Jump to content
** March Poker League Result : =1st Bridscott, =1st Like2Fish, 3rd avongirl **

Tennis Betting with Mean Regression Algorithm


Recommended Posts

OK, I am back again, this time with real money instead of paper trading. Using the same method of mean regression for each players performance over their career as explained in older threads. 

Long term I still see it working, the data shows it works and maybe the last thread of paper trading (which roughly broke even) just hit a quiet patch. 

This thread starting with roughly £500 (a little bit over)  2% per bet so £10. 

I will try to post picks with time if I can. As soon as I have done a few I will share. 

 

morning.png

Link to comment
Share on other sites

There was one more that I forgot to add to the spreadsheet, sorry guys! I only just saw it on the Betfair list with a bet matched and I checked to see who I had backed but couldn't find it. 

Backed Milojevic at 2.56 who is about to win the first set. I have attached a screenshot so I am not just after timing, i haven't forgotten to add it to the spreadsheet before and will double check going forward. 

update.png

match.png

Link to comment
Share on other sites

17 minutes ago, Smiles Tennis said:

Last ones on this painful day...lol

 

update.png

It's just one day. You could be 90 points up at this point instead of down and it wouldn't make any difference. The chances are if this works then the edge is tiny and so that means you'll need thousands and thousands of bets for the edge to show and stick. Keep going - as long as you can afford to of course if you're staking real money. :ok

Link to comment
Share on other sites

11 minutes ago, Torque said:

It's just one day. You could be 90 points up at this point instead of down and it wouldn't make any difference. The chances are if this works then the edge is tiny and so that means you'll need thousands and thousands of bets for the edge to show and stick. Keep going - as long as you can afford to of course if you're staking real money. :ok

Yeah it's real money at the moment, and i expect the edge to be small. I can keep going with the £500 i started with, then paper trade if that runs out. 

Link to comment
Share on other sites

OK, the original research and thinking even from before covid etc was mean regression. For example, player A is 1.5 but checking their stats and results shows they win 72% of the time at those odds. I thought they must regress to the mean towards 66%, and I would therefore bet against them. 

We always hear that the markets (mainly Pinnacle closing line and Betfair) are very efficient, the most efficient. After the other day with the losses and further research, I might have my doubts. 

I am always trying to do this in my spare time and throughout my workday (i have my own business so can jump on the computer whenever, this will be ending soon though as we've been taken over). So I spend most of the time making the selections rather than further research. 

Well I found a good Pinnacle article (they have some great ones for those that haven't read them) about the efficiency of tennis markets here: https://www.pinnacle.com/en/betting-articles/Tennis/how-efficient-is-the-tennis-betting-market/U522ZLSEPQJA8Q2N.

So this, coupled with further research makes me think mean regression might not happen with tennis. 

Seppi for eg, played the other day. I checked his career stats as he's been around since about 1925 playing tennis and he has consistently underperformed the odds throughout his career, winning 62% instead of 66% at 1.5 average odds. It's wins and defeats mixed about, not huge runs of wins that take it back to 66% then he drifts from this. 

I checked a few more players with long careers and lots of stats, and at different odds, Nadal (1.3 odds, wins more than expected by about 3%), Federer and others. 

I am now thinking that certain players will always under/overperform compared to the odds. With some overperforming and some underperforming, you can take the results from 10,000 tennis matches and overall it looks efficient (well, as it can be until Pinnacle article arrived in my thoughts). But if you dig down in those 10,000 matches then you have certain players above and some below. 

Now I will be backing/laying compared to if someone has overperformed or not. I will go back to paper trading but keep the same stakes and spreadsheet going. 

Other question I have is for lower ranked players. How many matches do they need to have played before we can see if they are over/underperforming? Do we have to wait until they're halfway through their career before trusting they're over performing? Something to think about. 

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...