Sleeper Picks Player Props
Pricing
from $0.15 / 1,000 player props
Sleeper Picks Player Props
Collect NBA and NFL Sleeper Picks props with paired sides, lines, multipliers and native IDs. Filter by player, team, stat and game date; monitor new, line, payout, status and removed props with a saved baseline.
Pricing
from $0.15 / 1,000 player props
Rating
0.0
(0)
Developer
Philip Kirkbride
Maintained by CommunityActor stats
0
Bookmarked
2
Total users
1
Monthly active users
3 days ago
Last modified
Categories
Features
Share
Collect NBA and NFL Sleeper Picks player props for DFS research: one record per prop with both sides, the line, side multipliers, native player and game IDs, teams and game start time. Filter a snapshot or retain a baseline and receive subsequent changes.
Use cases
- Shop DFS lines across saved Sleeper snapshots and other licensed data sources.
- Compare player props with sportsbook lines using player, event, stat and period.
- Feed current lines and multipliers into models while keeping fantasy scoring separate.
- Schedule alerts for line moves, payout changes, suspensions and disappearing props.
Also available
The player props suite: pick the job.
| Actor | Use when |
|---|---|
| Sleeper Picks Props (this Actor) | Sleeper lines, multipliers and changes |
| PrizePicks Props | PrizePicks lines and projection changes |
| Underdog Props | Underdog higher/lower lines and payouts |
| DraftKings Pick6 | Pick6 props, multipliers and alternates |
| Sports Injury Reports | Injury status changes across leagues |
| BetRivers Odds & Props | BetRivers and Unibet lines plus player props |
Quick start
Run with the Console prefill to collect up to 20 NBA/NFL props:
{"leagues": ["nba","nfl"],"includeAlternates": false,"includeLive": false,"maxResults": 20,"onlyChangesSinceLastRun": false,"proxyConfiguration": {"useApifyProxy": false}}
A bare {} runs an announced sample of up to 20 available NBA/NFL props and sets OUTPUT.sampleRun: true. It tries NBA first, then NFL if needed, and stops enrichment as soon as the sample is full. If no pre-game props exist, bare input may return currently available live offers with an explicit warning and OUTPUT.sampleLiveFallback: true; an explicit includeLive: false always excludes them. Between supported slates it succeeds with zero rows and an explanation. Explicit scope uses up to 5,000 rows when maxResults is omitted.
Input
Omit optional filters to include everything in scope. Explicit empty lists, empty strings and nulls end the run as no_results and name the field. One unusable or duplicate entry is skipped and reported in OUTPUT.skippedSearches; the rest still runs. Lists combine with OR internally; different filters combine with AND. Each list accepts at most 100 entries of 1–200 characters. Filters ignore case, accents and punctuation; use nba/nfl for leagues and the documented sport names. Strings that normalize to no letters or digits end the run as no_results.
| Field | Code behavior when omitted | Meaning |
|---|---|---|
sports | Both supported sports | basketball, american_football, or Sleeper codes nba, nfl. Must agree with leagues if both are given. |
leagues | NBA and NFL | Select nba and/or nfl. Other Sleeper competitions are outside this release. |
statTypes | All stats | Exact native or normalized keys, such as threes_made, threes, passing_yards. Unknown keys honestly match zero. |
periods | All periods | Normalized output values: game, first_period. Full game is game; an unknown value is skipped and reported in OUTPUT.skippedSearches; if no usable period remains the run ends as no_results. |
players | All players | Exact full names, ignoring accents, punctuation and case, or native Sleeper player IDs. |
teams | All teams | Player team abbreviations as Sleeper writes them, for example TOR or NYG. |
gameDateFrom | No lower date bound | Inclusive UTC game date, YYYY-MM-DD. |
gameDateTo | No upper date bound | Inclusive UTC game date, YYYY-MM-DD; must follow or equal the lower bound. |
nextDays | No relative date bound | 1–30 UTC days, including today. Cannot combine with explicit dates. |
includeAlternates | false | Include alternate lines supplied on the same board. Does not guarantee additional ladders. |
includeLive | false | False keeps pre-game offers. True also retains other game states present on the board; complete in-play coverage is unverified. |
maxResults | 20 for sample; 5,000 for explicit scope | Maximum 1–10,000 delivered rows. Observed undelivered changes are saved in a separate pending queue. |
onlyChangesSinceLastRun | false | First check emits baseline; later checks emit new, line, payout, status and removed. |
stateKey | default | History name, up to 200 characters. Same name plus same filters continues a feed; a new name resets it. |
proxyConfiguration | Direct access | Optional connection settings. Omit for normal use. |
To collect only full-game props, use:
{"periods": ["game"]}
Period values ignore case and surrounding whitespace. An explicit list excludes rows with an unknown (null) period. Period filtering happens before enrichment and delivery; excluded props incur no record fee. The run start fee still applies.
Output
A dataset row contains stable propKey and recordKey, a content snapshotKey, sport, league, game, player, native and normalized stat type, period, line, lineRaw, sides, optional alternates, status and suspension flags, source update time, collection time, and first/last observation times. First seen is scoped to this run in snapshot mode and to observed history delivered by the feed in change mode, including the original observation time of queued rows. Missing information is null. An unavailable side remains explicit with a null multiplier.
The quote block follows the same version 1.1 shape as our sportsbook player props, and canonical follows envelope 1.0. DFS multipliers stay in sides and the Sleeper extension; sportsbook odds and implied probabilities remain null. First-quarter props use first_period. Unmapped stats, including Sleeper fantasy points, have a null normalized stat and keep their native key. Overtime and settlement rules are unknown and must be checked before comparing prices.
espnPlayerId is populated only when supplied with the native player record. A captured NBA sample had 0/131 player IDs; NFL had 60/194. matchKey is <league>:<folded player name>:<folded player team>; folding removes accents, punctuation and case. It stays stable across option-ID and line changes, but a player name or team change changes this matching aid. It is used for matching possibilities, and is not a verified cross-source identity. ESPN event IDs remain null. Match both teams, game time, player, stat, period and settlement before comparing; different lines can be compared as lines, while odds comparisons require identical lines and rules.
The OUTPUT summary includes pending counts, queued deliveries and whether the board was refreshed. It reports delivered counts by sport, league and stat, source market counts, requests, compressed body bytes, decoded bytes, failures by reason, caps, baseline/change counts, state-save status and summaryLine. Delivery counts refer to written rows; source, matching and pending counters describe their named stages. Completed league batches are saved before collecting the next league. Source failures write no diagnostic dataset rows. A collection deadline reserves time to save partial results and a terminal summary. Partial failures name the affected league and preserve its history.
Output example
Selected fields from a real captured NBA record on 2026-11-11; no values below are invented:
{"propKey": "sleeper:nba:2026:regular:1393723855800893440:1718:assists:over_under","league": "nba","game": {"nativeId": "1393723855800893440","startTime": "2026-11-12T00:00:00.000Z","home": "TOR","away": "BOS","espnEventId": null,"status": "pre_game"},"player": {"nativeId": "1718","name": "Jakob Poeltl","team": "TOR","position": "C","espnPlayerId": null,"sportradarId": "e1b3d3d5-b4b1-481a-b5b9-efe4231f48bf"},"statTypeNative": "assists","statType": "assists","period": "game","line": 1.5,"sides": [{"role": "over","nativeOutcome": "over","nativeId": "1426347781999566866","line": 1.5,"payoutMultiplier": 1.65,"payoutMultiplierRaw": "1.65","status": "active","available": true,"payoutInfo": {}},{"role": "under","nativeOutcome": "under","nativeId": "1426347781999566867","line": 1.5,"payoutMultiplier": 1.93,"payoutMultiplierRaw": "1.93","status": "active","available": true,"payoutInfo": {}}],"espnPlayerId": null,"matchKey": "nba:jakobpoeltl:tor","sourceUpdatedAt": "2026-11-11T20:52:01.910Z","collectedAt": "2026-11-11T21:14:48.282Z"}
Pricing
$0.00015 per prop record ($0.15 per 1,000) plus $0.01 per run start.
Examples: a 20-prop sample costs $0.013; 1,000 props cost $0.16; a 2,000-prop pull costs $0.31. An unchanged check with no delivered records costs $0.01 for the run start. Under pay per event there is no separate compute or proxy charge.
One dataset item is one prop with both sides; alternates are nested in that item. Only delivered dataset items incur the per-record charge; the run start fee also applies. Unchanged checks deliver no rows. A charge cap reports how many rows were allowed and leaves remaining changes pending. Include both record charges and run starts when budgeting a schedule.
Scheduling and monitoring
Use a fixed scope and state name, then run the following input twice or on a schedule:
{"leagues": ["nba"],"players": ["Jakob Poeltl"],"onlyChangesSinceLastRun": true,"stateKey": "nba-poeltl-monitor"}
The first observed board labels its baseline; capped baseline rows retain that label when delivered later. Later observations return changed rows with previous values and changeType; baseline/new rows have no previous record. Quiet checks succeed with zero rows. Status changes are included even if the line stays fixed. Removed means absent from the selected scope, including a pre-game prop leaving that scope at game start. A whole empty league slice with saved props is treated cautiously: it preserves history and reports a warning instead of inferring mass removals.
Changing filter meaning, including adding, removing or changing periods, starts a separate baseline even with the same state name. Changing only maxResults keeps the same history. Relative nextDays windows roll with UTC dates. Avoid overlapping schedules using the same state name and scope; state updates are not transactional across concurrent runs. Schedule at a cadence suitable for the source's freshness and your budget. No automatic notifications are sent; consume changed dataset rows in your own alert workflow.
Limits and freshness
NBA and NFL only. This Actor observes the currently available board, not every sport Sleeper offers. At most 10,000 delivered props per run. Saved state, including delivered records and queued observations together, is limited to 10,000 records and 64 MiB; malformed or incompatible older state fails before source requests and requires a fresh state name. No historical backfill, settled results, certified closing prices, complete entry-size payout tables, power/flex rules or derived fair probabilities are provided. Native option IDs can change while propKey remains stable. No alternate offers were observed in the captured NBA/NFL board; support is limited to alternate offers present there.
Snapshot runs and change runs without a remaining backlog check the board anew. Change runs drain saved pending observations first; if a delivery or charge cap stops the run, it reports that the current board was not refreshed. Queued records retain their original observation timestamps, even if the source has since moved or removed them. They matched the filters when captured and may fall outside a later rolling nextDays window when delivered. Unobserved historical moves are not reconstructed. Player metadata can be cached for 24 hours and game metadata for 15 minutes. sourceUpdatedAt reflects the source's own timestamp and is not an update-frequency guarantee; timestamps can refresh without line movement. This is a snapshot service and offers no real-time latency guarantee. Missing native game/player enrichment causes that league to be skipped with a named failure. A fully failed run ends FAILED; a genuinely empty scope is a successful zero-row result.
FAQ
Can I collect only one league or stat? Yes, set leagues or statTypes. Sports and league filters must agree.
Does a multiplier mean decimal odds? No. A pick multiplier is retained as source payout information. It does not establish slip payouts or fair win probability.
Can I join ESPN or sportsbook props automatically? Native Sleeper IDs join within Sleeper. External IDs are incomplete; matchKey alone is insufficient. Verify event and settlement semantics before merging.
Why did a quiet monitor produce no rows? No selected prop changed relative to the observations delivered under that state name. Check OUTPUT for failures, caps and state-save warnings.
Can I export the data? Use Apify dataset exports such as JSON, CSV or Excel. JSON preserves nested sides, games and prior values most clearly.
Can I recover last month's closing lines? No. Schedule forward snapshots; this Actor does not retrieve a historical archive or certify an official close.