Odds compiling

swedish chef

swedish chef

At the Start
Joined
Jun 28, 2010
Messages
4,847
Location
London
A friend of mine has asked me this question - can anyone answer it? I can't.


"A snooker match, Trump vs Advani. If Trump is 63% to win a frame. How can you calculate match odds from just that information.

The answer is
Trump = 0.79; =1/0.79 = 1.26
Advani = 0.21; = 1/0.21 = 4.76

But i dont understand how you get the answer"
 
rorydelargy said:
You can't get that answer without knowing how many frames need to be won to land the match.
Click to expand...

My mate said - whatever this means?

"sorry, best of 9 frames mate.

These are the 4 questions, and 4 answers below. I just need to know how you work out the answers from the questions, if you could copy and paste that and see if any of the lads could talk me through the answers.

1. Match Odds
2. Correct Score
3. First 4 frames winner
4. First 4 frames correct score

1. Match odds
Trump = 0.79; =1/0.79 = 1.26
Advani = 0.21; = 1/0.21 = 4.76

2. Correct score
T A
0.099244 5-0 0.006934
0.183601 5-1 0.021843
0.203797 5-2 0.041284
0.175945 5-3 0.060687
0.130199 5-4 0.076466
0.792786 0.207214

1. First 4 frames winner
Trump 0.528
Advani 0.146
Tie 0.326

2. First 4 frames correct score
T A
0.1575 4-0 0.0187
0.37 3-1 0.1276

0.326 2-2

3. Advani +2.5 frames
= the sum of all Advani’s chances + Trump 5-3, 5-4
= 0.207 + 0.13+0.176 = 0.513% = 1.95

4. Trump +1.5 frames
= the sum of all Trump’s chances = Advani 5-4
= 0.793 + 0.076 = 0.869 = 1.15"
 
The long boring brute force way that makes you understand it:

Write out all the possible ways that Trump can win a best-of-9 match. For example:

Win 5-0:

W W W W W = 0.63 * 0.63 * 0.63 * 0.63 * 0.63 = 0.0992436543

Win 5-1:

L W W W W W = 0.37 * 0.63 * 0.63 * 0.63 * 0.63 * 0.63 = 0.036720152091
W L W W W W = 0.63 * 0.37 * 0.63 * 0.63 * 0.63 * 0.63 = 0.036720152091
W W L W W W = 0.63 * 0.63 * 0.37 * 0.63 * 0.63 * 0.63 = 0.036720152091
W W W L W W = 0.63 * 0.63 * 0.63 * 0.37 * 0.63 * 0.63 = 0.036720152091
W W W W L W = 0.63 * 0.63 * 0.63 * 0.63 * 0.37 * 0.63 = 0.036720152091

Add them all up = 0.183600760455

(Bear in mind that Trump must always win the last frame)

Do this for 5-2, 5-3 and 5-4. You'll notice each answer is the same as the correct score probabilities in the answer. Add them all up to get the match probabilities.

The better way that is much quicker (once you understand it) and allows you to scale your calculations to any length of match:

Learn about the binomial distribution and then plug the appropriate numbers into the BINOM.DIST formula in Excel, bearing in mind again that Trump must always win the final frame.
 
Thanks Chef for posting up and cheers lads for the help.
 
Last edited:
J_C said:
Thanks Chef for posting up and cheers lads for the help.

I am a trader for a firm in the U.K, relatively new i should add and have literally had one days training in the 3 months ive been here. I just wanted to try and do abit more homework and improve my knowledge.

Gareth, if you have the time, could you just show me how to get to first 4 frame winners and first 4 frame correct score.
Click to expand...

Surely the firm just uses models to workout all these prices? The snooker compiler probably only inputs outright prices. There is very little compiling done these days.
 
It is - if you're just going under another firm then that's a recipe for failure.

Edit - Assuming your client base isn't 100% loyal to your brand.
 
Last edited:
There are of course instances when models will throw up wrong prices. It happens in darts quite a bit. Lately I've been listening to a lot of the ESPN podcasts on betting and US pros have a massive horn at the prospect of getting on in play with off shore sportsbook on NFL. Models are only as efficient as the model builder.
 
So your firm trusts Bet365 prices above Betfair, it's models, it's compilers opinions and its apparent trading strategy. Interesting!
 
There has to be a reason they are doing it. It's probably cost saving from employing the number of actual of people to come up with a number they trust. I'd imagine they don't employ a big pool of compilers or risk personnel. It's not a firm I'd want to work for.
 
Seems daft to build a model and then rely on Bet365's compilers and models to do everything - if they suddenly employ a bad batch of traders/compilers then the firm would be kippered.

Martin
 
An angle in copying Bet365's prices! I've no doubt I could beat the firm he works for. Bet365 strive to deliver product not price efficiency.
 
If I were you I'd worry less about what your firm does and concentrate more on raping other firms mistakes. Also I'd delete every post you've made on the thread!
 
SlimChance said:
If I were you I'd worry less about what your firm does and concentrate more on raping other firms mistakes. Also I'd delete every post you've made on the thread!
Click to expand...

Done.

I did for about 6 weeks but quickly got limited, can only bet with marathon and betfair at the moment. Will try and open up another 365 account in future.
 
Back
Top