Calling BS on 15k airplanes / day


#1

One of the nearby receivers has cracked the code and is somehow pulling in 15,000 planes a day. Ref site 51904, most of them are showing up in one sector 250+ miles away. How is this possible?


2 Geographical locations combined for 1 site?
#2

It looks like a poorly-configured site that is aggregating data from more than one location.

(The majority of the aircraft are “other”, i.e. no position data; however they are reporting speed/heading, which means that the distance-from-receiver-location sanity check is throwing out the position data)


#3

No, I’d say it’s a rather cunningly configured site. It’s been in existence for all of 11 days and has climbed to #23 on the global listing. Unless he pulls the plug it’ll be #1 in another day. Personally, I think he found a way to record a days worth of messages from his two radios and somehow inject it into the NW 250+ segment the next day while still collecting and recording messages for recycling on the following day. I watched the site ramp up and it seemed to increase in that pattern. I really don’t understand that NW 250+ segment though. Is there a good explanation for why there are 250,000 positions in it? Is that something you recognized as a side effect of aggregating data?


#4

Feeling that cock-up should always be considered before conspiracy, is it possible that it could be achieved by accident - build the receiver and run it for a while without connecting to FA, then move to the target site, connect and it uploads the backlog?

This raises the question of whether the FA server(s) have sanity checking to eliminate absurd data and ignore rogue sites?


#5

I have been in touch with the site owner and it is legitimate aggregated traffic.

The northwest sector will probably be SFO traffic, which is close enough not to get thrown out.


#6

“Cock-up … before conspiracy”, I’ll have to remember that.
wOw, impressive! If he’s really getting that much SFO I need to look into some upgrades myself. Thanks for following up.


#7

Can that site owner post his hardware setup? I am more than curious about it.

Sent from my C6903 using Tapatalk


#8

Curiouser and curiouser…


#9

At what point did “legitimate aggregated traffic” appear on the table here? Seems to run contrary to the spirit of this community where one can stand up a receive site and watch its performance referenced to other like receivers. It appears there is another clown in the St. Louis area running the same aggregation con. He’s #1 after a few days. There’s a trustworthy stat.


#10

It appears that these increases have not happened because of only Hardware Upgrades.
If that were the case, then many of us would be lining up for these Super Antenna Systems!
Looking at the history of #1 User: Two weeks ago his site was pulling in 9 positions at 200+ NM.
Today this position is pulling in Over 350,000 at that distance.


#11

It’s accurate, it’s just not telling you what you want to know.

As I said above, what you are seeing are sites that are aggregating data from multiple receivers into a single feed.
We’d rather have those as separate sites, if only because most of the data is being thrown away currently, but we’re not going to refuse to accept the remaining data just because of that.


#12

Looking at the stats and standings, you appear to reward noise when you want signal.
Your system, your call.
It’s all part of the cost-of-doing-business.


#13

Everyone has a different opinion about what stats should measure. I’m not getting into those arguments again.


#14

Good :exclamation:

“When I use a word,” Humpty Dumpty said, in rather a scornful tone, “it means just what I choose it to mean—neither more nor less.”
ATTRIBUTION: LEWIS CARROLL (Charles L. Dodgson), Through the Looking-Glass, chapter 6, p. 205 (1934). First published in 1872.


#15

C’mon OBJ…I’m so disappointed in you! You are the recognized rockstar in this subject matter and you’ve chosen to run away from what would ordinarily be a slam dunk in your worldwide court. These data vandals are making a mockery of what has been a years long example of your pristine record keeping for thousands of contributors who correctly collect and transmit their information. Surely you can’t sit back and accept that these new jokers practicing creative data routing and aggregation of (by your own admission) low quality data deserve the #1 spot in your authoritative list after only a few days of acceptable trickery.

So far, only one gentlemen in this thread has exhibited any sensibility on this subject (myself excluded, of course), and that is the original poster who correctly described this new phenomenon as bullshit. I hope it can be sorted out…I cast a vote of no confidence in the Top 10 today.


#16

Uh, no, you have entirely the wrong end of the stick. There is no vandalism going on here, the data is fine. As I said above I’m not interested in revisiting arguments about what stats should show - past experience is that it’s never going to please everyone and I have lost interest in trying to make a square peg fit in a round hole.

If you think something should change with stats, you could raise that with Eric or the ADS-B support team, but I’m the wrong person to try to convince.


#17

Nah. Still not buying it.

I don’t think the upgrades exist that could bring KSFO hits in any significant quantity to my neighborhood. I get some KSFO-bound traffic along the coast and across my northern edge of reception but nothing crazy. This is typical of the range I get now:
https://c1.staticflickr.com/5/4286/35251743302_d12f98bd12_c.jpg
I have pretty good reach along the N to NE arc with 300+ pretty common and 324.5 the best I’ve witnessed. As you can see, not so much towards KSFO. 250 is about it.

Modeling it on HeyWhatsThat with a 1000’ mast for my antenna still doesn’t get me there. I tried 5000’ too, saw that didn’t make it either and threw in the towel. There are some mountains in the way that cause problems. I only see the high flyers past the Tehachapi’s.
https://c1.staticflickr.com/5/4240/35288506821_c5b2813bc4_c.jpg

This screenshot doesn’t show the aircraft ID but the light purple track under Modesto is a U2 at 60000+ ft. This is the closest I’ve seen anything get to KSFO.
https://c1.staticflickr.com/5/4227/35378813916_e72ac5632a_c.jpg

Which leads me to this.

No, we’re not getting that level of KSFO traffic anywhere around here (and judging from the kind of numbers I see the LAX locals get, they probably don’t see that much KSFO traffic in San Francisco either). I checked the rogue site using HeyWhatsUp to see if there was anything occluding me but not him. No, it’s pretty similar. I have a 400’ altitude advantage that helps me a little but nothing significant. The airport is 345 miles away from me and perhaps 360 miles away from him. Consequently, I don’t think your statement can be correct.

I’m not familiar at all with aggregating data for airplane tracking purposes but I can ponder the concept and make a few suppositions. I hope this logic holds up to scrutiny.

Aggregated data will have to reflect the site(s) the data originated at. Without getting into the philosophical debate, “legitimate aggregated data” should therefore show the arithmetic sum by heat-map location of the various data sets that were aggregated. If I combined two data sets and both contained 100 positions in the North 250+ sector I would expect to see 200 positions in that sector when looking at the aggregate. Aggregated data should look like regular data, just more of it. This does not look like regular data to me.
https://c1.staticflickr.com/5/4223/35289304711_f6d71fbba0_b.jpg

I don’t care how many sites you stack together, a KVCV locale won’t see a 260k spike of hits in that location. I really question the validity of the data you’re getting from your #2 (my neighbor) and #1 (Mr St. Louis) contributors. I get that you don’t want to miss any data but this data is flawed. As I said earlier, I watched the site ramp up. The site operator had two radios, one with a good setup and one with a punk setup (as do I and many others, no disrespect for that). The good setup is still in operation and getting the same kind of numbers as always. The punk setup was punk for a couple of days and then appeared to get aggregated with the good one. Good numbers + punk numbers added together and were distributed properly across the heat-map. That caught my attention. After a few days of that tomfoolery is when it went completely rogue. The rest of the map kept the two-receiver aggregated numbers seen before but the NW 250+ sector started exploding. It sure seems like additional manipulation of the data began then and continues now. From my perspective, you’re getting illegitimate aggregated data and your system is getting gamed.


#18

Can it be a high gain Yagi antenna? The directivity pattern is about right (for two receivers overlapped).
I saw a number of sites in the same condition.


#19

I’ve said this twice so far and I don’t know how else to say it.

It is data aggregated from several receivers
One of the receivers is probably close to SFO. It is not just a bunch of receivers all at the same place!
The positions near SFO are close enough to the nominal “site location” that those positions are not thrown out.

Think of it as plotting all the positions received from many receivers on a map, then drawing a circle around one “site position” and only counting positions that fall into that circle.


#20

I agree 100%

This type of aggregation of data should not be permitted at least as far as applying to the running statistics.

JMO