DOTA DRAFT

Use ChatGPT to Write Captains Draft Rules - Prompt

Updated 2026-07-13

How do you use ChatGPT to write captains draft rules?

You use ChatGPT to write captains draft rules by describing your format and asking for a short, plain-language document your group can pin in Discord. ChatGPT is well suited to this specific task because writing a rules doc is pure text generation — no live data, no player state, just clear wording covering what happens in each situation.

This is the ChatGPT captains draft rules workflow: get a first draft, then make it stick with a tool that enforces every Dota 2 draft the same way. Paste this prompt in to get a first draft:

What does ChatGPT get right about drafting rules?

ChatGPT writes clean, scannable rules docs — short headers, plain sentences, no legal-sounding filler — which is exactly the tone a Discord pin needs. Ask it to trim the draft to a strict word count or rewrite a section in friendlier language and it does that instantly, which saves the back-and-forth of drafting a rules doc from a blank page yourself.

It also handles edge cases you might forget to write down: what happens if a captain disconnects mid-pick, what happens if a drafted player leaves before the game starts, whether captains can trade picks after the draft ends. Ask it directly and it will produce reasonable defaults for each, even if you have to adjust them to match how your group actually plays.

Where do ChatGPT's rules fall apart once the draft actually starts?

A written rules doc is only ever a description of what should happen — nothing about the text itself makes it happen. If your rules say "30 seconds per pick," nothing on the page counts down, and nothing stops a captain from taking three minutes anyway if no one enforces it live. Try asking ChatGPT to show the actual pick sequence:

Why does the written pick order still need a live enforcer?

Anti-Mage, a common position 1 example used when discussing pick order enforcement

For a simple 8-pick sequence, ChatGPT will typically produce a correct A, B, B, A, A, B, B, A pattern, since it is a short, well-known structure. The problem isn't that it gets this one case wrong — it's that a written pattern in a Discord pin has no way to stop a captain from picking out of turn, and a long conversation full of edits and re-asks can drift the output in ways that are hard to catch by eye. The rules doc says what should happen; it cannot make it happen. If a captain jumps the queue to grab the friend who mains a classic pos 1 like Anti-Mage, something other than a pinned message has to catch it.

That gap between a written rule and an enforced one is the actual risk in a captains draft. A rule that exists only as text relies entirely on the group noticing and calling out violations in real time, which is exactly the kind of social friction a rules doc is supposed to prevent in the first place.

How does DOTA DRAFT make the written rules unbreakable instead of just written?

Divine rank medal, one tier the tool can use when auto-selecting captains by highest MMR

DOTA DRAFT enforces the exact rules a ChatGPT-written doc describes, instead of just stating them: the pick order runs A, B, B, A, A, B, B, A automatically, so a captain physically cannot pick out of turn. Every pick carries a real 30-second timer visible to the whole call, and if it expires, the tool auto-picks a random player from the pool rather than letting the draft stall while someone decides what the "3 minute" rule from the doc actually means in practice.

Captains can also be auto-selected as the two highest-MMR players in the pool — in a stacked lobby, say your two Divine players — removing the "how were captains chosen" question your rules doc otherwise has to answer in prose. Once the draft is enforced this way, the printed rules and the live draft say the exact same thing every time, because the tool is the rule rather than a description of it.

What should the final rules doc actually include?

Keep the ChatGPT draft, but make sure it names the tool that enforces each rule rather than just describing the behavior in the abstract.

Frequently asked questions

Can ChatGPT actually enforce captains draft rules?

No, ChatGPT can only write the rules down; it has no way to watch a live draft or stop a captain from breaking them. Enforcement requires a tool that runs the actual pick order and timer, such as DOTA DRAFT, rather than a document describing what should happen.

What should captains draft rules include?

At minimum: how captains are chosen, the exact pick order, what happens when the pick timer expires, and what happens if a confirmed player drops before the draft starts. A short, plain-language version of these four points covers nearly every dispute that comes up mid-draft.

Why isn't a written rules doc enough on its own?

A written rule only describes intended behavior — it doesn't stop anyone from ignoring it. Without a tool enforcing the pick order and timer live, the rules rely on the group noticing and calling out violations themselves, which is the exact friction a rules doc is meant to avoid.

Should the pick timer rule be strict or flexible?

Keep it strict and automatic rather than a suggestion. A 30-second timer with an automatic random pick on expiry, as used on DOTA DRAFT, removes any judgment call about whether a captain gets "a little more time," which is usually where informal rules break down.

Can ChatGPT write different rules for a smaller draft, like 8 players?

Yes, just describe the smaller group size in the prompt and ask it to adjust the pick count accordingly. The underlying tool handles this automatically too — the snake order simply continues until the player pool is empty, whether that pool has 8 players or 10.

Open the free DOTA DRAFT tool — no download, no signup