Season 19 by the Numbers

Season 19 is in the rearview mirror, and with a couple of weeks distance, we can look back on some of the key facts and figures with some clarity. What are the patterns in the data that we can learn from and use in our own stable's racing strategy? How does your strategy compare to what everyone else is doing?

Juveniles in Season 19

One of the key trends within the web3 game photofinish.live (PFL) is the expansion of the horse racing population. At the forefront of this is the class of new juveniles (2 year olds). In season 19 over ten thousand juveniles hit the track for the first time. That's 12.7% more than Season 18.

I feel that people give up on their juveniles too quickly too often. This obviously presents the keen eyed with opportunities in the marketplace, but it's a shame that owners don't explore different distances to determine whether a horse could be competitive under its optimal distance and preferred conditions. Without racing it, how would you know? The stud barn is full of horses with amazing attributes that were never run at the appropriate distance and therefore underperformed on the track. Great horses come through in small proportions, and the last thing you want to do is miss out on a would be champion.

Of the 10,835 juveniles that ran at least one race in Season 19, 1,563 ran exactly once. 14.4% of all juveniles were given up on after one run. Another 1,315 ran twice only, which is an additional 12.2%. So a quarter of all juveniles are only run a maximum of twice. This is very early to determine whether a horse is going to be profitable for you.

I know what you are thinking, yes a typical negative binomial distribution. The bars represent the number of horses that have run at least the number of races indicated on the x-axis. We start with 10.8k racing at least once, 9.3k getting to the second race and 8k getting to their third race. The line represents what percentage of the remaining horses make it to the next race. What we see is that for each race we lose about 12-13% of runners (the retention rate is 87-88%) until we get to about 12 runs and then less and less horses make it to the next race. You have to be pretty on it to run a juvenile 24 times in a season, but only 54% of all juveniles make it to their 5th race.

So about 75% of juveniles make it to their 3rd race and 50% to the 5th. This is evidence of giving up too early in my book, and I am sure there are bargains to be had amongst the prematurely sold horses.

If we look at the last day each horse raced in a season we can plot a survival curve, starting with 100% still active on day one of the season to about 5% of those horses racing on the last day of the season. We all know that week 4 represents a fairly large fall off when people start to conserve resources and get ready for the next season. However it maybe more surprising that it's in week 1 where we see the other largest falloff where we lose about 25% of active runners. Couple this with the previous data about juveniles and I feel the week 1 fall off is probably larger than it should be. Maybe there could be more reasons for people to keep running their new juveniles through week 1. The grade specific mini-majors definitely help but this is something to keep an eye on. Maybe Season 20 saw an improvement in these fall offs with the introduction of the popular new claimer races.

Overall 6.367 of the active juveniles didn't win a race in Season 19, that's 59% of the total. 22% won one race, 9% won two and 10% won three plus. So your chances of getting 3+ wins from a random new juvenile in its first season is about 10%. Again it would be interesting to see how the claimers affect this number in season 20. More importantly, Testamatta from Wolaf the Bandito's stable won 15 races in her juvenile season, three wins clear of her nearest rival and should probably be crowned rookie of the season.

Another way to look for the best juveniles is to apply a rating system, such as Microsoft's Trueskill rating. I post these semi-regularly to Twitter/X, often with little explanation. So let me just explain that Trueskill was originally designed for online pvp games. Horse racing is a kind of online pvp game, so it can be applied to PFL. It doesn't look at fastest times or other factors - it just determines whether a horse beat or lost to every other horse in a race as a series of pairs. From this it calculates a score that starts at zero, and over time goes up or down depending on the quality of opposition it is beating or not. I use the Python implementation of it found here and the wiki page has more details. Here is a list of the top 40 juveniles from season 19 according to their trueskill rating:

This table is based on the best trueskill rating achieved by each horse. Whilst Testamatta ended the season with the highest rating, other horses achieved higher peaks during the season. It's an interesting metric, and always good to see your horses on the list but its not as predictive a measure of future performance as something like FF2 from PhotoFinishEdge. Some people ask why A+ horses are amongst the S+ and the reason is if an A+ horse has only ever raced against other A+ horses and dominated them, then its score will be high. Should that A+ horse run into a lower rated S+ horse, of course the S+ would probably win. So think of it as a relative measure, and don't go buying at top end A+ thinking it will hold its own against S+ rivals, it won't.

A shout out to All Gin No Tonic, who despite being ranked 314, was the number 1 ranked LTF horse in this much smaller population archetype. Since retirement this horse was bred 4 times in Season 20, maybe a stud worth considering if LTF is your bag.

Beyond Juveniles

Continuing the theme of trueskill, let's take a look at the non juvenile racers in Season 19.

Another way of looking at trueskill is plotting the score over time. This illustrates how scores fluctuate as races are won and lost. A poor performance, from running your horse off preference for example, can have a significant downward pressure on the rating.

In Season 19, Disco Dan (S-) and the enigmatically named Brycebane (A+) achieved the highest end of season scores. Others had higher scores earlier in the season but these horses end the season on top.

Looking at racing activity across the age bands, you can see the sheer volume of juveniles running compared to older horses. Obviously as the population grows, we will see larger juvenile cohorts. These accounted for 52.6% of all active horses in Season 19, and 58.8% of all runners. Juveniles are run 6.5 times on average, higher than any other age band. With the changes to the cost of breeding, and the viability of continuing to run your horses in the revamped claimers maybe we will see fewer horses retired at the first chance. This table demonstrates that older horses still win at the same rate as their younger counterparts. Maybe it is harder to find their favourite races at the same frequency, but there are new older races to increase these opportunities and encourage owners to keep racing their horses.

In Season 19, less than 10% of active horses were aged 5 or older. For those looking for opportunities, maybe this underraced segment is an area to explore. Older horses are generally cheaper to buy, and a maybe a good place to learn the game for newer players.

Another aspect that some stables may be unaware of is the relative competition level in each archetype. The above table looks at the distribution of active horses and their high level performance. Currently LTF is the smallest archetype, supplying only 1.8% of horses and 1.2% of all runners in Season 19. This group also make the least number of runs on average, perhaps impeded by the lack of opportunities given the small population size. There are clearly prizes awaiting those that can generate a high preference runner in that archetype. LTS has a higher win rate than LTF, as its stronger cohort of horses are better equipped to compete in yielding conditions and even into good conditions.

At the other end of the spectrum RDF represents almost 22% of horses and over 24% of all runners. This is probably the most difficult archetype to dominate in given the level of competition. Some canny breeders may consider the jump to the much smaller RDS community in search of less competition and an advantage on slower ground.

Financial Performance

Ultimately, PFL is about profitability. For this post we are going to focus purely on on-track performance. The economy obviously supports many different ways to make a profit in the game, but that's another post for another day. In order to keep things simple, I am going to convert Crown to Derby at a USD price of $0.40. There is no science behind this beyond looking at the exchange rate and finding a round number that looks to be accurate for Season 19. You can't look at profitability without including Crown and its relative value. The problem is, if Crown is now $0.6 or even $6, this retrospective will always look wrong. So bear that in mind if you read this in the future when Crown has mooned.

At a global level, racing in Season 19 was profitable as an entire ecosystem. Almost 120,000 race entries spent 171.9MM Derby on fees. In return 155MM Derby was paid out as winnings alongside 37.4MM Derby in Crown stimies (converted at $0.40). This translates to 20MM more Derby being paid out than paid in, or +171 Derby per horse entered into a race. Now, it may not feel like everyone is profiting from the game because, clearly, the profits are concentrated in the elite horses who win the majors. That said, it's a good sign for the economy that we are playing a game where the house edge is currently returned to the punters.External sponsorship will take this to another level.

As an owner you need to figure out, fairly swiftly, which one of these three groups you horse is in.

The somewhat unfortunate reality of that 20MM Derby profit is it is not distributed equally. Like most distributions in life there is a Pareto effect, the 80:20 rule. In our example here, the top 10% of horses (2k horses) generate 50MM in profit. The next 20% just 5MM between them. Then we have the bottom 70% who basically are unprofitable. A question you should ask yourself, when persevering with a horse, is "is this horse a top 10%, a top 10-30%, or a bottom 70%?". The thing about Pareto effects, is that they are fractal in nature. If you dig into the top 20%, you will find that the top 20% of the top 20% generate 80% of 80% of value, like Matryoshka dolls. Who is at the very top of this? Well the major winners are at the top because the prizes for winning are much larger than any alternative race. The top 25 horses on their own (top 0.1%) account for 8MM of the overall profit generated. 376k is prize for winning a major.

Based on the calculations I explained earlier Spee-Dee, the Artemis Cup and Rolling Thunder S+ 10F winner, generated the most net profit of just under 600k Derby. Retired shortly afterwards, and maxed out for breeding in Season 20, half of which being external 24k breeds, he will continue to pull in profit long after Season 19 ended. If I had the choice I would rather be at the top of this table than the Trueskill one!

For contrast, Lawson creeps onto the top 25 with the lowest profit per race, but a high profit nevertheless. I successfully made several unsuccessful picks on Lawson in Season 19, so its good to see he was profitable for his stable, if not the betting public...

A final table to round out this post of complicated charts and tables. We can take broadly the same data and aggregate it by grade. Obviously there are a few C's and D's hanging around, but the bulk of the action is in the A and S classes. Season 19 saw a rise in S grade and similar levels of S- horses, representing over 60% of all active horses. S- as a grade in particular is a competitive level, so much so that is less profitable than any of the A grades at an aggregate level. The S- grade level grew by 500 horses from Season 18 to 19, but the S grade increase by over a 1,000. See the below chart for how this trend has evolved over the seasons.

Quick Recap

Season 19 is in the books. There were some significant changes to the schedule and introduction of new races. Season 20 has seen further changes to that schedule, particularly the revamped claimer races, which may affect the patterns we have explored in this article. It used to be that if a horse won 20% of its races and had a podium rate of 50% it was legit. But winning multiple low value claimers can alter those ratios and you need to look a bit deeper to ensure a horse's benchmark is not being artificially boosted by success in those races.

That's the beauty of the game I guess. With an evolving schedule, introduction of external sponsors, improved distribution (vital to any start-up) via the mobile and Roku app, the economy is likely to go through several more evolutions.

Some observations from this analysis:

1. Don't give up on your juveniles too quickly

2. 25% of juveniles who have been raced once don't even make it to their 3rd start

3. Juveniles represent over 50% of active horses and almost 60% of all races

4. Less than 10% of active horses are aged 5 or older

5. Some archetypes are more competitive than others, LTF represents only 1.8% of runners

6. The economy was healthy in Season 19 with a net positive contribution to owners

7. However, that global figure disguises the fact that there is a strong Pareto effect (80:20 rule) and the majority of profits go to the best horses

8. The top 10% of horses generated 50MM Derby in racing profit, the next 20% 5MM and the lower 70% contributed to a 35MM loss

9. Spee-Dee was the most profitable horse with circa 600k Derby profit (if you convert Crown to Derby)

10. The population is shifting towards S grade as the proportion of As as a % shrinks every season

I don't intend to write up every season as a regular report, this was more of a one off look at a season in isolation. I am sure I missed many important events, trends and insights. But now you have some of these insights you can conduct your own investigation into how you get onto the path to profitability. Just don't look at my profitability for any tips!

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=PADDOCK

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

Other blog posts:

• A race selection process flow: https://aipaddock.com/selecting-the-right-races-for-your-horses

• How good is your new juvenile: https://aipaddock.com/how-good-is-my-new-juvenile-horse

• Tips for breeding strategies: https://aipaddock.com/charting-the-uncharted-in-breeding-within-pfl

• Know your odds of breeding success: https://aipaddock.com/know-your-odds-before-you-roll-the-dice

• Further down the rabbit hole of ML: https://aipaddock.com/deeper-down-the-morning-line-rabbit-hole-we-go

• What can ML prices tell us: https://aipaddock.com/skill-luck-and-morning-lines

• Finding the optimal distance: https://aipaddock.com/understanding-optimal-distances

• Understanding breeding: https://aipaddock.com/understanding-the-most-successful-breeding-horses-in-pfl

• What are subgrades: https://aipaddock.com/what-are-subgrades-and-how-do-they-work

• Fastest horses: https://aipaddock.com/who-are-the-fastest-horses-in-the-game

• Racing profitability: https://aipaddock.com/how-much-profit-do-you-make-racing-horses

• Trueskill pvp ratings: https://aipaddock.com/which-horse-is-the-goat-in-pfl

• Breeding: https://aipaddock.com/trying-to-understand-breeding-in-pfl

• Evaluating horses: https://aipaddock.com/how-good-is-my-horse

• Understanding performance: https://aipaddock.com/understanding-the-true-performance-of-your-horse

• Do stars matter: https://aipaddock.com/how-much-do-preferences-matter

• FF Rating vs Finish Time: https://aipaddock.com/the-difference-between-ff-rating-and-finishing-time

• Are horses getting faster: https://aipaddock.com/are-pfl-horses-getting-faster