What is Bookie Bullies' MLB picks track record?

9 graded picks as of 2026-06-07, with a 55.6% hit rate (5-4-0). On the 7 bet-tracked picks (those with logged American odds + Kelly stake), we're at +0.94u units P/L and +3.1% ROI. Updated daily, no cherry-picking, every pick logged win or loss.

How does Bookie Bullies grade picks?

Every pick is auto-graded against the final game result. Wins, losses, and pushes are tracked separately. Units are calculated using the actual American odds at pick time and the recommended 1/8-Kelly stake. Pushes return the stake. Picks logged before the Kelly stake system went live in late April 2026 count toward hit rate but not toward Units P/L.

Are Bookie Bullies picks profitable long term?

Across 9 graded picks, the model hits 55.6% — above the 52.4% breakeven for -110 odds in moneyline/run-line markets. On the 7 bet-tracked picks (those with logged American odds + Kelly stake), we're at +0.94u units P/L and +3.1% ROI. Long-term profitability depends on the model maintaining its edge and the user betting at recommended stakes. CLV (closing line value) per pick is published as a more reliable predictive metric than recent W-L.

Why do you publish losing picks?

Because hiding them is the most common red flag in this industry. Every pick we've ever published is graded honestly. If the model goes cold, you'll see it here first. That's how a verifiable track record works.

● TRACK RECORD · UPDATED DAILY · NO HIDING

Lock Track Record

Every 5-star Lock we've ever published, fully audited. Lower-conviction picks don't count toward our win rate, only Locks do. 9 Locks graded as of 2026-06-07.

Honest note: 0% of these Locks are scored at 1u flat (Kelly sizing was added mid-season), and 0% use -110 default odds (the book price at publish time wasn't captured for every pick). Hit rate is unaffected; the units number is a backtest estimate, not actual realized P&L.

Total Locks Graded

5-4-0

Hit Rate

55.6%

all 9 graded picks

Units P/L

+0.94u

7 bet-tracked picks

ROI

+3.1%

on bet-tracked sample

How to read this: Hit Rate covers all 9 graded picks (wins ÷ (wins + losses), pushes excluded). Units P/L and ROI cover only the 7 "bet-tracked" picks where we logged both real American odds and a Kelly-sized stake at pick time. The remaining 2 picks were logged before the Kelly-staking system was wired in late April 2026, so their results count toward hit rate but not P/L. Every new pick from 2026-06-07 onward is bet-tracked.

Cumulative Units P/L

Live running total of units won/lost across every graded pick. Climbing line = positive ROI accumulating. Updated daily.

By Sport

Sport	N	W-L-P	Hit %	Units	ROI
MLB	5	3-2-0	60.0%	+1.09u	+3.6%
NBA	2 (0 bt)	1-1-0	50.0%	—	—
NHL	0	ungraded — picks logged today won't appear here until tomorrow
WNBA	2	1-1-0	50.0%	-0.15u	-49.8%

By Pick Type

Type	N	W-L-P	Hit %	Units	ROI
ML	9 (7 bt)	5-4-0	55.6%	+0.94u	+3.1%

By Confidence Tier

Stars	N	W-L-P	Hit %	Units	ROI
5★	9 (7 bt)	5-4-0	55.6%	+0.94u	+3.1%

Probability Calibration

When the model says "60% to win," it should hit ~60% of the time. The "delta" column is how far off the prediction was in percentage points. Wide bands or large deltas = model needs recalibration.

Model Prob Bucket	N	W-L	Mean Prob	Actual Hit %	Delta
50-55%	112	54-58	52.1%	48.2%	-3.8pp
55-60%	61	24-37	57.1%	39.3%	-17.8pp
60-65%	83	46-37	62.1%	55.4%	-6.7pp
65-70%	17	10-7	67.5%	58.8%	-8.6pp
70-75%	14	9-5	71.7%	64.3%	-7.4pp
75%+	22	16-6	83.0%	72.7%	-10.2pp

Brier score: 0.2496 (over 400 graded picks, baseline coin-flip = 0.25, lower is better). Interpretation: weak (0.22-0.25).

Latest Graded Picks

Every row links to a permanent page with the pick's signals, ESPN comparison, and outcome. Showing the 20 most recent.

Date	Sport	Matchup	Pick	Odds	R
2026-06-05	WNBA	PHX @ POR	POR ML	-170	L
2026-05-30	WNBA	SEA @ TOR	TOR ML	-198	W
2026-05-29	MLB	MIA @ NYM	MIA ML	-105	L
2026-04-29	NBA	ORL @ DET	DET ML	—	W
2026-04-28	NBA	PHI @ BOS	BOS ML	—	L
2026-04-23	MLB	NYY @ BOS	NYY ML	-156	W
2026-04-23	MLB	LAD @ SF	LAD ML	-163	W
2026-04-23	MLB	CHW @ ARI	ARI ML	-156	L
2026-04-23	MLB	ATL @ WSH	ATL ML	-136	W

Older picks indexed at /picks-sitemap.xml.

Rolling Windows

Window	N	W-L-P	Hit %	Units	ROI
Last 7 days	1	0-1-0	0.0%	-0.20u	-100.0%
Last 14 days	3	1-2-0	33.3%	-5.15u	-97.2%
Last 30 days	3	1-2-0	33.3%	-5.15u	-97.2%
Last 90 days	9 (7 bt)	5-4-0	55.6%	+0.94u	+3.1%

Calibration disclosure: a probabilistic model that's well-calibrated should hit its 60% picks ~60% of the time, its 55% picks ~55% of the time, etc. If our hit rates by confidence tier diverge from those targets, that's drift we need to fix. Tracking this publicly means we can't hide from it.

Cite This Page

Writing about AI sports betting models? Use this attribution + auditable numbers:

APA-style citation:
MarcDuck. (2026). Bookie Bullies Track Record. Retrieved 2026-06-07, from https://www.bookiebullies.com/track-record.html

Citable facts as of 2026-06-07:

9 graded picks · 5-4-0
Hit rate: 55.6% (all graded picks)
Units P/L: +0.94u · ROI: +3.1% (over 7 bet-tracked picks)
Updated daily at 8 AM Pacific Time
Public picks log: /picks_log.jsonl