Race Cafe

ANOTHER BOARD. ANOTHER COST. WHEN DOES IT STOP?
Here we go again. NZTR announces another Board, another layer of administration and inevitably more cost to an industry already struggling to survive. They say this is about creating a “more sustainable and efficient industry.” The irony is unbelievable. Their answer to reducing costs appears to be creating another Board to talk about reducing costs. How much is this going to cost? Who is paying? What measurable savings will it actually produce?
Enough of the boards, committees, consultants, reviews and endless bureaucracy.Stop spending and start delivering.
Put the money back into racing, stakes, clubs and the people actually keeping this industry alive.
The costs keep exploding. Where are the results?

The absolute shambles of the Cody Cole case should raise serious questions about the competence of the RIB and the way this investigation was handled.
The RIB left so many holes in its investigation that the case was ultimately thrown out on appeal. That is absolutely shambolic.
The fundamental question remains: why was Cody Cole ever allowed to leave without taking the test? And once he had left, why was he not required to return and be tested at the earliest possible opportunity, including the very next day?
If somebody leaves when they are supposed to undergo a test, surely the most basic investigative response is to establish where they are and obtain that test as soon as possible. Why wasn’t that done?

Then there is the explanation involving the horse. How could this horse supposedly have been so difficult that if Cody Cole had to leave, not a single other person could take over and hold onto it while he completed the test? Was that explanation properly challenged? Was everybody involved questioned? Was the veterinary evidence thoroughly tested?
These are not minor details. They go to the very heart of the case How can this be called integrity when racing is paying around $18 million for the RIB, and this is the type of investigation being conducted? For that sort of money, the industry is entitled to expect investigations that are thorough, professional and capable of standing up when challenged.Instead, we have ended up with a case riddled with holes and ultimately overturned on appeal.

From the beginning, this case has created the perception of friends in the right places and questions about whether everybody is treated equally. Whether that perception is right or wrong, it is precisely why the RIB needed to conduct an absolutely watertight investigation. Instead, its own investigative failures have only fuelled those questions.This is an absolute disgrace.The people running racing continually talk about integrity. Well, integrity has to mean something. It requires proper investigations, proper procedures, accountability and the same standards being applied to everybody.
In this case the RIB has shown itself to be not fit for purpose. Those responsible for these failures should be held accountable, and serious consideration should be given to replacing the people responsible.

Eighteen million dollars for this? Give me a break.

Today is a very sad day as we remember the loss of, Taiki Yanagida, a super young man who tragically lost his life in a race fall that should never have happened. Taiki moved to New Zealand to chase his dream of becoming a jockey, and he was right in the middle of fulfilling that dream when his life was so tragically cut short. He will always be remembered as the smiling, positive and dedicated young man he was full of life, chasing his dream and with so much ahead of him. He was taken far too soon, while still in the prime of his life. Today we remember Taiki, his smile, his passion, his dedication and the dream he had that brought him to New Zealand.

Fly high, Tykie. You will never be forgotten . Rest easy, young man. ❤️

NZTR trying to cut around $10 million off the prize money this season with little to no consultation with anyone at the coalface of the racing industry, and yet they’re still advertising for more staff.WTF is going on. And there’s another new person just started, and now there’s another role being advertised. What planet are these guys on? There seems to be absolutely no talk about cutting their own administrative and running costs. Instead, they just want to cut the prize money for the very people who do all the work, put on the show, and ultimately keep these people in a job.

They all need throwing out now. The clubs need to get together and say this cannot continue. These people simply cannot stay.The industry must stand up, and we all need to work together to get enough clubs on board to vote these people out. Then we can get back to scratch and start again with people who actually understand the industry, understand the grassroots, and understand where the money should be going, instead of continually wasting it while apparently thinking they’re doing a great job.The industry must stand together and demand change.

Over the last two seasons, 5 clubs have all raced at Trentham: Wellington, Manawatū, Marton, Feilding and Rangitīkei.
Over 3 seasons, since 2023/24 season,...12 of the 41 meetings held at Trentham were transferred there from another club.

Trentham has become the Central Districts' contingency venue.
Project Stamina says that venue changes must protect race days, field sizes and wagering revenue. (have a look at the graphs relating to these 3 points)

NZTR's published stats (in Annual Reports) record race meetings against the club licence...and no acknowledgement of the venue where they were run.
Trentham's role in keeping racing going in the region is invisible.
Does the network have enough resilient venues to keep racing?
Thanks Shah Amner

The New Zealand thoroughbred industry .. do you think its a value proposition problem or a product problem?

Every industry survives because if it creates value for those involved.
Racing needs to create value for ...breeders, owners, trainers, jockeys, punters and race clubs.
But, the numbers below are showing that it isn't really happening:

· The foal crop has fallen 22% over the past decade and is projected to fall another 18% by 2035, from a base already 49% below the 1994/95 peak.

· More than 500 breeders have left since 2015.

· The trainer population has roughly halved over the past 20 years.

· Around 37% of foals are exported as yearlings (although this is good for the breeding sector)
These are some signals that the industry isnt generating sufficient returns across its supply chain…to solve the value & product problem.
Businesses don't lose suppliers because they have too many buildings….They lose suppliers because the economics doesnt stack up.
The response so far (there have been major reports over past 16 years) has largely been structural reform.

· 2010, NZTR Directions Paper and Venue Plan.

· 2017, IER Size and Scope Study.

· 2018, Messara Report.

· 2024, IER Economic Impact Report.
· Structure doesn't breed a foal.

· Structure doesn't convince someone to buy a yearling.

· Structure doesn't keep a trainer in business.

· Structure doesn't improve the owner's return on investment.

· Structure doesn't give punters.. bigger, more competitive fields

It's not simply about producing more horses.

Does the industry create enough value for every participant in the supply chain?

· If breeders can't make money, they stop breeding.

· If owners can't justify the cost, they stop buying.

· If trainers don't have enough horses in work, they leave.

· Fewer horses mean smaller fields.

· Smaller fields produce a less attractive betting product.

· Punters bet less. Turnover falls.

· Stakes come under pressure.

· That makes breeding, ownership and training even less attractive.
Then, instead of fixing the economics that caused the decline,…the intent is focused on “How to get more turnover landing into the betting pot e.g: move prestige’s races to another venue in the hope that betting profitability improves.
Would changing where a race is run, create another horse?

· Does it bring back a breeder?

· Does it encourage another owner to invest?

· Does it keep another trainer in the industry?

· It redistributes the existing product across fewer venues.

· All the while....the underlying supply problem remains.
Until the industry improves the economics of participation, every structural change will reshuffle & redistribute a shrinking pie.

When another track fails, which venue picks up the racing?



Over the last two seasons, 5 different racing clubs held race meetings at Trentham….Manawatū Racing Club, Marton Jockey Club, Feilding Jockey Club and Rangitīkei Racing Club and of course; Wellington Racing Club.

10 of the 30 meetings scheduled at Trentham across 2024/25 and 2025/26 were run under another club's licence.


Trentham is not just Wellington Racing Club's home course. It is the venue the Central Districts has been using as a fall back option.


Awapuni's course proper has been unavailable for most of the last four seasons…Reopened on ANZAC Day 2025 and then the meeting abandoned after one race when a horse slipped.


Project Stamina

Pg. 92:"Ensure that venue exits are sequenced so that receiving venues can demonstrate that the displaced demand can be absorbed without degrading field sizes, race days or race quality before the exiting venue is formally retired."

Pg 78:"Transition timing decisions must preserve (or improve) average field sizes, reduce abandonments, and improve broadcast reliability."

Pg 35:"Every cancelled or unavailable race day is a direct wagering revenue loss."

NZTR's published statistics, record the race meetings to the club holding the licence, not the venue where the racing happened.

When Marton Jockey Club runs a meeting at Trentham, it appears in Marton's line.

Every meeting that Trentham has absorbed for another club is invisible when counting the venue data.


Project Stamina's Table 9 lists every setting used: metropolitan centre count, horse-to-people weightings, training centres, race day thresholds, starts per horse, races per meeting, scratchings, field size floor and ceiling.


There is no resilience variable. No abandonment history. No measure of a venue's ability to take racing at short notice. The model doesnt ask which venue you need when a track goes down.


Track failure is a system-wide risk, and Trentham isnt immune to this either.

Dec 2021, abandoned a Group 1 raceday in when jockeys reported horses slipping.

Jan 2024, partial abandonment in on unirrigated ground

Jun 2026, a postponed a meeting due to continuous rain.


No New Zealand track has a clean sheet. Riccarton abandoned its 2022 NZ Cup Day after race three …how much would that’ve cost or lost?

Ellerslie has had partial abandonments on its brand new $55 million surface in April 2024.

Tracks are failing…more so now, than before… the grass needs renovation windows and rest periods,


Project Stamina pg. 35, says all-weather surfaces are described there as "complementary rather than substitutional". And go google what Bart Cummings views were about all-weather surfaces.

A network that cannot absorb a failure is a network that loses race days. NZTR 2024/25 annual report = 27 abandonments, postponements, transfers and reschedules in a single season..



You do not judge a spare tyre by how often you drive on it. You judge it by whether it is there when you need it.





Sources: Project Stamina pp.35, 66, 78, 83, 92; One Equine View p.13; NZTR Annual Report 2024/25, Key Financial and Performance Targets p.37 and Statement of Service Performance p.48; official race calendar 2024/25 and 2025/26; race card 28 March 202.

Everything in this series is compiled from publicly available sources. Other than being a small thoroughbred owner and breeder with one broodmare, I have no commercial, governance or financial interest in these matters, and I do not represent any club, organisation or stakeholder group.
Thanks Shah Amner

The Model That Cannot See Money

Project Stamina's venue recommendations rest on a P-median
optimisation model. The report describes it at pgs 52 and 108 - 110,
and sets out its parameters in Table 9 on pg. 66.
P-median is one of the oldest & best understood tools in operations
research, first created in the 1960s.

It was built for, and good at, problems like where to locate ambulance
stations, fast food resturants (supermarkets & petrol stations may use
it too). Cases where the service delivered is identical to wherever
you put it (think McDonalds), where the users have no real choice but
to attend the nearest one, and where the thing you are trying to
minimise really… is travel.

It cannot tell you how many facilities you should have; The p in
P-median is an input, not an output. You must decide the number in
advance. The model places that number optimally. Ask for 4 and you get
4 Ask for 6 and you get 6. The model has no view on which is correct,
and it will confidently produce an optimal 4-facility network for a
system that needs 6.

It cannot distinguish between facilities; In the classic formulation
every facility is identical, the standard model has no variable for
what a facility produces, what it costs, how good it is, or what it
earns.

It cannot value anything; There is no revenue term.. The objective is
distance. A P-median model cant tell you that a facility is worth
keeping, because worth is not a concept it contains.
A P-median model optimises for physical proximity. It assumes that
being close to the facility is what generates the value.

Distance to a racecourse is close to irrelevant to a punter in
Melbourne, who is opening an app at 9pm to have a punt. What matters
to that punter is whether the racing is worth betting on.

Project Stamina states that wagering has decoupled from on-course
attendance & that the majority of domestic wagering is now off-course.
It says it in the narrative & then builds the decision on a location
proximity model anyway.

The second structural problem is how the model's output was used.
P-median assigns demand to facilities as a consequence of where it put
the hubs. Race day allocation is not a measurement of anything.

Project Stamina focused the Central Districts on Awapuni as both a
metropolitan venue and a training centre. Race days were then added to
that anchor by construction. Trentham was left with four.

Pg. 73:: "Trentham, insufficient demand to sustain minimum 8 race day
threshold, only allocated 4 race days."
The model produced the four. The four was then treated as evidence of
insufficient demand.
That is the result of the model reading back its own input. No
performance measures that Trentham could achieve would change the
allocation, the allocation was determined by the anchor decision
before any performance was considered.
Every one of the eight venues recommended for exit failed the eight
day threshold.That is not eight venues failing a test. That is one
modelling decision producing eight consequences.
Page 97- The model is "a decision-support tool, not a predictive instrument."
It "has not been independently audited." "All quantitative outputs are
sensitive to the assumptions on which they are based, and small
changes to key assumptions may produce materially different results."
The report told everyone that small changes to the assumptions may
produce different results, and then never varied the key assumptions.

A P-median model is the right tool: if you were building a racing
network from scratch.
The better approach would be to use the model to generate a direction,
then overlay everything the model cannot see
e.g what each venue produces,
what it costs to run,
what condition it is in,
what programme it carries,
what happens to the network when a track fails, and
what the wagering consequence of each change would be.

For 5 of the 8 venues recommended for exit, NZTR did apply judgement
and then overrode the model, retaining Whanganui, Ashburton, Te Aroha,
Ōtaki and Timaru on network capacity, jumping role and regional
balance.The other 3 incl. Trentham, have no reason/criteria of
assessment has been published…yet.

What is needed:

#1 Publish the projected Central Districts wagering turnover with
Trentham retained vs. Trentham exited, and reconcile it against three
seasons of NZTR's club statistics.
#2. Publish the criteria that separated the five venues retained
against the model's advice from the three that were not.
#3 Commission the independent audit of the optimisation model that the
report notes has not been done.

[Everything I post is compiled from publicly available sources. Other
than being a small thoroughbred race owner and breeder with one
broodmare, I have no commercial, governance or financial interest in
these matters, and I do not represent any club, organisation or
stakeholder group.]

How we got here & where "Optimised" might take it

New Zealand's first recorded race meetings were at Epsom in Auckland and on Petone Beach, both in 1842. Clubs created everywhere through the 1860s and 70s (nearly every town of any size laid out a course, because the logic then was purely geographic): get the horses, the riders and the crowd together with the least possible travel.

By 1905 there were 51 venues mapped across both codes, both islands. Most of what still exists today is that same footprint, built for a population, a settlement pattern and a technology that stopped existing decades ago.

The new plan...Project Stamina, admits:

Population urbanised.
Horse populations concentrated in Auckland, Waikato and Canterbury.
Wagering moved almost entirely off-course.
The venue network didn't move with any of it.

The standardbred foal crop fallen from over 3,000 a year in the 2000s to just over 1,000 in 2024 - that gives you an indication of the number of horses that'll turn up to race in two to three years' time.

This isn't the first time someone's been sent in to fix it...The Messara Report in 2018 said what Project Stamina is saying...fewer, better-quality tracks, reinvest the surplus, stop spreading capital so thin it can't build anything.

The 2018 report adoption was "slower and less comprehensive than recommended," and some of the surplus land it flagged still hasn't been sold...8 years later. Urgency faded before the plan finished...

and now it's happening again... bigger and more detailed.

Project Stamina calls the current network "over-venued and under-capitalised"...(in simple terms... too many venues).

TAB NZ is sitting on reserves the codes need to balance next season's books and won't release them without structural reform.

Admin costs are up 45% and 38% since 2023. Combined admin now eats 58% of what actually reaches stakes.

Project Stamina sets out what the "optimised" network looks like if it's fully implemented.

For harness racing: 3 metropolitan venues, nationally (Cambridge Raceway, Addington, Winton).

For thoroughbred racing: 4 (Ellerslie, Te Rapa (or its Waikato Greenfields replacement), Awapuni, Riccarton Park.

Everything else drops to provincial or community tier, competing for what left.

Rationalising the venues doesn't address the actual constraint...the product...the horse. None of those 4 thoroughbred hubs or 3 harness hubs need to exist at metropolitan scale if there isn't a horse population left to fill them (and the supply side is shrinking faster than the venue network is). The plan's figures show the thoroughbred foal crop down 49% from its 1994/95 peak and the standardbred foal crop down 59% over two decades.

That matters more than the track closures do, the cost to own / produce a horse keeps rising while more of what the industry takes in goes to 'admin and facilities' rather than through to the people funding the product. ...the owners,

The masterplan confirms that black-type racing (G1, G1, G3, Listed races) should be "concentrated" at metropolitan venues, and that "community venues are generally not appropriate for black-type delivery."

One club president has already said publicly that his club tops up NZTR's stakes out of its own pocket because the baseline isn't enough.

Put those two things together and the pattern is straightforward: the meaningful prize money concentrates into fewer, bigger races at fewer, bigger venues, and the owners and studs with the depth to compete at that level are the ones best placed to keep winning it.

Everyone else is left racing for less, at a venue further away, for a cheque that buys less than it used to.

TRENTHAM WASN’T ASSESSED. IT WAS ALLOCATED.

“Trentham… insufficient demand to sustain minimum 8 race day threshold, only allocated 4 race days.” pg.73 Project Stamina

4 race days allocated. Below the 8-day floor criteria for a provincial venue = an Exit status for Trentham.

How was that 4 allocation produced?… it wasn’t a measurement of anything Trentham did.

The analytical engine of Project Stamina is called a “P-median optimisation model” (pg.52). Simply: you tell “the model” (who we’ll call P-Median) how many hub venues you want, where the horses and people are, and how important each of those matters. It then places the hubs to minimise total travel, and then distributes race days accordingly.

The settings for thoroughbreds were 4 metropolitan centres and 4 training centres. For the Central Districts, P-Median named Awapuni as both the metropolitan venue and the region’s training centre (p.73).

Once that’s done, the race days flow to the anchor automatically.

Trentham is then measured and allocated on the residue…4 race days allocated…..and having 4 race days = an Exit status.

There was nothing Trentham could do to change P-Median’s calculation….. Better crowds wouldn’t. Bigger fields wouldn’t.

Higher turnover wouldn’t… Trentham’s turnover is already among the best in the country (more on this in another post). Trentham’s result was determined the moment Awapuni was made the anchor.

Think of it like a chain retailer moving all the stock from one shop to another, then closing the first shop for poor sales.

P-median, is a planning aid. It answers “where would you put hubs if you were starting from scratch?”

It does not answer “which existing venue should close?”…P-Median has no memory of what those venues actually deliver. (pg.97 says: the model is “a decision-support tool, not a predictive instrument.”)

A decision-support tool supports a decision. It doesn’t make one.

Before any exit is confirmed, the decision should be tested against Trentham’s actual performance (e.g attendance, turnover per meeting, average field size, nominations, and meetings hosted) on the race days it has really held, including the transferred ones (more on this in another post).

Not decided against a residue the “P-median optimisation model” created.

Should a racecourse be identified to close based on a number the model handed out?

8 May 2026. Project Stamina plan finalised. pg.97 records the date and adds that the report “has not been updated to reflect matters arising after that date.” Its optimised thoroughbred network names 4 metropolitan venues: Ellerslie, Te Rapa, Awapuni and Riccarton Park. Trentham is marked for exit.

Tuesday 19 May 2026. NZTR provides an update: Every racecourse in New Zealand is reclassified for the 2026/27 season. 5 metropolitan venues: Ellerslie, Te Rapa, Hastings, Trentham, Riccarton. 18 provincial tracks. 11 community venues.

Wednesday 20 May 2026. NZTR releases the racing fixture for 2026/27, on the classifications above.

Tuesday 28 July 2026. Project Stamina published, endorsing Trentham’s exit and confirming Awapuni as the Central Districts metropolitan venue and training centre.

Project Stamina pp.52, 73, 97, 108–110

Everything I post is compiled from publicly available sources — racing publications, reports, financial statements, meeting minutes, council and LINZ records, and media articles. Other than being a small thoroughbred owner and breeder with one broodmare, I have no other commercial, governance or financial interest in these matters. I don’t represent any club, organisation or stakeholder group. The purpose is to encourage informed, transparent, evidence-based conversation on matters with significant implications for the industry’s future.
Thanks Shah Amner.