Would love to see consistent ROI calculation in racing media
 

When I watch/consume racing, depending on the source, I might get:

A) An ROI based on a $2 wager with the $2 wager included as part of the ROI (using the total return). This number will never be negative. The lowest it can be is $0.00. (Example: ROIs on NYRA's "Talking Horses")

B) An ROI based on a $2 wager that does not include the wager itself. This ROI can range from -$2.00 to +???.?? (Example: BRISnet's trainer statistics)

C) An ROI based on each $1 wagered, with the $1 included. (Example: can't point to a specific one, but have heard results expressed as something on the order of "you get back $1.23 for every $1 bet")

D) An ROI done the way I've seen most "standard" investment ROI expressed (where you can't lose more than your original investment). Nothing would go lower than -100%, and the positive could obviously go as high as warranted. (Example: STATS RaceLens when Christina Blacker posts an ROI for an angle)

Side by side confusion examples:

Trainer Jones wins 1 race in 4 starts with a mutuel of $10.00
NYRA: $2.50 ROI
BRISnet: 0.50
That guy: "Returns $1.25 for every $1 wagered"
STATS: 25% ROI

Trainer Smith wins 1 race in 5 starts with a mutuel of $8.00
NYRA: $1.60 ROI
BRISnet: -0.40
That guy: "You get back 80 cents on the dollar"
STATS: -20% ROI

I would like to see racing move towards the more "standard" way of expressing ROI (of course others may tell me that I'm mistaken and that way is not so "standard")

I personally use return per $1, but I'd be all for using whatever it is that would be named the standard. I agree it can be confusing.

I obviously do it a lot, and use $2 because that's what Formualtor does and I use that for stats, but I kind of wish it was for every $1 bet. It actually annoys me and I'm fairly responsible for it being widespread.

It gets even more confusing on exotics. Dime super, what do you use? Dollar exacta, what do you use?

I use the $1 return in my calculations. In the two examples you gave the first one would be a $net of 1.25 the second example would be a $net of .80. Just another and better way to me of saying a 25% gain or a 20% loss.

An ROI standard makes sense. That doesn't sound like a tough change for software developers.

We should also have a standard for "Probables" and "Will Pays". Most ADWs probably just display whatever data they get from the sources, but when you look at payoffs you sometimes have to double check whether it's for $1 or $2 or even 50 cents because it's not the same everywhere for every bet type.

Yes, very frustrating. And some don't reveal what the basis is.

On a similar note, I'd like to see standardized parameters and definitions of running styles, for purposes of profiling and comparing different surfaces.

The ones I use are personalized and quirky, thus rendering track-to-track comparisons impossible, so a standard "language" would be helpful. Some might think Brisnet has already become the standard?

And in one context, I mean "language" literally. The terms "stalker" and "presser," for instance, tend to get conflated, confused, and reversed, having no standardized definition-at least that I'm aware of.

We have that in TimeformUS. Not sure I'd want to share for standardization purposes though. :)

One of our styles is "Plodder" which apparently offended some people the beginning but I stuck to my guns. Horses with that label return less than 50 cents on the dollar last I checked.

I use my own categories.

Pure Speed
Front Runner
Near the Lead
All others

I’m only interested in the horses that may contribute to the pace.

Each horse is defined by a style. They also have a number so I can separate horses in the same category. The one thing I don’t have is a good “late” rating. It’s very complex to create a good late rating.

Me, too. And a lot of it involves the sub-classification of speeds. But I dig a lot to determine if a certain distance or race-shape might have caused a false read on somebody's running style.

I've been storing all my information for the last 6-7 year. That way I can keep going back and testing what works and what doesn't for each set of conditions with larger and larger sample sizes. There's still some value in running style and race flow type information, but it's very difficult to find anything that will yield a lot of plays (at least for me).

I make my strongest stands when I disagree with prevailing opinion on how a race will play out. Everybody bets the race-shape, and lots delude themselves that others cant see what is painfully obvious.

I am glad to have moved into the elite class of those who've seen their thread hijacked! :lol::lol:

Apologies, sir. I will desist.