@idefix37
Well, really you're so way off, beyond help.
I like the old charts you found on Google, nice bit of research. I think you nailed it.
In case we end up discussing chemistry I decided to play the trump card early, because I've got this, which by its vintage has to evidence of something, right?

There's such a list of reasons as to how come your wrong, but on casual examination apparently right, that there not time in a productive day.
I would say though, you got a bunch of pilots laughing our asses off.
I was thinking to donate something useful to this outfit, but after seeing the (now shrouded) unethical conduct in 2017, and the determination to stay wrong/in-toytown, the taste seems all gone.
The right thing to do if so obvious, but somehow invisible for some because operating in a standardised manner is some kind of anathema.
The little arrows you're so keenly advocating for are just hokey little things someone has just made up in their head.
It's a toy and something like this it's diagnostic.
Now here's a question for you. Given we are at an impasse, do you need to have the last word? I'd say I've got money on it, but nobody here will bet against.

@idefix37
You do understand the pennants go to windward, right? The little circle isn't the head of an arrow or something like that.

@WXcycles
Lols

@idefix37
Makes us, formally speaking, polar opposites ha

43.gif/84a928a8-3389-4690-8453-005a9c15f906?t=1369746553556)
..or:

..or:

..or, well, maybe that's enough.

@xiyakl
BTW, here's the wind barb image from NOAA, reflecting WMO standard (https://www.wpc.ncep.noaa.gov/html/stationplot.shtml), which an earlier author posted as explanatory. I think this is the source of earlier confusion. Note the text ('outer' is relative to station position, in other words, the foot of the barb is at the station, the barb points 'out' into the wind, not along) above the image reads:

"Wind is plotted in increments of 5 knots (kts), with the outer end of the symbol pointing toward the direction from which the wind is blowing. The wind speed is determined by adding up the total of flags, lines, and half-lines, each of which have the following individual values:

Flag: 50 kts
Line: 10 kts
Half-Line: 5 kts

If there is only a circle depicted over the station with no wind symbol present, the wind is calm. Below are some sample wind symbols:"

@idefix37
Sorry but this last response is utterly incorrect, and the barbs example image given has been completely misconstrued by the author.
I came across this thread trying to find anyone who'd raised the issue before, having been a bit confounded by it myself.
@svsecondchance is right, Windy.com vanes are 180° out of phase; @idefix37 is very wrong, vanes_always_point into the direction of flow, this is a convention since first use of the wind rose. Anyone who's ever sailed beyond passive crewing could not confuse the rule for reading wind barbs, especially given the powerful feedback even a motor launch gives from actual winds.
I think the mistake here is confusing vectors for vanes. On a chart displaying vectors there are arrows, yes. N.B. however, vector arrows have a set form, and representing a vector they convey direction and magnitude, i.e. velocity.
For Windy.com to correctly display winds using arrows pointed along flow direction would require adhering to the international conventions for displaying meteorological data, which are published by WMO, in other words, thin arrows with relative velocities represented by center line length.
Note also, vector maps have equidistant data plots. Anything with direction display set to station locations is not a vector chart.
It's just a hokey idiosyncracy given the app/site is a toy, but it's wrong and as such, should be fixed.
Showing colleagues the app, everyone points out this as the first deficiency they notice.
It's a bit surprising that there could be any diversity of views on something so elementary as the global standard for representing wind data, but as we all know there's really no limit to how wrong each of us will be about things at various times of the day.
The arrow mixup should be fixed pronto. It's not hard, just reverse/change the symbol. If anyone needs help with the coding my neighbour's 12yr old can do it, I'm pretty sure.
I bet, though, the issue is well known, but leadership paralysis set in at the '..but what will users do when the symbols they're used to suddenly reverse polarity?'. I say this after finding other discussions about wind direction issues here, where people fear users might confuse 300° for 300°F (I'm not joking, just search this forum).