Built-in autosave feature (for the TE version).

A couple of well-organized surveys, such as "What factors guide you when choosing a Castle?", "Which Castle do you usually avoid letting your opponent have?", "Which Castle would you definitely choose for yourself if your opponent allowed it?", "Which Castle do you try not to choose for yourself?", and "Which Castle would you like to give to your opponent when choosing Castles?", followed by analysis, along with an analysis of tournament game statistics, can reveal imbalance nuances that can often be adjusted.

Everything has already been studied before us. And partially implemented.

About a year ago, I proposed a system of betting on a Castle/color on the Bay (which was quickly implemented in tournament practice) precisely to record the public's opinion regarding the balance of the Town Editor.

The criterion for balance is as follows: if, in a pair of Castles, players prefer to bet on the color, then the difference in the strength of the Castles does not exceed the difference in the strength of the colors, which can be considered a sufficient criterion for balance.

Experience (I was often asked to accept bets) showed the following: 3 groups of Castles emerged, within which bets are often placed on the color, or a small bet on the Castle. And between the groups, fairly significant bets on the Castle are possible.

1) Necropolis-Swamp-Fortress (closer to the 2nd group)
2) Stronghold (leans towards the 1st group)-Dungeon-Castle-Rampart-Inferno (a wide range of opinions)
3) Tower

The same applies to starting heroes. When generating identical Castles, bets are usually placed on the color for the following Castles: Castle, Rampart, Tower, Inferno. For the remaining Castles, the bet is only on the hero.

This is where many suggestions for balancing the Town Editor come from: to nerf chip leaders, to buff Tower, to nerf overpowered builds.