Percentage Predictor
-
This is meaningless, difficult to interpret by the average user and at best a distraction. I think it just creates unnecessary clutter on the screen and should be removed. Who agrees?
-
@In2deep said in Percentage Predictor:
This is meaningless, difficult to interpret by the average
I don’t agree. This is the predictably, expressing the certainty of the forecast. So it is really useful.
See a related thread. -
I think it should be renamed "confidence".
-
I am not so sure about that. The term "confidence" seems more appropriate for a model where the forecast confidence level is calculated based on the spread of results among its ensemble members.
The “predictability” criterion appears to factor in—in addition to the spread across multiple models—the inherent difficulty of the forecast based on weather type (e.g., anticyclonic conditions versus rainy or windy conditions, or, worse, thunderstorms). This refers daily to the ease or difficulty of forecasting the weather. And it tries to answer the question: is the weather easy to predict today?
-
@idefix37
With that explanation, I agree with you on that, confidence is the wrong word.
However...now I'm leaning with @in2deepAre people asking that question everyday?
- How easy is today to forecast?
If they have the answer, are they more or less confident that a model is correct for their location?
If the predictability of every model run is 80% for the next 3 days (which it is starting today for my location)...how is that useful to know? I still need to determine which model is representing my location the best...and that number doesn't help me at all lol
I don't know...I'm sure there was thought and excitement behind this addition, but I don't understand it.
I'd love to hear the thoughts behind it.Maybe I will better understand the usefulness when the What's In This Update article shows up.
-
@In2deep I disagree, as a probably-average user.
Much of Windy’s power goes over my head — I almost never change models, for example, since the error I would introduce by making a bad choice there is surely worse than the error of sticking with the default model. I move between areas constantly so re-evaluating which model to use every few days is just too much.
But Windy provides a great deal of information that, even if I don’t interpret it entirely right due to my obliviousness, has been valuable vs what I get on “normie” weather apps.
When I first started using Windy it seemed like the forecasts were spot-on, most of the time. In the last year it has felt much less so, presumably because it’s just a more difficult job these days. When the predictability scores appeared, it felt like Windy had heard my thoughts and was trying to address them.
For an unsophisticated user, these scores tell me “you can make confident plans based on this information” vs “it’s a crapshoot, don’t plan anything critical, but here’s one way it might turn out.” For photography and drone pilot work, that’s really helpful context to have. I lack the skill to make my own judgement call, lack the time to learn how, and appreciate a dumbed-down “stoplight” I can use to temper my decisions.
I’m currently waiting out a rainstorm that was supposed to be a clear morning. I pulled airspace authorization for today, but also got a backup for next week because today’s forecast was “red” predictability when I was submitting the applications. As a result, my client gets the rescheduled flight sooner.
I’ll grant you, even those tiny dots are just more noise in an already-busy user interface. I can see why a skilled user would want them gone. But it does have value to others.
Maybe they can make it hide-able?
