"Two, and that's his exact count through nine of the ten games. So as of everything verifiable: no, he hasn't hit 3.
(Small thing — he's The Tragedy of Russell. Liquipedia spells it "The Tragedy of Russel," one L, which may matter if a market's resolving on string match.)
The count is tight, not an estimate
Community Preliminary 4, August 6: 1st place plus two king snipes, $150. That's his only appearance on the payout ledger, and the arithmetic closes exactly — $100 community first-place base plus $50 for two heads. Every snipe pays $25 whether or not you place, so a third kill anywhere would have left a trace. There isn't one. 2, confirmed, no ambiguity.
He's also locked out of the obvious extra opportunity: the Snipers Final is for the top eight non-winning players by king kills. He won his game, so he was never eligible for it.
One game left, and it's the one nobody can see
Winning CP4 put him in the Winners Final, the eight prelim winners in one lobby — the game that still hasn't been logged anywhere I can reach. Liquipedia's indexed snapshot shows the field with Margougou, then 6th Popovici, 7th Schlumpf, 8th The Tragedy of Russel — all on 0 points, i.e. seeded and unplayed at the time of that crawl. Russell is the bottom seed.
So the whole question resolves on that single game, and I still can't load the page (429s, repeatedly).
If you want a number
Snipe rates across RR7's nine logged games run around six kills per eight-player lobby, but they cluster hard — typically only three or four distinct players get credited with any kill at all. That puts a generic player's chance of registering ≥1 snipe at roughly 40%.
I'd shade Russell below that, to around 30% — possible but unlikely:
He's the bottom seed in a lobby of eight game-winners including Kamigawa, Blve and Margougou. Lowest elo also means he picks his civ first, which is a real disadvantage.
More tellingly: he won Community Preliminary 4 by building a Wonder. His demonstrated style is kill a couple early, then turtle and monument up. That's a placement strategy, not a kill-farming one, and it gets more attractive, not less, against a stronger field.
Counter-argument worth holding: the bottom seed in a diplomacy FFA is often the useful ally nobody bothers to attack, and useful allies get handed kills. And two of RR7's community prelims were won by Wonder — the tactic evidently works, and a Wonder push tends to force chaotic all-in attacks that generate snipes in both directions.
One resolution caveat: the rules say that where it's unclear who killed a king, admins make the final call. So the official credited number can differ from what the VOD looks like — if this is a market, Liquipedia's per-game kill log is the thing that counts, not the footage."