

MI _ 1111.0_ 4.9_ 3.7_ 6.4_ 5.6_25.7___7.2_ 7.2_12.4_9.1 30.2___8.5 _4.4_ 7.3 _ 1.2 11.3____2.1 14.8_ 9.0 _ 9.6_48.1__232.7 

CI _ 883.5 __9.7 _4.7 _4.9 16.0 49.3 __10.7 _9.3_ 3.1_ 6.7_ 1.7___2.9_ 1.7_10.6_ 7.8_ 7.1___14.8_17.1_15.9_ 8.5_33.3__236.4 

IW _1145.0_  1.7_3.3_ 2.1_ 9.0_21.3___ 2.7_ 9.3_16.1 13.3 15.3__17.9 _ 8.3_ 4.1 _8.8 16.9___ 2.2 _ 8.1_ 9.9 22.5 44.7__241.5

DA _1128.3_ 0.6_ 0.6_ 3.8 _ 8.2_28.4___11.5_ 5.3_ 5.1 _5.4 21.6___8.8 _0.5_ 1.9 _1.1 36.6___ 8.6_17.8_10.1 15.1 47.0__242.0

RS _ 851.0 _ 2.7_10.7 20.1 21.0 40.3__19.7_ 2.7 _8.1_ 2.3_14.3___ 6.1 11.7_ 9.1_12.2 _1.9__ 15.8_23.1_ 7.9_10.5_ 0.7__ 244.9 

CY _1052.0_  2.7_0.3_4.1_13.0_35.3__ 16.7_12.3_ 4.9 12.3_ 1.7___ 5.4_ 3.0 15.3_4.2_15.9___14.8_21.1_14.9 22.5 57.7__277.7 

DM _ 787.6 _ 1.1 16.0 23.2 30.5 35.8__ 27.7 _ 0.5_ 1.5 19.9_ 7.3___8.5_11.7 _9.4_11.6_ 7.8___18.1_24.8_ 5.4_ 9.9_ 7.8__282.5

IT _1506.0_   9.3_0.7 _7.9 22.0 _8.7__16.3_15.1_60.9 28.3 79.3___29.4 27.0 _6.7 11.8_ 3.1___31.2 15.9_44.9 17.5 40.7__479.9

__________PAH CMH PIA DTW MSP___MLI FWA YYZ GRR YXU___ ORD MKE SDF IND_ CLE ___LSE_GRB_STL_APN MQT

Most recent _ 13.7 30.7 32.1 61.0 78.3_ 41.7 32.7 43.1_77.7_65.7 _36.1_46.7_19.1_23.2_53.1 __49.8_68.1_7.1_82.5_175.1_1037.7 total

______________________________________________________________________________________________

snowfall to date _ (Apr 21) _ 1037.7" (102.34% of 1986-2015 avg)

Apr 19 _ CLE 1.1" ... FWA 0.8" ... CMH 0.3" ... ORD 0.1"

Apr 18 _ LSE 6.0" ... MKE 3.2" ... ORD 0.4" ... GRR 0.3"

Apr 17 _ MQT 3.1" .. GRR 1.4" .. APN 1.0" ... DTW 0.8" ... CLE and CMH 0.3" YYZ 0.1"

Apr 16 _ MQT 7.8" .. APN 4.9" .. MKE 0.9" GRB 0.5" GRR 0.5" MLI 0.4" ORD 0.3" PIA 0.3" IND 0.3" SDF FWA and DTW 0.2", MSP 0.1" and LSE corrected from 0.1" to trace. 

Apr 15 _ YYZ 1.5" MQT 15.8" GRB 10.2", MSP 3.5", MKE 3.2", APN 2.7"  GRR 0.7", LSE 2.1" ORD 0.2" MLI 0.1" 

Apr 14 _ MSP 11.1", GRB 11.0", APN 7.3", LSE 3.3", YYZ 1.4" (sleet and snow), YXU Tr, MKE 0.5" and GRR 0.1"

Apr 13 _ 3.4" APN, 2.5" GRB (tr LSE, 0.0" MQT) MSP 1.1" 

Apr 12 _ 4.8" MQT

 

Station leaders ... subject to revision. 

... * tied (shared leaders -- YXU, IND, and STL at present)

... italics, location has less snow than this minimum forecast (bold = same amount)

... ^ leader has minimum as well as closest forecast to date but location now has more snow

... # leader has maximum as well as closest forecast so cannot lose the lead now

________________________________________________________________________

numbers in brackets show errors, + means forecaster is above actual snow, - means below actual snowfall to date.

... 0.0 means forecast equals snowfall to date

DAFF _____4 __ IND* (-1.1), SDF (-1.9), PAH (+0.6), MKE (+0.5)

madwx ___ 3 __ GRB (-6.1), LSE (+1.2), FWA (+0.3)

Hawkeye __2 __ YYZ (-0.1), STL* (+2.9) 

IWXwx ___ 2 __ PIA (-2.1), MLI (-2.7)

VPBob21 __ 2 __ CLE (+1.0), IND* (+1.1)

contest avg_1 __ IND (+0.3)

slowpoke __1 __ GRR (+0.3)

cmillzz ____1 __ STL* (+2.9) 

Cyclone77_ 1 __ YXU* (-1.7), CMH (+0.3)

IndTenn __ 1 __ MSP# (+8.7)

Central IL _ 1 __ YXU* (-1.7)

Michsnow _ 1 __ DTW (-5.6)

Thomp2m _1 __ APN (--1.4) 

Roger Smith 1 __ MQT (+0.7)

Stebo _____1 __ ORD (+0.9)

 

As of April 14th, one of the twenty locations had not reached minimum forecast, as APN made it 19 above minimum on Apr 14th, LSE became the 16th location on April 3rd and later GRB 17th. MQT joined on April 4th passing minimum by 0.2" mid-afternoon. STL still needs about three inches. SDF became the 13th location to pass minimum forecast on March 11th. CLE was the 14th to pass minimum forecast on March 13th. (IND 15th on March 24th).

8615 normal is not currently closest at any location. Contest average is closest at IND. (closest forecasts also shown in table)

_________________________________________________________________________

Alternate scoring system -- your percentage of actual snowfall

This contest has always been scored by departures, but as 20% of the contest snow happens at MQT and another 20% at three other locations combined, this may not tell the whole story of accuracy. This measure will compare your prediction to actual in percentage terms and average out the 20 scores. The rankings are compared to the first scoring table above, and ranks for contest average and 1986-2015 normal are in brackets, they are ranked as if they were lone additions to the contest (so not counting each other) and lower forecaster ranks are not shifted by these values. For a while these ranks were between two actual forecasters but now there is one contest entrant between them.

I have also added average percentage error (bearing in mind that the ideal percentage at this point may be about  +10%). 

This table is now ordered by AVG % error (col 2) as this is becoming the better measure of accuracy towards the end of the contest. 

 

FORECASTER ________ MEAN %  __ AVG % error ___ Position in total dep table above

 

hawkeye _____________ 95.9 ______ 19.1 _____________4

madwx _______________96.3 ______ 21.9 ____________ 1

slowpoke _____________ 97.5 ______22.2  ___________ 3

___ 1986-2015 avg ____ 101.9 ______22.3 __________ (2 a )

___ contest avg ________96.7 ______ 22.6 ___________ (2 b )

cmillzz _______________ 81.0 ______ 24.0 _____________ 5

Thomp2mp ___________ 88.2 ______ 24.2 _____________ 2

DAFF ________________108.7 ______25.2 ____________ 11

Michsnowfreak ________107.1 ______ 26.2 _____________ 8

Stebo ________________ 85.0 ______ 26.7 _____________ 6

IWXwx ______________ 110.3 ______27.6 _____________10

vpBob21 ______________90.9 ______ 28.7 _____________ 7

Roger Smith ___________82.0 ______ 31.9 _____________12

dmc76 _______________ 75.9 ______ 33.2 _____________14

cyclone77 ____________101.4 ______ 33.8 _____________13

Central Illinois _________ 85.1 ______ 34.3 _____________ 9

IndTenn _____________ 145.1 ______ 76.3_____________15

_____________________________________________________________________

____ all files are updated daily when snow happens __ latest update through April 21, 2018.

(how to read this table ... the first column is the average of your 20 forecasts in terms of % of actual snow that you predicted. Taking contest avg as an example, the average forecast so far is 108.2% of the actual snowfall. This is not quite the same as taking your total snowfall and comparing that to the contest total, it's an average of twenty locations. The second column which orders the table is your average percentage error. Once again taking contest average, the average forecast is 25.5% different from the actual. Why is it not 8.2% -- because some of those individual values that average 108.2 are higher or lower. At the moment, the highest contributor is STL where the average forecast is 255% of the actual snowfall. The lowest contributor is SDF where the average forecast is 69% of actual. Those are 155% and 31% percentage errors. If all the errors were on the same side of actual snow, the two figures would be similar. This leaves DAFF with the most similar values because their forecasts are mostly a bit higher than actual snowfall. Since the season is not quite over, the ideal values for somebody closing in on perfect forecasts would be something like 105 and 10%. )

Rank per location

new table ... your rank for each location as we near the end of the contest ... ranks of normal and contest average do not bump any higher ranks up (higher ranking numbers, that is, I suppose those are lower ranks). These appear in the order of lowest (best) average ranking. Ties are marked with an asterisk * .

Forecaster ____ APN_ORD_CLE_CMH_DTW_FWA_GRR_GRB_IND_LSE_YXU_SDF_MQT_MKE_MSP_MLI_PAH_PIA_STL_YYZ__Avg

 

slowpoke ______ 02*_ 03*_12*__ 03*_11*__05 _ 01 _ 07 __ 05 _ 04*_09 __ 03*_08 _ 02 _ 06 _ 04 _ 03*_ 08 _ 12 _05*__ 5.65

Hawkeye ______ 05*_  02 _ 07 __ 03*_ 08*__03 _ 02*_03 __ 06 _ 04*_10*__03*_09 _ 06 _ 08*_ 11*_ 09 _ 12 _01*_01 __ 5.65

madwx ________02*_ 10 _ 14 __ 11*_ 07 __ 01 _ 04 _ 01 __ 02*_01 _ 04 __ 14 _ 06 _ 05 _ 05 _ 02 _ 12 _ 10 _ 03*_03*__ 5.85

 

__8615 normal _ 02 _ 03 _ 11 __ 06 _ 10 __ 03 _ 02 _ 04 __ 05 _ 04 _ 09 __ 06 _ 08 _ 05 _ 06 _ 04 _ 07 _ 07 _ 13 _ 02 __ 5.85

__contest avg __ 04 _ 06 _ 07 __ 10 _ 05 __ 06 _ 04 _ 07 __ 01 _ 07 _ 10 __ 06 _ 07 _ 02 _ 08 _ 07 _ 06 _ 08*_13 _ 08 __ 6.65

 

Thomp2mp_____ 01 _ 07 _ 04 __ 10 _ 05 __ 13 _ 08 _ 14 __ 07 _ 07 _ 06 __09 _ 02 _ 07 _ 02 _ 08 _ 10 _ 05 _ 06 _ 10 __ 7.05

DAFF __________12 _ 13 _ 15 __ 02 _ 03 __ 07 _ 05 _ 09 __ 01 _ 08 _ 13 __ 02 _13 _ 01 _ 07 _ 07 _ 01 _ 02 _ 11 _ 09 __ 7.05

Stebo _________ 04 _ 01 _ 06 __ 13 _ 02 __ 10*_ 10 _ 05 __ 02*_09*_06*__12*_07 _ 03*_14 _ 05 _ 13 _ 12*_05 _ 03*__7.10

cmillzz ________ 05*_ 08*_05 __ 14 _ 10 __ 06 _ 07 _ 11*__08*_04*_ 05 __ 06 _ 03 _ 10 _ 08*_11*_07 _ 11 _ 01*_05*__ 7.25

Michsnowfreak __ 09 _ 11*_ 10 __ 08 _ 01 __ 08 _ 09 _04 __ 02 _ 02 _ 14 __ 07 _ 14 _ 08 _ 04 _ 03 _ 08 _ 06 _ 08 _ 13 __ 7.45

IWXwx _________14*_14 _ 13 __ 07 _ 04 __ 10*_ 13 _ 02 __ 12 _ 03 _ 12 __ 03*_12 _ 11 _ 03 _ 01 _ 03*_01 _ 09*_14__ 8.05

Central Illinois ___07 _ 03*_08 __ 09 _ 08*__ 10*_ 06 _ 08 __ 10*_09*_01*__12*_10 _ 03*_15 _ 06 _ 15 _ 04 _ 14 _ 05*__8.15

VPBob21 _______ 08 _ 03*_01 __ 06 _ 14 __ 09 _ 11*_15 __ 10*_09*_ 03 __ 01*_05 _ 12 _ 13 _ 09 _ 11 _ 09 _ 09*_ 11 __8.45

Cyclone77 ______ 14*_ 06 _ 11 __ 01 _ 06 __ 14 _ 11*_10 __08*_ 09*_01*__15 _ 15 _ 09 _ 10 _ 11*_ 05*_03 _ 13 _ 08 __9.00

Roger Smith ____ 11 _ 08*_ 02 __ 11*_ 11*__04 _ 02*_11*__15 _ 13 _ 10*__08 _ 01 _ 13*_12 _ 14 _ 05*_14 _ 07 _ 12 __ 9.20

DMC76 _________10 _ 11*_ 09 __ 15 _ 15 __ 02 _ 14 _ 13 __ 13 _ 14 _ 08 __ 10 _ 04 _ 13*_11 _ 15 _ 02 _ 15 _ 03*_02 __9.95

IndTenn ________13 _ 15 _ 03 __ 03*_ 13 __ 15 _ 15 _ 06 __ 14 _ 15 _ 15 __11 _ 11 _ 15 _ 01 _ 10 _ 14 _ 07 _ 15 _ 15__11.30

 

Forecaster ____ APN_ORD_CLE_CMH_DTW_FWA_GRR_GRB_IND_LSE_YXU_SDF_MQT_MKE_MSP_MLI_PAH_ PIA_STL _YYZ__Avg

 

[Roger Smith]

Contest standings will continue to be updated in the previous post.

March snowfall is tracked in the post before that, and any additions in April and May will appear in the same table (updated).

Cmillzz is in the lead in both total departure and average percentage error per station. 

The standings can still change as MQT errors are a large part of the total, and Central Illinois will be the first to start accumulating errors there as total snowfall is just below that minimum forecast now. Comparing forecasts is made somewhat easier by checking whether one line has similar type (red or black) departures, where they are the same as each other, no imminent changes are likely. 

The contest will run through May which is still snowfall season at MQT and sometimes a few other locations as well. Check back and see how the tables are changing as we get closer to the end of the contest. 

[michsnowfreak]

Good work. Interesting is that I was a bit low at DTW (I said 55.4, they are at 58.6) but too high in most other places.

[Roger Smith]

(posted earlier today)

Looking back to the standings (Mar 7 post edited daily) the lead is now with Thomp2mp by just 1.6" over cmillzz. 

You could figure this is really 0.2" if cmillzz gets to use his remaining credit of 0.7" at YYZ (which seems likely, although not a guarantee).

From that point, the differentials would be at the following locations (any place that both are in the black is not going to affect the contest results for them):

Thomp2mp has the advantage at YXU (5.3") and IND (4.3") which if completely used would place the differential at 2 x 9.6" or 19.2" plus the 0.2" left after YYZ.

Also both have a credit left to use at MKE and STL but Thomp2mp has 2.7" more at MKE and 4.7" more at STL. Using those requires getting into the differential zone first, which requires 1.6" at MKE and 2.9" at STL. Those seem possible although unlikely for STL. 

The opposite situation exists only at LSE so that the only chance cmillzz has to win would be to get the YYZ snow and more than 4.2" snow at LSE after which the next 2.7" would be a further advantage before both begin to accumulate. Cmillzz also has 1.1" credit left at GRB. 

This is an example of how plausible snowfalls at the two locations would affect the outcome:

YYZ 1.5" leaves the differential at 0.2" in favor of Thomp2m

YXU 2.0" restores the differential to 4.2"

IND 0.5" increases the differential to 5.2"

MKE 2.0" increases the differential to 6.0"

STL 0.2" leaves the differential at 6.0"

LSE 7.0" reduces the differential to 1.6"

GRB 1.5" changes the outcome to 0.6" win for cmillz

So between these two, the outcome depends on what happens at these stations (APN and MQT are also differentials at present, in favor of Thomp2mn) but appear to be beyond the range of the

contest totals).

Currently third place Central Illinois can easily win also, if there is substantial snow (3" should suffice) in the ORD-MKE-FWA-IND-PIA region, because they have large credits to use in all of those locations and the two ahead do not. A smaller 1-2" snowfall in this region will make things very close all round. However, CI has a disadvantage for MQT where errors are already increasing while others have 10 to 20 inches or more left to reduce their totals. 

I don't see an easy path to contest victory for any other forecaster but IndTenn can still win with massive snowfalls throughout the region. Stebo has an outside chance with heavy snowfalls in Ontario, Alpena and Green Bay. 

(posted later today)

Snow at some of those locations mentioned, up to 5 p.m., has already moved Central Illinois into second place but it's very close among the top three. I would now expect the lead to keep changing (when MQT alone gets snow, this will be a disadvantage for CI). The main question now is, how much more snow can fall before the season becomes mainly MQT additional snow? And how much will that be? 

 

(as of April 13th)

Yesterday's 4.8" of snow has moved Central Illinois back into third and the current separation at the top is only 1.8" ... with snow expected in parts of WI (MSP is not in play for the leaders, they are both accumulating errors there), and second place cmillzz holding some advantages over leader Thomp2mp at both LSE and GRB, the lead could change hands depending on what happens at a few other locations. Only last place IndTenn can now benefit from any further snow at MSP or MLI, the total there has passed the rest of our forecasts so we're all adding on new snowfall as the same amount of errors whenever it snows there. At DTW, SDF and PIA, two forecasters have yet to be passed (DAFF and IndTenn at DTW,  VPBob21 and IWXwx at SDF and Central Illinois and IndTenn at PIA). Two locations remain with seasonal snowfall below our minimum forecast, STL and APN.

[Roger Smith]

The 2.5" of snow that fell at GRB on 13th changed the lead with cmillzz back in front by a mere 0.4" ... so this has eliminated GRB as a location that can change the outcome with both cmillzz and Thomp2m now accumulating errors there at the same pace. The remaining locations where they have any leverage on each other (not both in black numbers) are:

YXU and IND _ Thomp2m can gain about 5" at YXU and 4" at IND, if that much fell, the total advantage would be 18" (one goes up 9, the other down 9). However, doubtful whether much more snow will fall at these locations. Still, a total of 0.3" would change the lead if nothing else happens in the following list. (0.3" x 2 being the total change). That much seems possible at YXU in particular.

APN and MQT both present Thomp2m with chances to move ahead. At APN, that will start to happen if 8.5" more snow falls, and at MQT, if 19.5" more falls. Up to those amounts, both are gaining error amounts. Both of these propositions seem to be about 50-50 to me. 

STL is the same story for 3.0" or more, anything less does not matter, and that's the most likely outcome. 

LSE on the other hand can improve cmillz25 after 3.7" falls, which seems quite possible. The total improvement available is 5.8" (2 x 2.9" differential). 

MKE snow will help Thomp2m from 1.2" to 3.8" (before that they both decrease error totals, after 3.8" they both gain errors). That is a 5.2" differential.

It's probably too close to call now. Central Illinois is not in a good position with MQT accumulating errors for him while everyone else still has a cushion, although if this MQT snow is balanced by GRB snow the first four inches won't count against his overall total. Stebo can still do well if there's a lot of snow at the two Ontario locations and FWA. 

You can check the tables a few posts back and see if you have any chances of overtaking the leaders. You need higher red numbers than their red numbers, or red numbers where they have black numbers, and the differential has to be half of your deficit, for you to have a chance. If you can only find a differential of 50" and you are 120" behind now, you can't catch up. 

(added Apr 14 1530z)

MSP has posted 1.1" for 13th and added 0.2" since midnight. Since forecasts call for 10-15" there, I will leave the tables for a later edit when I expect MSP seasonal total to be higher than now shown although at this point in time the total is actually 63.8" ... MSP is not a factor in the contest since all but IndTenn have already been passed so with that exception we are all gaining the same error scores at MSP now. MLI is the same story. Even if IndTenn gains the full amounts available at these two locations, their score cannot catch any other forecasters on those two alone. (has 39.3" credit in total, catch-up value is 78.6", currently 288.2" behind 14th place. IndTenn needs snowfalls of stupendous proportions at a number of other locations to catch the pack). 

Overnight snowfalls include 0.7" additional at APN (all forecasters gain that much), 0.6" at LSE, and 5.0" at GRB. As there has been no snow at MQT, Central Illinois is back in the mix now, having used his 4.5" credit shown in the table to get to within 0.3" of cmillzz and passing Thomp2m by 0.1" ... however, these two still have some snow credit at LSE which is forecast to use up their credits today, so the situation in the table will be restored fairly soon since GRB is no longer going to be in play among the three of them. At some point later today, Stebo will also make a pass at the lead since he had 6.6" of credit to use at GRB and still has 1.6" of that left, but he's in the same situation regarding LSE. 

It looks as though APN is in line to pass minimum forecast during this storm with just 4.7" left now. Most of you have enough credit at APN to keep gaining but I will start to accumulate error points there about near the end of this storm (min forecast as stated 4.7" left to go).

This storm is bringing several forecasters in the middle of the scoring table into contention, for example, madwx has a lot of snow credit to use at LSE, GRB and APN.

The table will be updated around 23z with the afternoon subtotals. At the moment I have not adjusted it for any of this new snowfall reported this morning and MSP will be passing the value shown during the day, expect that to increase also. I have edited in some estimates with the end of day changes shown in brackets. Those will be replaced with the new actual scoring totals later today. If the storm adds slight amounts at MQT this will be a differential for CI who is already gaining error points there. The next to be passed at MQT would be Stebo who has 8.4" credit left. 

Contest is very much "up in the air" until this storm ends and we get some indication from the models about any future potential snowfalls. 

(Apr 15 08z)

Using actual values for 14th (all close to earlier estimated totals for 14th) the table has changed yet again and Stebo now has a narrow lead over cmillzz and Thomp2m. It looks like Stebo can hold off both Thomp2m and cmillzz if snow falls at YYZ, YXU and FWA, also some at MKE, but Stebo is vulnerable at APN compared to Thomp2m. 

More snow appears possible on future dates so this contest is by no means settled yet. 

[Roger Smith]

Lead has changed hands twice during the day (15th), first Stebo moved in front, then cmillzz, but it's back to Thomp2m now due to added snow at APN and MKE.

Looking at these three now, the remaining locations in play would be any with red numbers for any of them, those with all black numbers are settled as far as their differentials are concerned.

These are arranged in order of Thomp2m advantage over cmillz, red numbers mean they have this much credit left to use, black means errors are going to increase with all new snow.

Location _______ Thomp2m __ cmillzz __ Stebo ____ when advantage begins ___ 

MQT ____________ 23.0 _____ 8.5 _____2.5 ______ after 8.5" more snow _____

APN _____________ 6.2 _____ 0.1 _____ 2.1 ______ after 0.1" more snow _____

STL _____________ 7.4 _____ 2.9 _____ 6.9 ______ after 2.9" more snow _____

YXU _____________ 4.3 _____ 10.7 ____ 4.3 _______from now on ____________

IND _____________ 4.1 _____ 3.9 _____ 0.9 ______ from now on ____________ 

MKE _____________ 2.1 _____0.6 _____ 4.4 ______ from now on ____________  

LSE ______________ 0.6 ____ 2.3 _____ 7.4 ______ cmillz adv from now on ___

FWA _____________ 9.5 ____ 3.7 _____10.3 ______ Stebo adv from now on ___

YYZ _____________ 3.6 ____ 1.1 ______ 3.9 ______ Stebo adv from now on ___

ORD _____________ 4.6 ____ 5.1 _____ 1.9 ______ Stebo adv from now on ___

 

So what does all this mean? 

Let's plug in some plausible additional snowfalls and find out. (Thomp2m currently leads cmillzz by 0.8" and Stebo by 2.1"

30" more at MQT means Thomp2m reduces error by 16", cmillzz increases error by 13.0" and Stebo increases error by 30.0"

__ net effect alone is Thomp2m leads cm by 29.8" and Stebo by 48.1" ... leads some others with more leeway at MQT by 20 instead of 30. these others (Madwx, slowpoke, Hawkeye) move past cmillzz and Stebo.

5" more at APN means Thomp2m reduces error by 5", cmillzz increases error by 4.9" and Stebo increases error by 0.8"

__ net effect alone is Thomp2m leads cm by 10.7" and Stebo by 7.9", Stebo moves into second ahead of cmillzz.

STL not likely to see more snow, so no estimates.

IND could see 1.0" more, this would increase Thomp2m lead by 1.8"

MKE could see the full 4.4" that Stebo has left as a credit. If this is all that happens, Stebo moves back into the lead by 0.2" (cmillzz moves to third).

LSE could see the full 2.3" that cmillzz has left as a credit. This is the only place left for cm to move back into the lead, and if this is the only thing that happens, he would lead by 1.5". This may be the first change and cm could be temporarily back in the lead as long as MQT doesn't add 8.5" more first.

The three last entries are Stebo's best chance to move back into first, if more than an inch of snow happens at each of them and nothing else happens, Stebo goes back in front.

So it's complicated, looks to me like Thomp2m has a big advantage over cmillzz because LSE is not going to be the only place to gain snow, and a considerable advantage over Stebo because MQT is more likely to gain snow than all the others combined. 

I will revisit this proposition in late April to see if there's any shift in thinking. Remember, if all are in the black, it doesn't matter what happens at those locations, all errors will increase in lock-step. 

(added later)

LSE has been slow to report confirmed amounts but from Regional totals I could see evidence for an increase of at least 0.4" today which ties Thomp2m and cmillzz for the lead, Stebo currently 2.1" behind in third. 

(later)

Thomp2m is back in front of cmillzz now, with Stebo just behind in third. The table above is already somewhat out of date especially for LSE and MKE. I may revise it tomorrow.

(16th) 8 pm

Madwx and Slowpoke have moved ahead of Stebo today, so these three are 3rd, 4th and 5th now. Once this storm is totally done end of today, I may look at their chances relative to the two leaders, but Thomp2m has opened up a larger lead now. 

Also of note, we are within 2" of reaching seasonal normal and passed contest average today. 

(17th) 8 am

New calculations based on end of day snowfalls scrambled the order a bit more, now it's Thomp2m, madwx, slowpoke, cmillzz and then Hawkeye with Stebo dropping down to sixth. 

Contest average and normal have moved up, contest average is between first and second forecasters, and Normal is between 2nd and 3rd forecasters. 

With the big storm just about done, it's possible that there won't be much further activity other than MQT so if you still have any credit there, you aren't likely to drop down the table now. 

There's also a new scoring paradigm to check, rankings for each location, and average of those rankings, which shows Slowpoke with the best average ranking (5.75) with Hawkeye at 5.9 and Thomp2m tied with madwx at 6.30. 

So now there's three ways to score the contest, all with slightly different results (total error, average percentage error, average rank order).

 

 

[Roger Smith]

The tables have been updated through April 17th, will not be updated again until around Monday 23rd as we go on a short road trip, not planning to be on line much so if it does snow across the region around Wed-Thurs these additions won't appear until next Monday. You can see from the first of the scoring tables where you still have "credit" (higher forecast than actual) which is good for your position in the contest -- if it snows there -- and those places are in red type in the first table (the one where TH is the leader). The other tables are interesting to compare, not quite the same results, although it's the same group of about half a dozen forecasters who are jostling for top spots in each table and the same group near the lower end (including the host so no worries about bias). 

Will be back to update all the tables on Monday then ... after which hopefully it will be all MQT and easier to foresee changes if any. 

Right now I am very close to the MQT total so that means it will snow like crazy there in May. 

[Roger Smith]

Tables all updated now for snow 18th and 19th, lead in the main scoring table (total error) has changed as a result (madwx now ahead of Thomp2m with slowpoke also very close in third). Hawkeye is in fourth place but leads in terms of average percentage error. 

Slowpoke has a slightly better chance than Thomp2m now to catch madwx at the final buzzer (which is whenever MQT stops snowing). This is because slowpoke has a larger credit left to use at MQT and one or two other locations, Thomp2m has few places where credit exceeds either of the other two. 

Will comment further once I get a handle on the pattern evolving, have been away for several days. 

[Roger Smith]

I think we can declare this contest final although a slight chance remains for snow at MQT into early June, and so these are the current standings (see posts that I made on March 2 and March 7, edited since, for all the details, and pretty much ignore anything I posted in April because those situations got demolished by the freak heavy snowfall event).

There were three ways that I scored the contest. Going by the main contest scoring function, the winner is madwx with a current total error of 159.0" and 12.7" left to use up at MQT (in other words, madwx stays ahead by at least current margin through the first 12.7" of snow that might fall at MQT later, and that seems higher than what will probably happen there). The second and third place error totals belong to 1986-2015 "normal" (average -- 164.6") and contest average (166.0") values. It should be noted that 1986-2015 normal has a mathematical chance to finish lower than madwx if more than 16.2" of snow falls at MQT because that non-forecast has 24.5" inches to use up.  But ignoring those two, actual second place goes to Thomp2mp with error of 171.9" -- Thomp2mp cannot catch madwx using any outcome at MQT since there is only 7.2" left in their account and so they would start to fall further behind after that 7.2". Third place at the moment belongs to Slowpoke (winner last two contests), and their total error is 176.5" with 24.7" left to use at MQT. If that much snow actually falls at MQT, slowpoke could win since their total would then fall to 151.8" while madwx would go to 158.3" -- the break-even point is 20.8" at MQT. However, 1986-2015 normal would still be lower at any point than slowpoke who can only get 0.4" closer between 24.5" and 24.7" snowfall at MQT. Fourth place goes to hawkeye with 185.1" (29.7" credit at MQT) which means that hawkeye could also pass Thomp2m and madwx if most of that credit were to materialize at MQT, and can even catch slowpoke as the final 5.0" of the credit if used would result in a gain of 10.0" while starting from 8.4" behind. There would need to be 28.9" of snow at MQT to create a tie then anything more would move hawkeye ahead of slowpoke.

You can see the rest of the standings in the March 7th post. Cmillzz, Stebo and VPBob21 are the next in line and none of them have any MQT leverage, in fact cmillzz and Stebo are already gaining errors with any new MQT snow. Then we find Michigansnowfreak with a large enough MQT credit to move from 8th place to almost any position ahead except first or second, if by some miracle 48" of snow falls at MQT. That would not be quite enough to overtake hawkeye or slowpoke on the rebound. However, I don't foresee MQT getting more than perhaps 3 to 5 inches more snow if even that, so all of this discussion is probably academic.

As for any snow at other locations, seems very unlikely but in any case GRB and MSP have no credits left for anyone except IndTenn who is too far down the standings to benefit much even if all that available credit were to materialize. I did mention scoring the contest without YXU since we had to use an alternative location but most of the errors at YXU are small anyway and I don't distrust the numbers we were able to obtain from a nearby location.

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

A second metric that I used to score the contest was average percentage error of the 20 forecasts. This is not quite the same as total error because if you made a lot of good forecasts for heavy snowfall locations and not so good at light snowfall locations, you might be down the list comparatively speaking. The best performance (so far) in average percentage error (table two in the March 7th post) belongs to hawkeye (19.1%) then madwx at 21.9%. Next is slowpoke (22.2%, ahead of 1986-2015 normal at 22.3%), then contest average at 22.6%, fourth place in the actual contest for cmillzz at 24.0% then Thomp2mp at 24.2% and DAFF at 25.2%.

Then a third way to look at the results was to add up rankings for each location, as in golf, low score wins. That method shows hawkeye and slowpoke tied at an average rank of 5.65 and madwx close behind at 5.85. 

Finally, in terms of closest forecasts to actuals, DAFF had four out of twenty and madwx had three. You can see the rest of the list in that same March 7th post. There were some ties involved so the numbers could be shown differently if we gave decimal values for ties (0.5 or even 0.33). The only one of these subject to change would be MQT. This is currently my lone closest to the pin as I have 0.7" left to give. The closest forecast above current value would be Thomp2mp who is 7.2" above, so the lead changes hands when MQT gets to 179.3" which is 4.0" above where it stands now. I wouldn't be too surprised if that happened. 

I will keep an eye on freak late snowfalls if any and declare the contest entirely done on June 30th. But I wouldn't expect much to change now. 

 

 

[madwx]

Can't believe I actually won of the metrics.  Thought I didn't stand a chance at the beginning of April.  I got lucky with a couple big snowfalls in the middle of the month. 

Excited to do this again next year!

[Roger Smith]

To be more precise, you will likely win the contest, the other two metrics were not official scoring but just two more ways of looking at performance. 

[thomp2mp]

YAY I GOT THE SILVER MEDAL

[Roger Smith]

MQT continues to have a snow-free May, the normal amount cited in the daily climate report for 21st is 1.5" and the normal amount cited last May 31 was 1.5" so I surmise that it normally does not snow after this date. The normal June snowfall is cited as zero, I imagine there have been small amounts on a few occasions in the historical record. The current GFS output shows no strong chances for snow in the next sixteen days. So I think we can wrap this contest up.

FINAL RESULTS declared. Drive safely. Don't forget to play again next late October into early November. 

[madwx]

Wooooo!!!  Can't wait to play again next year