Be The Smart Money - August 25th, 2022

The Myth

“You must bet horses that are bet down from their Morning Line odds.”

We have all seen it…. 8:1 ML horse is 3:1 at post-time.  Should we care as handicappers?  Is the proverbial ‘smart money’ driving this horse down because they have some insider information; and are those horses are more likely to win and produce consistent higher ROI?



The morning line (ML) at Saratoga is produced by Aragon since 2018.  The ML cannot account for scratches or changes in weather/track conditions as it is produced a few days before race day.  For this analysis we adjust the ML to account for scratches by using implied probabilities of the ML.  This isn’t perfect but it is the best way to compare ML to the Final Odds of any horse.


Data Speaks Louder than Words

We segmented horses into 3 groups: 

Group A (Final Odds are significantly lower than the adjusted ML (aML))

Group B (Final Odds are significantly higher than the adjusted ML (aML))

Group C (All Other)



The Data Set (Same Data)

  • 4.5+ years of data from Saratoga (2018 through August 22, 2022)
    • Aragon produced MLs
  • All races on FAST DIRT or FIRM TURF
    • Eliminates betting odds changes due to weather
  • Excluding races with dead heats, entries and DQs
  • Excluding races that switched surfaces
  • Excluding races that went from Inner to Outer Turf (and vice versa)
  • Clean & rich sample of 1,261 Races with 9,896 horses ran



The Analysis & Results (with the Raw Data)


                The Segments

Group A:  806 Horses (average ML 8:1; average Final Odds 4:1)

Group B:  1,030 Horses (average ML 15:1; average Final Odds 41:1)

Group C:  8,060 Horses (average ML 9:1; average Final Odds 11:1)


Note:  As you can see from the averages lower odds ML horses are more likely to drop lower than higher ML Odd horses



To illustrate

Group A Examples from this past Saturday (8/20/22):

*IP = Implied Probability to win


Group B Examples from this past Saturday (8/20/22):

Note:  Race2 #5 and Race9 #5 both are 5:1 ML but have slightly different aML IPs.  This is due to adjusting the ML to account for scratches.  The #8, #9 and #10 were scratched from Race 2; and only #5 was scratched from Race 9.


Ok…. Enough set-up …. Show me the money



Expected Wins vs. Actual Wins


Based on the Implied Probabilites we can calculate the number of winners we expect out of each group by the adjusted ML (aML) and the Final Odds (FO)



So – the Final Odds are more accurate in predicting the number of winners when there is a large difference between aML and Final Odds.


Therefore, when odds drop on a horse it is more likely to win than the ML suggest; and if the odds increase on a horse, it is less likely to win than the ML suggests.  


Not Surprising – the betting pool is a better predictor than the ML maker.


Thanks for nothing


Wait – there is more



Payouts & Final Analysis


Here is where this analysis may change your betting/handicapping strategy….



Large swings from the morning line, up or down, produce significantly lower ROIs.




They are over-bet…  

Gambler A sees a 12:1 ML horse going off at 30+:1 throws a few bucks at the longshot to win.

Gambler B sees a 6:1 ML horse being bet down to a favorite and throws a few bucks on the horse to win.


Surprisingly, these bets create a bias in the parimutuel.




So… don’t chase the ‘smart money’ or the longshots floating up to higher odds

                … well, unless you know something.



Myth Busted?  (YES)


“You must bet horses that are bet down from their Morning Line odds.”

You will pick up more winners betting the horses that drop significantly from the ML; but you are betting into a significantly negative ROI group of horses.



Be The Smart Money

Transforming How to Handicap for Value

Get your Smart Sheets for Saratoga and Del Mar at