Letter Grade ROI at Oaklawn


A satisfied customer (their words not mine) has been compiling this data all meet.  We waited until Arkansas Derby week to ensure large enough sample size to post this blog.  My sincerest thanks to him for pulling this data together all meet and allowing me to write this blog.

First of all this data covers 467 races all meet at Oaklawn.  Oaklawn is a tough circuit even though its only on dirt with large competitive fields.  Additionally there was a substantial rail bias for a good portion of the meet.  This has cooled off lately but the rail ROI is still 0.97 (1.0 is break even).

We will be discussing this screen shot below and leverage some learning.  The ABC letter grade is a handicapping short cut that looks at all the chart data and gives points to different horses.  Then different letter grades get assigned at various point levels.  The ABC algorithm is logical but also arbitrary.  CHV could assign points or letter grades differently.

The PLUS or + designation goes to a single horse per race with best combination of ABC letter grade & VALUE.  VALUE is determined by comparing the minimum of PFDS & BEST Odds to the Morning Line.  Again, this algorithm has been in place for over 2 years.  It could be improved with machine learning at some point.  If someone asks, which horse should they bet on, the PLUS horse is the first one they should look at.

At Oaklawn, this meet, all K horses win at 37.3% rate, with an ROI of 0.84.  Although these horses have high win %, the ROI is lagging due to often being short priced favorites.  With that said, a K horse at 2-1 odds or greater is a good wager.  K horses should be used in serial wagers defensively.

A "s" before the letter grade designates a spot play (see blog on this topic).  These horses have strong VALUE and no obvious warning signs.  sK+ win at 47.4% rate with 1.53 ROI (sample size is small).  K+ horses on other hand have a disappointing ROI of 0.73.  This is because these horses usually have negative VALUE but earn the PLUS because there is no other horse worth wagering on in the race.

All A horses win at 27.5%, with a solid ROI of 0.95.  In particular, A+ horses have a ROI of 1.21.  sA+ horses surprisingly have a ROI of 0.63 (sample size is small).  This is probably just a small sample size anomaly.

B horses win at 17.9% with an above average ROI of 0.88.  If you combine B+ & sB+ you have a ROI of 1.19 with a large sample size.  Both the A+ & B+ grades are finding overlaid horses you can wager on profitably.

C horses win at 12.6% and a poor 0.75 ROI.  You should only be using these horses as back-ups if they are long-shots.

The PLUS + designation is really working.  First of all, look at the average prices.  The K+ and sK+ prices are better than K.  The A+ and sA+ prices are better than A.  The B+ and sB+ prices are better than B.  Even the C+ price is better than C.  And although there have only be 24 blank PLUS horses, three have won at an average price of $25, ROI 1.56.

All PLUS horses are winning at 22.6% rate and 0.99 ROI.  This is a sample size of 426 horses.  Every race has a PLUS horse unless it is scratched.  Having an automatic method to selecting a horse in every race get a 0.99 ROI is phenomenal.  Still think the PLUS algorithm can be improved but this is very encouraging.

Since C+ horses are performing so poorly, if we remove them and the blank + horses (to be consistent) from the wagering menu, you improve the ROI to 1.04 (4% profit).  This is still 323 races (69% of races).  Finally, if you filter out K+ from your win wagers you still bet on 226 horses (48% of races) with a ROI of 1.17 (17% profit).  Let me know if you find an alternative handicapping method that can drive that kind of profitability with such a high volume of wagers.

Finally all spot play horses (sX) also have a 17% profit level.  The sample size of 43 is still small.  

If you got any questions on this data or learnings, post below or send me a message on twitter.