pnclENGINEGet API access

CAPACITY NOTES

What one more sport does to your request bill

A new league inside a sport you poll costs nothing. A new sport costs a board per phase per interval. Price the marginal math before enabling it.

Coverage decisions feel like product decisions, but every one of them lands on your request bill. The rule that keeps the bill predictable: a new league inside a sport you already poll costs nothing, and a new sport costs one board per phase per interval. The capacity planning guide states it flatly: covering four soccer leagues or every soccer league is the same request, while covering five sports costs five, checked 2026-10-24.

Leagues ride along, sports do not

A board request answers one sport, every event in it. Whether that sport's response carries ten matches or three hundred, you paid one call. So the cheapest coverage expansion is horizontal inside a sport you already fetch: same request, more matches, and the since cursor keeps each answer to what changed.

A new sport is different. It adds its own board to every cycle, on every phase you poll it, at whatever interval you run. That is the whole marginal cost, and you can price it before you enable anything.

Price the sport before you add it

The formula is the planning guide's one line: requests per second equals boards divided by the interval in seconds, and requests per day is that rate times the active seconds. Run it for the sport you are considering.

A prematch board for one more sport, polled every 30 seconds around the clock, adds about 2,880 requests a day. The guide notes the prematch refresh sits around 30 seconds, so polling that board faster buys nothing. A live board is the real spend: every five seconds during eight active hours is 5,760 requests a day, and it adds 0.2 to your per-second load all by itself.

Put that against the guide's worked scale: five sports live at one second is 5 requests a second and 432,000 a day. One added live sport at one second adds a full request per second to that total, the single largest line item a scope change can create.

The plan line is where it bites

The per-second figure is the one with a hard edge, because plans are per-second allowances and crossing the ceiling returns 429 with a Retry-After, as the rate limits guide documents, checked 2026-10-24. A sport that fits comfortably inside your daily volume can still be the one that tips the per-second sum over the ceiling during the busiest hour.

So the marginal question is not "can we afford this sport" but "can the peak minute afford it". Sum boards over intervals for everything that runs live when your slate is fullest, then add the candidate. If the new total leaves no headroom, the sport costs a plan, not a board.

Seasonal sports are a dial, not a fixture

Coverage you no longer need should leave the bill the way it came. When a sport's season ends, stop its board and keep the stored cursor; when it returns, resume from the cursor and the first delta read is small. The sport's cost exists only in the months it earns its place.

The habit that makes this safe is recording the marginal math next to the scope change: which sport, which phase, which interval, what it added per second and per day. Six months later, when the bill looks heavy, that note is the difference between a review and an archaeology project.

Coverage is a product call. The request bill is just where it signs its name.