-
Posts
803 -
Joined
-
Last visited
-
Days Won
16
Posts posted by Smiles Tennis
-
-
-
-
7 hours ago, Torque said:
You need a big sample before drawing any conclusions. A very big sample. I'd say five figures. The smaller the dataset the more likely that conclusions are drawn that aren't necessarily correct, like the one you mention about Slams.
yeah, the picks come thick and fast when multiple tournaments are on. They did before anyway. I did keep records for a long time but got rid of them as I thought I wouldn't need them again when covid came. Even after covid, the players didn't have any sort of form for 18 months so I decided to just wait. I wish I kept the records now. Maybe they will show up in a corner of my hard drive one day lol
-
-
-
-
-
One more selection for the Sabalenka match. I have now added the expected and actual % for the match. I always go off of the fave. So for this first one, Sabalenka is 1.37 (73%), her actual win rate when around these odds is 77%, so I expect (hope) for her to come closer to 73% over time. Hence backing Kanepi.
-
On 1/23/2022 at 1:59 AM, Torque said:
I'm curious why you're using level stakes, unless it's just for the sake of simplicity. If you're working out a player's historic performance within a particular odds band then surely there's times where the odds are only just above that performance and other times where the odds are way away from where you think they should be. Surely those times where the odds deviate particularly severely from what you think they should be demand a bigger stake than those where the deviation is smaller.
Yes, level stakes just for simplicity at the moment. You're right in that there are some differences in the level of performance when betting, and these could be used to stake via Kelly in the future. I might start noting down the win percentages (actual and expected) and see if larger differences result in more wins. I will add it to the spreadsheet.
-
-
-
-
-
-
-
-
-
Ok, the tennis is back so I am back. My business is manic at the moment so I will try to get selections done when I can but I might miss some as it's constant at the moment.
Current results total is visible in the top right of the image.
I have made selections up to the 06.00 matches, (stopping at Kerber match) and will do those in the morning if they're still to play.
Still paper trading, Betfair prices used at the time of research. Let's see how it continues to fare.
-
Still no bets. Very little money in the markets on Betfair. I am usually looking for at least a few thousand to be matched to have a roughly settled price to work with.
-
-
-
7 minutes ago, alexcaruso808 said:
It is very impressive!
It will be if it works long term, it's not impressive if it doesn't make any money lol
-
1 hour ago, alexcaruso808 said:
Are you getting your stats from somewhere and feeding them into a spreadsheet which calculates what bets you should place? I know you most likely don't want to give specifics, but I'm interested if you're using a script, or spreadsheet, or software? I'm just a nerd who likes to know how this is done.
I have software a friend built that collates all results, odds, stats etc for all matches. I check current markets for matches, feed info into the software that pulls the historical data for similar matches, then shows whether I should bet or not. That's all I can say really without going into specifics too much.
-
Tennis Betting with Mean Regression Algorithm
in Betting Systems & Strategy
Posted
Two starting shortly.