LineCrush Graded Pick Record
As of September 1, 2026 (Pacific time)
Since yesterday, from August 31 games: Signal picks went 2-4-0 (-1.8u, -30.7% ROI) · Noise picks went 24-27-0 (-8.3u, -16.2% ROI).
Best picks (Optimal Strategy): 5-9-0 (-1.4u, -9.7% ROI, avg odds +117) from August 31 games. These are the picks the LineCrushBot leaderboard account tracks, selected by LineCrush's live betting strategy.
LineCrush only recommends Signal picks. Everything else on this page is tracked and graded for transparency, not recommended.
This is LineCrush's self-graded record for its published picks. Signal picks, produced by the V4, V2, and V1 models, are the only picks LineCrush recommends. Noise picks from the V4 and V2 models are the broader consideration tier: they tend to be better than average and are worth a look, but they are noisier and not recommendations, so their record is shown separately from Signal in every section (the V1 model's Noise-tier picks are not counted). Team picks, the moneyline, spread, and totals leans from the Games page, are a separate pick class with their own record. Every player-prop table shows wins, losses, and pushes with flat one-unit staking, the average opening odds, ROI, and the edge over random picking. Losses are included and nothing is filtered by outcome. Esports picks are not counted in this record. Windows cover Pacific-time calendar days ending with the most recent completed day.
See today's picksLast 7 Days2026-08-25 to 2026-08-31
By pick tier and model
| Tier / Model | W-L-P | Win % | Avg Odds | Units | ROI | Edge vs Random |
|---|---|---|---|---|---|---|
| Signal picks (recommended) | 52-50-0 | 51.0% | +108 | -1.1u | -1.1% | +11.4 pts |
| Signal · V4 model | 19-22-0 | 46.3% | +111 | -1.0u | -2.4% | +10.1 pts |
| Signal · V2 model | 29-24-0 | 54.7% | +107 | +2.2u | +4.2% | +16.7 pts |
| Signal · V1 model | 4-4-0 | 50.0% | -106 | -2.4u | -29.9% | -17.4 pts |
| Noise picks (not recommended) | 216-222-0 | 49.3% | +114 | -21.1u | -4.8% | +7.7 pts |
| Noise · V4 model | 1-0-0 | 100.0% | -641 | +0.2u | +15.6% | +28.1 pts |
| Noise · V2 model | 215-222-0 | 49.2% | +114 | -21.2u | -4.9% | +7.6 pts |
Signal picks by sport (recommended)
| Sport | W-L-P | Win % | Avg Odds | Units | ROI | Edge vs Random |
|---|---|---|---|---|---|---|
| All Signal picks | 52-50-0 | 51.0% | +108 | -1.1u | -1.1% | +11.4 pts |
| WNBA | 28-20-0 | 58.3% | +102 | +4.4u | +9.2% | +21.7 pts |
| MLB Pitchers | 19-22-0 | 46.3% | +108 | -2.0u | -4.8% | +7.7 pts |
| MLB Batters | 2-6-0 | 25.0% | +111 | -4.0u | -50.1% | -37.6 pts |
| Tennis | 1-2-0 | 33.3% | +197 | -1.3u | -42.0% | -29.5 pts |
| MMA | 2-0-0 | 100.0% | -122 | +1.6u | +82.1% | +94.6 pts |
Noise picks by sport (not recommended)
| Sport | W-L-P | Win % | Avg Odds | Units | ROI | Edge vs Random |
|---|---|---|---|---|---|---|
| All Noise picks | 216-222-0 | 49.3% | +114 | -21.1u | -4.8% | +7.7 pts |
| MLB Batters | 124-138-0 | 47.3% | +127 | -12.8u | -4.9% | +7.6 pts |
| MLB Pitchers | 32-38-0 | 45.7% | +101 | -6.5u | -9.3% | +3.2 pts |
| Tennis | 36-26-0 | 58.1% | -111 | -3.5u | -5.6% | +6.9 pts |
| WNBA | 18-16-0 | 52.9% | -103 | +2.4u | +7.2% | +19.7 pts |
| Boxing | 4-1-0 | 80.0% | -480 | -0.5u | -10.0% | +2.5 pts |
| MMA | 2-3-0 | 40.0% | +145 | -0.2u | -4.9% | +7.6 pts |
Team picks (moneyline, spread, totals)
| Market | W-L-P | Win % |
|---|---|---|
| All team picks | 147-133-4 | 52.5% |
| Moneyline | 40-44-0 | 47.6% |
| Spread | 33-25-0 | 56.9% |
| Total | 74-64-4 | 53.6% |
| Sport | W-L-P | Win % |
|---|---|---|
| MLB | 97-87-4 | 52.7% |
| WNBA | 23-11-0 | 67.6% |
| NFL | 14-18-0 | 43.8% |
| MLS | 13-17-0 | 43.3% |
Priced at captured odds (publish-time since August 22, 2026; verified opening or publish-time market odds for earlier leans): 261 of 284 picks in this window, +4.8u at flat one-unit stakes, +1.8% ROI, average odds -107.
Team picks are the team leans LineCrush publishes on its Games page: typically one side lean (moneyline or spread) plus one totals lean per game, refreshed on game-day morning with the latest vibe context. They are graded automatically against final scores. Since August 22, 2026 the odds for each lean are captured when it is published, using the best available book price whenever it can be verified for the exact lean and line, otherwise the market price; earlier leans are priced retroactively from verified opening or publish-time market odds for the same lean and line. Units, ROI, and average odds cover only the priced picks; leans whose odds could not be verified show in the win-loss record only. Team leans are a separate pick class: they are not Signal picks, and LineCrush only recommends Signal picks.
Night Before vs Morning
| Tier · Pick timing | W-L-P | Win % | Avg Odds | Units | ROI | Edge vs Random |
|---|---|---|---|---|---|---|
| Signal · Night Before | 52-54-0 | 49.1% | +149 | +18.8u | +17.7% | +30.2 pts |
| Signal · Morning | 73-66-0 | 52.5% | +139 | +28.0u | +20.2% | +32.7 pts |
| Noise · Night Before | 215-209-0 | 50.7% | +121 | -17.4u | -4.1% | +8.4 pts |
| Noise · Morning | 289-292-0 | 49.7% | +121 | -20.9u | -3.6% | +8.9 pts |
Taking Signal Morning picks instead of their Night Before versions gained 2.4 ROI percentage points in this window.
Taking Noise Morning picks instead of their Night Before versions gained 0.5 ROI percentage points in this window.
Last 30 Days2026-08-02 to 2026-08-31
By pick tier and model
| Tier / Model | W-L-P | Win % | Avg Odds | Units | ROI | Edge vs Random |
|---|---|---|---|---|---|---|
| Signal picks (recommended) | 343-320-1 | 51.7% | +108 | +19.5u | +2.9% | +15.4 pts |
| Signal · V4 model | 106-115-1 | 48.0% | +110 | -6.3u | -2.8% | +9.7 pts |
| Signal · V2 model | 185-165-0 | 52.9% | +109 | +13.0u | +3.7% | +16.2 pts |
| Signal · V1 model | 52-40-0 | 56.5% | +101 | +12.8u | +13.9% | +26.4 pts |
| Noise picks (not recommended) | 827-868-0 | 48.8% | +112 | -97.2u | -5.7% | +6.8 pts |
| Noise · V4 model | 4-1-0 | 80.0% | -726 | -0.4u | -8.4% | +4.1 pts |
| Noise · V2 model | 823-867-0 | 48.7% | +112 | -96.7u | -5.7% | +6.8 pts |
Signal picks by sport (recommended)
| Sport | W-L-P | Win % | Avg Odds | Units | ROI | Edge vs Random |
|---|---|---|---|---|---|---|
| All Signal picks | 343-320-1 | 51.7% | +108 | +19.5u | +2.9% | +15.4 pts |
| WNBA | 165-158-0 | 51.1% | +105 | +4.8u | +1.5% | +14.0 pts |
| MLB Pitchers | 102-100-0 | 50.5% | +103 | -4.6u | -2.3% | +10.2 pts |
| Tennis | 40-30-0 | 57.1% | +123 | +9.0u | +12.8% | +25.3 pts |
| MLB Batters | 26-29-0 | 47.3% | +132 | +3.1u | +5.6% | +18.1 pts |
| MMA | 9-3-1 | 75.0% | +110 | +7.1u | +54.7% | +67.2 pts |
| Boxing | 1-0-0 | 100.0% | -517 | +0.2u | +19.3% | +31.8 pts |
Noise picks by sport (not recommended)
| Sport | W-L-P | Win % | Avg Odds | Units | ROI | Edge vs Random |
|---|---|---|---|---|---|---|
| All Noise picks | 827-868-0 | 48.8% | +112 | -97.2u | -5.7% | +6.8 pts |
| MLB Batters | 430-471-0 | 47.7% | +117 | -47.1u | -5.2% | +7.3 pts |
| MLB Pitchers | 148-169-0 | 46.7% | +101 | -18.1u | -5.7% | +6.8 pts |
| Tennis | 130-108-0 | 54.6% | +117 | -12.2u | -5.1% | +7.4 pts |
| WNBA | 101-103-0 | 49.5% | -101 | -12.5u | -6.1% | +6.4 pts |
| MMA | 11-10-0 | 52.4% | -106 | -0.9u | -4.4% | +8.1 pts |
| Boxing | 7-7-0 | 50.0% | +193 | -6.3u | -44.7% | -32.2 pts |
Team picks (moneyline, spread, totals)
| Market | W-L-P | Win % |
|---|---|---|
| All team picks | 537-505-16 | 51.5% |
| Moneyline | 203-176-0 | 53.6% |
| Spread | 83-74-0 | 52.9% |
| Total | 251-255-16 | 49.6% |
| Sport | W-L-P | Win % |
|---|---|---|
| MLB | 411-375-16 | 52.3% |
| WNBA | 85-81-0 | 51.2% |
| MLS | 22-26-0 | 45.8% |
| NFL | 19-23-0 | 45.2% |
Priced at captured odds (publish-time since August 22, 2026; verified opening or publish-time market odds for earlier leans): 966 of 1058 picks in this window, -36.2u at flat one-unit stakes, -3.7% ROI, average odds -113.
Team picks are the team leans LineCrush publishes on its Games page: typically one side lean (moneyline or spread) plus one totals lean per game, refreshed on game-day morning with the latest vibe context. They are graded automatically against final scores. Since August 22, 2026 the odds for each lean are captured when it is published, using the best available book price whenever it can be verified for the exact lean and line, otherwise the market price; earlier leans are priced retroactively from verified opening or publish-time market odds for the same lean and line. Units, ROI, and average odds cover only the priced picks; leans whose odds could not be verified show in the win-loss record only. Team leans are a separate pick class: they are not Signal picks, and LineCrush only recommends Signal picks.
Night Before vs Morning
| Tier · Pick timing | W-L-P | Win % | Avg Odds | Units | ROI | Edge vs Random |
|---|---|---|---|---|---|---|
| Signal · Night Before | 342-367-1 | 48.2% | +118 | -24.0u | -3.4% | +9.1 pts |
| Signal · Morning | 437-455-1 | 49.0% | +125 | +7.2u | +0.8% | +13.3 pts |
| Noise · Night Before | 781-779-1 | 50.1% | +120 | -81.2u | -5.2% | +7.3 pts |
| Noise · Morning | 1033-1108-1 | 48.2% | +119 | -135.4u | -6.3% | +6.2 pts |
Taking Signal Morning picks instead of their Night Before versions gained 4.2 ROI percentage points in this window.
Taking Noise Morning picks instead of their Night Before versions gave up 1.1 ROI percentage points in this window.
Past 3 Months2026-06-03 to 2026-08-31
By pick tier and model
| Tier / Model | W-L-P | Win % | Avg Odds | Units | ROI | Edge vs Random |
|---|---|---|---|---|---|---|
| Signal picks (recommended) | 861-825-2 | 51.1% | +111 | +43.2u | +2.6% | +15.1 pts |
| Signal · V4 model | 290-289-1 | 50.1% | +111 | +9.7u | +1.7% | +14.2 pts |
| Signal · V2 model | 461-453-1 | 50.4% | +112 | +11.5u | +1.3% | +13.8 pts |
| Signal · V1 model | 110-83-0 | 57.0% | +107 | +22.0u | +11.4% | +23.9 pts |
| Noise picks (not recommended) | 2757-3006-0 | 47.8% | +118 | -342.5u | -5.9% | +6.6 pts |
| Noise · V4 model | 7-1-0 | 87.5% | -873 | -0.2u | -2.4% | +10.1 pts |
| Noise · V2 model | 2750-3005-0 | 47.8% | +118 | -342.3u | -5.9% | +6.6 pts |
Signal picks by sport (recommended)
| Sport | W-L-P | Win % | Avg Odds | Units | ROI | Edge vs Random |
|---|---|---|---|---|---|---|
| All Signal picks | 861-825-2 | 51.1% | +111 | +43.2u | +2.6% | +15.1 pts |
| WNBA | 398-393-0 | 50.3% | +104 | +5.7u | +0.7% | +13.2 pts |
| MLB Pitchers | 184-201-0 | 47.8% | +104 | -23.2u | -6.0% | +6.5 pts |
| MLB Batters | 122-100-0 | 55.0% | +122 | +32.9u | +14.8% | +27.3 pts |
| Tennis | 85-77-1 | 52.5% | +151 | +8.9u | +5.5% | +18.0 pts |
| NBA | 45-41-0 | 52.3% | +103 | +10.6u | +12.3% | +24.8 pts |
| MMA | 23-9-1 | 71.9% | +103 | +9.2u | +27.8% | +40.3 pts |
| EPL | 2-4-0 | 33.3% | +284 | -1.5u | -24.3% | -11.8 pts |
| Boxing | 2-0-0 | 100.0% | -336 | +0.6u | +29.8% | +42.3 pts |
Noise picks by sport (not recommended)
| Sport | W-L-P | Win % | Avg Odds | Units | ROI | Edge vs Random |
|---|---|---|---|---|---|---|
| All Noise picks | 2757-3006-0 | 47.8% | +118 | -342.5u | -5.9% | +6.6 pts |
| MLB Batters | 1660-1960-0 | 45.9% | +124 | -248.7u | -6.9% | +5.6 pts |
| MLB Pitchers | 474-492-0 | 49.1% | +102 | -21.4u | -2.2% | +10.3 pts |
| Tennis | 305-229-0 | 57.1% | +123 | -25.5u | -4.8% | +7.7 pts |
| WNBA | 214-236-0 | 47.6% | +102 | -33.6u | -7.5% | +5.0 pts |
| NHL | 32-49-0 | 39.5% | +117 | -17.2u | -21.2% | -8.7 pts |
| MMA | 38-24-0 | 61.3% | +101 | +1.0u | +1.6% | +14.1 pts |
| Boxing | 22-7-0 | 75.9% | +111 | -1.0u | -3.5% | +9.0 pts |
| NBA | 11-8-0 | 57.9% | -104 | +2.2u | +11.7% | +24.2 pts |
| EPL | 1-1-0 | 50.0% | +238 | +1.7u | +83.5% | +96.0 pts |
Team picks (moneyline, spread, totals)
| Market | W-L-P | Win % |
|---|---|---|
| All team picks | 1439-1347-47 | 51.7% |
| Moneyline | 334-275-0 | 54.8% |
| Spread | 385-364-0 | 51.4% |
| Total | 720-708-47 | 50.4% |
| Sport | W-L-P | Win % |
|---|---|---|
| MLB | 1171-1085-47 | 51.9% |
| WNBA | 170-176-0 | 49.1% |
| NBA | 51-34-0 | 60.0% |
| MLS | 22-26-0 | 45.8% |
| NFL | 19-23-0 | 45.2% |
| NHL | 6-3-0 | 66.7% |
Priced at captured odds (publish-time since August 22, 2026; verified opening or publish-time market odds for earlier leans): 2495 of 2833 picks in this window, +12.6u at flat one-unit stakes, +0.5% ROI, average odds -106.
Team picks are the team leans LineCrush publishes on its Games page: typically one side lean (moneyline or spread) plus one totals lean per game, refreshed on game-day morning with the latest vibe context. They are graded automatically against final scores. Since August 22, 2026 the odds for each lean are captured when it is published, using the best available book price whenever it can be verified for the exact lean and line, otherwise the market price; earlier leans are priced retroactively from verified opening or publish-time market odds for the same lean and line. Units, ROI, and average odds cover only the priced picks; leans whose odds could not be verified show in the win-loss record only. Team leans are a separate pick class: they are not Signal picks, and LineCrush only recommends Signal picks.
Night Before vs Morning
| Tier · Pick timing | W-L-P | Win % | Avg Odds | Units | ROI | Edge vs Random |
|---|---|---|---|---|---|---|
| Signal · Night Before | 843-880-2 | 48.9% | +124 | -46.0u | -2.7% | +9.8 pts |
| Signal · Morning | 1100-1178-2 | 48.3% | +129 | -31.0u | -1.4% | +11.1 pts |
| Noise · Night Before | 2631-2960-2 | 47.1% | +126 | -497.4u | -8.9% | +3.6 pts |
| Noise · Morning | 3479-3982-2 | 46.6% | +127 | -591.8u | -7.9% | +4.6 pts |
Taking Signal Morning picks instead of their Night Before versions gained 1.3 ROI percentage points in this window.
Taking Noise Morning picks instead of their Night Before versions gained 1.0 ROI percentage points in this window.
How this record is graded
Every pick LineCrush publishes is graded automatically against the final result of its game or match. The record counts the final published version of each pick and scores it as a flat one-unit stake at the best available odds recorded when the pick was published: a win pays the listed odds, a loss costs one unit, and a push returns the stake. Win percentage excludes pushes. Avg Odds is the average of those best available opening prices, shown as American odds. ROI is net units divided by total units staked, counting pushes as staked units.
Edge vs Random compares each ROI with the roughly -12.5% ROI a bettor should expect from picking randomly at typical sportsbook odds, where the bookmaker's vig guarantees a long-run loss. A positive edge means the picks outperform random selection by that many percentage points.
These are published results, not a backtest. The page refreshes about once an hour as new games finish grading. LineCrush does not publish the model internals behind the picks; the graded output on this page is the complete public record of its recommended Signal picks and the tracked Noise tier across the counted sports.
Looking for the record of the picks LineCrush posts publicly on X? That smaller day-by-day set, graded at the exact odds quoted in each tweet, lives at linecrush.com/scorecard.
Frequently Asked Questions
Which picks are counted in this record?
The record covers the Signal and Noise confidence tiers, always shown separately: Signal is the only tier LineCrush recommends, and Noise is the broader consideration tier, tracked for transparency. Signal picks come from the V4, V2, and V1 models; Noise picks come from the V4 and V2 models only. It covers the final published version of each graded pick and includes every win, loss, and push with no outcome-based filtering. Static-tier picks and the V1 model's Noise-tier picks are not recommendations and are not counted on this page, and esports picks (VALORANT, CS2, Dota 2, League of Legends) are excluded because LineCrush does not count esports in its public record. Team picks (the moneyline, spread, and totals leans shown on the LineCrush Games page) are a separate pick class with their own table.
Does LineCrush recommend every pick shown here?
No. LineCrush only recommends Signal picks. Noise picks tend to be better than average and are worth a look, but they are noisier and are not recommendations; their record is published separately in every section so the two are never blended. Team leans are their own pick class and are not Signal recommendations either.
Are team picks (moneyline, spread, totals) included?
Yes, in their own separate table. LineCrush publishes team leans on its Games page: typically one side lean (moneyline or spread) plus one totals lean per game, refreshed on game-day morning with the latest vibe context. They are graded automatically against final scores. Since August 22, 2026 the odds for each lean are also captured when it is published (the best available book price whenever it can be verified for the exact lean and line, otherwise the market price); leans graded before capture began are priced retroactively, from the best opening book price verified for the same lean and line (or, for moneylines, the market price recorded when the lean was published). The priced portion of the record shows flat-stake units, ROI, and average odds; leans whose odds could not be verified show in the win-loss record only.
What do Signal and Noise mean?
They are LineCrush's confidence tiers. Signal is the highest-confidence tier and the only tier LineCrush recommends. Noise picks tend to be better than average and are worth consideration, but they are noisier and not straight recommendations.
What are the best picks shown for yesterday?
The best picks line is the record of the pick set selected by LineCrush's live betting strategy for the most recent completed day. The strategy chooses a subset of the published picks, mostly from the Signal tier plus a small number of proven Noise-tier markets, and it is exactly the set the LineCrushBot account tracks publicly on the LineCrush leaderboard, so the two always match. Like everything on this page, it is graded automatically with losses included and no outcome-based filtering; the strategy's selection rules themselves are not published.
What does Night Before vs Morning compare?
LineCrush publishes each day's picks in batches: a night-before batch the previous evening and a refreshed morning batch on game day. The Night Before vs Morning table grades each timing's version of the day's picks, separately for the Signal and Noise tiers, so you can see the edge gained or lost by waiting for the morning refresh instead of taking the night-before picks. Each row is the board as it stood at that batch, so a pick published the night before and never refreshed still counts as its own morning version. Morning covers more picks than Night Before because it also includes everything the morning batch added.
How are units and ROI calculated?
Each pick is scored as a flat one-unit stake at the best available odds recorded when the pick was published. A win pays the odds, a loss costs one unit, and a push returns the stake. ROI is net units won divided by total units staked, counting one staked unit for every graded pick including pushes. For team picks, units and ROI cover only the picks with a captured and verified price (publish-time odds since August 22, 2026, verified opening or publish-time market odds before that).
What does Avg Odds mean?
Avg Odds is the average of the best available opening odds across the graded picks in a row, shown as American odds. It is the same average-odds calculation used on the LineCrush performance dashboard. A win rate only means something next to the odds it was achieved at: a 55% win rate is excellent at +110 but loses money at -150.
What does Edge vs Random mean?
Because sportsbooks charge vig, picking randomly at typical odds loses about 12.5% of the amount staked over time. Edge vs Random is the pick ROI minus that -12.5% baseline, in percentage points: a positive edge means the picks beat random selection even when their raw ROI is negative.
How often does this page update?
Picks are graded automatically against final results throughout the day, and this page refreshes about once an hour. Windows use Pacific-time calendar days and end with the most recent completed day.
Why might these numbers differ from other LineCrush pages?
Other LineCrush views, like the performance dashboard, can apply different filters, for example a single model, sport, or signal tier, and may include populations this page excludes (the Static tier, the V1 model's Noise-tier picks, and esports). This page covers the Signal tier from the V4, V2, and V1 models and the Noise tier from the V4 and V2 models, with losses included and no outcome filtering, so its totals can differ from any differently filtered view.
Does LineCrush publish how the picks are generated?
No. LineCrush publishes the graded results, not the model internals. The record on this page is the complete graded output of the recommended Signal tier and the tracked Noise tier across the counted sports.
Records by sport and model
MLB Batters · MLB Pitchers · NBA · NFL · NHL · WNBA · MMA · Boxing · Tennis · V4 Model · V2 Model · V1 Model