Exchanges - Spread
Hi. Sorry about this, but I'm a relative newbie and have a stupid question that is really bugging me. Why is it that the lay odds and back odds must never be the same, that the lay odds are generally slightly higher and that the lay market has an overround under 100% while the back market has an overround over 100%? I understand that, hypothetically, if I offer to lay a horse for £10 at say 3.5, what I am doing is laying a backer's bet and I'm liable for the backer's £25 payout should the horse win. However, that being the case, I would expect the backer to be able to back at 3.5 in order to have a chance of winning my £25, but of course because of the 'spread' it's only available to back at 3.3 or 3.4 in which case surely the backer can only win 23 or 24 of my £25? Therefore, why the discrepancy and where does the difference go? Apologies, I know I'm dumb and haven't fundamentally grasped how these markets work but I can't find a simple explanation anywhere! Any help and explaining like I'm a 5 year old is much appreciated!
- ShaunWhite
- Posts: 9731
- Joined: Sat Sep 03, 2016 3:42 am
In your example of laying at 3.5, the person you're matching with has offered to back at 3.5. The whole of the £25 transfers from one to the other. You're selling at the same price they are buying (or visa versa).
Betfair doesn't have a middleman, if you back then someone else is laying your bet and visa versa. How much the spread affect you will depend on if you're taking or offering prices. Buy and sell by taking prices and you lose a tick, but if you offer both sides then you gain a tick.
Betfair doesn't have a middleman, if you back then someone else is laying your bet and visa versa. How much the spread affect you will depend on if you're taking or offering prices. Buy and sell by taking prices and you lose a tick, but if you offer both sides then you gain a tick.
Thank you. I think your reply has made something click with me. So am I right in saying that essentially, the odds on offer for laying always appear slightly higher because they are offers made by backers who are obviously looking for the highest odds possible? And likewise, the odds on offer for backing are always slightly lower because they are offers made by layers who obviously want the lowest odds (to give the lowest liability) possible? And that being the case then, early in the life of a market, back and lay odds and overrounds will be miles apart because there is little money in the market at that point and people will be making pretty speculative offers for unrealistic odds in many cases, but as the market gets closer to the off, it becomes more efficient as the 'correct' price is decided upon and thus the overrounds and back/lay odds get closer together? I think that makes sense at least...
The prices are determined by the interaction between supply and demand. In the context of Betfair Exchange, supply refers to the amount of money being offered to back or lay a particular selection, and demand refers to the amount of money people match from the supply. So, even if supply and demand are in balance the odds will remain stable until there is a change in the market conditions such as new information becoming available or a shift in the betting patterns.
To back and lay at the same price without the price changing you have to be part of both supply and demand.
To back and lay at the same price without the price changing you have to be part of both supply and demand.
- ShaunWhite
- Posts: 9731
- Joined: Sat Sep 03, 2016 3:42 am
Same is true of large scale HFT, EMH is true at mid price so offers are value, taking isn't. That was my lightbulb moment. And it isn't sport or market type dependant, and it will always be that way. I've just clocked up day 1,500 doing nothing but that.
So yep, offering is where its at unless there's a strong move you'd miss unless you took. And adverse selection affects offers too sometimes so as usual everything is nuanced.
- jamesedwards
- Posts: 2325
- Joined: Wed Nov 21, 2018 6:16 pm
I find that offering bets always works best for pre-off, where the market moves relatively slowly. But for very fast moving markets like in-running horse racing I find taking bets works best, otherwise I tend to get picked off for anything -ev and left hanging for anything +ev.
Interested in your thoughts on this. Given your experience do you think I'm leaving money on the in-running table?
I would say if the odds are steaming/drifting quickly in-play then 'Back' or 'Lay' at 'Best Market Price'jamesedwards wrote: ↑Thu Aug 03, 2023 8:52 pmI find that offering bets always works best for pre-off, where the market moves relatively slowly. But for very fast moving markets like in-running horse racing I find taking bets works best, otherwise I tend to get picked off for anything -ev and left hanging for anything +ev.
Interested in your thoughts on this. Given your experience do you think I'm leaving money on the in-running table?
- ShaunWhite
- Posts: 9731
- Joined: Sat Sep 03, 2016 3:42 am
Offering, taking best or taking worse than best all depends on the individual strategy. I rarely depart from offering with a variable margin above current price for a variable duration. Others will be completely different.
Quite strategy dependent as if you can get value by taking a price at an extreme, then that's OK. There is an element of either value taking, or value creating (offering). Offering pre-off works well as the markets are slow moving, but there probably are a lot of cases in-play where taking a wrong price is OK as the volatility is probably creating those opportunities.jamesedwards wrote: ↑Thu Aug 03, 2023 8:52 pmI find that offering bets always works best for pre-off, where the market moves relatively slowly. But for very fast moving markets like in-running horse racing I find taking bets works best, otherwise I tend to get picked off for anything -ev and left hanging for anything +ev.
Interested in your thoughts on this. Given your experience do you think I'm leaving money on the in-running table?
The thing I noticed about in-play is there are opportunities for both. That's not so much the case pre-off, apart from extremes.
- jamesedwards
- Posts: 2325
- Joined: Wed Nov 21, 2018 6:16 pm
Makes sense. Thanks for your thoughts.Euler wrote: ↑Fri Aug 04, 2023 11:57 amQuite strategy dependent as if you can get value by taking a price at an extreme, then that's OK. There is an element of either value taking, or value creating (offering). Offering pre-off works well as the markets are slow moving, but there probably are a lot of cases in-play where taking a wrong price is OK as the volatility is probably creating those opportunities.jamesedwards wrote: ↑Thu Aug 03, 2023 8:52 pmI find that offering bets always works best for pre-off, where the market moves relatively slowly. But for very fast moving markets like in-running horse racing I find taking bets works best, otherwise I tend to get picked off for anything -ev and left hanging for anything +ev.
Interested in your thoughts on this. Given your experience do you think I'm leaving money on the in-running table?
The thing I noticed about in-play is there are opportunities for both. That's not so much the case pre-off, apart from extremes.
- ShaunWhite
- Posts: 9731
- Joined: Sat Sep 03, 2016 3:42 am
Matching at -EV can't ever be a good idea. This 'volatility' opportunity presumably creates net +EV but it would be even greater without the -EV open and just doing the *EV 'closing' bet.
It's always difficult to have a sensible debate on here because the conversation just ebbs away but in short, how can taking a "wrong'" price ever be OK? ("wrong" presumably meaning - ev). Stripping out the - EV bets is the key to refining your process.
Here's a test for you experts.
From the following log data can you guess whether I 'offered' or 'took' the following bet
and did I use 'Best Market price' or Reverse Market Price' and was it +ev or -ev
04/08/2023 14:03:45: [G_Auto 1] : Store Value (Shared) for Buford: BP1 = 1.24
04/08/2023 14:03:45: [G_Auto 1] : Store Value (Shared) for Buford: LP1 = 1.29
04/08/2023 14:03:45: [G_Auto 1] : Store Value (Shared) for Buford: LTP1 = 1.29
04/08/2023 14:03:45: [G_Auto 1] : £ 1.87 Back bet placed on Buford at 1.07. Fully matched at 1.24
From the following log data can you guess whether I 'offered' or 'took' the following bet
and did I use 'Best Market price' or Reverse Market Price' and was it +ev or -ev
04/08/2023 14:03:45: [G_Auto 1] : Store Value (Shared) for Buford: BP1 = 1.24
04/08/2023 14:03:45: [G_Auto 1] : Store Value (Shared) for Buford: LP1 = 1.29
04/08/2023 14:03:45: [G_Auto 1] : Store Value (Shared) for Buford: LTP1 = 1.29
04/08/2023 14:03:45: [G_Auto 1] : £ 1.87 Back bet placed on Buford at 1.07. Fully matched at 1.24
I guess that was a bit of a difficult test for all you experts. There doesn't appear to be any clues in the 'log' data to say whether you have 'taken' or 'offered' a betBobajob wrote: ↑Fri Aug 04, 2023 9:41 pmHere's a test for you experts.
From the following log data can you guess whether I 'offered' or 'took' the following bet
and did I use 'Best Market price' or Reverse Market Price' and was it +ev or -ev
04/08/2023 14:03:45: [G_Auto 1] : Store Value (Shared) for Buford: BP1 = 1.24
04/08/2023 14:03:45: [G_Auto 1] : Store Value (Shared) for Buford: LP1 = 1.29
04/08/2023 14:03:45: [G_Auto 1] : Store Value (Shared) for Buford: LTP1 = 1.29
04/08/2023 14:03:45: [G_Auto 1] : £ 1.87 Back bet placed on Buford at 1.07. Fully matched at 1.24