How much do preferences matter?

In a game of partial knowledge, PFL tells you every horse's preferences right up front. But how much do they matter? Let's explore the role of preferences and try to dispel some of the folklore and myths that surround them today.

What are preferences?

Preferences are a horse's preference for track direction, surface and condition. There are two directions, left or right. Two surfaces, turf and dirt. Five actual conditions, but horses overtly only prefer two, firm and soft (but the actual range is fast, good, yielding, soft & sloppy).

Each of those attributes is rated on a 3 star scale with half star steps - meaning there are 7 steps per range and two opposites per attribute (the aforementioned right v left, turf v dirt and firm v soft).

Horses are predominantly run on preference, that is a horse that prefers right turning tracks on turf and soft conditions (or RTS) will mostly be run in those types of races. This is encouraged and if you use the in game Prefered Race Finder it will default to races that are RT and leave you to determine where you want to enter your horse based on the more variable aspect of weather. This is not to say that you can't run your horse against its preferences. Maybe you feel a certain combination of preferences (say LDF) is weak and you fancy your chances. Or your horse is so strong it can overcome its rivals in certain strength fields.

Some people believe preferences affect different horses to different extents. Maybe they are just seeing the randomness in performance, or maybe they are onto something. Let's check out some basic scenarios with some cold hard facts.

How much do preferences matter?

Let's start by taking a look at off preference versus on preference. All the data is taken from the recent season 13 and represents the full season.

Firstly almost 97% of runs are made on preference for surface and direction. Only 0.5% (1 in 200) runs are made completely off preference. Perhaps that's a good thing because off preference runners are significantly less likely to win. Semi-preference runners (can't think of a better term) podium on a par with on-preference runners. However the real difference, and the more important factor, is seen on ROI (return on investment, which here is just winnings divided by entry fees with no considerations for CROWN). Rather than get into the overall profitability of horses (we will save this topic for another day), I have indexed profitability and from this you can see the off-preference runners of any type generate less than half the ROI of on-preference runners.

In summary off-preference running doesn't happen very often and when it does, as a population, it is less profitable. We haven't discussed condition here because it has more nuance so consider this analysis - "all other things being equal" (which they might not be, but we are sweeping this under the statistical carpet for now).

So out of left field recommendation number 1: run your horses on preference unless you have a very good reason not to.

"We are all in the gutter, but some of us are staring at the stars"

You often hear on Discord discourse about the number of stars a horse has. There are maximum of 9 stars and you get the star total by adding up the stars for direction, surface and condition. If you look at the marketplace and filter by grade, you will see horses are broadly ordered by the number of stars they have where more stars equals higher cost. You can understand why, stars are very visible on the listings, and for people who can't be bothered to review the bloodlines they might appear to be a key decision factor in whether to buy.

How important are stars? Well, let's take a look at the same dataset as before (full season 13) and compare the performance of all runs by horses grouped by how many stars they have.

What we see is a similar pattern to off-prefence. Here maybe surprisingly horses with less than 2 stars for direction preference represent only 10.8% of all runs. They win at about 2/3s the rate of high star horses and 0.5-0.6 of the ROI (where 1.0 = equal). Three star horses win 13.8% of their races versus 12.5% for 2.5 stars - so a small but significant difference

You will have to take my word for it that surface stars (whether turf or dirt) behave in a very similar way, with the performance improvement with every half star. Condition stars however behave slightly differently. Given that condition itself is more variable that is perhaps not a surprise. Regardless, condition stars are less sensitive to success (wins and ROI). A canny player maybe able to overcome a shortfall of condition stars with some careful race planning and a bit of luck. Whilst high stars do make a difference, the advantage is less marked here.

Firstly, only 10.7 of runs are made by horses with 3 stars for firm or soft (compared to almost 30% for direction and 25% for surface). Secondly, the win rate increases but much more slowly than for direction. 1 star condition horses win 12% of their runs versus 12.2% for all horses and 13.6% for 3 star condition horses. Finally, the ROI index is almost flat across star amounts.

Pulling it all together

So one final cut and paste table. Let's look at the same performance indicators against the total number of stars a horse has. This is a worthwhile exercise because it is a key building block in our understanding of the game. From this we can figure out is it worth it to pay a premium for 9 star horses on the marketplace? Should we be claiming 9 star horses that other stables have foolishly made available to us if the price is right? Under what circumstances would I be willing to buy a low preference horse?

I would say this data mostly speaks for itself. There is a clear straight line improvement; the more stars you have the higher the performance across all three metrics. 9 star horses win three times as frequently as <1 star horses. In a game where every little margin counts, it should be clear from this post that more stars equate to better performance. Yes there will always be amazing horses that don't have 7 stars let alone 9. But maybe your horse isn't amazing, just very good, and needs all the help it can get to beat amazing horses.

There may be some other factors at play. Are low preference horses perservered with less and given up on before the optimal racing conditions are found? Are low preference horses more likely to be used as CROWN miners? There could always be other factors at play but I think this broad logic still applies; the game is giving you information about every horse and you should use that information to your advantage, in racing, in breeding, in buying and in claiming.

Finally, as of the time of writing, that fine 3 match generation zero stud from the first image of this post (The Show Must Go On) is available on the marketplace for all you bloodline purity experts out there.

Join the fun and put these insights into practice at PhotoFinish.Live and if you are considering starting your own stable please consider using my referral code: PADDOCK or just click on this link: https://signup.photofinish.live/?referralCode=8EUMC4P2

Please remember this is a web3 game where your spend your own money. Nothing I write about should be considered financial or investment advice.