# Does a Low Bet Angel Risk Meter Improve the Reliability of In-Running Drifts?

[prettysoul86]

I've been working on an in-running strategy and would appreciate some feedback from those with experience using the Bet Angel Risk Meter.

My hypothesis is as follows:

* The pre-race market has a **low (green) Risk Meter**, indicating a stable market with relatively orderly price discovery.
* The horse starts as one of the market principals (for example, BSP below 10.0).
* The horse then drifts significantly in-running (for example, to 20.0 or higher) after around 50% of the race has been completed.

My reasoning is that a green Risk Meter suggests the pre-race market had a relatively strong consensus about each horse's chances. If that's true, then once the race is underway, a substantial in-running drift is more likely to represent new information from the race itself (position, pace, travelling, etc.) rather than unresolved pre-race uncertainty.

In other words, I'm wondering whether:

> **Given a stable pre-race market, does a significant in-running drift have greater predictive value than the same drift occurring in a more volatile pre-race market?**

I'm **not** suggesting the outcome is certain, only that the conditional probability of the horse eventually losing may be higher when the pre-race market has been stable.

Has anyone:

* Back tested something similar?
* Compared green vs amber/red Risk Meter markets?
* Found that the Risk Meter adds predictive value once the in-running price is already known?

[megarain]

Never used the Risk meter, but its an interesting angle.

Correlation between pre-race and in-run markets is something I do trade - but only given certain characteristics.

You might be able to rig something using TPD output.

[rosecruz]

I am not sure if the Risk Meter produces data that is currently saved anywhere.

You should be able to figure this out quite easily using pre-off market data—ranging from 15 minutes to BSP, or even 5 minutes to BSP.

At the same time, the races that do not tend to move much are mainly competitive handicaps and competitive listed races.

Once you have the data, you will need to determine which races represent low risk and classify them.

You then need in-play data to figure out if the drifters lose more often than their in-play prices suggest and/or if favourites win more in those races than their in-play odds suggest. It is doable.

I have not tested this specific theory, but I have tested some in-play scenarios, and even in-play, the market is quite efficient.

I have found it incredibly hard to find repeatable and profitable scenarios based solely on big, general ideas.

[Dallas]

The race meter is solely for use on pre-off markets (when sentiment is all that's driving the price) and if it is behaving normally using tens of thousands of historical pre-off markets
Once in-play odds are moved by how they are performing at which point the risk meter is redundant

[jamesedwards]

The race meter is solely for use on pre-off markets (when sentiment is all that's driving the price) and if it is behaving normally using tens of thousands of historical pre-off markets
Once in-play odds are moved by how they are performing at which point the risk meter is redundant

The hypothesis is plausible though. That the correlation between in-play price movement and race outcome is linked to the stability of the pre-off market.

It would be useful if the race meter value could be available to call on as a stored value in Guardian?

[Dallas]

The race meter is solely for use on pre-off markets (when sentiment is all that's driving the price) and if it is behaving normally using tens of thousands of historical pre-off markets
Once in-play odds are moved by how they are performing at which point the risk meter is redundant

The hypothesis is plausible though. That the correlation between in-play price movement and race outcome is linked to the stability of the pre-off market.

It would be useful if the race meter value could be available to call on as a stored value in Guardian?

I guess it could be added, just with the pseudo its values can only be relied on during pre-off markets (tbh the fact its limited to pre-off UK & IRE horse racing only is probably the reason it hasn't ever been added)

PS, To avoid confusion to anyone were both referring to the 'risk' meter in our posts above

[jamesedwards]

The race meter is solely for use on pre-off markets (when sentiment is all that's driving the price) and if it is behaving normally using tens of thousands of historical pre-off markets
Once in-play odds are moved by how they are performing at which point the risk meter is redundant

The hypothesis is plausible though. That the correlation between in-play price movement and race outcome is linked to the stability of the pre-off market.

It would be useful if the race meter value could be available to call on as a stored value in Guardian?

I guess it could be added, just with the pseudo its values can only be relied on during pre-off markets (tbh the fact its limited to pre-off UK & IRE horse racing only is probably the reason it hasn't ever been added)

PS, To avoid confusion to anyone were both referring to the 'risk' meter in our posts above

Yes, sorry, I meant risk meter.

If people find the risk meter useful then it would seem sensible to have the value available for use in automation.