Today's Horse Racing

[Jukebox]

If you check the total you'll see that Euler's total counts the Lays as negatives on a back strategy - and comes up with his £225

[Dallas]

Actually enjoying today

[ruthlessimon]

Returns to £100 stakes when backing the favourite yesterday and trading out at post time were £225.88

Hex - 12:20 - Back to open @ 1.61 to 1.49 for 12 ticks & £8.05
Ling - 14:10 - Back to open @ 5.70 to 5.20 for 5 ticks & £9.62

Why do you prefer this analysis technique over a variable stake?

[brimson25]

This is a great question, since I think you advocate tick-based staking?

(This still makes me feels very uncomfortable)

[ruthlessimon]

This is a great question, since I think you advocate tick-based staking?

(This still makes me feels very uncomfortable)

My question may be incorrect.

The answer could be "It's better to look at it with various staking techniques"

1. Variable stake
2. Fixed stake
3. Volume stake (i.e. most accurate in theory)

Problem is, I wouldn't call that a model, I'd call that 3 backtests

[Euler]

I just express it this way because that's what people want to see. The most important thing is the underlying market activity.

[ShaunWhite]

Returns to £100 stakes when backing the favourite yesterday and trading out at post time were £225.88

Hex - 12:20 - Back to open @ 1.61 to 1.49 for 12 ticks & £8.05
Ling - 14:10 - Back to open @ 5.70 to 5.20 for 5 ticks & £9.62

Why do you prefer this analysis technique over a variable stake?

Obv I can't speak for PW but I do the same....you know where you stand with a flat £100.

What ype of variable stake do you mean, back/lay liability staking or amount per tick? Amount per tick is a inconsistent as a back at 1.99 and one at 2.02 vary so much depending on them steaming or drifting. Any variable staking also makes your RoI harder to work out as you need to know what you (would have) invested for each trade rather than it being simply 100 * #favs.

Flat stakes are a nice simple benchmark and so long as you're doing the same every time, why does it matter?

[ruthlessimon]

What ype of variable stake do you mean, back/lay liability staking or amount per tick? Amount per tick is a inconsistent as a back at 1.99 and one at 2.02 vary so much depending on them steaming or drifting. Any variable staking also makes your RoI harder to work out as you need to know what you (would have) invested for each trade rather than it being simply 100 * #favs.

Flat stakes are a nice simple benchmark and so long as you're doing the same every time, why does it matter?

Actually yeah, that's a good point - because variable staking does have some real issues (i.e. near XOs, a 1 tick move & you're halving/doubling the stake potentially etc)

I thought I might as well knock up what those three techniques look like, & yup you're right, fixed tracks it fairly well for the simplicity

[frankob20]

Last few races were hard to read....
nothing special for me today

[ShaunWhite]

I thought I might as well knock up what those three techniques look like, & yup you're right, fixed tracks it fairly well for the simplicity

These are all favs too so the price variation shouldn't be too wild. If you were looking at anything from 2.0 to 500 then fixed would be a pretty terrible idea and be massively skewed towards what happened on the biggies.

I can't disagree though that there's probably a more accurate way to track it, scrap the cash aspect and just use Move% ? That's if we can find a consensus about what that means

[ShaunWhite]

I thought I might as well knock up what those three techniques look like

Or 4...How about measuring it with "predictive tick size"? ....ie work out your amount per tick from the closing price band rather than the opening one? Could be b*ll*cks but I know you like your alternative methods. I still think the best thing is to remove the whole issue of step sizes and stakes and go for a %age.

[ruthlessimon]

Or 4...How about measuring it with "predictive tick size"? ....ie work out your amount per tick from the closing price band rather than the opening one? Could be b*ll*cks but I know you like your alternative methods. I still think the best thing is to remove the whole issue of step sizes and stakes and go for a %age.

Umm yeh.. I'll have to try get my head around that one

[jimibt]

Or 4...How about measuring it with "predictive tick size"? ....ie work out your amount per tick from the closing price band rather than the opening one? Could be b*ll*cks but I know you like your alternative methods. I still think the best thing is to remove the whole issue of step sizes and stakes and go for a %age.

Umm yeh.. I'll have to try get my head around that one

you guys need to exchange skype contact details

[PDC]

Some terrible field sizes tomorrow, Doncaster in particular with 3 x 3 runner races and the largest field being 7 runners and that is all before any late withdrawals, chuck in 15 minute gaps after each Cheltenham race and it is looking very uninspiring.

[spreadbetting]

Or 4...How about measuring it with "predictive tick size"? ....ie work out your amount per tick from the closing price band rather than the opening one? Could be b*ll*cks but I know you like your alternative methods. I still think the best thing is to remove the whole issue of step sizes and stakes and go for a %age.

Umm yeh.. I'll have to try get my head around that one

you guys need to exchange skype contact details

+1

All the data is there if you want to assess it by another method. Displaying it to £100 is just the simplest, I doubt Euler has some devious motive for doing it otherwise .