Best basket wins

1 · Daily snapshot

Once a day, scheduled for 00:15 UTC, a bot records every holder's balance and picks and every basket token's USD price, then publishes it at /data/comp. Scores come from those files and never from a live feed, so yesterday's standings can still be re-checked tomorrow. Both sources are public too: the chain itself, and DexScreener for prices.

2 · Your entry day

Your entry is the first day of your current unbroken streak: same picks, at least 100M $MXMX, on every daily snapshot. Change your basket and the streak restarts. Dip under the bar and it restarts. Streaks that began before the season count from day 1. A league basket needs at least 2 tokens — one-coin baskets are lottery tickets, and someone running a wallet per token would win every time.

3 · Beat the index

Each pick's move from your entry price to the closing price, averaged equally, minus the whole index over your own window. The index is what you would have got by not picking at all, so what is left is the part your picking actually added.

Up 25% while the index is up 30% is a loss to the default. Down 5% while the index is down 20% is one of the best baskets in the league. Nobody wins for entering on a lucky week.

4 · Both prices are medians

Your entry price is the median of your first snapshots and the close is the median of the season's last ones, over the same number of days at both ends. A median cannot be moved by a single snapshot, so pumping a token for the minute the bot reads it does nothing — you would have to hold the price up for most of the window.

That window grows to 3 days as the season does: one snapshot each end at the start, three from day six on. It is always the widest that still leaves your entry window clear of the close, because a window that overlapped the close would have no move in it to measure.

5 · Score

5000 points is the index. Above it you beat the market you were actually in, below it you did not. Ten thousand and zero are the far ends and nobody reaches them; the scale bends rather than stopping, so an extraordinary season never gets flattened onto the same number as a merely great one.

The square root keeps conviction rewarded but leaves the door open: a late joiner with a genuinely better basket can still win, while a day-20 tourist cannot. On the way down a loss divides by that factor instead of multiplying, so holding longer is always better in both directions. If every basket ends the season red, the one that lost the least wins.

Your $MXMX balance adds a thumb on the scale, and only a thumb. Ten times the entry bar is worth about a tenth of a point, against roughly 0.13 between neighbouring real scores — so holding more decides between wallets that did the same thing, and cannot buy a place off one that picked better. Over a whole season it moves a score by about a fifth of a percent. It counts per ten times the bar rather than per token, or the largest wallets would simply be the leaderboard. Balances are not published here; the score is.

6 · Prize and ties

The pool splits across the top five — 40% / 24% / 16% / 12% / 8% — and it grows with the field: $250 to start, plus $10 for every 10 NEW wallets that qualify after the season opens, up to $1,000. Wallets already holding a basket on day 1 set the baseline, so the pool only moves when someone new arrives.

Balance is already inside the score, so two wallets that picked the same tokens on the same day are ordered by who holds more without anyone's holdings being put on screen. A true tie needs the same picks, the same entry day and the same balance to the token, and those share a rank rather than being put in an invented order. Blatant on-chain price manipulation of a basket token disqualifies.