I think you're right - but for the wrong reasons.

The reason isn't because martingales are better value than single bets in any absolute sense, but because of the way in which the house edge is applied on J-D (it's 1% of amount staked not a percentage of the winning chance).  So on a 50/50 bet the player wins 1.98 times their stake - i.e. 2% less than they would in a game with no edge.  On a 98% win bet the player only wins just over 1.01 their stake - i.e. nearly 50% less than they would in a game with no edge.

It's the impact of that which produces the effect you describe - as a multi-bet martingale to simulate a single bet has to be at higher odds which represent better value on J-D.  

You can see this point if you consider a bet that isn't allowed on J-D - and apply J-D's method of applying edge.  Consider a bet with a 99% win chance.  Such a bet would pay exactly what you wagered any time you won - and would give precisely the standard 1% house edge.  Obviously that bet isn't allowed as it's plainly unfair (you could never make a profit from it) - but the reason WHY it's unfair applies to other 'odds-on' bets as well, just not to the same extent  (and applies all the way down the spectrum with the bets with a tiny chance to win being the best value).  And that is why J-D is martingale-friendly - because martingales use better value bets to achieve the same target profits from same starting capital.  

My instinct tells me that the best value series will be the longest martingale you can afford to BR that will reach your target profits in 1 successful cycle - once it take more cycles your expectation reduces as you end up placing more in total wagers than you would in a single bet.  As house edge is a percentage of amount wagered you need to minimise the amount wagered whilst ending either at your target or at zero.  Your example does that in 2 bets - on average wagering significantly under 1000 (71% of time you wager 1000, other 29% of time you only make the first bet).  If you then applied same principle to split the other two bets into 2-step martingales amount wagered would decrease and percentage of the time you won increase.  In practice this is constrained by your target profits having to be lower than max win and number of steps being limited by the minimum bet size (which is reached pretty quickly as at each division the odds for previous step have to be higher for the bet to be optimal).

Martingales are NOT a 'winning' strategy - there's no such thing in general.  But there are specific circumstances where betting on a house-edge game CAN make sense (where you have X coins and if you had Y more you do something with more extra value to you than the cost of the house edge on making that extra).  And in those circumstances optimised martingales CAN be the best option - with the goal always being to reach one of bust or your target with the smallest average amount wagered.

Below quote is from usagi's reddit post - just to show how retarded he is.


PURCHASE does NOT take a management fee on income from investment - only from initial purchases.

The 'mysterious' 0.4% isn't actually all that mysterious at all.  It's a 0.2% BTC-TC fee for buying PURCHASE plus a 0.2% BTC-TC fee for SELLING it.  It has nothing whatsoever to do with anything else.  So long as it makes enough to cover the site fees you don't make a loss - and it's made plenty more than that.

I'm not sure what the fascination is with investment into LTC-ATF.B1 anyway.  DMS has nearly 2K BTC in capital.  Just over 100 BTC is invested in LTC-ATF.B1 - and nothing has EVER been invested in LTC-ATF itself or LTC-ATF.B2 (even though usagi seems to think it has).  The main source of loss (or not making a profit) for investors in PURCHASE (which isn't recommended as an investment anyway - it only exists because there's no other way to easily sell pairs of MINING/SELLING) is the majority of capital which just sits totally unused in its account (which is verifiable due to the API url to view it being published - which anyone can check at any time).

Eventually you - and other retards - will get it into their heads that DMS.Mining is a bet with those who hold DMS.Selling.  My role is mainly to act as an escrow allowing them to speculate against one another.  I don't sell DMS.Mining OR DMS.Selling - so holders of neither are investing (to any large extent) with me.  I qualify it with "any large extent" as they ARE investing with me to the degree that I invest some of the capital to generate modest returns - but I have no control over the prices traded at or the amount of dividend each receives.

So when I refer to people as retards or idiots it's for entering into a bet at a bad price with whoever they bought their MINING from.  It has nothing to do with investing in me - as they AREN'T investing in me : nothing I do is going to change what those holding MINING receive (unless I do something really stupid with the capital).

I could have made a lot more profit if I'd set the prices myself rather than allowing anyone who wants to market-make (i.e. kept MINING priced just under the lowest-priced competitor).  But I don't run investments like that - anything I run which relies on my judgment is going to be designed to make all investors profit.

As for your question of whether I'm breaking the asset-issuers' terms of service.  No I'm not.  Read them more carefully - hint: the rule isn't a blanket ban on investing in securities with same issuer (there's a qualifier).

Rest of your reddit post is just you showing your usual lack of understanding : somewhere you seem to have got the mistaken impression that LTC-ATF/B1/B2 invest in coinlenders - they don't.  You also seem to believe that DMS mainly invests in my own securities - it doesn't (it has less in mine than in either of Coinlenders or J-D).  You also seem to believe investors in PURCHASE make a loss because of my 3% management fee - they don't (you can check what it started selling at, take off dividends to date and compare to current price - you'll find it's made a profit).  In short you've got none of your facts correct on which you base your conclusions.  Which is about par for the course.

Haven't you got better things to be doing?  Like:

Paying back the Nyan.A investors who were promised their money back in Q1.
Continuing your 'buy-back' offer on BMF which strangely results in outstanding shares count increasing.  As with previous buy-backs seems it's just a feeble attempt to pump the price for you to sell into.
Try to devise ways to explain why an investment fund that is only mandated to invest in Mining securities ends up with way more in non-mining than mining-related.  All BMF seems to be is a vehicle to invest across bonds issued by other people with the occasion spell of chucking away some cash on overpriced PMBs.
Try to find ways to pump your valuation of assets.  Valuing fixed-rate bonds at the compulsory buy-backs was a good start - but that's only inflated it a little bit.

And if my investmetns are so dodgy why is it you always seem to end up holding them?

I've patched the code for now and auto-transfer is running again - had already manually caught up on the few missed since it stopped.  Will check in a few minutes in case anyone sent just as it started.

Automated transfers are down at the moment.  Burnside just changed output for transaction history - adding in a unique transaction ID.  Whilst that's great longer-term (allows far more rigorous checking) it breaks the code for now.  Am having a look to see if I can quickly bodge around it for now (by ignoring it when parsing) until the developer can update to handle the new format.

Sold   1236
Swapped   0
Total   1236
Price   0.02817
Total   34.81812
Less Fee   34.74848376
Man Fee   1.042454513

BTC Balance (BTC-TC)    1,367.84405354
10164 LTC-ATF.B1    101.64000000
Coinlenders CD 28/8    201.63351930
Coinlenders CD 13/8    101.77455515
Just-Dice Balance    159.49605424
TOTAL ASSETS    1,932.38818223
   
Outstanding MINING   65193
Outstanding SELLING   65193
Outstanding PURCHASE   6804
Effective Units   71997
   
Block reward   25
Difficulty   37,392,766
Hashes per MINING   5000000
   
Daily Dividend    0.00006725
50 days (Min Liquid)    0.00336232
100 days (Forced Close)    0.00672465
365 days (Buyback)    0.02454497
405 days (IPO)    0.02723483
400 days (Post SELLING div)    0.02689860
410 days (Pre SELLING div)    0.02757106
   
NAV Post MINING Div    1,927.54663608
NAV/U Post MINING Div    0.02677260
Days Dividend Post Div   398.13
SELLING Dividend    -         
NAV Post SELLING Div    1,927.54663608
NAV/U Post Selling Div    0.02677260
PURCHASE selling price    0.02811
PURCHASE buy-back price    0.02624
   
J-D House profit at report   4205

Same here - can't login at all.

Even did a password reset (which seemed to work) - despite 100% knowing my password and having 2FA on.  Still no luck afterwards.  Guess it's just down at moment.

Sold   2685
Swapped   0
Total   2685
Price   0.02822
Total   75.7707
Less Fee   75.6191586
Man Fee   2.268574758

BTC Balance (BTC-TC)    1,338.89664703
10164 LTC-ATF.B1    101.64000000
Coinlenders CD 28/8    201.50588306
Coinlenders CD 13/8    101.71159795
Just-Dice Balance    159.47092745
TOTAL ASSETS    1,903.22505549
   
Outstanding MINING   64234
Outstanding SELLING   64234
Outstanding PURCHASE   6527
Effective Units   70761
   
Block reward   25
Difficulty   37,392,766
Hashes per MINING   5000000
   
Daily Dividend    0.00006725
50 days (Min Liquid)    0.00336232
100 days (Forced Close)    0.00672465
365 days (Buyback)    0.02454497
405 days (IPO)    0.02723483
400 days (Post SELLING div)    0.02689860
410 days (Pre SELLING div)    0.02757106
   
NAV Post MINING Div    1,898.46662601
NAV/U Post MINING Div    0.02682928
Days Dividend Post Div   398.97
SELLING Dividend    -         
NAV Post SELLING Div    1,898.46662601
NAV/U Post Selling Div    0.02682928
PURCHASE selling price    0.02817
PURCHASE buy-back price    0.02629
   
J-D House profit at report   4198

The only ones I've seen running so far on BTC-TC/LTC-Global are ones that just trade the spread with tiny quantities.  Some of those are exploitable when the spread gets wide enough (which doesn't happen on DMS - you generally need a 40%+ spread because of the parameters they seem to run with).

It was written specifically for this.  It's in C#. Here's a screen-shot of it :



Grayed out section to the right is the relevant part - allows me  to run in test mode (where it reports what it would have done but doesn't do it), only process trades for 1 user (used in testing), set the frequency it runs at, set it to only do automatic transfers below a certain amount and set a start time for it to catch up on back-log from (by default it only process trades that occur after it's started).

I sent in a few of my own PURCHASE just before the screen-shot so it showed what it looked like when a transfer actually occurred (last real one was some hours back).

Auto-transfers are now fully enabled - though being watched carefully by me.  So if you send PURCHASE and don't receive your MINING+SELLING within 5 minutes drop me a PM (though check BTC-TC hasn't gone down first).

If BTC-TC does become inaccessible (which happens occasionally - it was unavailable for me for a while yesterday night) then the bot should catch up with any back-log as soon as connectivity returns.  This net connection is also not the most steady and, for various reasons, it isn't set to reconnect when connection drops.

From time to time it'll be turned off as it's running on the computer I'm using (so I can monitor it) which tends to get restarted a few times each day.  Once it's been goign a few days without problems I'll sling it on a new computer with its own internet connection so it doesn't have problems whenever I max out bandwidth usage on this connection myself (I have a backup mobile broadband connection that's currently unused which it can have for itself).  At that stage it'll be set to reconnect automatically and downtime other than when BTC-TC is down should be far less frequent and for shorter periods.

It may also be down occasionally whilst I test software upgrades (LTC-Global integration is being worked on to allow cross-exchange transfers that are needed for the next securities LTC-ATF will be launching).  Functionality is now added (and tested) so that after any period offline I can set it to clear back-logs.

One last point is that it does NOT attempt to retry transfers that fail (just reports them for me to manually fix).  That has to be the case as I can't rely on everyone being honest and returning double-sends (on occasion BBTC-TC has been known to report a failure when an action actually gos through - e.g. yesterday it 3 times failed on transferring my DMS management fee with ''wallet lock" errors but two of them actually transferred.)

Auto-transfers back off whilst I'm out for a bit.  So any transfers now won't get processed for a few hours.

Auto transfers will be turned on for next half hour or so - but only for quantities of 25 or less.

If anyone buys any PURCHASE in that period and wants to help test, then please split them up over a few smaller transactions and we'll see how the bot handles it.  Will post again once its turned back off (I'm heading out in half an hour and not gonna risk leaving it unattended yet) so after that DON'T split transfers up - as I'll have to catch up manually (bot has no current functionality to clear a back-log).

It's definitely true that the profit potential on SELLING has dropped.  But IF the price is correct then it's not as bad as it may at first seem - as for its value to really be that high the bulk of that capital would need to be returned pretty promptly anyway.  Part of the design of DMS was to try to ensure that even in this scenario (anticipated very high difficulty rises, so low MINING price and high SELLING price) SELLING didn't have to leave capital tied up excessively.

But if the prices are precisely correct then NEITHER is worth holding.  If total information were available and time-cost of capital and CP risk ignored then MINING would trade at slightly under total life-time expectation of dividends and SELLIng slightly above (reflecting that SELLING has the EV from investment).  The margin of profit on both would be so tiny as to make them fairly worthless as investments due to any cost placed on CP risk making them -EV.

DMS relies on the situation not arising where everyone agrees on exactly what the prices should be - but if that ever happens then I'd expect a mass sell-back as DMS would no longer be serving any purpose.

Well it just did its first actual transfer (I've got it running in a mode where it only actually does transfers to my personal account - everything else it still just reports but doesn't act).

From its log:

Most recent entry in trade history at 08/08/2013 18:05:50 (server time)
Update started at 08/08/2013 18:07:57
TX-IN: 20 x DMS.PURCHASE <- DeprivedPersonal
TX-OUT: 20 x DMS.MINING -> DeprivedPersonal
TX-OUT: 20 x DMS.SELLING -> DeprivedPersonal
TX-IN: 30 x DMS.PURCHASE <- DeprivedPersonal
TX-OUT: 30 x DMS.MINING -> DeprivedPersonal
TX-OUT: 30 x DMS.SELLING -> DeprivedPersonal
Update completed at 08/08/2013 18:08:06
Most recent trade was at 08/08/2013 18:07:00 (server time)

My transaction history last entries:

2013-08-08 18:08:07    DMS.SELLING    transfer-out (DeprivedPersonal)    30    0    0    0
2013-08-08 18:08:06    DMS.MINING    transfer-out (DeprivedPersonal)    30    0    0    0
2013-08-08 18:08:04    DMS.SELLING    transfer-out (DeprivedPersonal)    20    0    0    0
2013-08-08 18:08:03    DMS.MINING    transfer-out (DeprivedPersonal)    20    0    0    0
2013-08-08 18:07:00    DMS.PURCHASE    transfer-in (DeprivedPersonal)    30    0.02864    0    0.8592
2013-08-08 18:06:44    DMS.PURCHASE    transfer-in (DeprivedPersonal)    20    0.02864    0    0.5728

So we're basically there now - just got to test it a bit more thoroughly and make sure it handles things like network connection dropping etc.

It won't ever be quite that quick.  There's a limit of one call per minute to that API and it's also cached.  Would guess average time for transfers will be 2-3 minutes once it's all running - which will hopefully be by the weekend.

Have been ironing out some glitches today (more testing to go) then tomorrow will be running it in test mode where it'll report transfers and what it would have done but not actually do it (that'll speed up transfers anyway as I'll see more promptly when I need to do some).  If that works fine tomorrow then on Saturday will probably turn it on for smaller transfers (under 100) and do large ones manually then let it go fully live once I've checked all the safeguards are working properly.

I didn't develop it - someone else did (up to them if they want to take credit in public or not).

So you raised thousands of BTC from IPO and hold $800 of stock?  What's the rest of capital being used for - or has the rest all vanished?

Yeah - there are exceptions where someone has a specific competitive edge.  Manufacturing your own ASICs being the obvious one - but even that will cease to be profitable for new entrants at some point: when the returns cease to repay NRE costs in a timely manner.  And even then they still face the risk of making a BTC-denominated loss if BTC rises strongly vs USD at an unfortunate time.

"Reinvestment in more advanced hardware can make an operation more profitable." is only true if the operation is already profitable.

To an extent we're already at the point where small-scale miners (and companies) can't make a profit - in part because there's only one manufacturer with a monopoly on shipping actual hardware (everything other than ASICMINER is either pre-orders or vaporware).  So ASICMINER are in the lucky situation of being able to sell hardware at a price where the recipients would be lucky to ever break even if their power was free and they had zero other costs.  But when other manufacturers finally get their act together it won't change anything - as cheaper prices will just increase demand and hash-rate will rapidly rise to make those new cheaper prices unprofitable as well.

All the time issuers can sell over-priced shares to the ignorant they'll do so.  And in the process they suck any chance of profit away from the few investments that may otherwise have had a chance to be profitable for investors - and also ensure there's no urgency for manufacturers to lower prices.  Mining will only be profitable for those who either gamble (first preorders from a new manufacturer who are selling preorders cheap to gain capital without giving away equity rights) or have a significant edge.

Note that J-D profits are estimated.  I KNOW exactly what our balance is (higher than reported) and have roudned it down to approximate for the management fee due (to dooglus) on our profits.  I don't want to divest+reinvest to get it exact - as that potentially costs us if J-D gos into a dive (as we paid fees we wouldn no logner have to).  The value I've given is accurate to within .1 BTC - I could try to calculate it exactly but no real point as it moves around by more than that every second anyway.

Sold   1135
Swapped   0
Total   1135
Price   0.02828
Total   32.0978
Less Fee   32.0336044
Man Fee   0.961008132

Coinlenders CD 28/8    201.37435494
Coinlenders CD 13/8    101.64672106
Just-Dice Balance    159.60000000
TOTAL ASSETS    1,834.38349418
   
Outstanding MINING   61381
Outstanding SELLING   61381
Outstanding PURCHASE   6695
Effective Units   68076
   
Block reward   25
Difficulty   37,392,766
Hashes per MINING   5000000
   
Daily Dividend    0.00006725
50 days (Min Liquid)    0.00336232
100 days (Forced Close)    0.00672465
365 days (Buyback)    0.02454497
405 days (IPO)    0.02723483
400 days (Post SELLING div)    0.02689860
410 days (Pre SELLING div)    0.02757106
   
NAV Post MINING Div    1,829.80562155
NAV/U Post MINING Div    0.02687887
Days Dividend Post Div   399.71
SELLING Dividend    -         
NAV Post SELLING Div    1,829.80562155
NAV/U Post Selling Div    0.02687887
PURCHASE selling price    0.02822
PURCHASE buy-back price    0.02634
   
J-D House profit at report   4256

Anyone who looks carefully will find that the number of PURCHASE paid a dividend is actually 25 less than in the report above.  reason for that is I've been testing software to automate the transfer process - and as part of it have been sending PURCHASE from my personal account to the fund (and then returning them).  I forgot to send the last 25 back before the dividend processed so I missed out on 25 MINING dividends on those shares.  I only spotted the mistake when I did the audit just now - which immediately highlights any missed transfers as the total outstanding doesn't tally with the previous outstanding and sales (every report so far it's exactly matched - though on a few occasions I've had to go through transaction logs to find missed transfers).

Reinvestment is a huge red herring - if something is making a loss in the first place then all reinvestment does is compound that loss.

The fundamental error OP is making is working on the assumption that there IS a way to profitably invest in mining.  In general mining is, at best, marginally profitable (exceptions being those that get in EARLY on new technology - early has now passed for ASICs) - that leaves very little scope for a manager to take a fee and investors to still make a profit.

The basic theory is simple : all the time mining is profitable more people will mine until it ceases to be profitable.   That then gets compounded because when BTC rises vs USD mining becomes profitable (in USD terms) again - so mining capacity again increases.  That not only removes the profit from new miners but guarantees a loss (in BTC terms) for existing miners/investors - whilst they end up making a USD profit (sometimes) they're inevitably going to be worse off than if they'd just held BTC throughout.

If you believe BTC will rise vs USD there's NO long-term profitable way to invest in mining (by which I mean ending up with more than if you'd just held BTC).  All the companies doing reinvestment achieve is compounding their losses (in BTC terms) whilst at the same time giving the illusion of making more profit by paying out more dividends whilst their book value falls far faster.  And, of course, if BTC doesn't rise vs USD then that means it's stagnating and dieing anyway - as it HAS to rise if demand/use increases (due to supply definitely not increasing much and quite possibly falling in terms of BTC available for circulation/use).

The reason PMBs/PMCs perform worse than mining shares is two-fold:

1.  A real reason : the massive markups often placed on their price.
2.  An illusory reason : The visibility of it.  Even when they aren't performing worse than other mining investments they give the impression of doing so as their value (the dividends they pay) are visibly dropping.  In comparison mining companies don't depreciate their hardware or even give book values in BTC - so the (often as bad or worse) losses occurring there are hidden and so don't get reflected in market price.  That's until someone notices what the emperor's new clothes actually are - and the share price crashes.

Any mining share valued at above its Book Value is unlikely to ever make a profit if you invest and hold - as the manager's fees will inevitably outweigh any marginal profit made and reinvestment will turn out to be throwing good money after bad.  And if BTC rises at all significantly a loss (in BTC terms) is assured.

There's a few simple questions to ask when looking at a mining investment:

1.  Will the manager take fees/shares even if investors don't make a profit?
2.  Where's the manager's projection of future difficulty - and how has it compared so far to what actually happened?

If they can't answer 2 then they don't even claim to know it'll make a profit - not many businesses get away with raising capital without even a token effort to establish that there'll be a profit.  Which explains why the answer to 1. is nearly always "Yes - they get a cut of revenue not profit : turning a small loss for investors into a large one".

There's only a small set of securities that should be allowed to operate without some reasonable effort to determine that investors will make a profit - investments which are by their very nature speculative.  Things like:

Precious metals funds - which are explicitly speculation on something outside the control of the issuer (the price of PMs).
Options, Futures and similar - where investors effectively bet with one another on things (DMS is this type of operation - those holding SELLING and MINING are basically betting against one another on what the future dividends from 1 MH/s will be.  It's explicit that some of them will win at the expense of others who lose).

Anything which actually plans to operate a business should determine whether it'll make a profit before asking for cash.  Mining investments don't do this - with the general excuse being that it's hard to work out how difficulty will change.  That's true - so look back rather than forward - and see how horribly mining investments did in the past.  Now try to explain why the same won't happen again - why hashing power will stop being added before it becomes unprofitable.

You DON'T need to predict future difficulty to work out that mining investments in general won't make a profit.  Common-sense and history tell you that they'll make a loss in BTC terms unless BTC falls vs USD.