Your best Anti-Sicilian depends on your ELO. 6.6M games show how.

@armaanrao said ^

I play 2. a3.

It is SO fun.

Also after Nc6, the b4 ideas are crazy

Just pulled it; 2.a3 Nc6 3.b4 is actually trending up in popularity over time, and the numbers back up your enthusiasm: 52% White / 44% Black overall, a real edge at the population level. The fun part is why it works: 72% of Black players grab the pawn with 3...cxb4 and score only 42% => the gambit succeeds because most opponents take the bait.

The catch is 3...Nf6, which scores 59% Black / 37% White. That's Black's correct theoretical reply, and Wikibooks calls it accurately, but only 2% of players actually find it. Classic information asymmetry weapon; theory says Nf6, practice says cxb4, and your scoreline benefits from the gap.
Curious: do you ever face 3...Nf6 in your games? That's the one tactical reality check the data flags.

According to opening tree, (https://www.openingtree.com/) I have never faced Nf6. It isn't in my prep either, so it is a good hole to patch.

Hi!

@Jisu101 said ^

You forgot the Mengarini's and Wing Gambits existed lol

Anyways, cool stats! I personally play the Alapin, but I have to say, I feel like most players at my level do know that 2...d5 exists. I also know the lines so that's not a big issue, but there are a few weird statements in the blog.

Overall though, these are some cool stats!

Ha, fair on Mengarini's and Wing Gambit, both fell below my popularity floor (sub-2% at every ELO) so they got cut. That said, Mengarini's gets interesting one move deeper: 2.a3 Nc6 3.b4 scores 52% W / 44% B, mostly because 72% of Black players grab the pawn with 3...cxb4 and score only 42%. Deserves a "rare weapons" follow-up at minimum.
On Alapin and 2...d5: fair pushback. The 18.6% number from my article is a population stat across all 1600-2000 players, not "what informed Alapin players know." Your pool almost certainly plays it more often, anyone choosing the Alapin as White, and anyone facing it regularly as Black, is filtered for opening knowledge. The 18.6% reflects the general distribution where most Black players just don't specialize against 2.c3.
What statements felt weird? Genuinely curious, if something's actually wrong I'd want to fix it.

Accepting the Morra gambit is taking dxc3, not not taking on d4.

Also, where are you taking your statistics?

In Lichess database, 1600+, White wins more than black with the gambit, specially when accepted.