18th November 2005, 02:12 PM
|
|
Member
|
|
Join Date: Jan 1970
Location: Mt Tamborine
Posts: 574
|
|
Quote:
|
Originally Posted by stebbo
Ok Duritz....
I have whipped up a quick application and run some numbers....
I have done 5,000 separate tests of a run of 2,000 bets. I didn't bother going out to 1,000,000 bets unlike others, as I don't expect to ever have 1,000,000 bets for any one of my strategies.... I can fully expect to have 2,000 bets though, so it's a little more realistic.
I calculated Longest Run of Outs and Maximum drawdown based upon the original 22% Strike Rate and $4.80 dividend. I then averaged these over the 5,000 separate tests. For the "average" I used the arithmetic mean. The simulation isn't ideal as we don't have a spread of dividends, but it's probably close enough.
The average Longest Run of Outs for 2000 bets was 26.47.
The average Maximum Drawdown for 2000 bets was 73.20 units.
The ratio of Maximum Drawdown to LRO averaged out to 2.77. Therefore, Bhagwan's ratio of 3.5 is not too far from the mark. If you expect to have a LRO of 30, then you should expect your MaxDD to be around 100.
What does this mean for staking????? Simply calculate your MaxDD and then structure your betting bank to withstand that drawdown. For a 100 unit MaxDD, then you should only be betting 1% of your bank if you decide to bet flat stakes.
Following on from your most recent post, if your plan is to accumulate a bank, then I would recommend that you bet 1% of your bank non-reducing. This means that as your bank increases you increase your bets, but when it decreases you do NOT decrease your bet. I have been using this method of staking for quite a long period of time now, and it has stood me in good stead.
Cheers,
Chris.
|
Stebbo,
I feel you are doing our friend Duritz a mis-service by offering this advice. Using JFC's simulator I did only 5 runs of 2000 bets and got maximum drawdowns of 46, 120, 66, 57 and 65. The average maxDD here is only 49 so you could say I was "luckier" than you who averaged 73.
Now poor Duritz who was probably thanking you for your advice after the first 2000 bets would still be using 1% of his bank non-reducing during the second batch of 2000 bets and whammo. His drawdown of 120 units would have sent him back to his day job.
Averaging a bunch of 2000 bet batches is not SAFE. Look at your 5000 figures for maxDD and count how many times you would have blown your bank at 1%. On my five figures you have a 1 in 5 chance of losing ALL your money for each batch of 2000 bets. Maybe that's OK if you are one of these people who have a lot of banks running concurrently (and I'm not knocking that) but if you rely on one bank like I do, no thanks. I doubt if I'll ever have a million bets either but it gives a more realistic view of what can happen when "Sod's Law" comes in to play.
KV
|