For 30 years, anti-torpedo nets were an effective defense. But by WWI, they were folly.

Anti-torpedo nets were originally designed not to cause a torpedo to explode, but rather they were designed to render the torpedo harmless by catching it before it could hit the ship.
As torpedoes became faster, it became impossible to prevent their exploding when they struck the nets. In 1887 the Royal Navy swapped to a heavier net, and performed some experiments. They found that as long as the torpedo was exploded at least 25 feet from the ship, the only damage would be a torn net.
Newly designed spars allowed the nets to be deployed in a little as a minute.
The Russo-Japanese war showed both the strength and weaknesses of the nets. A Russian battleship was at anchor in Port Arthur, and the Japanese launched at least 104 torpedoes at her over the course of 5 nights. Most missed or were hung up in the nets, and a few were exploded by the nets causing no damage to the ship. However, the nets did not completely encircle the ship, so eventually a couple of torpedoes found the unprotected stern and the ship was sunk.. Starting with Dreadnought, the new battleships had nets that covered 3/4 of the hull length. The nets were used at anchor, or while the ship was at low speed doing shore bombardment work, but the nets would tangle if the ship went too fast.
In 1907, a new generation of torpedo was deployed. It had much stronger motors and faster speed, so the torpedo was capable of punching clean through a torpedo net. The world’s navies stopped putting torpedo nets on new construction soon afterwards, but retained them on existing ships.
At the beginning of the Gallipoli campaign in spring 1915, three British pre-dreadnoughts were sunk by torpedoes, even though they had their nets deployed. The nets were progressively removed from all British ships starting soon after this. The Germans retained theirs, but found out at Jutland that damaged nets would tangle in the ship’s screws, so they removed all of theirs immediately.