TrackSmart AI: Cumulative Performance Report - NYRA Tracks

Reporting Period: December 26, 2025 – March 8, 2026

Track: Aqueduct (AQU)

  1. Executive Summary: March 8 Update

Volume Expansion: The tracking dataset has grown to 303 total races with the addition of 32 races run across four cards from March 5 to March 8.

Strike Rate Resilience: The cumulative Top 4 Strike Rate increased slightly to 78.9%. The model was incredibly accurate this week, hitting on 84.4% of its races, capped off by an impressive 88.9% strike rate on March 7 and an absolutely perfect 8-for-8 (100%) card on March 8.

Value Normalization: The Average Payout for AI hits saw a downward shift from ~$8.05 to ~$7.87. This was primarily driven by extremely chalk-heavy, formful days on March 5 and March 8 where logical favorites dominated, though it was beautifully offset by catching a massive $37.16 upset winner on March 7.

Alternate/Scratch Protection Strength: The AI's alternate picks and scratch logic proved essential again this week. The protocol nailed a $37.16 winner (Victory Hall) on March 7 and seamlessly elevated winning horses (Ah Ca Ira) into the active selections following late scratches to preserve a perfect card on March 8.

  1. Daily Performance Breakdown

 

 

Date

 

Track

Top 4 Strike Rate

Total Races

1st Pick

2nd Pick

3rd Pick

4th Pick

Avg Payout (AI Hits)

Prev. Totals

 

AQU

 

78.2%

 

271

 

81

 

50

 

52

 

29

 

~$8.05

Mar 5

AQU

75.0%

8

5

0

0

1

$4.31

Mar 6

AQU

71.4%

7

4

0

0

1

$8.02

Mar 7

AQU

88.9%

9

3

3

1

1

$8.76

Mar 8

AQU

100.0%

8

5

2

1

0

$4.86

NEW TOTALS

 

AQU

 

78.9%

 

303

 

98

 

55

 

54

 

32

 

~$7.87

 

  1. Winner Distribution by Rank (Cumulative)

 

AI Rank        Wins      % of        Trend Note

Hits

 

Top Pick (#1)

 

98

 

41.0%

Absolute Dominance: The top pick exploded this week with 17 winners, accounting for over 60% of the week's total hits. Showed incredible precision on Mar 5 and Mar 8 with 5 top-pick winners each day.

2nd Selection

 

55

 

23.0%

Weekend Surge: Went completely cold on Mar 5 and Mar 6, but bounced back late in the week with 5 hits across the weekend cards.

 

3rd Selection

 

54

 

22.6%

Losing Share: Dropped to 3rd place in overall volume, contributing only 2 winners this entire week. The model's stalkers and mid-pack runners in this slot failed to capitalize.

4th Selection

 

32

 

13.4%

The Value Safety Net: Continues to deliver exactly when needed, hitting 3 winners this week—highlighted by an incredible $37.16 payout on March 7.

 

  1. Cumulative Statistics Summary

 

Metric          Stat         Description

Overall Accuracy

 

78.9%

(239/303 Races) The AI has successfully placed the winner in its Top 4 selections in 239 of the 303 recorded races.

Top 2 Accuracy

 

50.5%

(153/303 Races) The primary or secondary choice now wins over half of all races, surging significantly after a dominant week of top-pick performance.

Avg Payout

 

~$7.87

The average payout adjusted slightly downward below the $8 mark, driven by chalk-heavy, predictable pace scenarios on March 5 and March 8.

 

  1. Aggregate Observations: Mar 5 - Mar 8 Insights

Top Pick Precision on Front-Runners (All Week): The model was incredibly dialed in when flagging horses with a 40%+ Win Probability and tactical/front-running speed. Top selections dominated the week with blowout victories, successfully isolating logical class plungers and heavily favoring tactical speed on a track bias that rewarded forward placement.

Wet Track vs. Pace Meltdown Complexities (Mar 5): The AI outsmarted itself on a sloppy track in Race 7 by heavily penalizing a high-speed horse (What's Up Bro) due to a projected pace duel. The duel never materialized and the speed held. Corrective Action: Simple speed and rail bias adherence needs to override complex pace-meltdown algorithms on wet, speed-favoring tracks.

Stalker Miscalculations (Mar 6): Misses on Friday occurred in races where the winner employed a stalking/closing running style. While the AI perfectly predicted the contested pace scenarios that would set up a closer, it missed on the individual horse ranking algorithm. Corrective Action: Refinement is needed in weighting tactical stalkers who have demonstrated success in contested pace matchups.

Upset Detection & Scratch Protection Excellence (Mar 7 - 8): The AI's safety net protocols performed flawlessly. The system correctly snagged the massive $37.16 longshot of the day on Mar 7 (Victory Hall) via the alternate spot. On Sunday, the dynamic scratch protocol protected the perfect 8-for-8 ticket by automatically shifting horses like Ah Ca Ira into active winning positions after late scratches ruined the original pace structure.

THE MACHINE (AKA) TrackSmart AI isn’t a tip sheet or a picks website — it’s a professional-grade handicapping engine that works directly with the Past Performances (PPs) you already use from DRF, Brisnet, and Australia's Racing and Sports. Each version of TrackSmart AI — DRF, Brisnet, Harness, and International — is custom-engineered to read and interpret every printed detail in those PPs. No guessing. No databases. No shortcuts. The AI performs the deep work for you — comparing figures, detecting trainer intent, modeling pace, and uncovering hidden value plays — with the speed and accuracy of a professional analytics team. You stay in control of your handicapping while AI handles the heavy lifting. Pace projections, class and form analysis, TPN rankings, and betting strategy — all built from the same data you’d normally spend hours studying.