ShaunWhite wrote: ↑Fri Mar 20, 2020 2:10 am
Bog wrote: ↑Thu Mar 19, 2020 9:16 pm
spreadbetting wrote: ↑Thu Mar 19, 2020 7:23 pm
Always worth holding onto value when you can get it
Can you give an example SB?

If I'm matched at 7.00 and dog steam to 5.00 BSP, it's good value worth holding onto? How about only 6.00 BSP?
Very good liquidity today, some races with 50k+ matched
If you look at the data then you'll see that over a very large sample BSP is sort of, just about, near enough, right. So you're not gaining anything by closing at BSP vs letting it run except for a less volatile balance on a 'typical' dog. The additional trick SB and others
might be pulling off re value, is identifing where BSP
isn't correct for
that specific runner, and then making a judgement about whether closing or running is better.
Morning Shaun
That is spot on.
I've been working on something recently and it is looking good. A really good tip I had from Xitan was to run a monte Carlo simulation on the data/results
(you probably know this already Shaun, but for the benefit if people new to this)...if you have a set of results, illustrating BSP, you can run a simulation say 500,00 times against you data set and determine whether your results are down to just good luck or how it would have performed (on average) at BSP
To run the simulation, tabulate the data in excel, (incl BSP and whether it won or not), adjacent have a random generated number between 0-1. If a dog had a BSP of 2 for example, then you would expect it to win 50% of the time. If the random generated number is above 0.49 then consider a loss, 0.5 and above consider it a win.
Compare the Monte Carlo on your actual data and then see how many times in 500,000 (or more) iterations it matched or beat your actual results and if it is very low (ie<5% of the time as an example) then you may have something
Regards
Peter
Edit : Just doing this course
https://www.udemy.com/course/statistics ... 0#overview
Which covers things that I thought I knew enough about (but didnt

such as Bernoulli; Sampling, Hypothesis testing, Regression etc, worth looking at as a refresher. More strings to your bow as they say...