Pinned Tweet
the full formbook in your agent
claude mcp add --transport http formbook mcp.formbook.io/mcp
codex mcp add formbook --url mcp.formbook.io/mcp
opencode mcp add formbook --url mcp.formbook.io/mcp --global
ask it anything
i appreciate RTV providing stride data, but I think the way they do it is wrong
the numbers they give bar max length and max frequency are not useful without context
in the RTV example we have no idea where the min and max took place, and its particularly important for the min
if the min sps was taken from a legless horse shortening stride in the final furlong, that tells us absolutely nothing about the horse
stride guys see a low number for min and think this horse might be suitable for longer distances but the only way thats an accurate indicator is if its taken from an even pace sectional
were looking for horses who can drop low early and through the race, and putting a race min number up from the final furlong completely misses the point
1
174
bloodbath today, Sans Bruit seemed to hit an invisible wall in the last 20 metres when looking certain to get back up, and Skelton ought to have won on Supreme Malinas, the rest were value soup at best
heres tomorrows slop
Petticoat Bay and Hereafter two interesting Obrien backend fillies for the Oaks, neither quoted in the market unfortunately
So Lovely available at 50/1
pedigree cant get much better and the 2yo campaign is just a foundation as weve seen with multiple Obrien fillies stepping up massively after looking exposed at 2
she moved nicely into the lead in the Montrose and seems way overpriced with such a bad favourite
the optimal distance model produces a distribution of probabilities on what a horses best trip will be
stride data turned out to not be as valuable as hoped for predicting 2yos optimal distance at 3+
it doesnt add much signal over traditional pedigree, trainer and 2yo distance data, although we only have a few years data so that could change as time goes on
we look at the last 10 winners of the Oaks and Derby and see which 2yos for this season are a good fit
you can interact with all this data right now through your agents on mcp.formbook.io/
couple of winners today
Old Habits 5/2 > 6/5 🥇
Hardy Buck 13/2 > 9/2 🥇
Persian Art 10/1 > 6/1 🥈
Frantic 9/4 > 11/10 🥈
Diplomatic Ash 16/1 > 14/1 🥉
no joy today
Smile John Boy ended up 1.93 and was done at 1.07 ir
most of the time youre not bumping into a Crookshaven in that spot
Savvy Disko must have traded short as well, another one I think wins most of the time the way it picked up at that level
few for tomorrow
12:27 Warwick - Smile John Boy 9/4 (model 1.54)
14:25 Thurles - Allo Al Khawaneej 10/1 (model 6.25)
15:35 Newcastle - Savvy Disko 13/2 (model 3.71)
4
420
some solid gains tonight, very hard earned
think ive about squeezed every drop out the data so this is as good as it gets
never going to beat the bsp but still a very useful guide for studying races a day or two before
and identifies some wrong prices in softer early market
4
305
few for tomorrow
12:27 Warwick - Smile John Boy 9/4 (model 1.54)
14:25 Thurles - Allo Al Khawaneej 10/1 (model 6.25)
15:35 Newcastle - Savvy Disko 13/2 (model 3.71)
the early data is telling us to avoid low grade all weather muck, you need more than a model to swim in that water
maybe Kempton and Dundalk tomorrow will strike back for the surface
1
4
486














































