Ask most Minecraft admins what they do when a player is being a problem and they'll tell you they IP ban them. It's the intuitive response - block the address, block the person. The problem is that IP banning is, in practice, one of the least effective moderation tools available. Understanding why takes about five minutes of thinking about how IP addresses actually work.
Why IP Banning Fails
IP addresses are not stable identifiers. Your home internet connection almost certainly changes IP every time your router restarts, and ISP-assigned addresses often rotate on a daily or weekly cycle. Most players have been assigned dozens of different IP addresses over the course of a year.
On top of that, mobile data assigns an IP from a large shared pool. Every time someone switches from Wi-Fi to mobile data, they get a new address. A player who primarily plays from a phone can evade an IP ban within seconds just by switching networks. And VPNs make the whole exercise even more pointless - there are free VPN services that provide a new IP with two clicks.
The final problem is collateral bans. In a student dorm, a shared office, or any location where many people share the same public IP, banning one person blocks everyone else behind that IP. ISP carrier-grade NAT means that in some countries, hundreds of people share a single public IP address.
What UUID Banning Actually Does
UUID banning targets the Minecraft account itself. Every Mojang/Microsoft account has a unique identifier that never changes, regardless of username changes, IP changes, or device changes. When you ban a UUID, you're banning the account - and the only way to evade it is to buy a new one.
This is meaningfully harder than changing an IP. A new Minecraft account costs around €27 at full price. For casual rule-breakers, that's a significant deterrent. It doesn't stop serious cheaters who will happily spend the money, but it eliminates the trivial evasion that IP banning completely fails to prevent.
The Alt Account Problem
The remaining weakness of UUID-only banning is alt accounts. Someone who really wants back in will buy a new account and join with a fresh UUID. Against this, UUID banning alone offers no protection.
This is where alt account detection fills the gap. When the new account joins, the system checks whether that account's IP has ever been associated with a banned UUID. If there's a match, the new account is flagged before it can cause trouble. The combination of UUID banning and IP-based alt detection is substantially more effective than either approach alone.
Should You Use IP Bans at All?
There's one legitimate use case for IP banning: blocking automated bots that flood your server from a specific range. Bots often come from known datacenter IP ranges, and blocking those ranges at the network level is reasonable. For individual problem players, though, stick to UUID bans with alt detection enabled. You'll catch more actual bad actors and create far fewer problems for legitimate players.
Getting Alt Detection Right
For alt detection to work well, you need a large IP history database. A single server only sees players who have visited before. A network-level IP hash registry - covering all servers in the network - dramatically increases the chances of catching a new alt on first join, even if they've never visited your specific server before.