2018-2019 Great Lakes / Ohio Valley snowfall contest

[Roger Smith]

STL has now passed four forecasts and PIA just overtook the tied lowest two forecasts. MLI is quite close to the lowest forecast there now also. Most other stations are not very close to the minimum forecast yet. (new snowfall amounts are edited into a post that was entered on Dec 17th). 

[Roger Smith]

After the widespread snowfalls of the past few days, the contest grid has now passed the 40% mark (table is back in the thread, first posted Dec 17th and edited up to date). STL has passed its average for 1986-2015. MQT has roughly one quarter of all the snow recorded and is a little over half of its seasonal average at 105" now. Plenty of snowfall opportunities appear on the charts so we should at least keep pace with normal snowfall for a while yet. I think that 40% on January 21st is probably a touch below average pace but not much, as February and March contribute at least 40% of a normal winter's snowfall and April perhaps 5%, so we should be at about 55% at the end of this month to be on a normal pace. That would only take 150" of snow over the 20-station grid or an average of 7.5" per location in the next ten days. The location doing worst so far is YXU sub Tillsonburg although they have yet to report the second half of the snowstorm and could go up from current 18% of normal. MSP and LSE are also in a snow drought relatively speaking in the low 20's for percentages. The highest performers behind STL (now at 103%) are MLI and PIA (both at 93%). That is largely because of their totals in November. (MLI took over the lead on Jan 23rd).

Forecasts are generally still ahead of actuals, with a few exceptions for STL, MLI and PIA. But these are some of the error totals you now have locked in from those locations (updated for Jan 22nd snow at MLI, none reported at STL or PIA) ... this has changed as of Jan 23rd, see the scoring table two posts down for updated info as MLI added 5.9" and PIA 0.1" bringing in contest normal (MLI) and adding to all of these totals. The table below was valid after Jan 22nd but won't be updated.

dmc76 _____ 7.7" __ (MLI and PIA, still 6.0" to give at STL)

Jackstraw ___6.2" __ (PIA and STL)

Stebo ______ 5.7" __ (PIA and MLI, almost caught at STL with 1.0" left to give)

Roger Smith _4.5" __ (STL only)

madwx _____ 2.8" __ (STL only)

DAFF _______2.2" __ (STL only)

DonSutherland 1.2" __ (MLI only)

and the rest of the field are so far not into red numbers on any of these. ... The entire field have amounts left to use up everywhere else. 

 

 

[Roger Smith]

Snowfalls to date -- to May 9, 2019 

-- this post will be updated whenever new snow is reported ... 

 

LOCATION _______ Oct __Nov __Dec __Jan _ Feb _Mar _Apr+May Contest total __ % Normal (full season)

APN ______________ 0.2__11.2 __11.5__33.6 __24.7__ 8.1 __10.1 ____ 99.4 _______ 124

ORD ______________ Tr __12.7 __ 1.4 __18.2 __ 9.0 __ 0.3 __ 7.9 ____49.5 _______ 130

CLE _______________0.0 __ 4.4 __ 2.8 __16.0 __ 5.4 __ 6.6 __ 0.0 ____35.2 ________ 52

CMH ______________0.0 __ 2.1 __ 0.4 __11.3__11.6 __ 2.0 __ 0.0 ____ 27.4 ________ 96

DTW ______________ Tr __ 6.7 __ 0.5 __14.2 __ 7.3 __ 2.6 __ 0.0 ____ 31.3 ________ 70

FWA ______________0.0 __ 2.1 __ 0.1 __14.3 __ 5.4 __ 2.7 __ 0.1 ____ 24.7 ________ 73

GRR ______________ Tr __14.4 __ 3.2 __30.5__ 20.4 __ 7.9 __ 4.9 ____ 81.3 _______105

GRB ______________ Tr __ 3.8 __ 8.1 __22.7 __ 28.4 __ 6.5 __ 4.5 ____ 74.0 _______137

IND _______________Tr __ 0.4 __ 0.5 __ 11.7 __ 4.5 __ 2.3 __ 0.0 ____ 19.4 ________ 75

LSE ______________ 0.1 __ 1.9 __ 4.0 __15.1__ 31.1 __ 5.9 __ 6.2 ____ 64.3 _______ 139

YXU ______________ 0.0 __ 3.2 __ 2.0 __16.0 __ 9.4 __ 2.8 __ 0.6 ____ 34.0 ______ 45

SDF ______________ 0.0 __ 0.5 __ Tr ___ 7.5 __ 0.6 __ 2.6 __ 0.0 ____ 11.2 _________81

MQT ______________4.2 _ 37.7 _ 39.0 __34.1 _ 89.9 __ 8.2 __ 8.6+5.8_227.5 ____ 114

MKE ______________ Tr __ 6.6 __ 1.5 __20.2 __18.7 __ 1.6 __ 7.5 ____ 56.1 ________ 114

MSP ______________0.3 __ 4.0 __ 6.7 __ 6.8 __39.0 __10.5 __ 9.8 ____ 77.1 _______ 151

MLI ______________ 0.2 _ 18.4 __ 1.1 __30.2 __ 8.8 __ 2.1 __ 0.0 ____ 60.8 ________ 181

PAH ______________0.0 __ 2.3 __ Tr ___ 3.2 __ 0.3 __ 0.3 __ 0.0 _____ 6.1 _________ 66

PIA _______________Tr __ 9.8 __ 0.4 __17.9 __ 2.5 __ 0.8 __ 5.0 ____ 36.4 _________146

STL ______________0.0 __ 5.2 __ 1.7 __12.5 __ 2.4 __ 2.4 __ 0.0 ____ 24.2 _________132

YYZ ______________1.0 __ 6.2 __ 2.2 __24.9 __17.3 __ 3.3 __0.2 ____ 55.1 _________130

________ _______ _______ _______ _______ _______ _______ _______ _______ _______

Total _____________________________________________________ 1094.6 _____ 108.0%

_____________________________________________________________________________

 

_ All locations updated to May 2nd. ORD had 2.5" MKE 1.7" and LSE 0.3" on Apr 27th, YYZ 0.1" on 29th and MQT 5.4" on May 1st and 0.4" May 7-9. 

... Canadian locations are converted to inches from cms. 

... LSE and GRB became 6th and 7th locations to pass contest average on 20th. MSP is now the 8th and MQT the 9th station to pass 1986-2015 contest averages.

... APN surpassed seasonal average but the addition came from a revised November total. That makes ten stations above contest average now.

... MKE edged past contest normal on April 10th. GRR reached contest normal on April 11th and passed it on the 14th.

... CMH is within 4% of seasonal average but is running out of time to increase.

... The least snow relative to normal has fallen at YXU (45%) and CLE (52%).

... The remaining five locations (DTW, FWA, IND, SDF, PAH) are in the 66 to 81 per cent range. 

TOTAL SNOWFALL PASSED THE CONTEST NORMAL VALUE ON MARCH 30th. 

 

 

[Roger Smith]

Table of Departures

This table is based on the earlier table of forecasts compared to the ongoing snowfall to date table above. Entries in this table in black are amounts left to use up. Entries in red are already accumulating errors as the station total has passed the forecast value. As of Feb 27th, the table of departures is placed in contest order. The table of forecasts is repeated below this table of departures for easier reference. The total contest snowfall has now passed the three lowest forecasts of total snowfall.

Updated through May 9th.

 

FORECASTER ________ APN_ORD_CLE_CMH_DTW_FWA_GRR_GRB_IND_ LSE__YXU_ SDF__MQT _MKE_MSP_MLI__PAH_PIA_STL_ YYZ __TOTAL_rank

 

Roger Smith __________8.9_14.5_44.8__0.4_13.7__5.3 _ 8.7__4.0__7.6_ 9.3__53.0__0.7_27.0_10.6_11.1_21.8__5.9__6.4__ 9.7__0.1___263.5 __ 1

vpbob21 ____________ 23.0__5.4_14.3__7.3_11.9_ 8.5_15.4_27.0_12.0_25.5__36.1_13.4_ 11.5_11.4_24.2_15.7_ 10.8_ 0.1__2.4__3.1___278.8 __ 2

slowpoke ___________ 27.4__7.5_21.8 _1.6__6.7__5.3_13.3_25.0__5.6_22.3__32.0__0.8__39.5__9.1_30.1_20.8__ 5.9__7.4__2.2__7.1___291.4 __ 3

Mississauga Snow _____ 3.4_10.5_ 5.8__5.4_18.7__6.3_11.3_29.0__2.4_29.3__46.0_ 0.2__27.5_22.1_22.1_20.8__8.9_10.4__5.8__ 8.1___294.4 __ 4

cyclone77 ____________9.4__5.5_23.8__1.6__ 9.7 _1.3__8.3_27.0__4.6_22.3__44.0__2.2__52.5_13.1_30.1_14.8_ 4.9__5.4_ 0.2_14.1___294.8 __ 5

___ Contest normal ___ 19.0_11.4_32.0__1.0_13.5_ 9.3__4.2_19.8__6.5_18.2__41.7__2.7__27.7__6.9_26.5 _27.3__3.1_11.5__5.8_12.6___300.7 _ (6)

___ Contest median ___22.4__7.5_30.9__0.1_13.7_ 8.5 __6.3_25.0__6.8_23.7__36.1__2.8__39.5_13.1_29.4_21.8__8.9__7.4__2.2__5.9___312.0*_ (6)

DAFF _______________24.4_12.5_14.8__1.4_ 7.7__5.3__ 8.3_23.0 _0.6_19.3__31.0__0.8__55.5_15.1_29.1_22.8__7.9__7.4__7.2_18.1___312.2 __ 6

madwx ______________14.9__4.1_30.9_ 0.1_20.8_11.4__6.2_24.2__6.8_23.0__31.1__3.0__46.3_13.8_28.8_20.1__5.3_10.1__8.0__3.8___312.7 __ 7

Jackstraw ___________ 27.4_19.5_55.8__0.4__ 9.7_11.3__6.3_25.0_11.6_33.3__30.0__2.8__15.5_21.1_47.1_26.8_15.9_16.4__8.2__9.1___393.2 __ 8

Stebo ______________ 34.4_14.5_47.8__5.4_ 16.7_12.3__4.3_29.0__5.6_29.3__26.0__6.8__42.5_16.1_42.1_29.8_11.9_16.4__4.2__5.1___400.2 __ 9

dmc76 ______________22.4__3.7_47.2_11.5_27.1_23.1__0.7_ 18.4_ 9.4 _23.7__70.4__6.5__38.5__2.1_29.4_33.8_14.9_14.4__0.8_ 8.1__ 406.1 __10

DonSutherland.1 ______17.7__0.5_48.7_17.6_25.2_16.8__1.1_ 25.1_23.3_26.1__75.4_11.8__82.3__9.0_36.7_28.5_12.4__1.4__1.0__5.9 __470.5 __11

 

note: bold entries are now current location winners in the contest. 

All forecasts passed by actual at MQT, MLI, LSE, MSP, GRB and on April 14, APN, PIA and MKE.

... (at MQT and PIA vpbob21 had high forecast, at MLI slowpoke had high forecast, at MKE the winner is now dmc76, at APN high forecast and the win goes to Mississauga Snow, and at the other three GRB MSP and LSE, Roger Smith). As a result, those eight locations now have a confirmed winner. ORD also has a confirmed winner (DonS) although he has 0.5" more predicted than actual. 

YYZ has been passed by all but one now, dmc76 still has 8.1" but needs 4.0" more to take that one away from current low departure 0.1" (RS)

STL has passed all but four forecasts. 

These locations have not yet passed any forecast (so minimum forecast could win) ... CLE, DTW, FWA, YXU, PAH. ... IND has passed one forecast so far.

_______________________________________________________________________________

*contest median total is sum of twenty median forecasts, the median total is 984.0

... ranks of contest normal and median do not change ranks of forecasters in contest, and they are independent of each other's rank. 



Tiebreakers (same order as table of entries) -- winners in green when settled

-- amounts shown here are forecasts not errors or departures

 

dmc76 ______________ Dec ORD 10.2" __ Jan IND 11.0" ___ Feb DTW 12.0"

Roger Smith __________Dec ORD 13.5" __ Jan IND 10.0" ___ Feb DTW 12.5"

DonSutherland.1 ______ Dec ORD 11.7" __ Jan IND 13.7" ___ Feb DTW 14.2"

vpbob21 _____________ Dec ORD 3.6" ___Jan IND 9.4" ____ Feb DTW 13.1"

madwx ______________ Dec ORD 9.2" ___ Jan IND 8.4" ____Feb DTW 18.3"

Mississauga Snow ______ Dec ORD 4.1" ___ Jan IND 11.0"___ Feb DTW 16.0" ___ closest Dec ORD

cyclone77 ____________ Dec ORD 6.7" ___ Jan IND 9.6" ____ Feb DTW 13.4"

Jackstraw ____________ Dec ORD 4.5" ___ Jan IND 12.0"___ Feb DTW 10.5" ___ closest Jan IND

slowpoke _____________ Dec ORD 6.0" ___ Jan IND 7.0" ___ Feb DTW 8.0" ____ closest Feb DTW

Stebo _______________ Dec ORD 6.7" ___ Jan IND 10.2" ___Feb DTW 15.2"

DAFF________________ Dec ORD 10.7" ___Jan IND 13.8" ___Feb DTW 14.9"

(mean of above entries) _________7.9" __________ 10.6" ___________ 13.5"

actual snowfall _________________1.4" __________11.7" ____________ 7.3" 

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

Table of forecasts (repeated from earlier post)

FORECASTER _________ APN_ORD_CLE_CMH_DTW__FWA_GRR_GRB_IND_LSE__YXU_SDF__MQT_MKE_MSP_MLI__PAH_PIA__STL_YYZ __ TOTAL

 

dmc76 _______________77.0_45.8_82.4_38.9_58.4_47.8_82.0_55.6_28.8_40.6_104.4_17.7__189.0_54.0_47.7_27.0_21.0_22.0_25.0_63.2__1128.3

Roger Smith __________ 90.5_35.0_80.0_27.0_45.0_30.0_90.0_70.0_27.0_55.0__87.0_10.5__200.5_45.5_66.0_39.0_12.0_30.0_14.5_55.0__1109.5

DonSutherland.1 _______81.7_50.0_83.9_45.0_56.5_41.5_82.4_48.9_42.7_38.2_109.4_23.0__145.2_47.1_40.4_32.3_18.5_35.0_25.2_49.2__1096.1

vpbob21 _____________ 76.4_44.1_49.5_34.7_43.2_33.2_65.9_47.0_31.4_38.8__70.1_24.6__216.0_44.7_53.1_45.1_16.9_36.3_26.6_52.0__1049.6

 

___ Contest normal ____ 80.4_38.1_67.2_28.4_44.8_34.0_77.1_54.2_25.9_46.1__75.7_13.9__199.8_49.2_50.6_33.5__9.2_24.9_18.4_42.5___1013.9

 

madwx _______________84.5_45.4_66.1_27.3_52.1_36.1_75.1_49.8_26.2_41.3__65.1_14.2__181.2_42.3_48.3_40.7_11.4_26.3_16.2_51.3__1000.9

 

___ Contest median ____77.0_42.0_66.1_27.3_45.0_33.2_75.0_49.0_26.2_40.6__70.1_14.0__188.0_43.0_47.7_39.0_15.0_29.0_22.0_49.2___998.4*

 

Mississauga Snow _____ 96.0_39.0_41.0_22.0_50.0_31.0_70.0_45.0_17.0_35.0__80.0_11.0__200.0_34.0_55.0_40.0_15.0_26.0_30.0_47.0___984.0

cyclone77 ____________ 90.0_44.0_59.0_29.0_41.0_26.0_73.0_47.0_24.0_42.0__78.0__9.0__175.0_43.0_47.0_46.0_11.0_31.0_24.0_41.0___980.0

Jackstraw ____________ 72.0_30.0_91.0_27.0_41.0_36.0_75.0_49.0_31.0_31.0__64.0_14.0__212.0_35.0_30.0_34.0_22.0_20.0_16.0_46.0___976.0

slowpoke _____________72.0_42.0_57.0_29.0_38.0_30.0_68.0_49.0_25.0_42.0__66.0_12.0__188.0_47.0_47.0_40.0_12.0_29.0_22.0_48.0___963.0

Stebo ________________65.0_35.0_83.0_22.0_48.0_37.0_77.0_45.0_25.0_35.0__60.0_18.0__185.0_40.0_35.0_31.0_18.0_20.0_20.0_50.0___949.0

DAFF ________________75.0_37.0_50.0_26.0_39.0_30.0_73.0_51.0_20.0_45.0__65.0_12.0__172.0_41.0_48.0_38.0_14.0_29.0_17.0_37.0___919.0

______________________________________

*contest median total is sum of twenty median forecasts, the median total is 984.0



Tiebreakers (same order as table of entries)

 

dmc76 ______________ Dec ORD 10.2" __ Jan IND 11.0" ___ Feb DTW 12.0"

Roger Smith __________Dec ORD 13.5" __ Jan IND 10.0" ___ Feb DTW 12.5"

DonSutherland.1 ______ Dec ORD 11.7" __ Jan IND 13.7" ___ Feb DTW 14.2"

vpbob21 _____________ Dec ORD 3.6" ___Jan IND 9.4" ____ Feb DTW 13.1"

madwx ______________ Dec ORD 9.2" ___ Jan IND 8.4" ____Feb DTW 18.3"

Mississauga Snow ______ Dec ORD 4.1" ___ Jan IND 11.0"___ Feb DTW 16.0"

cyclone77 ____________ Dec ORD 6.7" ___ Jan IND 9.6" ___ Feb DTW 13.4"

Jackstraw ____________ Dec ORD 4.5" ___ Jan IND 12.0"___Feb DTW 10.5"

slowpoke _____________ Dec ORD 6.0" ___ Jan IND 7.0" ___ Feb DTW 8.0"

Stebo _______________ Dec ORD 6.7" ___ Jan IND 10.2" ___Feb DTW 15.2"

DAFF________________ Dec ORD 10.7" ___Jan IND 13.8" ___Feb DTW 14.9"

(mean of above entries) _________7.9" __________ 10.6" ___________ 13.5"

 

 

[Roger Smith]

About 100" of snow was added to the contest grid by the recent storm. A fourth location, ORD, has now passed at least one forecaster (MLI, PIA and STL have passed quite a few, in fact MLI has passed all but cyclone77 who has 0.6" left before he gets passed as well (but will have best result for MLI with any new snow). Cyclone77 is the only forecaster left without any red numbers, or stations with more snow than forecast. 

The order of total departures is still fairly similar to the reverse order of the total snowfall forecasts, only Stebo and jackstraw have moved even one position down in relative terms by having larger total departures than one or more forecasters with higher seasonal totals. Most of us have plenty left to use up at lake effect locations GRR, CLE and YXU. The lake effect portion of the total snowfall at those places is fairly small so far, so they are not running much ahead of the pack. We are probably right around where a normal season would be, 55% of total snowfall by the end of January seems about average. 

When we reach March 15th or so, and the season is approaching a conclusion, I will reorder the table above but for now it's almost in order from bottom to top as far as the contest goes. 

[Roger Smith]

IND finished with 11.7" in January, Jackstraw at 12.0" wins this tiebreaker. 

We just passed 600" total and are very close now to 60% of contest normal snow. 

Ranks of total departures are now shown, the table will be placed in order when it gets closer to the final countdown. 

[Roger Smith]

 

Table of departures -- as percentages of actual snowfall to date

 

_ to May 2nd _ this table was same as earlier contest table which has been placed in rank order, so has now been changed to this new format

_ this is an alternate scoring system, the first table above is the official contest scoring table.

Numbers represent differences of forecast and actual as % of actual. For example, if a station had 200" snow then a forecast of 180" would read -10 and a forecast of  220" would read +10.

The absolute values of these differentials is then averaged to establish rankings. Red numbers can increase in absolute value, black numbers can decrease.

(To calculate what percentage of actual snow your forecast is, simply add 100 to the number. For example, -10 means 90%, 2 means 102% of actual snowfall). All numbers can only decrease over time but the absolute value of red numbers can only increase.

 

FORECASTER _______APN.ORD.CLE.CMH.DTW.FWA.GRR.GRB.IND.LSE.YXU. SDF.MQT.MKE.MSP. MLI..PAH.PIA. STL. YYZ ___ AVG _ rank

 

slowpoke ____________-28 _-15_62 _ 6_ 21 _ 21 _-16_ -34 _ 29 _-35 _ 94 __ 7 _-17 _-16 _ -39 _-34 _ 97_-20 __-9 _-13 ___ 29.2 __ 1

cyclone77 ____________ -9 _-11_68 _ 6 _ 31 __5 _-10_ -36 _ 24_ -35_ 129_ -20 _-23 _-23 _-39 _-24_ 80 _-15 _ -1 _-26 ___ 29.7 __ 2

DAFF _______________-25 _-25_ 42_ -5 _25 _ 21 _-10_ -31 __ 3 _-30 _ 91 __ 7 _-24 _-27 _-38 _-38 _130_-20 _-30 _-33 ___ 30.2 __ 3

Roger Smith __________-9_-29 _127 _-1 _ 44 _ 21 _11 _ -5 _ 39 _-14_156 __-6 _-12_ -19 _-14 _-36 _ 97 _-18 _-40 __0 ___ 32.0 __ 4

___ Contest normal ___-19 _ -23_ 91__4 _ 43 _ 38 _-5 _-27 _ 34 _-28 _123 _ 24 _-12 _-12 _-34 _-45 _ 51_ -32 _-24 _-23 ___ 32.2 _ (4)

Mississauga Snow ______-3_-21 _16 _-20_ 60 _ 26 _-14_-39 _-12 _-46 _135 _ -2 _-12_ -39_ -29 _-34_146 _-29 _ 24 _-15 ___ 34.0 __ 5

madwx ______________-15 _-8 _ 88 _ 0 _ 66 _ 46 _ -8 _-33 _ 35 _ -36 _ 91 _ 27 _-20_ -25 _-38_-33 _ 87 _-28 _ -33 _ -7 ___ 35.3 __ 6

___ Contest median ___ -23 _-15_ 88 _ 0 _ 44 _ 34 _-8 _-34 _ 35 _-37 _106 _ 25 _ -17 _-23 _-37 _-36 _146_-20__ -9 _-11 ___35.9 _ (7)

vpbob21 ____________-23 _-11 _ 41 _ 27 _ 38 _ 34_-19_-36 _ 62 _-40 _106 _120 __-5 _-20 _-31 _-26 _177 _ 0 _ 10 _ -6 ___ 40.5 __ 7

Stebo ______________-35 _-29_136_-20_ 53 _ 50 _ -5 _-39 _ 29 _ -46 _ 76 _ 61 _ -19 _-29 _-55 _-49 _195 _-45_-17_ -9 ___ 46.9 __ 8

Jackstraw ___________-28 _-39_159 _ -1 _31 _ 46_ -8_ -34 _ 60 _-52 _ 88 _ 25 __ -7 _ -38 _-61_ -44 _261 _-45 _-34_-17___ 49.8 __ 9

dmc76 _____________ -23 _-7 _134 _42 _ 87_ 94 __ 1 _-25_ 48 _-37 _207 _ 58 _ -17 _-4 _ -38 _ -56 _244 _-40 __ 3 _15 ___58.2__10

DonSutherland.1 ______-18 _ 1 _138 _64 _ 81_ 68 __1 _-34_120_-41 _222 _105 _ -36 _-16_ -48 _ -47 _203 _ -4 __ 4 _-11___63.1 __11

___________________________________________________________

... ranks of contest normal and median do not change ranks of forecasters in contest, and they are independent of each other's rank. 

... confirmed best forecasts are in bold type. 



[Roger Smith]

As you may have noticed, I posted the contest standings in order recently (previous post) and it looks like we're entering a volatile phase of the contest now. 

The numbers of stations with at least one forecast passed by reality of 2018-19 snowfall has increased with this past event to ten out of the twenty. For quite a while, we were holding at three (MLI, PIA, STL). Then ORD and MSP joined the list a couple of events back. More recent additions are APN, LSE, MQT, GRB and YYZ. 

We all went quite high on CLE and YXU, those two snow belt areas have been a general dud this winter, while the Lake Michigan snow belt (GRR) has done a bit better  (both from lake effect and synoptic scale snow) and our forecasts are already not way too high there on average. Our current leader DAFF has been approaching the red zone (accumulating errors) at quite a few locations recently and may come under some pressure from slowpoke and cyclone77 who are both about 30 inches behind in total departures. Cyclone77 has a big advantage for APN where his forecast can absorb another 25". Anyone further back is going to need to see large additions to various locations where they have some excess capacity. 

When you look at the table, and compare to any other forecaster, you only have a potential advantage for any stations where your black numbers are larger than theirs. If you have similar numbers, you have no advantage. If you have red numbers of any size, same story (you'll both be adding errors at the same pace). A black minus red differential can be black-doubled to establish advantage (example, I am in the red by 0.5 at ORD, DonS has 14.5 left in his account, so his advantage is 29, as for the next 14.5" he will be decreasing his error total and I will be increasing on mine. ... the size of my red number plays no role, just the black number doubled is in play in those situations.)

I think anyone can still win this depending on where the rest of the snow actually falls. However, you don't want to be accumulating at MQT which has a higher ratio of percentage errors to total errors. Later in the contest I will calculate other scoring metrics such as ranks for station errors and average percentage error. Last year we found these were somewhat different from the principal method of total accumulated error. 

 

Good luck and remember, April can be very snowy as we saw a couple of years ago (or was it three, time flies). 

[Roger Smith]

ORD is now our fourth location to surpass the contest normal value. GRB missed out and remains just below, as does YYZ until perhaps data for 18th ... confirmed, YYZ is the fifth location to pass contest normal.

LSE has increased its total to the point where most forecasters are now accumulating errors there, although still not at the seasonal average. MKE is closing in on that range. In general there has now been about 10% more snowfall than normal across the grid, considering what stage of the season normally shows 82% (closer to Feb 24th than 17th). 

Nobody has much to worry about (yet) at YXU or CLE. From there towards DTW and FWA, even IND, snowfall totals have been rather low this winter. Relative to normal, it picks up further west and even south into Kentucky. 

 

 

[Roger Smith]

LSE and GRB have now become 6th and 7th locations to surpass contest normal snowfalls. MSP will join them after its next snowfall event. 

Good locations to have surplus snow in your forecast would seem to be APN, GRB, MSP, LSE, MQT and YYZ. The contest is getting very close now in terms of projected total errors. If I assume that each location will add 20% to its current total, then the order would be considerably altered from the current contest order. These are the total "locked in" errors that you have to date (where your forecast already exceeded by actual snow, sum of red numbers in the tables). But it all depends on where you have the snow available to use up, if most of it is CLE and YXU, then everyone else has similar totals there. 

This list shows the total locked in errors, amounts left to use up, and total amount left to use up at APN+MQT+GRB+LSE+YYZ+MSP which are still in play for some. 

vpbob21 ___ 31.2" (has 210.8" left at 16 locations). ... 47.7" at the six (APN etc)

Roger Smith_32.2" (has 271.7" left at 17 locations). ... 84.2" at the six (APN etc)

slowpoke ___36.6" (has 129.6" left at 14 locations). ... 10.2" at two of the six (APN etc)

cyclone77 __ 39.9" (has 151.9" left at 13 locations). ... 15.3" at APN only

madwx ____ 43.2" (has 174.1" left at 12 locations). ... 17.8" at two of the six

dmc76 _____48.4" (has 306.3" left at 16 locations). ... 29.1" mostly at YYZ

DAFF ______ 58.8" (has 107.8" left at 10 locations). ... 0.3" at APN

Miss. snow __59.9" (has 174.3" left at 13 locations). ... 48.2" at four of the six. 

DonSutherland 88.5" (has 314.6" left at 15 locations). ... 12.9" at APN and YYZ

Stebo ______ 96.9" (has 177.9" left at 10 locations). ... 9.2" at MQT and YYZ

jackstraw __105.7" (has 211.7" left at 11 locations). ... 2.7" at YYZ

______________________________________________________________

It seems very likely that all (perhaps most for YYZ) of those six-location totals will come off the total errors so with my total of 84.2" coming off, it means you will be trading at double the differential of your total (applied to current differentials) so that only looks good for vpbob21 and Mississauga Snow who will lose about 80" but are ahead by more than that now. I can catch just about anyone else ahead of me on these stations. However, this does not factor in when MKE, DTW or perhaps CMH flip to accumulating errors, I don't have anything left at MKE. So there's one scenario, if it snows heavily across the north and not much in places where I'm behind, such as ORD, then I have a chance. You can look over the errors and see what's your best bet for a high finish. I don't think there's any point in looking at CLE or YXU, they aren't going to reach anybody's forecasts (so your errors there in relative terms will not change). MLI is used up for all forecasters and PIA soon will be (vpbob21 has a bit more to use there, an advantage of about 12").

 

 

[Roger Smith]

The above discussion about the impact of snow at the six locations mentioned is now about two-thirds used up by the snow that just fell at MQT, MSP and LSE (with 2" estimated for APN, a total of 30" at these locations). Later edit, 2.5" also added at GRB, so now a total of 32.5" most of which is confirmed. Most who had some surplus there have now used that up and are accumulating. I still have small amounts left to use up at MSP and LSE, and others have some to use at MQT. The new standings show that the race is tightening up near the middle of the pack. 

[Roger Smith]

With the updated table of departures (back in the thread) you can see that it will be quite a finish to this contest. I have used up just about all my previously mentioned advantage snowfall but have about 7" left at MSP+GRB+YYZ and some at APN. However, other contenders have more stored up at other locations. So it will all come down to where it snows from now to end of the contest, probably anyone in the top six at present has a chance. 

If you're trying to work out your chances, if you're in the red and your competitors are also in the red, you can ignore that location, you will gain errors equally from now on. If you're black and they're red, you have twice that differential advantage (but it has to snow to get paid off). If you're both in the black, your advantage is the differential but it depends on whether it snows or not to claim it. YXU is probably not a factor, it likely won't reach minimum forecast (so all errors will change equally going forward). CLE, PAH and SDF may be similar but the lowest forecasts there are in some danger of losing an advantage if there are some snowfalls. For PAH and SDF the season is almost over though. (edit Mar 9th _ SDF has now passed two lowest forecasts).

[Roger Smith]

At this point, it may look like the host is favored to win, but vpbob21 currently in third place has a clear path to first if it snows at ORD, PIA and/or STL later this month. The remaining amounts that he has at MQT would move him closer but in addition to that (which seems likely, only 4") there needs to be some snow at ORD, STL and/or PIA to close the deal. I don't see any way that any other forecaster can pass vpbob21 or myself, and slowpoke has run out of places with any advantage no matter how much snow falls, with three exceptions that don't add up to the required amount to change the order of finish. Also, vpbob21 has an advantage if any major snowfall were to hit the Ohio valley with more left at those locations than any of the other leaders. 

Even if vpbob21 can move ahead on all of those outcomes, he then has to avoid getting a return blow from added snow at GRR where I've still got 14.5" left to use. So it would need to be rather specific storm tracks that do well across the south and not at GRR. I've also got a bit left at APN and YYZ. 

Contest could therefore remain undecided until well into April. 

[Roger Smith]

Small amounts of snow on March 30th have put the contest over the 1986-2015 average of 1013.9" by a fraction of a percentage (1014.4"). 

The contest still depends on whether more snow falls at stations such as PIA, STL, APN and YYZ where the various leaders are in different positions relative to snowfall they can absorb with lower departures. 

See the tables above, the percentage table just above this is an alternate scoring system, the main contest scoring has the date January 23rd and has been edited since then. 

We won't know until the end of April who wins the contest. 

[Roger Smith]

Contest updated through April 11th, still some chance for vpbob21 to overtake the two ahead of him if enough snow were to fall at PIA, STL, ORD and possibly in the Ohio valley, without too much at GRR to negate the gains made. (see updates of scoring tables posted above, the first one back is an alternate scoring system using percentages, the main contest scoring has the date Feb 2 attached). 

[Jackstraw]

If I end up beating a meteorologist that lets just say mentioned bird farts being added to atmospheric models, do I get to be a Pro-forecaster for a day?  

[Roger Smith]

You never know ... here's the result of the snowfall yesterday and into early hours of today, now updated into tables above. 

For the main scoring table, vpbob21 moved into second but with no snow at STL or Ohio valley stations where he has an advantage, he is still 15.3" out of the lead, which means that 7.7" would have to fall at the right locations (plus twice whatever fell at GRR) so if he got the 2.4" to fall at STL still to his advantage plus any amount at CMH and SDF  plus amounts greater than my residuals at IND (7.6") and PAH (5.9"), any of those results could move him into first, problem being that no snow is apparent on forecast charts to end of April in any of those places. This is the only route to get past my lead that I can see from the numbers. Anywhere else, we are either both in the red (most of the northern stations) or I have more left to give (CLE, DTW, YXU) so no help is available there. 

Slowpoke has fallen to third place and has no mathematical route back to second or first (other than an implausible very heavy snowfall at CLE to edge into second). I can't see any realistic paths for any other forecasters to move much further ahead than where they are now. So the contest is almost settled but there remains a slight chance of the needed results as discussed above. 

The percentage tables are also probably near their final outcomes. The order of finish there is somewhat different, mainly because the stations that underperformed like YXU, CLE and PAH have large enough percentage errors to swing the results, so the forecasters who went low there have a boost compared to their average departures. 

[Roger Smith]

This graphic shows how actual snowfall compared to the eleven forecasts made for the contest.

It is in pseudo-map format to make the regional breakdown a bit easier to see. 

The number at each station tells which forecast (ranked 1 to 11 with 1 being the most snow) is currently closest to the actual value. Any entries that show 1 as the value cannot change with any new snowfall. All other numbers 2 to 11 could change. When it's 11, that means low forecast is closest. 

(map edited at ORD after Apr 27)

 

......................................................... MQT 1 ................................

.................................................................................. APN 1 ..............

.......................MSP 1 ............GRB 1 .......................................................... YYZ 2

................................LSE 1 ...........................................................YXU 11 .......

.............................................. MKE 1 .........GRR 3 ........................................

................................MLI 1.......ORD 1 .......................DTW 11 .......

.................................................................FWA 11.......................... CLE 11

.........................................PIA 1 .........................................................

...........................................................IND 10..................CMH 6 ......

................STL 5 ........................................SDF 9 ...............

................................PAH 11..........................................

 

This table shows who has the closest forecasts for each location (and notes who could take over the lead with additional snow)

 

LOCATION __ Amount __ Leader _____ Forecast __ error __ ranked ___ lead changes possible with more snow

__ MQT _____ 227.5 __ vpbob21 ______ 216.0 ____-11.5 ____ 1 _______ none

__ APN _______99.4 __ Mississauga snow _96.0 ____-3.4 ____ 1 _______ none

__ GRR _______81.3 __ dmc76 _________ 82.0 ____+0.7 ____3 _______ DonSutherland1 at 83.1, Roger Smith at 90.0

__ MSP _______77.1 __ Roger Smith _____66.0 ___-11.1 ____ 1 _______ none

__ GRB _______74.5 __ Roger Smith _____70.0 ____-4.5 ____ 1 _______ none

__ LSE _______ 64.3 __ Roger Smith _____55.0 ____-9.3 ____ 1 _______ none

__ MLI _______ 60.8 __ cyclone77 _______46.0 ___-14.8 ____ 1 _______ none

__ YYZ _______ 55.1 __ Roger Smith _____55.0 ____-0.1 ____ 2 _______ dmc76 at 63.2

__ MKE _______56.1 __ dmc76 _________ 54.0 ____-2.1 ____ 1 _______ none

__ ORD ______ 49.5 __ DonSutherland1 __50.0 ____ +0.5 ___ 1 _______ none

__ PIA _______ 36.4 __ vpbob21 _______ 36.3 _____-0.1 ____ 1 _______ none

__ CLE _______ 35.2 __ Mississauga Snow_41.0 ____+5.8 ___ 11 _______ vpbob21 at 49.5, DAFF at 50.0, all others higher

__ YXU _______ 34.0 __ Stebo __________ 60.0 ___+26.0 ___11 _______ Jackstraw at 64.0, all others higher

__ DTW _______31.3 __ slowpoke _______ 38.0 ___ +6.7 ___ 11 _______ DAFF at 39.0, cyclone77 and Jackstraw at 41.0, all others higher

__ CMH _______27.4 __ madwx _________ 27.3 ___ -0.1 ____ 6 ________ slowpoke and cyclone77 at 29.0, vpbob21 at 34.7, two others higher

__ FWA _______24.7 __ cyclone77 _______ 26.0 ___ +1.3 ___11 ________ slowpoke, DAFF and Roger Smith at 30.0, seven others higher

__ STL ________24.2 __ cyclone 77 ______ 24.0 ___ -0.2 ____ 5 ________ dmc76 at 25.0, DonSutherland1 at 25.2, vpbob21 at 26.6, Miss' snow at 30.0

__ IND ________19.4 __ DAFF __________ 20.0 ___ +0.6 ___10 ________ cyclone77 at 24.0, slowpoke and Stebo at 25.0, six others higher

__ SDF _______ 11.2 __ Mississauga Snow _11.0 ___ -0.2 ____ 9 _______ slowpoke and DAFF at 12.0, six others higher

__ PAH ________ 6.1 __ cyclone77 _______ 11.0 ___ +5.9 __ 11 _______ madwx at 11.4, slowpoke and Roger Smith at 12.0, seven others higher.

____________________________________________________________________

Best forecast count ...

Roger Smith,  cyclone77 _____________4

Mississauga snow __________________ 3

vpbob21, dmc 76___________________ 2

Stebo, madwx, DAFF, slowpoke, DonS __1

__________________________________________

 

 

[Roger Smith]

I don't see much chance of snow anywhere except possibly MQT (where it won't affect the contest) in the next ten days, but if any does fall, I won't be around to post any new scoring, will be back to the internet around April 30th. 

[Roger Smith]

Some snow fell in the last two weeks and tables have been updated. The only significant outcome is that DonSutherland.1 takes the top spot for ORD. 

If the only future changes are more snow at MQT, that will be noted in edits here and tables.