> ## Documentation Index
> Fetch the complete documentation index at: https://docs.ballista.gg/llms.txt
> Use this file to discover all available pages before exploring further.
> ## Agent Instructions
> Ballista is a trading client for Polymarket, not an exchange. Every order Ballista places goes to Polymarket's CLOB.
> There are three end-user surfaces — the Telegram bot, the web terminal and the Chrome extension — and they share ONE account and ONE balance. Never attribute a feature to a surface without checking that surface's source.
> Every user registers in the Telegram bot first. The terminal and the extension sign in with the Telegram login widget. There is no other sign-up path.
> Deposits arrive through EVM, Solana and Bitcoin bridge addresses and a Polymarket proxy wallet, with a $10 minimum. There is NO withdrawal UI — Export Keys is how a user takes custody of their funds.
> Feature parity is uneven and must never be assumed. The full ~30-field copy-trade form exists only in Telegram; the terminal's copy form has 7 fields; the backtester, leaderboard and trader profiles are terminal-only; alerts and notifications are extension-only; referrals are Telegram-only.
> Slippage means two different things. On a live copy trade it is a fill tolerance (default 30). In a backtest it is the simulated fill price (default 1). Deploying a backtest does not carry its slippage over to the live copy trade.
> Take-profit and stop-loss run on live copy trades but are NOT simulated in backtests — the backtest API rejects those fields.
> A backtest never places an order and never moves funds. Its fidelity caveats are published with every result and must be quoted, not paraphrased.
> Leaderboard PnL is own-tape: harvested by Ballista from its own trade tape. The leaderboard sorts by one transparent metric at a time and there is NO composite score or grade. The sparkline exists only for harvested wallets — a blank sparkline means unmeasured, never zero.
> Copy friction (slip) and a toxic flag DO exist and are own-engine: the wallet's own 30-day fills replayed as a taker (1% slippage per fill plus the real Polymarket taker fee schedule) against its actual 30-day PnL over the same fills, so the delta is pure friction. slip_pct is already a percentage, not a 0-1 fraction. The toxic flag is server-computed under evidence guards and is never re-derived client-side. Slip is not a sortable column. Blank means unmeasured — never zero and never clean.
> Copyability is a taker share, reported in aggressive (≥60%), passive (≤30%) and mixed bands. It is never a score.
> Do not document Feeds, the Twitter Tracker, the Telegram Settings button, percentage buy buttons, GTD orders, the Discord bot, or withdrawals.
# Choosing a trader
> Taker share tells you whether a wallet can be copied at all — which is a different question from whether it makes money.
A copy trade places market orders. That single fact decides which wallets are worth
following, and it is not the same question as which wallets are profitable.
## Taker share
**Taker share** is the fraction of a wallet's volume that **crossed the spread** —
that is, took liquidity off the book rather than resting on it and waiting to be
filled. It is weighted by notional, not by fill count, so one large take counts for
more than a hundred tiny ones.
Ballista names three bands:
| Band | Taker share | What it means for a copier |
| -------------- | ------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Aggressive** | 60% or above | Most of the volume was taken, which is exactly what a copy trade can follow — your market order fills against the same book the leader was hitting. What is left between you and them is timing and the price of crossing, not the shape of the strategy. |
| **Mixed** | Between 30% and 60% | A meaningful share was filled as a maker: someone else crossed to them. You reproduce the taking half and pay the spread on the rest, so expect the copied version to earn less than the leader even with perfect timing. |
| **Passive** | 30% or below | Most of the volume was made, not taken — the wallet was quoting and getting filled. |
The percentage is always shown beside the label. It is a description, not a score:
nothing blends it with P\&L or win rate into a grade, because a blended number would
hide exactly the disagreement between the two that you need to see.
## Why a passive wallet copies badly
A passive wallet's edge lives in **resting orders that a copy trade cannot
reproduce**. They posted a bid, waited, and got filled by someone in a hurry — and
collected the spread for the service. Your copy trade does the opposite: it crosses
the spread to get in, paying the very thing the leader was earning.
So a passive wallet with a spectacular P\&L is not a wallet to copy. Copying it does
not reproduce the strategy, however good the record looks. That is the single most
useful thing a trader profile can tell you, and it is why Ballista puts the number
at the top of one.
When you type a passive wallet into a copy-trade or backtest form, Ballista says
so under the address field before you submit. The notice is dismissable and never
blocks you — finding out how much of a passive wallet's P\&L survives a copy is a
perfectly good reason to run a backtest.
## When the number is blank
Taker share is computed from a **trade tape Ballista holds for that wallet**, and
pulling a full tape is heavy, rate-limited work shared with live copy trading — so
merely viewing a profile does not start one.
| What you see | Why |
| ---------------------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------- |
| **Copyability not measured yet**, with a *Backtest this wallet* button | No tape has been harvested. Running a backtest harvests it, and the metric appears afterwards. |
| **Copyability not measured yet**, with no button | A tape exists but has nothing in it to weigh — cached fills carrying no notional have no taker share to report. |
A blank is never a zero, and it is never "this wallet is passive". It means nobody
has measured it.
## Copy friction
Taker share says *whether* a wallet's edge is the kind a copy trade can follow. **Copy
friction** says *how much of it survives being followed*, in the wallet's own dollars.
Ballista measures it with its own engine. The wallet's last 30 days of fills are
replayed one for one, as though a copier had taken every one of them — same market,
same direction, same size, same moment — but always as the **aggressor**:
* every fill executes at a slippage-adjusted price, **1% per fill**, which is exactly
the execution cost the [backtester](/guides/backtesting) charges by default, so the
board and a backtest of the same wallet rest on the same assumption;
* every fill pays the market's **real Polymarket taker fee**, charged per fill at the
schedule for that market's category;
* settlements pay face value with no friction, because redeeming a resolved market is
an on-chain call rather than a trade.
That replayed P\&L is then compared against the wallet's **actual** 30-day P\&L, which is
computed from the **same fills over the same markets**. Because the inclusion set is
identical, the difference between them is friction and only friction — not a coverage
difference.
Two figures come out of it:
| Figure | What it is |
| ----------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Slip rate** (the board's `Slip` column) | The share of the wallet's own 30-day profit that friction takes. Already a percentage — 23.4 means 23.4%. |
| **Slip amount** | The same loss in dollars. A 30% slip on a wallet that made $400 is a different sentence from a 30% slip on one that made $400,000, and only the dollar figure tells them apart. |
The profile shows both next to the replayed P\&L and the wallet's actual P\&L, so you can
check the subtraction yourself.
This is the question a P\&L column cannot answer. A wallet can be up six figures and
still be uncopyable, because its edge was collected on the side of the trade a market
order pays for.
### The toxic flag
At or above **25%** of recent profit lost to friction, a measured wallet carries a
**toxic** badge. That is the signature of an edge earned on the maker side or on speed,
neither of which a mirrored market order can capture. The P\&L beside it is real; it
simply is not transferable.
The verdict is computed and stored **server-side** and is never re-derived from the
percentage on screen — publication is guarded by the evidence rules below, and a
client-side comparison would light up wallets the engine deliberately declined to
judge. A badge only ever appears for a wallet that was measured *and* judged.
### Blank means unmeasured, never clean
Ballista publishes nothing rather than a number the next dollar of P\&L would swing by
tens of percent. Any one of these suppresses it:
| Guard | Why |
| ---------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------- |
| Fewer than **5 settled markets** in the last 30 days | One market's friction is an anecdote, not an execution profile. |
| A 30-day P\&L too close to zero to divide by | Near zero the ratio describes the rounding rather than the wallet. |
| No fee schedule for some market in the window | A friction number that charged zero fees on it would read cleaner than the truth. This one resolves itself as the tape is swept. |
| Nothing in the window that could be valued | Every market still open, or opened before the tape starts, leaves no P\&L to measure friction against. |
The profile names which guard fired rather than showing a bare dash.
A **negative** slip is real, and it is not a compliment: the replay came out level
with or ahead of the wallet's own fills, which mostly happens on a wallet whose
losing positions a taker would have been out of sooner. It is a statement about
execution over those 30 days, not an endorsement of the P\&L beside it.
## What else to weigh
Taker share and copy friction answer "can this be copied, and what would it cost".
These answer "is it worth copying":
* **Recent P\&L over lifetime P\&L.** An eight-figure lifetime number belonging to a
wallet that stopped trading two years ago is a fact about the past. Ballista leads
with the 30-day figure everywhere for that reason.
* **The execution profile** — median hold, average entry price, how often the wallet
hedges by buying both outcomes, and its record over the last 20 markets.
* **A backtest.** The only way to see what *your* configuration would have done with
their flow, rather than what they did with theirs.
Every one of these stays a separate number. Ballista publishes no composite score
and no overall grade for a wallet: a blended figure would hide exactly the
disagreement between two of them — a great P\&L and a toxic slip rate, say — that
you most need to see.
## What to read next
Lifetime stats, 30-day stats, and the execution profile.
Who appears on it, and what it ranks on.
Replay their real trades through your configuration.
The filters that shape what you actually take from their flow.