Forecasts have been collected from our participants and sorted into three categories, events, sky conditions and temperature. Over the next few days, we’ll see who most accurately forecasts the weather from Thursday through Monday.
The original rules have been amended in order to make this a bit easier to tabulate. See below for details. As mentioned, a control case has been added for comparison.
This is very simply created by averaging the forecasts for the last several years and assuming it will rain every day, the sky will be mostly cloudy, and the temperature is the average of the previous ten years’ temperatures on the particular day.
For those of you who are interested, here is the ten year weather history for the five day period. As you can see, the chance of some form of precipitation occurring pretty good. Comparatively, over 60% of the days in February see rain. So my friends in Texas don’t panic, we historically average about 1/10” of rain a day, enough to keep the banana slugs slimy, but nowhere near enough to warrant building an ark. Snow occurs very rarely, less than 2% of the time.
Sky condition was not available on a daily basis, but looking at the aggregate historical data for the month, you’ll understand why Seattle has an unusually high suicide rate in the winter: only 17% of February days are sunny.
The temperature stays within a fairly tight band, ranging from lows of around 37°F to highs of 51°F. There’s no ice fishing here, but we’re unlikely to be working on our tans in the backyard, either.
Because there is some normalization of the forecast into the matrix, I have saved screenshots of the original forecasts in the event the judging committee (e.g., me) needs to resolve disputes. In tabulating the data, it was hard to map sky conditions beyond the absolutes, and some sources lacked any precision. So, in order to keep this similar and work around any interpretation issues, sky condition points are as follows:
- One point will be awarded for accurately predicting partly cloudy, mostly cloudy, partly sunny, overcast.
- Two additional points will be awarded for accurately predicting mostly sunny and/or fog.
- Three additional points are awarded for accurately predicting sunny.
Fog is considered an “event” by weather underground, though I judge it is a sky condition.
I think the url you meant to trackback got inserted as this entry’s excerpt instead… or at least that’s what my newsreader seems to be implying.
that said i’m awfully excited to see the results of the experiment.
Click here for Calculation of the points.
KING TV 5 is the winner of round one with six points. KCPQ TV 13 is the runner up with five. “Control” and KIRO TV 7 both earned two points, while everyone else earned less than zero.
It should be noted that KING and KIRO were the closest in teperature and both had accurately called the sky conditions of Fog and Partly Sunny. However, the Judge has intpreted the rules in an aggressive manner:
Predicting rain is worth 1 point and predicting no rain is worth 3 points. Since there was rain, those who predicted no rain were penalized three points for inaccuracy. This is the difference between KING’s and KIRO’s forecast for today.
Furthermore: KCPQ TV 13 did not specifically call out any kind of sky condition. Because they noted rain, the Judge has interpreted this as falling within the wide boundaries of “partly sunny” through “overcast” and is therefore worth one point in this category.
Finally: No points were awarded for Smoke.
(The Judge is not used to referring to himself in third person.)
smoke? Were you standing too close to your chimney when you made the observation?
Friday was an easy leg in the Tour de Weather with very typical conditions, though the official reporting site did register a light thunderstorm.
Control edged out The Weather Channel for first place based on a more accurate temperature forecast. The peloton was behind by one point.
KING TV5 still retains the yellow jersey based on Thursday’s solid performance.
Ranking as of Saturday morning:
1. KING TV 5
2. KCPQ TV 13 (+1)
3. Control (+3)
4. KIRO TV 7 (+4)
5. TWC, KOMO TV 4, Accuweather (+7)
8. Seattle Times, Weather Underground (+8)
Saturday was overcast, with a little bit of rain. There was a six-way tie, which control wins based on the closest temperature (mechanically calculated using the last ten years on this date).
KING TV5 still retains the lead. KCPQ lost a couple of points for innacurately predicting fog. The big loser was KIRO TV 7, who predicted mostly sunny for today.
Overall ranking as of Sunday morning:
1. KING TV 5
2. KCPQ TV 13, Control (+3)
4. Weather Channel, Accuweather, National Weather Service (+7)
8. Seattle TImes (+8)
9. KOMO TV 4, KIRO TV 7 (+11)
10. Weather Underground (+12)
Sunday was partly cloudy to overcast, with some rain in the morning and early afternoon. Nearly everyone had predicted fair weather, and was penalized appropriately, which is why there was a big change in overall point values.
Overall ranking as of Monday morning:
1. Control
2. KING TV 5 (+4)
3. KCPQ TV 13 (+5)
4. Weather Channel, Accuweather, National Weather Service (+8)
7. Seattle Times (+9)
8. KOMO TV 4 (+12)
9. Weather Underground (+13)
10 KIRO TV 7 (+15)
Although Control is in the lead. KING TV 5 and KCPQ TV 13 will probably make up the difference if tomorrow’s forecast, mostly sunny, holds up, which it likely will. The whole week is looking good, which means I’m going to be spending time outdoors.
The original forecasts were stored here:
Seattle Times
Weather Channel
KIRO TV 7
Seattle PI
KCPQ TV 13
KING 5 and 10 day
Accuweather
National Weather Service
Weather Underground
(Any “Not Found” you see in the acrobat files is where I’ve been blocking specific ad servers.)
Boeing field was colder than normal. With the temperature and dewpoint equal, it stayed misty/foggy most of Monday before going mostly cloudy/overcast. (Surrounding areas cleared out late in the afternoon.) KCPQ TV 13 fared the best, followed by KING TV 5. The Seattle Times, KIRO TV7 and Seattle PI didn’t have a forecast available for today.
Control lost points because it predicted rain (as it had every day).
Final standings and a Tuesday morning armchair weatherman analysis will be posted in a separate blog entry.
This is just amazing. I commend the judge and his team for doing such a wonderful job!