What to include in a tennis betting system?

[gottogethelp]

I've started building a system for betting on tennis and have been reading avidly through these forums and a few other sites for ideas on what to include. I've compiled the following list:

• Matches/games played (for experience, fatigue, match fitness)
• Match/game wins
• H2H record
• In-play stats, e.g. serve, return, break point record etc
• Upset liklihood
• Momentum
• Home advantage
• Elo / Glicko / other rating systems
• Player matchups, e.g. big server vs weak returner, leftie vs rightie etc
• Player specifics, e.g. don't play well in hot weather or at altitude
• Tournament specifics, e.g. faster surface than average
• General news stories, e.g. personal life disruption, potential injury etc
• Sentiment analysis - the forums here included ?

Most of the above can then be further divided by:

• Time period
• Surface
• Tournament
• Rank
• Round

I was wondering if there was anything missing? I'm interested in absolutely anything... from the glaringly obvious right down to something small but usually very useful! Indeed there may be specific views on the importance of the various data points above. Would be great to hear peoples' thoughts...

[Smiles Tennis]

How long is it going to take you to make selections if you are looking at all that info for each player? All the common info is already included in the price. You need to find a way to find an edge, rather than just looking at H2H record etc. 

Also, upset likelihood? How are you going judge that? Just your opinion? a pundit's opinion? the markets opinion? 

A few things to think about. 

[The Accountant]

Fitness, motivation etc are huge as its an individual sport - take Osaka yesterday, all was geared towards her victory and she was beaten in SS. Weather also can have a big effect; if its very hot, or at altitude etc. There can also be value inplay where momentum changes. I was told by a pro tennis trader a few years ago that no one should bet or trade on tennis until they've paper traded and simultaneously watched 100+ matches. Most people I would suggest would barely do that for 20 matches and they'll more than likely lose. Just FYI tennis is also a sport with a HUGE amount of pro bettors and traders at it full time so you'd need to be v v good to consider going at it seriously.

[gottogethelp]

On 7/26/2021 at 11:19 PM, Smiles Tennis said:

How long is it going to take you to make selections if you are looking at all that info for each player? All the common info is already included in the price. You need to find a way to find an edge, rather than just looking at H2H record etc. 

Also, upset likelihood? How are you going judge that? Just your opinion? a pundit's opinion? the markets opinion? 

A few things to think about. 

Hey. Thanks for the reply and sorry for the delay in me replying!

I've written some software to pull the info from a range of sources and trained a machine learning model to output predictions in real time. Acknowledge that all the common info is in the price but market inefficiencies definitely still exist right up until closing. I've actually found speed seems to be one of the most important components of making a consistent profit.

On upset likelihood - we look at the predictions our model made vs the results and then run a statistical analysis to determine if a player is more or less likely to cause or suffer an upset.

[gottogethelp]

On 7/28/2021 at 3:26 AM, The Accountant said:

Fitness, motivation etc are huge as its an individual sport - take Osaka yesterday, all was geared towards her victory and she was beaten in SS. Weather also can have a big effect; if its very hot, or at altitude etc. There can also be value inplay where momentum changes. I was told by a pro tennis trader a few years ago that no one should bet or trade on tennis until they've paper traded and simultaneously watched 100+ matches. Most people I would suggest would barely do that for 20 matches and they'll more than likely lose. Just FYI tennis is also a sport with a HUGE amount of pro bettors and traders at it full time so you'd need to be v v good to consider going at it seriously.

Hey. Thanks for the reply and sorry for the delay in me replying!

We've definitely captured fitness and weather now - on a short and medium term basis. Motivation is an interesting one - were you thinking of things like breaking into the top 100 or qualifying for the end of season finals? I've been trying to think about other angles in this space but that's as far as I've got.

Yeah there are definitely some talented people in the market. We back test against 500k previous matches and then have a rule of paper trading for 300 before we put anything live!