Raid Report presents fantastic detail on teams completing every raid race event, including the player names, and duration of the raid instance they were in. I pulled data from the site's API that details unique instance and player IDs for each raid race, and deduplicated player IDs within each raid race to arrive at "Unique Players," which is not a stat reported elsewhere on the site.

Raid Report's numbers can include players who only completed the final encounter of a raid, but may not have completed every encounter of the raid. This can happen if someone joins an LFG team that's already partway through the raid, completes the rest, and doesn't go back to complete the other encounters. For this reason, Raid Report's numbers can be considered a ceiling - the "real" number of completions can't be higher than Raid Report's numbers because you'd need to finish the final encounter, and therefore show up in Raid Report's data, in order to qualify for finishing the whole raid.

Charlemagne presents huge scope across nearly all item and collection types in the game. It uses Destiny's API to determine across all accounts ever made, how many have unlocked each collection item. Only items with less than 80k account unlocks will have their absolute numbers published here though. For this validation phase, that means that since the Root of Nightmares raid race saw so many completions, we're relying on an estimated number generated using the average global population estimate discussed on the methodology page.

When players complete a raid race, there is usually an associated triumph they need to actively claim in-game to unlock the emblem in their collections. There are sometimes technical issues preventing players from claiming the emblems, and Bungie has had to provide technical support to fix that, such as with some teams that finished Crota's End challenges during contest mode. However, it's possible that a (very) small number of players never claim their raid race emblems, and therefore the emblem would not be found under account collections to be measured by Charlemagne. Therefore, Charlemagne's numbers can be considered a floor - no account can have the emblem unlocked if it didn't complete the raid race, but the "real" number of completions could be slightly higher due to the extra step needed to unlock the emblem.

Emblem Report does into great detail on all emblems in the game, offering precise counts of all players who have unlocked each emblem, at least from within its sample of 4M guardian accounts.

Because Emblem Report is also looking at account unlocks, it should ideally be very close to Charlemagne's numbers. However, the limited number of accounts under measurement means there's an unknown filter governing which accounts are measured and which aren't. We'd have to make an assumption that for their numbers to be accurate at all, they'd need to be measuring the most active accounts, but we still wouldn't know what that means for players who were active years ago and then stopped playing later.

Also, according to their About Us page, there is the potential for some duplication of players who had accounts on multiple platforms prior to cross-save being introduced to the game in August 2019. They mention a later fix for this, but there's no mention of that being implemented. If this were still a bug, it would likely have a very small impact, but could possibly duplicate some raid race completion numbers.

Braytech has a beautiful site design that closely mirrors the in-game UI and provides detailed accounting of player collections for every unlock in the game.

However, its data server system "Voluspa" only measures about 980k guardian accounts as of 4/7/2024, which is less than a quarter of Emblem Report, and less than 2% of Charlemagne. This means its absolute numbers for all raid race emblems fall far behind the other sites', and for Scourge of the Past and Root of Nightmares in particular, it reports less than 70% of Raid Report's numbers. Additionally, the trending features on the site looking at the "commonality" of emblems is helpful, but it's updated unevenly, with sometimes large gaps between measurement dates.

As a result of these issues, Braytech data is far too volatile and inaccurate to be used for comparison, and should be considered a significant lower-bound outlier.

Bungie is the authority when it comes to reporting internal metrics such as raid race completions, and they have on occasion taken up the habit of reporting that after raid race events. But they've not always done it, and the numbers they provide don't always line up with known quantities.

I was able to find Bungie sources for six of the raid races. For each of Vow of the Disciple, King's Fall, and Root of Nightmares, Bungie's numbers fell between Raid Report's and Charlemagne's, which is exactly where we'd hope they'd be given our understanding of those sites' methods.

However, for Garden of Salvation and Deep Stone Crypt, Bungie's numbers were slightly higher than Raid Report. They're still close enough to put stock in Raid Report and Charlemagne's numbers as rough estimates. But in the same article Bungie provided those numbers, they also said that Last Wish had only 4 player completions. This is verifiably false, as the Last Wish raid race is widely known in the community for having 2 teams (12 players) complete it within the first 24 hours, with popular content creator Datto missing the cutoff by 2 minutes, getting third place. It's inexplicable why Bungie reports 4 completions here, and lends a seed of doubt about whether the Garden of Salvation and Deep Stone Crypt numbers are off a bit as well.