These figures are crap aren’t they? They can’t be right. There simply isn’t that much air traffic in Spain, a look at the statistics just for Spanish receivers demonstrates that.
It makes a mockery of the FlightAware leaderboard statistics.
I agree. I’ve seen feeders higher in rank than mine with around 50 nm range and way lower messages and aircraft seen except for this mysterions “other” which IMO makes the ranking worthless.
This guy is known for fiddling around with his recievers, so I don’t take his readings seriously.
Saw him mentioned in some dutch ADSB forum by @alleyoopie so I guess he’s just a fraud.
And yes that contaminates the ranking indeed.
1 Like
Correct. He’s using a bunch of devices as a combination. Obviously to be #1 in Ranking.
Highest number of Aircraft is in the area > 400 km distance on both of his receivers.
Coincidence?
Same on Planefinder:
1 Like
Thanks for reporting this info. We’ll look further into this user and take appropriate actions.
6 Likes
Thank you.[spoiler][/spoiler]
I thought rankings have been worthless since Skyaware Anywhere was launched? Or does that double only some of the metrics?
Skyaware Anywhere increases reported positions but doesn’t affect the number of aircraft reported as each unique aircraft is only counted once per UTC day.
1 Like
This thread ratled my cage due to frustration of “positions” and the feeling for me is: What is the point in even recording Positions if they have no weight in the ranking?
My understanding was that Other positions (possibly eroneously?) are those that no other site has witnessed (so should be regarded suspect anyway, particularly if it is vastly over-inflated number) as in the case of that Spanish site.
How the #1 site can get just a daily average of 8 for MLAT positions is very fishy indeed.
In contrast, a high MLAT position number is probably a pretty reliable indicator that the same positions have been coroborated.
One would expect to see higher MLAT positions in Europe compared to US due to the density of the receiving stations all receiving the same broadcasts with sites having relatively high numbers of synchronized “nearby receivers”.
For instance, my 20 NEARBY SITES are all within just 7 miles - so it indicates a pretty high density of sites!
While my area doesn’t get especially high figures for ADS-B Mode-S Aircraft due to being a little too far west, there are plenty of positions for those of us lucky enough to have a favourable location and a decent low noise setup with high message rate.
Yet sadly there is no credit in the ranking for MLAT Positions.
Personally I’d be in favour of ditching Other completely and perhaps at least give some credit in the ranking for MLAT Positions. As you can see, it touched a nerve in my case because I’m getting a great Daily Average for MLAT Positions, but it counts for nothing 
Sorry: my rant over.
1 Like
I have no interest in ranking what so ever. In most cases, ranking has more to do with where you live than station performance.
It only takes one message to ‘claim’ an aircraft, but the number of positions tells how ‘solidly’ an aircraft is being tracked. As such, it give a better indication how well a site is performing.
3 Likes
That’s a very good point. You can live in the middle of nowhere with a stunning range, but if no aircraft are passing by, you will never have a significant poisition in ranking.
I stopped checking my ranking also. I only compare it to neighbor stations for deviations. There aren’t any? Good, then my station is operating as usual
1 Like
Another example of bad rank counting are sites like Flightradar or Radarbox.
They do not count aircraft, but a combination of range and uptime.
This station has a range of 331 NM which makes it rank on #7 globally .
Overall the amount of aircraft is pretty low.
So the only logical rating can be a combination of all parameters, not only a subset of them.
Or in short words: rating is useless 
Ups screenshot forgotten. Here it is
1 Like
I am in the top 30 on fr24, but ~11,000 on flightaware. I get amazing range with my setup(lots of ducting happening over the Arabian Gulf), but there just isn’t that much traffic in the middle east when compared with Europe or the US.
1 Like
To my line of thinking, that make your site more valuable than dozens of indoor sites near airports like LAX or LHR that get higher numbers.
3 Likes
‘Other’ is unique ICAO codes that have been received but have no position. Either they are mode S only aircraft, which it hasn’t been possible to calculate an MLAT position for, or they are bogus ICAOs that have been spuriously generated from bad decodes. Filtering can reduce those somewhat, but no doubt some will likely still be counted even on a well set up system.
You could argue for or against including aircraft without positions in the count - it is traffic that has been received, and the lack of an mlat fix could be completely outside of the control of any one particular receiver - perhaps there aren’t sufficient ads-b aircraft visible to synchronise with other receivers, or perhaps there aren’t enough receivers close by that can simultaneously see the same aircraft.
I find the whole concept of ranking to be fairly arbitrary, since the rank depends entirely on what metrics you choose to incorporate. Flightaware uses the number of unique aircraft seen in a day. Other sites use various combinations of maximum range, number of aircraft, number of messages etc. How useful any of it is as a practical measure is somewhat questionable, since location is probably the biggest factor in what traffic you see. That means rank is only really useful as a comparison with other receivers which are very close to you and largely see the same traffic.
I don’t think there is an easy metric you could choose that would be ‘fair’ to rank receivers worldwide given the differences in traffic - a perfectly set up receiver in a quiet part of the world would never come above a poorly performing one located next to Heathrow for example. I do think though that if there is going to be a ranking system, it should at least be self-consistent and not include bogus data in the results.
4 Likes
This is the important thing for me, as much as possible.
3 Likes
Perhaps using the number of messages received per aircraft and the range for ranking would be a fairer indication of a feeder’s performance, and forget all about number of messages and aircraft seen per day, number of unique ICAOs and all that jazz as those depend on the amount of air traffic where the feeder’s located, which is completely out of the feeder’s control.
It seems to me that FlightAware would benefit by getting more feeders in the dark areas of the coverage map vs. yet another feeder near LAX or Heathrow, so it doesn’t make much sense to me to reward the latter by ranking them way high. That’s also discouraging for the feeders located way out in Podunk, to know that they don’t stand a chance in the ranks no matter what they do simply because there’s not very much air traffic where they are. And it’s most definitely discouraging for them to see a feeder higher than theirs in the ranks with a crappy range and stats, with the only difference being that that higher-ranking feeder has a ridiculously huge count of “other.”
2 Likes
I wasn’t discouraged at all. I just ignored the ratings completely.
I had a receiver on Cocos Keeling Island in the Indian Ocean a few years ago. There was nothing else within 500 miles let alone another flightaware receiver.
When the winds were favourable it could see planes flying from the Gulf States to Australia and the twice a week supply plane. There were also Military and Coast Watch planes that were Mode-S only. As there were no other receivers within MLAT range there was no possibility of MLAT so those aircraft just showed up as Other.
I think Cocos Keeling Island and any other receivers providing unique information should be highest ranked.
The others should be ranked by comparing them to all the others in their area. The idea needs more work obviously.
I use the statistics to compare my sites with those around to see if I can do better but to me the rankings are of no use and hence meaningless.
S
1 Like
That’s a good idea. Factor in the number of nearby feeders in the ranking, and the fewer there are the higher the rank. However, if FA did that there may be a lot of feeders who stop feeding in disgust at losing their currently high positions in the overall ranking, so I think the simplest thing to do would be to remove “Other” from the ranking algorithm. AFAIA unMLATted mode-S and incorrect ICAO messages contain no useful information anyway. I recall that someone on the FA staff had already mentioned the idea of getting rid of “Other” and that they were considering doing that. In the example I gave in a previous message, the next-highest ranking feeder above mine had a very poorly performing feed with only 50nm of range and daily ADS-B and MLAT counts in the double digits yet had several hundred daily Other messages, so his feeder outranked mine with my 220nm range and a much higher count of aircraft seen and ADS-B messages received. This makes a good case for not permitting Other messages to affect the ranking IMO.
Such a value-based ranking mechanism would still just be ranking optimal locations like it does today. But you would change the definition of “optimal” from raw accumulated numbers to a site’s value to the entire mesh of data collection. That would be cool to see, but I bet hard to implement the score weighting consistently.
At the end of the day, assuming standard good hardware, probably 90% of what you can get from any given ADS-B site is predetermined by the surroundings of the antenna location. We only have a small margin to play with to get every possible message through skill and/or budget. Adding ranking weight to those who optimize tracking for constrained locations and/or under-tracked locations would probably be healthy for the community. (And more people might actually start to pay attention to the rankings)
2 Likes