Book - Mordin on Time

[Derek27]

I checked my local library...

What's that?

[Brovashift]

I checked my local library...

What's that?

What? You never know, if you don't ask you don't get lol

It is quite a good library tbf, I've had quite a few obscure trading books from my local library over the years. I just tried to get 'Statistics without tears' for an introductory level, but there was only 1 copy in the whole of North Wales and thats gone missing... So, no dice!
I've got a pdf version but it hurts my eyes reading via phone/tablet, can't beat a good book sometimes.

[Brovashift]

So, I thought I'd ask the Bing AI for info on the Nick Mordin book, hope it helps a little.


SpeedRatings1.png


Spped Ratings2.png


Applied to a race.png

Thanks WeatherFox, I am going to have a Bing later, these results seem way better than ChatGPT (Open AI version) that said it couldn't access websites to review data such as racecards. Unless I was asking the wrong questions of course...
Very useful

[ShaunWhite]

Just started perusing this book "Mordin on Time" (pdf) and was wondering if anyone has studied this...

What level maths do you have? The issue you might not be aware of is that horses don't have a predicted pace, they have a predicted distribution of paces. So unless you're comfortable with binomial distributions and converting a model output like the one below into probabilities, you might want to put your effort elsewhere.

Screenshot_16.jpg

[wearthefoxhat]

So, I thought I'd ask the Bing AI for info on the Nick Mordin book, hope it helps a little.


SpeedRatings1.png


Spped Ratings2.png


Applied to a race.png

Thanks WeatherFox, I am going to have a Bing later, these results seem way better than ChatGPT (Open AI version) that said it couldn't access websites to review data such as racecards. Unless I was asking the wrong questions of course...
Very useful

Yes, you need to ask open ended questions and re ask in slightly different ways.

Chat GBT is still a good model as it might offer a different perspective. I'm leaning towards Bing AI at the moment.

[wearthefoxhat]

Just started perusing this book "Mordin on Time" (pdf) and was wondering if anyone has studied this...

What level maths do you have? The issue you might not be aware of is that horses don't have a predicted pace, they have a predicted distribution of paces. So unless you're comfortable with binomial distributions and converting a model output like the one below into probabilities, you might want to put your effort elsewhere.
Screenshot_16.jpg

Ah, but what's the fun in that?

Most with a bit of excel knowledge and creative flair could put together a speed rating model that takes into account current course topography..etc

I also think there's variance either way in the approach, and maybe you have to accept you won't get the perfect set up.

[StellaBot]

It was myself that posted that book to Peter

[ShaunWhite]

Ah, but what's the fun in that?

Maths is the fun and $3bn gets you plenty of fun too. I prefer Benter's approach rather than Mordin who IMO was already out of date when his book came out. I'd guess Brovashift will be onto something else new next month anyway so the replies are (excuse the pun) academic

[wearthefoxhat]

Ah, but what's the fun in that?

Maths is the fun and $3bn gets you plenty of fun too. I prefer Benter's approach rather than Mordin who IMO was already out of date when his book came out. I'd guess Brovashift will be onto something else new next month anyway so the replies are (excuse the pun) academic

I know what you mean about Benter's approach. Has an easier ratings/measurable, go to approach.

Benter.png

[ShaunWhite]

I know what you mean about Benter's approach. Has an easier ratings/measurable, go to approach.

I could study for the rest of my life and only be 10% as smart as him but there's a lot about the general approach to like. The key realisation is that all indicators (pace, vwap, price, you name it) aren't points they're ranges.

[Derek27]

I've always thought speed figures are useless on their own. As Phil Bull used to say, speed figures tell you how bad a horse isn't, (not how good it is).

[ShaunWhite]

I've always thought speed figures are useless on their own. As Phil Bull used to say, speed figures tell you how bad a horse isn't, (not how good it is).

And there stands the best arguement for using distributions, a single pace figure doesn't tell you the liklihood of that pace being acheived, or what deviation there might be either side. It's also the case that no horse runs enough for any degree of statistical significance so you have to work with the generic type 'horse' and fit selections to type. You'll find systems for converting bare ratings into odds but they make too many assumptions to be the best solution.

[Kai]

Chat GBT is still a good model as it might offer a different perspective. I'm leaning towards Bing AI at the moment.

Quite scary how quickly we get used to AI like this

Human history from AI perspective must read like a blooper reel

It's almost embarrassing really

[Brovashift]

Lucky for me don't need maths now with Bing lol

I think this new AI might lead to a lot of lazy kids I was watching the news last night and it showed this AI recreating art masterpieces in seconds. It's absolutely nuts.

[ShaunWhite]

Has Google made you lazy or more knowledgeable? The thing AI might lack for some time is a bullshit-o-meter. And art isn't pretty pictures or clever drawing, it's the physical manifestation of a personality or an idea, and as far as I can see AI doesn't have one, or any.