ProphetAiQ YouTube Recap: Episodes 1 & 2 — How the AI Did on the April 28 and April 30 Slates
ProphetAiQ Team · 2026-05-01
A complete recap of the first two ProphetAiQ Sports YouTube episodes — featured AI picks for the April 28 and April 30 night slates, three-way grading (moneyline, spread, total) for every game, and full slate breakdowns across MLB, NBA, and NHL with links to every recap.
Executive summary: The ProphetAiQ Sports YouTube channel launched April 28 with Kendall covering the slate, then dropped Episode 2 on April 29 previewing the April 30 board with Drew making his debut as the second host. The model publishes three independent picks per game (moneyline, spread, total) and we grade all three. Across 33 games on the two episode slates the combined record was 17-16 ML, 13-18 ATS, 15-14 O/U. Episode 1 (Apr 28) was the better night — 12-9 ML / 8-11 ATS / 11-7 O/U. Episode 2 (Apr 30) was rougher — 5-7 ML / 5-7 ATS / 4-7 O/U.
Quick take
- Three-way grading: the model publishes ML, spread, and total picks independently for every game — and we publish all three results, not just the moneyline
- Episode 1 slate (Apr 28): 12-9 ML (57.1%) / 8-11 ATS / 11-7 O/U (61.1%) across 21 graded games
- Episode 2 slate (Apr 30): 5-7 ML (41.7%) / 5-7 ATS / 4-7 O/U across 12 graded games
- Combined two-day record: 17-16 ML / 13-18 ATS / 15-14 O/U — roughly break-even, with the under tier holding up best
- Featured picks went 4-5 ML, 4-5 ATS, 6-3 O/U — the totals tier outperformed even on a slate where headline calls missed
- Best night: Tigers @ Braves on Apr 28 — three-team sweep at 65% confidence
- Worst stretch: Celtics 0-2 ML at 80% confidence in back-to-back episodes (76ers swept the series)
- Drew Sterling debuted as the second AI host in Episode 2 — the Kendall + Drew duo is now live
The ProphetAiQ Sports YouTube channel went live Tuesday April 28, 2026 — and instead of running a launch trailer and disappearing, we ran the model live on the actual slate two episodes in a row. Episode 1 covered the Apr 28 night card. Episode 2 (with Drew Sterling debuting alongside Kendall) previewed the Apr 30 board.
This post is the full recap: what the AI picked, what actually happened, and how every market graded.
How ProphetAiQ Grading Works
Before the recap, a quick note on how we score the model. ProphetAiQ publishes three independent picks per game:
- ML (moneyline) — straight-up winner
- ATS (against-the-spread) — who covers the run line, puck line, or point spread
- O/U (over/under) — total runs/points/goals vs the posted total
These are run as separate Groq calls inside the prediction engine. The ATS pick is not a derived consequence of the ML pick — the model is allowed to (and frequently does) say "Team A wins the game but Team B covers the spread" or "Team A wins but the under cashes." That's why you'll see what look like contradictions in the tables below: a Celtics ML loss can sit next to a 76ers ATS cover and an under cash on the same row. That's three separate picks, three separate grades.
Every game gets a row in the slate tables with all three results. ⭐ marks the games featured directly in the YouTube episodes.
Episode 1: April 28 Night Slate
The launch episode dropped Tuesday morning. Kendall walked through four featured games — one MLB, two NBA playoff spots, one NHL — and the model published predictions on the entire 21-game board across the three sports.
Featured Picks vs Reality (Episode 1)
| Game | AI Pick | Conf | Final | ML | ATS | O/U |
|---|---|---|---|---|---|---|
| Tigers @ Braves (MLB) | ATL ML / ATL -1.5 / Under 8.5 | 65% | ATL 5, DET 2 | ✅ | ✅ | ✅ |
| 76ers @ Celtics (NBA) | BOS ML / 76ers +10 / Under 215 | 80% | PHI 113, BOS 97 | ❌ | ✅ | ✅ |
| Bruins @ Sabres (NHL) | BUF ML / BUF -1.5 / Under 5.5 | 70% | BOS 2, BUF 1 | ❌ | ❌ | ✅ |
| Knicks @ Hawks (NBA) | NYK ML / Hawks +6.5 / Under 214.5 | 80% | NYK 126, ATL 97 | ✅ | ❌ | ❌ |
Featured ML record: 2-2. Featured ATS: 2-2. Featured O/U: 3-1. The headline metric was an even ML split, but the under tier was the cleanest — three of four featured totals cashed.
What Actually Happened in Each Game
Braves 5, Tigers 2 — The cleanest hit of the episode. Martin Perez outdueled Casey Mize, Atlanta's offense cleared the closing total of 8.5, and the Braves covered the run line. ML, ATS, and O/U all cashed at 65% confidence. This is what the model is supposed to do on its medium-confidence calls — be right more often than the implied price suggests.
76ers 113, Celtics 97 — Boston was -450 at home and was the highest-confidence NBA pick on the board. Philadelphia walked into TD Garden and won by 16. The model's ATS lean (76ers +10) cashed by a wide margin and the under at 215 hit. Two-of-three on the three-way grade is a recovery, but the moneyline at 80% confidence is the loss that sticks.
Bruins 2, Sabres 1 — Buffalo was -172 at home, the model agreed at 70% confidence, and Boston grinded out a one-goal road win. NHL goaltending variance is the highest single-game variable in any major sport. ML and ATS both lost; the under at 5.5 cashed cleanly with just three combined goals.
Knicks 126, Hawks 97 — Vintage Brunson. The Knicks were -265 on the road and instead of a tight game New York hung 126 on Atlanta in a 29-point blowout. ML cashed easily, but the Hawks +6.5 lean missed (29-point loss does not cover) and the under at 214.5 ate dirt with 223 combined points. This game illustrates why we run independent ML, ATS, and O/U calls — the moneyline pick was correct and the under was wrong, and treating those as the same decision would have masked half the information.
Episode 1 Full Slate Breakdown
The model published predictions on all 21 games on the night of April 28. Here is the full graded record across all three markets.
#### MLB on April 28 — 9-6 ML / 6-7 ATS / 7-5 O/U
A 15-game slate.
| Matchup | Pick | Final | ML | ATS | O/U |
|---|---|---|---|---|---|
| Rays @ Guardians | TB ML | TB 1, CLE 0 | ✅ | ✅ | ✅ |
| Astros @ Orioles | BAL ML | BAL 5, HOU 3 | ✅ | ❌ | ✅ |
| Rockies @ Reds | CIN ML | CIN 7, COL 2 | ✅ | ❌ | ❌ |
| Cardinals @ Pirates | PIT ML | STL 11, PIT 7 | ❌ | ✅ | ❌ |
| Giants @ Phillies | SF ML | PHI 7, SF 0 | ❌ | ❌ | ✅ |
| Red Sox @ Blue Jays | TOR ML | TOR 3, BOS 0 | ✅ | — | — |
| Nationals @ Mets | NYM ML | NYM 8, WAS 0 | ✅ | ❌ | PUSH |
| Tigers @ Braves ⭐ | ATL ML | ATL 5, DET 2 | ✅ | ✅ | ✅ |
| Angels @ White Sox | LAA ML | CWS 5, LAA 2 | ❌ | ❌ | ✅ |
| Diamondbacks @ Brewers | ARI ML | MIL 13, ARI 2 | ❌ | ❌ | ❌ |
| Mariners @ Twins | SEA ML | SEA 7, MIN 1 | ✅ | ✅ | ❌ |
| Yankees @ Rangers | NYY ML | NYY 3, TEX 2 | ✅ | ❌ | ✅ |
| Cubs @ Padres | CHC ML | CHC 8, SD 3 | ✅ | ✅ | ❌ |
| Royals @ Athletics | OAK ML | KC 4, OAK 1 | ❌ | — | — |
| Marlins @ Dodgers | LAD ML | MIA 2, LAD 1 | ❌ | ✅ | ✅ |
The standout was the Rays at Guardians 1-0 walk-off win — the model nailed all three markets on that one (Tampa to win, Tampa to cover the run line, and the under at 7.5). Two games had no posted spread/total at the time predictions ran, which is why some rows show dashes for ATS and O/U.
#### NBA on April 28 — 2-1 ML / 1-2 ATS / 2-1 O/U
Three games on the playoff and regular-season board.
| Matchup | Pick | Final | ML | ATS | O/U |
|---|---|---|---|---|---|
| 76ers @ Celtics ⭐ | BOS ML | PHI 113, BOS 97 | ❌ | ✅ | ✅ |
| Knicks @ Hawks ⭐ | NYK ML | NYK 126, ATL 97 | ✅ | ❌ | ❌ |
| Trail Blazers @ Spurs | SAS ML | SAS 114, POR 95 | ✅ | ❌ | ✅ |
#### NHL on April 28 — 1-2 ML / 1-2 ATS / 2-1 O/U
Three games. The Sabres miss was the painful one; the Oilers/Ducks clean sweep on ML and O/U salvaged the night.
| Matchup | Pick | Final | ML | ATS | O/U |
|---|---|---|---|---|---|
| Bruins @ Sabres ⭐ | BUF ML | BOS 2, BUF 1 | ❌ | ❌ | ✅ |
| Wild @ Stars | DAL ML | MIN 4, DAL 2 | ❌ | ✅ | ❌ |
| Ducks @ Oilers | EDM ML | EDM 4, ANA 1 | ✅ | ❌ | ✅ |
**Episode 1 day total: 12-9 ML (57.1%) / 8-11 ATS (42.1%) / 11-7 O/U (61.1%)** across 21 graded games. The under tier was the strongest, the moneyline pick beat break-even, and the spread tier had a tough night — particularly because the model defaults to the road dog covering on most -1.5 / -10 lines and several big favorites covered comfortably.
Episode 2: April 30 Night Slate
Episode 2 dropped late on April 29 — Kendall handed the desk to Drew Sterling for his first appearance, and they previewed the April 30 board together. This was the model on the road, looking forward instead of backward, which is the harder test.
The featured games shown in Episode 2 were drawn from a stacked Thursday slate with three NBA series elimination spots (Celtics/76ers Game 6, Nuggets/Wolves Game 5, Knicks/Hawks Game 6) and two NHL contests.
Featured Picks vs Reality (Episode 2)
| Game | AI Pick | Conf | Final | ML | ATS | O/U |
|---|---|---|---|---|---|---|
| Knicks @ Hawks (NBA) | NYK ML / NYK -2.5 / Under 213.5 | 80% | NYK 140, ATL 89 | ✅ | ✅ | ❌ |
| Stars @ Wild (NHL) | MIN ML / DAL +1.5 / Over 5.5 | 60% | MIN 5, DAL 2 | ✅ | ❌ | ✅ |
| Celtics @ 76ers (NBA) | BOS ML / BOS -6.5 / Under 212.5 | 80% | PHI 106, BOS 93 | ❌ | ❌ | ✅ |
| Nuggets @ Wolves (NBA) | DEN ML / DEN -7.5 / Under 224.5 | 80% | MIN 110, DEN 98 | ❌ | ❌ | ✅ |
| Oilers @ Ducks (NHL) | EDM ML / EDM -1.5 / Under 6.5 | 65% | ANA 5, EDM 2 | ❌ | ❌ | ❌ |
Featured ML record: 2-3. Featured ATS: 1-4. Featured O/U: 3-2. A rough night on moneylines but three of five totals cashed — the under tier saved a lot of the night for full-slate three-way bettors.
What Actually Happened in Each Game
Knicks 140, Hawks 89 — The Knicks' second straight Episode-featured win, this time on the road in Atlanta. New York hit 140 points (the most by any team across either night) and the 51-point win was a generational ATS cover at -2.5. The under at 213.5 missed by 16 points — combined 229 — but ML and ATS were as decisive as it gets. When the model has the Knicks in this matchup, the answer has been the same both times.
Wild 5, Stars 2 — Minnesota was the AI's medium-confidence NHL play (60%) and they cashed two of the three markets. The Wild won straight up and the over hit, but the puck line on Stars +1.5 missed — Dallas needed to keep it within one goal and lost by three. Still, the ML + O/U combo at 60% is exactly what the medium-tier is supposed to deliver.
76ers 106, Celtics 93 — Boston, again. The Celtics were 0-for-1 in this series at the time of Episode 2, and the model still leaned them at 80% in Game 6 in Philadelphia. The 76ers won by 13 to clinch a Game 7 (May 2 at TD Garden). The under at 212.5 cashed (combined 199), so the total grade was 1-for-3, but the back-to-back Celtics ML losses at the highest confidence tier are a real data point. The model loved Boston's regular-season record and shooting percentages, and this series has been a reminder that playoff matchups can override regular-season signals — especially when one team's role players (Quentin Grimes, Tyrese Maxey, Kelly Oubre Jr.) are firing at home.
Wolves 110, Nuggets 98 — Denver was -350 on the road and the model agreed at 80%. Minnesota won by 12 at home and effectively closed the Western Conference series. Anthony Edwards's recent return from injury seems to have flipped the dynamic. The under cashed here too at 224.5 (combined 208), softening the loss but not erasing it.
Ducks 5, Oilers 2 — The cleanest miss of the night. Edmonton was -157 and the model leaned them at 65%, citing home ice against a Ducks team riding a 0.810 save percentage. Anaheim flipped the script — strong goaltending, three-goal margin, and the over hit too with 7 combined. Connor McDavid being out had been factored in but not aggressively enough.
Episode 2 Full Slate Breakdown
The April 30 board had 12 graded games across all three sports.
#### MLB on April 30 — 3-4 ML / 4-3 ATS / 1-5 O/U
A 7-game MLB slate (with two completed double-headers).
| Matchup | Pick | Final | ML | ATS | O/U |
|---|---|---|---|---|---|
| Giants @ Phillies — G1 | SF ML | PHI 3, SF 2 | ❌ | ✅ | ✅ |
| Nationals @ Mets | WAS ML | WAS 5, NYM 4 | ✅ | ✅ | ❌ |
| Diamondbacks @ Brewers | MIL ML | MIL 13, ARI 1 | ✅ | ❌ | ❌ |
| Royals @ Athletics — G2 | OAK ML | OAK 6, KC 3 | ✅ | ❌ | PUSH |
| Astros @ Orioles — G2 | BAL ML | HOU 11, BAL 5 | ❌ | ✅ | ❌ |
| Giants @ Phillies — G2 | SF ML | PHI 6, SF 5 (10) | ❌ | ✅ | ❌ |
| Blue Jays @ Twins | TOR ML | MIN 7, TOR 1 | ❌ | ❌ | ❌ |
The MLB O/U tier was rough — only one of six O/U calls cashed (with one push). This was a high-scoring slate where most posted totals went over the line, and the model's default lean toward unders during cool-weather April games hurt.
#### NBA on April 30 — 1-2 ML / 1-2 ATS / 2-1 O/U
Three featured playoff games.
| Matchup | Pick | Final | ML | ATS | O/U |
|---|---|---|---|---|---|
| Knicks @ Hawks ⭐ | NYK ML | NYK 140, ATL 89 | ✅ | ✅ | ❌ |
| Celtics @ 76ers ⭐ | BOS ML | PHI 106, BOS 93 | ❌ | ❌ | ✅ |
| Nuggets @ Wolves ⭐ | DEN ML | MIN 110, DEN 98 | ❌ | ❌ | ✅ |
#### NHL on April 30 — 1-1 ML / 0-2 ATS / 1-1 O/U
Two NHL games.
| Matchup | Pick | Final | ML | ATS | O/U |
|---|---|---|---|---|---|
| Stars @ Wild ⭐ | MIN ML | MIN 5, DAL 2 | ✅ | ❌ | ✅ |
| Oilers @ Ducks ⭐ | EDM ML | ANA 5, EDM 2 | ❌ | ❌ | ❌ |
**Episode 2 day total: 5-7 ML (41.7%) / 5-7 ATS (41.7%) / 4-7 O/U (36.4%)** across 12 graded games. Below break-even on all three markets — but the under tier still cashed on three of the five featured plays, salvaging the headline games even when the moneylines missed.
Two-Day Combined Performance
Across 33 graded games on the two episode slates the totals were:
| Market | Record | Hit Rate |
|---|---|---|
| Moneyline (ML) | 17-16 | 51.5% |
| Spread (ATS) | 13-18 | 41.9% |
| Total (O/U) | 15-14 | 51.7% |
The model finished roughly break-even on moneyline and totals across the two-episode sample, with the spread tier coming in below break-even. That's the honest scoreboard. The Episode 1 night carried the average — Apr 28 was a strong performance across all three markets, while Apr 30 was a clear losing day.
The featured-game scoreboard (the picks shown in the videos themselves) was 4-5 ML, 4-5 ATS, and 6-3 O/U. That under-tier outperformance on featured calls — even on a night the moneyline picks struggled — is the kind of asymmetry that's worth paying attention to. When the model says "team X wins" and "the under hits", those are two different bets, and a sharp full-slate bettor playing the model's three-way grade walks away with more cashed tickets than the headline ML record suggests.
This is exactly why the Record page exists. Every prediction the model has ever made is graded publicly with engine version breakdowns, sport splits, and timestamped publishing. Two great days do not make a model good. Two bad days do not make it bad. The full sample is the only honest scoreboard, and we publish all three markets on every game so you can pick the bets that fit your style.
The next test comes Friday May 2 with Game 7 between the Celtics and 76ers at TD Garden — a game the model will publish a fresh prediction on by tip. Whether the AI sticks with Boston (which would make it 0-for-3 in this series at 80% confidence) or finally pivots to Philadelphia after two straight losses is the question we'll find out together.
What's Next on the Channel
The Tuesday/Thursday/Saturday cadence continues from here. Expect:
- Slate preview videos anchoring each episode around 2-4 featured plays per night
- AiQ+ Picks recaps showing where the narrowed premium slate hit and missed for paid members
- Quick explainer shorts on confidence tiers, edge plays, and how the model thinks
- Drew solo episodes rotating in alongside Kendall as the second host
If you want to see the picks before the games go live, the full daily slate — with ML, ATS, and O/U picks for every game — is at prophetaiq.com. AiQ+ at $4.99/month for the first 200 founding members unlocks narrowed AiQ+ Picks plus Best Price line shopping across seven books.
Subscribe to the YouTube channel to get every new episode the day it drops, and follow @ProphetAIQ on X for between-show pick reactions and slate-day reminders.
The model is on the board every day. The receipts are public. Three-way grading on every game.
Frequently Asked Questions
Q: How did the AI picks actually do across the first two episodes?
A: Across all 33 graded games on the April 28 and April 30 night slates combined: ML 17-16 (51.5%), ATS 13-18 (41.9%), O/U 15-14 (51.7%). Episode 1 was the better night at 12-9 / 8-11 / 11-7. Episode 2 was rough at 5-7 / 5-7 / 4-7. On the 9 featured picks shown directly in the videos, the model went 4-5 ML, 4-5 ATS, and 6-3 O/U — the under tier outperformed even on a slate where the headline moneyline calls struggled.
Q: Why are there three picks per game?
A: ProphetAiQ runs the moneyline (ML), against-the-spread (ATS), and over/under (O/U) as three independent Groq calls inside the prediction engine. The ATS pick is not derived from the ML pick — the model is allowed to (and frequently does) say "Team A wins but Team B covers" or "Team A wins but the under cashes." This separation is intentional and lets users pick whichever market fits their style. We publish and grade all three.
Q: What was the worst pick of the two days?
A: The Celtics moneylines were a painful 0-for-2 across both episodes at 80% confidence each time. On April 28 Boston lost 113-97 at home as -450 favorites. On April 30 Boston lost 106-93 in Philadelphia. The 76ers won this matchup twice while the AI picked Boston twice. Both unders did cash as small consolation prizes (the games stayed lower-scoring), but the back-to-back high-confidence ML losses are the most notable miss of the two episodes.
Q: What was the best pick of the two days?
A: Two contenders. The Tigers @ Braves game on April 28 — the model called Atlanta ML / -1.5 / Under 8.5 at 65% confidence and the Braves delivered all three (ATL 5-2). The other was the Knicks @ Hawks game on April 30 — the model called NYK ML / NYK -2.5 / Under 213.5 at 80% confidence and New York delivered a 140-89 win. ML and ATS cashed by historic margins; the under at 213.5 missed because the Knicks alone scored 140.
Q: Where can I find recap pages for individual games?
A: Every game ProphetAiQ predicts has its own recap page at prophetaiq.com/game/{game_id}/recap, with the AI prediction, reasoning, final score, box score data, and three-way grading. All 33 games covered in this post are linked inline above.
Q: Did Episode 2 feature a new host?
A: Yes — Drew Sterling debuted as the second AI host in Episode 2. Drew joins Kendall on the channel rotation and will appear in roughly 30% of episodes going forward, with deeper dives on specific matchups planned.
Q: How is the Record page different from this recap?
A: The Record page tracks every prediction the model has ever made across all sports, with engine version breakdowns, sport-specific splits, three-way grading, and timestamped publishing. This blog post is just a 2-day snapshot framed around the YouTube content. The Record page is the comprehensive accountability layer.
Q: How often does the YouTube channel publish?
A: The current cadence is Tuesday, Thursday, and Saturday at roughly 9-10am CT. Episodes run 4-6 minutes typically, with longer formats for slate previews on busy nights and shorter explainer shorts mixed in.
Q: Where can I watch the videos and try the model?
A: YouTube channel: youtube.com/@ProphetAiQ. The full AI prediction slate (with ML, ATS, and O/U picks on every game) and the free BetTracker live at prophetaiq.com. AiQ+ at $4.99/month for the first 200 founding members unlocks narrowed AiQ+ Picks and Best Price line shopping across seven sportsbooks.