robot

yes its early days, but the gist of this thread seems to suggest that the bot hooverers on the underround IR horse racing markets are out of business - to the benefit of BF

Surprising we havent heard any moaning yet!

Still think NR's left in an IR market will cause big problems

X matching is the only decent thing to have come out of betfair towers in 10 years.

Battery eggs to be phased out by 2012 in EU scarecrow. Factory betting will go the same way once betfeck starts up.

I noticed that there were big fluctuations about say 1 2 or 3 horses once they had gone past the
2 furlong pole on the flat but that's just because anyone without fast pics abandons the markets at that point.
I had a nice little strategy that relied on price swings but not that late on in a race
Back to the drawing board

FINE AS FROG HAIR Joined: 12 Mar 07
Replies: 1563 24 Nov 10 11:16 
Maybe BF has decided that it's real future lies with satisfying and attracting the normal punter, who is unsophisticated in the ins and outs of bots, high speed bets, arbitraging value discrepancies etc etc etc.
Maybe they see no long term future in having a realitevly few highly skilled operators making guaranteed profits at the expense of others' naivety.
And you know, maybe they're even right.


good point for once there frog [;)]

Think far too many people are getting their knickers in a twist over this. I'm new to the whole in running horsey thing but it does seem to me that it's probably one of the few (if only) markets on betfair where it's possible to make big profits, as opposed to creaming off 0.000.2 percent by trading on some obscure east european women's volleyball match.

Quite simply, some of the odds that are matched in-running on the gee-gees are astounding. the level of stupidity that takes place on a daily basis is nothing short of breathtaking.

I wouldn't even dream of using a bot, presumably this would maximise my performance but when people talk about bots, ladders, gruss, spreadsheets etc my eyes just gloss over....

most other markets on here have been worked out to an almost mathematical precision using a combination of computer programs, statistics etc, but i don't know anyone who has ever invented a computer program that can assess how well a horse is travelling, and all other aspects of in running horse racing punting that is far beyond the reach of any bot, and until they do, it'll be possible to make profits on these markets.

besides I don't see how cross matching will stop the flood of punters on a daily basis who get so excited by how well a horse is travelling that they jump over each other to get on at as low a price as possible.

And I don't feel sorry for those punters who try to lay the field at 1.5, for example. I doubt they can last very long anyway given there are no click and win systems which work on betfair and I don't feel any sympathy for anyone employing such a crude approach.

mythical prince
25 Nov 10 03:55 Joined: 20 Sep 06 | Topic/replies: 2,071 | Blogger: mythical prince's blog
most other markets on here have been worked out to an almost mathematical precision using a combination of computer programs, statistics etc, but i don't know anyone who has ever invented a computer program that can assess how well a horse is travelling, and all other aspects of in running horse racing punting that is far beyond the reach of any bot, and until they do, it'll be possible to make profits on these markets.

Doesn't the market reflect how well a horse is travelling?  You can write a bot to analyse and profit from the market, so the bot can assess how well a horse is travelling and profit from it.  I don't think that in running horse racing is not beyond the reach of automated processes.

The bottom line now is your bets gunna get matched
So think about it for yourself

egner....agree.  Spikes are everywhere.  It means the logarithms must be reaching up or down to fill the cross-bets. 

On 'micro-bets' - thought about this more:  You lay at 6.20.  The 'cross-betting' wouldn't use a delay and would calculate immediately what was required once conditions were filled. 

This might also explain why my (in example) 6.20 'seemed to be' taken before the 6.4 and 6.6 that stacked in front of me shortly after.  The cross-matching had identified my 6.20, checked condition, taken THEN the bet showed up in MY BETS after the 6.4 and 6.6 appeared.  (I work with 1 second refresh rate on site and/or api).

Brodie mentioned how it scuppers his strategy.  Right now, I seem to be benefiting since I identify a value lay in-running and putting it up, it seems to getting filled quicker and allowing me to move onwards faster to the next target.

aye robot, I think you hit something - the cross-matching will trigger the bots who have a decision process based on "last price matched" since this information will be jumping around more instead of bets sitting to be matched.  Someone I worked with had a bot for in-running.  He had it long before most (a pretty costly custom beast but it is really good for programming pre-race and to change decisions in-running) and when other bots arrived and more and more traffic, etc, he found slowing it down (in his case, half second to .80 seconds) was enough to see through the static.  Without knowing your strategy, you might see if adjusting your speed might work.

dustybin - personally, all decisions to lay in-running are mine.  Prices are "set" but not set in stone since it depends on how the race is panning out (I go into the race with notes on 3 to 5 horses and will bet according to their performance as the race unfolds).

Right now I am in-running without even a bot to calculate close and finding this "old skool" approach is keeping me more aware of risk and reward as the race unfolds.

Will be interesting since the past few months I feel that people are chasing odds they shouldn't and placing positions in-running that make no sense.  Some punters, especially Saturdays, should step away from the keyboard once the race starts and accept their bets and positions instead of (essentially) chasing.

Assuming the horses work like all the xmatching that we've had in other sports for ages, if you use a regular betfair window rather than API to view the markets, you'll see the betfair money on display.

May or may not help you see what's going on better shapey if you compare one to the other.

Here's a strange one...kick off due at 1 AM...
Cucuta Deportivo v Nacional Medellin - in-play football, 25 Nov fixtures.

Over/Under 2.5 goals

Under 2.5 goals
Traded and Available
Price To back To lay Traded
1.01 £44     
1.02      £63
1.05      £1,014
1.10      £230


Over 2.5 goals
Price To back To lay Traded
1.01 £44     
1.02      £63
1.05      £1,014
1.09 £2     
1.10      £230


Both overs and unders state £1,014 traded at 1.05 (wow, someone's unhappy). So does that really mean that someone's backed £507 on overs only, or backed £507 on both overs and unders? I assume the latter, but maybe it has something to do with x-matching and he's only backed the one, as the other matched amounts are also indentical?

What a cute chatname.

Not really Trevh.

Commission is only paid on the gross profit made within any individual market. Therefore he/she/they would only pay commission on one account, as the other account would show a loss on the market.

Account 1:
Back Under 2.5 £507 @ 1.05
Back Over 2.5 £507 @ 1.05
Gross Profit (0.05*507)-507 = -£481.65
Commission paid = £0
Net Account Position = -£481.65

Account 2:
Lay Under 2.5 £507 @ 1.05
Lay Over 2.5 £507 @ 1.05
Gross Profit 507-(0.05*507) = £481.65
Commission Paid (assuming 5% rate) = £24.08;
Net Account Position = +£457.57

Thus £457.57 would be transferred from Account 1 to Account 2 regardless of the outcome of the match, other than void, and a fee of £24.08 would be paid to bf (in effect from account 1).

It would be no cheaper for Account 2 to lay £481.65 from Account 1 at 1.01 on a selection correctly priced at 1000. In that situation there would also be a 1/1000 chance that the transfer of funds would be in the opposite direction to that desired.

It's probable that some of these transfers, but by no means all, are made to avoid the scrutiny other forms of transfer, for example bank transfer, may be subject to. The holder(s) of the accounts may never previously have made any bets or paid any commission other than in relation to such transfer(s).

Most of the time in horses IR the cross-matcher will be executing rather than posting bets. In these instances bf will be offering a form of 'best execution' to the prices actually available on the other side. Of course, it will not be perfect cross execution, as bf will pocket the breakage they gain by not improving to a perfect reciprocal.

In some cases tho, as I believe Bayes highlights, in unformed or immediately post-off markets bf will effectively be taking 'one against the field' or most of the field. If so, this will really drain the pool and will have an adverse effect on any winning IR player.

With cross-matchig, market depth (for one runner) becomes market breadth (for the field). Sometimes even breadth will not be enough to sustain a competitive market, resulting in overrounds and out-of-line prices continuing to be matched.

Imv there is a category of people who will lose out for all this who haven't been mentioned yet, and that is the small backer who wins at a massive price due to something submitting into an empty market fraactionally before them.

Rather than sitting there momentarily, this lay will now be parcelled out between existing offers on the other side, meaning that the £5 bettor (or less) is less likely to win at 500.

The cross-matcher benefits panicky people going too far to correct an earlier position (a proportion of bf's client base almsot certain to be losing). It goes against OR scrapers, going in-play layers and maybe even winning IR price-taking layers.

As on the very occasional race I fall into the panicky category I shd not be negatively affected, but it still rankles w/ me that bf will not actually ofer best execution.

this will really drain the pool and will have an adverse effect on any winning IR player.

This is the bottom line, the cross matcher helps losers but it never helps winners.

Only a matter of time before they **** up and let one go in running, tomb raider, those button pushers aren't the brightest. Be interesting to see what they do then

jt45 : Not really Trevh

Yes you're right of course jt.

I spotted another today on the Mainz match, where under 2.5 had been matched pre-off at 32.0 for £553. I would imagine that was a horrendous mistake for the layer, showing a liability of £8571.

The first example that you posted would in many ways be the optimal way of making a transfer between accounts because there is no risk of funds being transferred in the opposite direction to that desired. Despite this, it is likely that some people may attempt to transfer funds between accounts using a variation of the second example that you posted.

One reason for this is that the second method is much more plausible as an error than the first method. I suspect that the first method would be more difficult to explain as an 'error' where it to be detected.

In the second case, if the transfer of funds did go in the opposite direction to that intended then I would imagine that the parties involved would simply try again on another market. The overall cost (commission) may be significantly greater than using the first method if this occurred but those involved may consider that to be a risk to be worth taking.

It is also possible that some of the people that wish to transfer funds between accounts may simply not have considered the 'optimal' method.


** I have used the bets that you posted above to explain how funds could be transferred between accounts using similar methods. I have no reason to assume that the actual bets that you posted were anything other than genuine errors. I believe that any transfers, of the type that I detailed above, are prohibited.

By the way, stop loss if not properly programmed will cause errors.

example.  LAY a player in tennis at 1.50 for 100.

For a stop loss, the proper way is to say if "LAST PRICE" is less than 1.40, take best offer.

I was aware of someone who told me he puts if "BEST OFFER" is below 1.40 to obtain amounts quick.  The problem here is when it goes from MARKET to "IN PLAY", if there are no keep bets or the keep bets are below 1.40, the bot jumps into action and closes out the stop loss when it wasn't necessary.