Forums

General Betting

Welcome to Live View – Take the tour to learn more
Start Tour
There is currently 1 person viewing this thread.
travelling_man..
30 Nov 09 11:01
Joined:
Date Joined: 04 Feb 09
| Topic/replies: 5,527 | Blogger: travelling_man..'s blog
Which is the better indicator of whether or not a system is working?
Pause Switch to Standard View ROI or points profit?
Show More
Loading...
Report The Betfairy November 30, 2009 11:11 AM GMT
Points make prizes.
Report ZEALOT November 30, 2009 11:20 AM GMT
any
Report FINE AS FROG HAIR November 30, 2009 11:34 AM GMT
They're pretty interrelated aren't they ?
Can't have a positive ROI if you're not in points profit, for example.
Basically, the higher the ROI the better the system for sure.
Report Alex the old wrinkled retainer November 30, 2009 12:07 PM GMT
Basically, the higher the ROI the better the system for sure.


It ROI is based on turnover then that is not true. It is a bit like saying I would not prefer to be a insurance company with a tiny margin and a big profit instead of a rag and bon man with a big margin but a tiny profit.
Report SHAPESHIFTER November 30, 2009 12:09 PM GMT
Alex is right. LSP (level stake profit) will create spikes in the progression whereas ROI will show the progression over time better.
Report kenilworth November 30, 2009 12:19 PM GMT
Both can be misleading unless level stakes are used.
Report SHAPESHIFTER November 30, 2009 12:29 PM GMT
LSP is an indicator. But more spikes

ROI factors in your staking plan and how effective it is incorporated to the approach.
Report travelling_man.. November 30, 2009 12:44 PM GMT
So ROI would seem a better way to evaluate.
Thanks guys - appreciated.
Report Cosmic Horizon November 30, 2009 12:56 PM GMT
What is the ROI and how do you calculate it?
Report Alex the old wrinkled retainer November 30, 2009 1:08 PM GMT
Cosmic Horizon 30 Nov 13:56
What is the ROI and how do you calculate it?



Profit/turnover/100.

It is a complete waste of time unless you are comparing it with somebody using virtually the same strategy. And you will never be sure that you are.


Back to the insurance company analogy.
Report My Way De Solzen November 30, 2009 1:37 PM GMT
Profit/turnover x 100
Report The Betfairy November 30, 2009 1:39 PM GMT
ROT for me.
Report SHAPESHIFTER November 30, 2009 2:17 PM GMT
er,

(RETURN DIVIDED BY TURNOVER)
minus
1
will give you the percentage return on your investment.
Report kenilworth November 30, 2009 2:29 PM GMT
Loser 1 pont
1/2 winner 10 points
Staked 11 points
Return 15 points
Profit 4 points
ROI +36%
Level Stake - 0.5
ROI - 25%

Crude example.
Report FINE AS FROG HAIR November 30, 2009 3:13 PM GMT
Of course Kenilworth, if the winning bet was placed before the losing bet, one could argue that the capital invested ( ie potential loss) was only ever 10 points, being the first bet of 10points which won, in which case the net ROI of the 2 bets was 40%.
The return on turnover ( ROT) was however the same in both cases, irregardless of which bet was placed first.
It just goes to show that on ROI calculations, a long series of bets must be used to quantify the true drawdown of capital which is required to implement the system.
Report SHAPESHIFTER November 30, 2009 8:00 PM GMT
Kenilworth, your example is based on someone betting anywhere from 1 to 10 points on an event.

I can't speak for anyone else but myself, but if a point = £10, I could not see me betting £10 on one event and £100 on another.

As well, the example given, as someone else mentioned, is over 2 bets. If someone is looking at a return on an approach, it would take more than two bets.

So, yes, a crude example on your part ;)
Report DFCIRONMAN November 30, 2009 8:12 PM GMT
If over the same period of time you have 4 betting strategies that can produce following-

NUMBER ................PROFIT/LOSS
OF ..........................RE STRATEGY
BETS FOR

-1,736..............................3,217.63 points

-484.................................1,513.95

-634................................. 231.63

-366.................................1,365.87


Which strategy would you go for ??????
Report aye robot November 30, 2009 11:20 PM GMT
Which is the better indicator of whether or not a system is working?

If what you're looking for is a good measure of whether your results are likely to be a "lucky streak" or show genuine statistical significance none of the measures sugested above are of much use- if any. What you need is a test that is specifically designed to look for statistical significance in small samples. ROI isn't that, as it's only meaningful for large samples, as is "points profit". Stats experts can tell you how to do this (I can't) but be warned that it gets quite complex quite quickly.

As a "quick and dirty" test of your results you could ask yourself some questions like this:

1: How many of my results would have to go the other way for me to be in the red now? If all your profits are attributable to a few wins then you may just have been getting lucky- you want to be able to show that even if you'd done much worse than you actually have done you would still be in profit.

2: How much comission am I paying as a percentage of profit? This is a pretty good measure of the efficiency of your strategy (less is better). For all people whinge about the PC your aim should be to get yourself into PC territory as that will mean you are betting efficiently.

3: If I had taken slightly worse odds on every bet would I still be ahead? This will give you a good sense of the size of your edge (if you have one). A good strategy will still at least break even at slightly worse odds. If your strategy doesn't stand up to this then your edge is precariously small.

Make up some more like the above that are appropriate to your strategy and take a very pessimistic view of the results, that will give you a much better feeling for your sucess than ROI.
Report kenilworth December 1, 2009 10:34 AM GMT
Shape, yes it is a crude example but it needed to be, to make my point. Some people bet randomly and therefore the ROI would be skewed, and that was my point, using an extreme example.
Report Northern Wizard December 1, 2009 10:56 AM GMT
A good way to test if a system is 'working' is to do a confidence interval test, usually to 95% confidence level. These tests are relatively simple to do and a quick search of google should point you in the right direction.
Report Alex the old wrinkled retainer December 1, 2009 10:58 AM GMT
FINE AS FROG HAIR 30 Nov 16:13
Of course Kenilworth, if the winning bet was placed before the losing bet, one could argue that the capital invested ( ie potential loss) was only ever 10 points, being the first bet of 10points which won, in which case the net ROI of the 2 bets was 40%.
The return on turnover ( ROT) was however the same in both cases, irregardless of which bet was placed first.
It just goes to show that on ROI calculations, a long series of bets must be used to quantify the true drawdown of capital which is required to implement the system.







They are a strange lot in the gambling world and don't use the conventional terms that people in finance would use. I have learned that the use of the term ROI is illogical and for ROI in gamblers parlance read: gross margin. Or as we are rapidly inventing a new term: ROT.

For ROCE in this thread they use the term points, erm, I think. :|
Report Alex the old wrinkled retainer December 1, 2009 11:04 AM GMT
In any case benchmarking is difficult, if not impossible for betfarians as there as so many different techniques used and the succesful people are unlikely to disclose to subdivide the techniques to allow useful benchmarking. (See above my insurance company/Steptoe and Son analogy.)
Report kenilworth December 1, 2009 11:09 AM GMT
I think the only way is betting each pick to RETURN a fixed sum, perhaps 100 points, therefore an 2.0 chance would carry a stake of 50 (100/2) or a 4.0 poke 25 points (100/4). Alternative to that is level stakes but that could be skewed by a big priced winner. Having said that I agree with what Alex says.
Report DFCIRONMAN December 1, 2009 11:39 AM GMT
"Big priced winners" is not "skewing" figures but essential to them if big priced winners is what you are trying to attain!
Report Alex the old wrinkled retainer December 1, 2009 2:17 PM GMT
DFCIRONMAN 01 Dec 12:39
"Big priced winners" is not "skewing" figures but essential to them if big priced winners is what you are trying to attain!




Which brings us back to the insurance company/Steptoe and Son comparison analolgy. Impossible to benchmark it is, to be sure, to be sure.

I am right most of the time. ;)
Report dlarssonf December 1, 2009 3:16 PM GMT
kenilworth 01 Dec 12:09
I think the only way is betting each pick to RETURN a fixed sum, perhaps 100 points, therefore an 2.0 chance would carry a stake of 50 (100/2) or a 4.0 poke 25 points (100/4). Alternative to that is level stakes but that could be skewed by a big priced winner. Having said that I agree with what Alex says.



is this not a classic case of too much on bad stock and too little on good stock?
Report Treble_Underscore December 1, 2009 3:46 PM GMT
ROI on a small bank. ROT on a large bank.
Report kenilworth December 1, 2009 4:46 PM GMT
dlarssonf 01 Dec 16:16
is this not a classic case of too much on bad stock and too little on good stock?


Haven't heard that expression before, could you please explain ?
Report FINE AS FROG HAIR December 1, 2009 6:10 PM GMT
Alex is always right except when he is wrong.
ROI is return on investment.
Investment means the max. amount of monies invested or risked.
This is not the amount of money cumulatively bet, which is turnover.
If you start off for example with a bank of 100 units, have never placed a bet or simultaneous bets where the liability totalled more than 100 units, you have never added to your capital in any form or manner apart from accumulating profits, then all you have ever invested or risked is 100 units.
Then your return on investment that is your accumulated net profit divided by 100.
Anything else is return on turnover.
How you calculate your net profit is also important.
As Alex said you should impute or allocate some cost for your time and labour involved in betting.
Then you should compare this ROI with other forms of comparably risky investing alternatives and see if your gambling is truly remunerative comparatively.
Report kenilworth December 1, 2009 6:14 PM GMT
If you back a team at, say 2.0 for £100, a goal is scored in your favour and you close at 1.33 for a profit of £50, what is your ROI ?
Report Zola's Back Heel December 1, 2009 6:19 PM GMT
Bank/StartingBank
Report FINE AS FROG HAIR December 1, 2009 6:43 PM GMT
50 % on that one investment.
What's your point ?
Report FINE AS FROG HAIR December 1, 2009 6:49 PM GMT
Imo a perfect , professional system will have a very high ROI and a very low ROT.
That is it will bet a lot in a regular consistent size.
Report stewards December 1, 2009 7:15 PM GMT
I did 306 soccer bets, each bet for 100$ with a bank of 2000$

I have 2000$ profit.

this should be if I understood it right

ROI = 100%

ROT = about 6.5%

ok, but does this mean? is it good or bad?

for sure good, because I am in profit right now :)
Report FINE AS FROG HAIR December 1, 2009 7:31 PM GMT
Looks real good to me.
Over what period may I ask ?
Report SHAPESHIFTER December 1, 2009 8:02 PM GMT
kenilworth 01 Dec 19:14
If you back a team at, say 2.0 for £100, a goal is scored in your favour and you close at 1.33 for a profit of £50, what is your ROI ?


kenilworth, if I am trading, I do not calculate ROI. A simple plus/minus and then I look at my average per events and type of events.

But in the example you gave, I would calculate my ROI on my max risk/position which, in the example you gave, was £100. Thus a 50% ROI.
Report Alex the old wrinkled retainer December 1, 2009 9:50 PM GMT
Whether you measure this by ROI or ROT or gross margin or ROCE this little ditty by Ray Davies always stands me in good stead when trading:

It should improve your ROI and your ROT if the philosphy is absorbed.



They seek him here, they seek him there,
His clothes are loud, but never square.
It will make or break him so he's got to buy the best,
'Cause he's a dedicated follower of fashion.

And when he does his little rounds,
'Round the boutiques of London Town,
Eagerly pursuing all the latest fads and trends,
'Cause he's a dedicated follower of fashion.

Oh yes he is (oh yes he is), oh yes he is (oh yes he is).
There's one thing that he loves and that is flattery.
One week he's in polka-dots, the next week he is in stripes.
'Cause he's a dedicated follower of fashion.

They seek him here, they seek him there,
In Regent Street and Leicester Square.
Everywhere the Carnabetian army marches on,
Each one an dedicated follower of fashion.

Oh yes he is (oh yes he is), oh yes he is (oh yes he is).
His world is built 'round discoteques and parties.
This pleasure-seeking individual always looks his best
'Cause he's a dedicated follower of fashion.

Oh yes he is (oh yes he is), oh yes he is (oh yes he is).
He flits from shop to shop just like a butterfly.
In matters of the cloth he is as fickle as can be,
'Cause he's a dedicated follower of fashion.
He's a dedicated follower of fashion.
He's a dedicated follower of fashion.
Report FINE AS FROG HAIR December 1, 2009 11:05 PM GMT
Get the kinks out of your betting by listening to Ray Davies.
I like it Alex.
Report stewards December 2, 2009 7:07 PM GMT
FINE AS FROG HAIR 01 Dec 20:31


Looks real good to me.
Over what period may I ask ?


about 3 month and about 650$ comiission paid.
Report FINE AS FROG HAIR December 2, 2009 7:29 PM GMT
Thus 400% pa ROI.
Way to go.
Onwards and upwards.
Report stewards December 2, 2009 7:38 PM GMT
FINE AS FROG HAIR 02 Dec 20:29


Thus 400% pa ROI.
Way to go.
Onwards and upwards.



yes, but if I want to bet levelstakes 200$ or more instead of 100$ than there will be less bets :(
Report FINE AS FROG HAIR December 2, 2009 8:44 PM GMT
Why don't you just set your bet to be ( say ) 5 % of your fluctuating accumulating capital base, and work your way gradually up to 200 units per bet.
Or put more capital in immediately to double it to 4000 units and double your bet size to 200 units.
Only do the latter if you strongly believe that your betting selection process is one that should continue having the amazing success it has had to date.
Otherwise, do the former and just increase your bet size slowly over time in correlation with your continuing success ( if any ).
Report Zola's Back Heel December 2, 2009 8:46 PM GMT
if you build it they will come.
Report FINE AS FROG HAIR December 2, 2009 8:50 PM GMT
Sorry Zola.
You're being far too subtle for me.
Care you enlighten me ( us ) a bit more ?
Report stewards December 3, 2009 4:26 AM GMT
FINE AS FROG HAIR

I will let it run with 100$ and see what will be in another 3 month.then I will see if it makes sense to bet higher. but as I mentioned it before. the higher the bet there will be much less bets because there is not much money available on those kind of bets.
Report joshuag December 3, 2009 11:51 AM GMT
Try using ARCHIE to evaluate the confidence you can have in a set of system results. There is an article showing you how to use it in this months smartersig.com magazine.
Post Your Reply
<CTRL+Enter> to submit
Please login to post a reply.

Wonder

Instance ID: 13539
www.betfair.com