Blood Red was on a decline watch last week. Its Horse Truth read now points the other way.

When we wrote about Ipswich on 16 October, Blood Red's page carried a decline watch. Its form was falling. It had just run third over 1965 metres.

Open that page now and the read has turned around. Horse Truth files Blood Red as a fast horse, quick on the clock with its class still to be proven against better company. Its form is rising. The flag on its page reads positive watch. Its best rated run now sits at 2150 metres, and the read rests on eleven rated runs.

Same horse, same third placing, a few days apart. The result did not change. The read did.

A finishing position stays in the form guide for good. A Horse Truth read is rebuilt as new race data arrives, and the way it moves is something the result alone will never show you.

What the placing hid

Blood Red's third looked ordinary on paper. Over the run home it lifted its stride rate second best of the nine runners, its cadence noticeably quicker than in the middle of the race.

The decline watch was a fair flag on the evidence the model had before that run. The run itself pointed the other way, and the read now does too.

None of this says where Blood Red will finish next time. It says the horse that ran at Ipswich was not the horse the old flag described.

A winner, relabelled

Night Market won race five at Ipswich the same afternoon. On 16 October its read called it a grade specialist, a horse that beats rivals rather than the clock, with its best rated run at 1000 metres from four rated runs.

The read now rests on five rated runs and files it differently. Horse Truth calls Night Market a class horse, one that beats its rivals consistently with figures to support it, though it is racing below its best. Its best rated run sits at 1200 metres. Its form is holding steady and its page asks for a review.

Both labels say Night Market beats the horses around it. What moved is the model's view of the figures behind those wins.

On a file of five runs, one more run can shift a horse from one type to another. Horse Truth marks that file as low evidence, and a reader should hold the label loosely for exactly that reason. The Ipswich signal still sits beside the new type. Night Market's stride that day was a lot shorter than in its last three runs, after allowing for the going. The page calls that a watch signal from race data, not a diagnosis.

A small step on a deep file

Mordecai shows the other end of the scale.

Its read rests on 24 rated runs, which Horse Truth marks as high evidence. On the evening of 18 October, the day it raced at Townsville, its type changed. It had been filed as an honest midfielder. It is now a class-edge midfielder, a horse that rates fairly against rivals for its grade without a standout figure. Its class measure rose with the change, and its form trend moved from rising to holding steady.

That is a modest step, and it should be. A horse with two dozen runs on file does not get rewritten by one afternoon. When a deep file moves at all, the new evidence has pushed past a lot of history.

The Townsville run fits the change. Mordecai held its stride late second best of the ten runners, and its stride in the run home was a little longer than in the middle of the race. Its running profile reads sustained finisher, with a top three last 400 metres in four of its last six runs.

A class turn this small on a horse this well measured will not make a form guide. Horse Truth records it anyway.

When the trend turns first

Colonel Parker won at Cairns on 11 October while its read said its form was falling and it was racing below its best.

The read now says its form is rising. It still files Colonel Parker as a class horse racing below its best, with its best rated run at 1400 metres, now from nine rated runs. The type held. The trend turned.

A form trend and a type answer different questions. The trend describes which way a horse is heading. The type describes what kind of horse it has been across its file. They do not have to agree, and when they part ways the next few starts will show which one was closer to the truth.

When a bad day changes little

Carnegie Hill shows the other side. It ran tenth in the last at Ipswich, the clear miss of that meeting, with its stride a lot shorter than in its last three runs after allowing for the going.

Its read held. Horse Truth still calls it a false figure horse, one whose numbers flatter it against class. Its form is still rising, its best rated run is still at 1600 metres, and it now sits on nine rated runs. Its page carries a positive watch.

A tenth did not sink it. A read answers to the measured run, and one result in either direction is one more line in the file.

How to read a horse page

The type is the headline, and on a deep file it is slow to change. The form trend can move sooner, as Colonel Parker shows. The number of rated runs tells you how much weight either one can carry.

Read them together. A thin file with a fresh type change is a horse the model is still learning. A deep file that steps up, even a little, is a horse whose level has moved.

Blood Red's current read is on its horse page, and every pick Horse Truth seals before a race stays on the public record. Any horse you are following can go through Horse Check.