It's All In Black and White (and Red)

I've been following, studying, and watching horse racing for a long time. Sure, I'm a huge fan of the sport, the majesty of it all, and of course THE HORSES; and I'd be lying if I said that wagering on the horses wasn't a part of my experience. For some background, my horse racing experience all started with going to Canterbury Downs with my dad as a kid. For years, and all the way through high school and college, my dad and I would develop spreadsheets to decipher the most obscure patterns to try to give ourselves an advantage at the track. Performing statistical analyses on horse racing has stuck with me all through these years. This statistical analysis and handicapping are my favorite aspects of the sport; whether I am actually wagering 'real money' or not. 

So what's the purpose of this blog then? (I'm glad you asked) I performed a thorough statistical analysis of all of my wagers over the last three-plus years (well since 1/1/2010 to be exact). I wanted to know where my strengths were, which types of wagers I win most, and where my best ROI exists. Now to qualify this, I only used data that I downloaded from my ADW and these stats do not include wagers actually made at the track. However the total number of wagers is statistically significant and I feel that the stats on the wagers made on site at the track would closely follow suit. Now I may not be as good or as thorough as others, but this analysis which I am about to share with you certainly opened my eyes. I now know, historically, where I have won big, and where I should just stop putting my money. I have consulted some very good handicappers with these results to help validate my self-analysis and I hope you enjoy this report!

(WARNING: This report is lengthy and may cure insomnia)

To explain the charts a little, there are three columns for each slice of data. BETS WON represent just that. The percentage of those kinds of wagers that actually won. BETS w/ PROFIT are bets that actually turned a profit. For example, if I wagered $5 on an exacta and it only paid $4.20, then that would represent a bet won, yet it was not profitable. The return on investment (ROI) is represented as a percentage. -100% would be a total loss, 0% would be break-even, and +100% would be doubling my money. 

First I did an analysis of the kind of wagers I have made. You know, bets to win, place, exactas, doubles, pick 4's etc. Whenever I would lose on exotics for extended periods (repeatedly finishing 2nd and 3rd for example in typical Sorenson style), I would get extremely frustrated for always being so close. I finally gave up on exotics when my friend BSM challenged me to bet to win only. And sure enough, when that happened I was hitting more winners! But was I making more money?

(Now, in these reports I will not reveal the 'N' as I would rather not reveal how much I am (or am not) playing, but remember this is three years worth of data.)

The answer to this question is a resounding NO! It is quite apparent that even though I was winning the same amount of win bets as exactas, I was losing money on win bets and profiting on the exactas! Take a look at the ROI on daily doubles! WOW! And the Pick 4's... well they are off the charts. Why the heck wasn't I playing these bets more often? For what it's worth, the 'N" on the win/show and place/show wagers is extremely small. My record on pick 5, pick 6, and pick 9 wagers is, well, bad! (really, really bad)

(Let me qualify something, sure I may do well on my own Pick 4's, but ask Brian and Scott...whenever I join others' tickets, they LOSE!)

OK, So then I decided to generalize the wagers a bit more and classify the wagers in this slice:

At the highest level, I lose money on all straight bets and profit on exotics. Apparently it didn't really matter which exotics I played, the profit was just there. However one thing did jump out at me. When I box the verical wagers, I make a profit; and when I attempt to ladder or key the veritcal wagers, I lose. Interesting. I had always thought that straight bets were more profitable, at least for me. Clearly that is not the case. Joe DePaolo and I had a discussion at the 2011 Breeders' Cup regarding the profitability of exotics (and beating the take out rates). All the way up to when I did this analysis I thought I was right and he was crazy. Joe, you were right!

Now here is another interesting slice. I took a look at ROI based on how much was wagered in any one bet (total stake). Yes a strange way to look at it yet I was curious.

Whoa. Apparently there is a rather large ROI when the total stake of the wager is greater than $20. Well that does make a little sense as most exotic wagers, especially pick 4's, do cost a little more per wager. I wonder what it means when the wagers between $2 and $10 are losing money?

I also wanted to see how I performed at each track. As you may know, Arlington was my home track for years and therefore had the highest sample of wagers.

Can it be? Can I be THAT bad at THAT many tracks? The results are astounding. Apparently I should just stick with Churchill and (the soon to be defunct) Hollywood. The Colonial ROI is skewed because of one large score on a very small sample size. I'm not that far off from break even at Hawthorne, Meydan, Fair Grounds, and Keeneland. My performance at Arlington is very disappointing considering it is the track which I know the most about.

I took another swipe at a track by track analysis, this time looking at exotic wagers only.

Well that sure changes things a bit. Again, throwing out Colonial due to the very small sample size, my performance is a little better at some tracks. And again how about Churchill? Sick. Canterbury is a minor loss even though I won 50% of my wagers there.

Finally, I took a very different slice of the data. I wanted to know how I performed when I handicapped races versus when I just did an 'impulse' or 'hunch' bet. Those bets are defined as ones where I did not handicap the race at all using a racing form, followed someone else's picks (Facebook, Twitter, etc.), or just decided to throw a few bucks on a horse or two while watching TVG. These results are amazing.

Whoa. Only 9% of 'impulse' bets turned a profit. Amazing. Still, I know I'm a better horseplayer than this. I have to be. C'mon now I've been following this sport for 25 years!

So when I removed all of the impulse bets all together and re-ran the analysis, this is what I found: 

• - Exacta ROI - jumps from +30% to +54% 
• - All exotics ROI - jumps from +24% to +40% 
• - When I box verticals ROI - jumps from +29% to +52% WOW! 
• - Keeneland ROI - jumps from -7% to +16% 
• - Hawthorne ROI - jumps from -1% to +7% 
• - Arlington ROI - jumps from -4% to +6% 
• - Daily Double ROI - jumps from +74% to +97%

There were improvements in MANY other categories as well!

So here are the lessons learned:

• - It's time to focus on wagers that are more proftable. For me they are doubles, exactas, and pick 4's. This can be dangerous. I must not change when I would normally use or not use these wagers. Rather I must...
  
  
• - Avoid straight bets for now. Yes, there is a time and a place for each kind of wager. Different wagers exist for different situations. However, in three years I have been unable to turn a profit on a straight wager. So...it's time to do something different.
  
  
• - And on all that is holy, STOP IMPULSE WAGERS! It's a serious detriment to ROI.

I have many people to thank for helping me with this analysis, and you know who you are. I realize that, as my friend Stats Master E says, there are infinite ways to slice the handicapping pie. What works well for me may not work that well for you, and vice-versa.

Well there you have it.  I thank you for your patience reading through this report. What do you think of this analysis? Do you keep a wager log? What is your ROI? How would you have done this differently? Let me know!

Part II of the series can be found if you click HERE.