You'll always have "flyers" in any statistical analysis. Take Battle of Longewala for example, a freaking 10:1 ratio
Those were definitely not outliers. And I was specifically talking about the Indian offensive undertakings, as the Pakistanis suck balls in that department. Take the 2nd battle for Dograi, for example, where a depleted 3 Jat (~550 men left for the assault at the beginning) managed to overrun the fortress town, which was crisscrossed with trenches, concrete pillboxes and was being defended by two full-strength battalions (one each from Punjab and Baloch regiments), supported by tanks and artillery!!
The casualty ratio was about 3-4 to 1 alright, but in India's favor!
Forget about that, take the battles of Kargil for example. Did the attacking Indian units suffer three times as many casualties as the Pakis?? No, they didn't. The same goes for the PLA units that overran ours in 1962 (except in few instances, Rezangla being one of those) or the NATO against the defending Iraqis.
I think, this whole 3 to 1 stuff is poorly understood by many of the so-called experts who've propagated this rather outlandish claim that an attacker is bound to suffer three times as many casualties as the defender, when the actual facts on the ground would prove otherwise.
As for Longewala, that was Pakistanis' genius on full display, lol. The outcome was more the result of utter incompetence on the Pakistanis' part than it was of the Indian defenders. Don't get me wrong, the Indian soldiers displayed a remarkable show of resolve and nerves of steel in the face of such odds, but had the situation been reversed, you can bet that the defenders would have had their asses handed to them.