The WxChallenge Rules

Items in red are rule additions or changes for or since the beginning of the 2007-08 contest.

Eligible Participants

All undergraduate, graduate, and Ph.D. students, faculty, and staff of a higher education institution are eligible to participate in the contest.

All participants must be full time students or employees of their respective higher education institutions.

To allow for fair competition, forecasters are separated into four categories (based on current enrollment status, not hours completed):

Category 1: Faculty/Staff/Post-Doctorate
Category 2: All graduate student (MS or PhD) forecasters
Category 3: Junior and senior level forecasters
Category 4: Freshman and sophomore level forecasters

At this time, alumni of participating institutions (that do not fall into an above category), private sector employees, or the general public are not allowed to participate in the contest.

Eligible Teams

An eligible team consists of five or more eligible forecasters from the same institution. Those institutions with less than five forecasters may still participate, but will not be eligible for team awards.
An eligible team consists of five student (i.e. CAT 2, 3, or 4) forecasters that participated throughout the entire semester. These students do not have to have an eligible score, but they must have participate in the contest for the year.
Teams from different Universities may not join together to form one team. However, teams from the same "parent" University may join together (e.g. Univ. of Oklahoma - Norman and Univ. of Oklahoma - Tulsa may (but are not required to) join together to make one Univ. of Oklahoma team).

Forecaster IDs

Forecaster IDs must not contain any vulgar, objectionable, or derogatory (to a person, race, sex, creed, or other school) in any manner.
Receipt of registration forms with duplicate forecaster ids will be handled by the local manager. However, a forecaster who is using their previous year's forecaster id will take precedent over any other requests for that forecaster id.
The local manager and the WxChallenge manager reserves the right to reject any forecaster id.

Fees

Participants pay $3.00 per semester or $5.00 for the fall and spring semesters combined. The WxChallenge does not run during the summer months.
All entry fees are non-refundable and are due to the local manager before the beginning of the first forecast period of the fall or spring semesters. The local manager is to get the fees to the WxChallenge Manager by the start of the second forecast period for each semester.
Entry fees are used to cover the cost of trophies, plaques, shipping costs, and other administrative costs to run the contest.

Forecast Periods

The challenge runs for twenty weeks, covering five (5) two-week forecast periods each semester. Forecasts are due every Monday through Thursday, so for each two-week period, you'll enter eight (8) forecasts. Each forecast period is set in a different city. We'll announce the ten cities before the contest begins.

Forecasts

Forecasts are due Monday evening through Thursday evening at 00 UTC (technically, Tuesday-Friday) and are valid for 06 UTC until 06 UTC for the following day. Each forecast consists of a high temperature (in F), a low temperature (in F), maximum wind speed (in knots; not gusts), and a cumulative liquid precipitation measure (in inches) for that forecast period.
Official temperature and precipitation verifications are taken from METAR observations and climatological records (i.e. a spike in temperature between hourly observations is the correct high temperature for that time interval.) Official temperature and precipitation verifications are taken from METAR observations (A spike in temperature between hourly observations is the correct high temperature for that time interval.) The climatological report will only be used as a backup if the METAR fails for temperature or precipitation reporting.
In extreme circumstances, where an ASOS is unable to reasonably record frozen precipitation, an estimate will be made by the Advisory Board and/or the Manager based on available observations.
If a METAR observation has a temperature greater than the climatological report high, less than the climatological report low, or a wind value greater than the climatological report values, then the METAR values will be used. The precipitation verification (liquid or solid) will come from the 06-06 UTC METAR reports. The climatological report will only be used as a backup if the ASOS fails and is unable to report precipitation. Care must be taken to ensure that the climatological values represent the forecast period since those reports are issued midnight to midnight local time and not 06 UTC to 06 UTC.
Note: The maximum wind speed is determined from the daily climatological report. This report is issued at midnight local time every night. As a result, in many forecast cities, this may be a few hours off 06 UTC. If a higher wind is reported on the hourly observation between the midnight report and 06 UTC, it will not be counted as the official result.
Up to three days of consecutive forecasts may be entered at one time, and forecasters may update/change their forecasts as many times as desired before the 00 UTC forecast deadline.
Know your types of forecasts!

Numerical forecasts: A forecast entered by the forecaster.
Guidance forecasts: This forecast is the ETA-NAM
high and low temperature, and the Extended ETA-NAM maximum (6-hour) wind speed and precipitation total. These numbers are taken from the 12 UTC model run of the day the forecast is due (i.e. the model run before the forecasts are due). In addition to the calculated error points, a 5 point penalty will be added to the forecast score for every forecast beginning on the second non-human forecast. This means that each forecast is allowed one "free" guidance forecast with no penalty, unless
they have already had a non-human forecast.
Persistence forecast: This forecast consists of the previous day's (6 UTC to 6 UTC) high temperature, low temperature, maximum wind speed, and cumulative liquid precipitation measure. In addition to the calculated error points, a 10 point penalty will be added to the forecast score.
Missed forecast: A missed forecast will receive the climatological high temperature, low temperature, average wind speed, and average precipitation for the forecast date to be scored as normal. In addition to the calculated error points, a 10 point penalty
will be added to the forecast score. The climatological values used in the scoring procedure are based on the 30-year mean from the National Climatic Data Center, and will be available before the forecast contest begins.

Scoring

Here's how the daily forecasts will be scored:

ROUNDING

Temperature measurements are rounded to the nearest degree Fahrenheit
Wind speed measurements to the nearest knots
Precipitation measurements to the nearest one-hundredth of an inch.

ERROR POINTS

The following error points will be accessed for non-perfect forecasts :

Max and Min. One error point for every degree difference between your forecasted high and low and the verified high and low temperatures.
Wind speed. 0.5 error points for every knot difference between your forecasted wind speed and the verified wind
speed.
Precipitation.

.4 points for each .01 of error in the verification range from 0.00 - .10 inclusive
.3 points for each .01 of error in the verification range from 0.11 to 0.25 inclusive
.2 points for each .01 of error in the verification range from 0.26 to 0.50 inclusive
.1 points for each .01 of error in the verification range over 0.50
This is NOT an absolute error range. The score depends where you are on the scale.
(Note: A verification of trace precipitation will be scored as 0.00" of precipitation.)

Examples:

If a forecaster gave a forecast of .17 and the verification was .30, the score would be (25 -17)*.3 + (30 - 25)*.2 = 3.4

If a forecaster gave a forecast of .17 and the verification was .65, the score would be (25 -17)*.3 + (50 - 25)*.2 + (65-50)*.1 = 8.9

If a forecaster gave a forecast of .17 and the verification was .02, the score would be (17 - 10)*.3 + (10 - 2)*.4 =5.3

If a forecaster gave a forecast of .90 and the verification was .70, the score would be (90 - 70)*.1 = 2.0

If a forecaster gave a forecast of .90 and the verification was .45, the score would be (90 - 50)*.1 + (50 - 45)*.2 = 5.0

If a forecaster gave a forecast of .90 and the verification was 0, the score would be (90 - 50)*.1 + (50 - 25)*.2 + (25 -10)*.3 + (10 - 0) *.4 = 17.5

If a forecaster gave a forecast of 1.07 and the verification was .80, the score would be (107 - 80)*.1 = 2.7

For each day, a daily raw score will be calculated by summing the error points from the high temperature, low temperature, wind speed, and precipitation forecasts. Any penalty points for non-human forecasts will also be added into the daily raw score.

PERIOD SCORES

For each period, a forecaster will have a raw score (the sum of the eight daily scores for that forecast city) and a normalized score. The normalized score indicates a forecasters standing among all forecasters (consensus) for that forecast period.

CONSENSUS SCORES

The consensus score is the average score of all numerical forecasts (non-human forecasts entered by forecasters are not used) entered for a forecast day, rounded to the nearest degree and 0.01" of liquid precipitation.

NORMALIZED SCORES

Normalized scores will be calculated using the following formula:

Normalized score = ((Raw score - Consensus score)*10)/Standard Deviation

Example:
Forecaster raw score = 59.23
Consensus raw score = 62.45
Standard Deviation = 3.13
Normalized score = ((59.23-62.45)*10)/3.13 = -10.29

A negative normalized score indicates a score better than Consensus. Every 10 normalized points (above or below zero) indicates 1 standard deviation of the population.
The standard deviation used to calculate the normalized score uses only forecasters who have 2 or less non-human forecasts at the point of the results. (i.e. Day 1 and Day 2 will use all forecasts since there will be no forecasters who have more than 2 non-human forecasts at those points.)

Opting Out

Forecasters cannot opt out of a forecast period, however the worst (highest) scoring forecast period for each semester will be dropped from each forecaster's cumulative score. Each forecaster's cumulative score is an average of all of their normalized scores minus their worst (highest) period score(s).

Team scores

A team's score is a combination of the following scores:

33.3% = The average of the five forecasters with the lowest normalized scores of eligible forecasters (note: this is the five best forecasters based on their cumulative score, not the five best forecasters for a forecast city).
33.3% = The average of the five forecasters with the 5 median cumulative scores of eligible forecasters (note: this is the five median forecasters based on their cumulative score, not the five median forecasters for a forecast city).
33.3% = The average of all of the eligible forecasters. (Note: This is the forecaster's scores based on cumulative scores, not the scores for each forecast city).

Eligible Forecasters

A forecaster is considered eligible only if they have a cumulative normalized score of 20 or less.
A forecaster who is participating in both the fall and spring semesters. If a forecaster changes from a "fall only" registration to a "full" registration, then they will become an eligible forecaster.
Only eligible forecasters are considered in any of the team scores.

Awards

City Awards:

We'll award a trophy to the forecaster in each category (1, 2, 3 and 4) with the lowest score for the period. In the case of a tie, we may award multiple trophies. The forecaster with the lowest score (regardless of category) will get special designation on the trophy. These forecasters are considered the winners of the respective cities.

Cumulative Awards:

Trophies will be awarded to the Top 3 forecasters in each division with the best cumulative score. These forecasters must have participated for a full year. These forecasters are considered the winners of the overall contest.

Tournament Awards:

The four best forecasters in the WxChallenge Tournament will receive trophies. This includes the tournament champion, the runner-up, and the other two final four forecasters.

Team Awards:

The team with the lowest score will receive a team trophy.

The top five forecasters for the winning team will have their names engraved on the trophy and will be presented with the trophy at the conclusion of the contest. The trophy will be kept at the institution of the winning team for a period of time until being returned to the WxChallenge main office.

Tournament

As an added bonus for participants who forecasted in both the Fall and Spring contests, a Collegiate Tournament contest is available. The top 64 participants may participate in the contest. There is no additional cost for the contest, but there will be an award for the winner of the tournament as well as the Runner-up and the other Final 4 participants. For more information, please see the Selection Criteria and Bracket Contest Rules on the bracket page.

Miscellaneous Rules

Any forecaster who feels that their forecast(s) listed on the official website are incorrect (for any reason) must contact their local manager and explain the scenario. The local manager has the right to dismiss a change if they feel that it is not warranted (e.g. missed deadline due to a late class, lost internet connection, etc.). If the local manager feels that the request is warranted and is in the spirit of the contest, they will contact the WxChallenge Manager to correct the forecast. Any forecaster who contacts the local manager directly will be instructed to contact the local manager about their issues.
Any forecaster(s) who feel that observations for a particular location are incorrect or highly suspect must first consult their local manager regarding the issue. If, after 24 hours of the end of the forecast day, the local manager feels that the recorded finalized verification numbers are still incorrect or highly suspect and do not represent meteorological conditions, but a sensor error, they (local manager, not individual forecasters) may submit a request for an investigation to the Advisory Board through the WxChallenge Manager.
All forecasts must be entered by the registered forecaster using the WxChallenge website and may not be entered by other forecasters, proxy forecasters, or automated methods.

Improper Behavior

Any forecaster purposefully attempting to modify forecasts or scores illegally, accessing other forecaster's accounts, or demonstrating any improper behavior will be immediately removed from the contest and will forfeit any plaques and/or scholarships they may be due to receive. All entry fees will not be refunded, and the forecaster will not be allowed into future competitions.

While forecast discussion amongst forecasters and/or teams is encouraged, each forecaster is responsible for making their own independent forecast. Submitting frequent identical or near-identical forecasts is considered inappropriate behavior and all forecasters involved will be subject to suspension from the contest.

"Tanking" a forecast in order to purposely ensure a high (i.e. bad) score is considered improper.
Trading accounts is considered improper and against the rules. Each forecaster must have their own account. Accounts can not be handed down from one forecaster to another for any reason. This includes, but is not limited to, full accounts being trade after a semester, accounts for a dropout being given to another forecaster, etc.

Disclaimers

By signing up to participate in the WxChallenge contest, the forecaster agrees to all of the rules stated. They also agree to take the contest seriously and behave properly while participating in the contest.
By signing up to participate in the WxChallenge contest, the forecaster agrees to have all of their forecasts available for public viewing and public use. The forecaster waives all rights to the ownership of any of their forecasts.
By signing up to participate in the WxChallenge contest, the forecaster agrees that his or her name may be released publicly, used in publicity ventures, and/or engraved on trophies.