Tennis article

[sniffer66]

Gong back to the article that Juicey posted re 3rd set lays on 3 setters, I've been running some live code, thresholds based on a little back testing, since the 8th June. No rank\surface filtering as yet, I'd have to pull some external data in to achieve that

3rd_set.jpg

Only 106 entries so far, so it's a small sample but it's not doing too badly.
I can get scores from ITF & Challenger in my code that aren't available in BA, those are actually doing better than ATP\WTA, with WTA performing quite badly to date. Running this in BA in the same period would have you at a slight loss, assuming WTA\ATP only

It's hands free so will leave it running, but looks like it might be an easy slow burner to leave on in the background.

[JuiceyJones]

Gong back to the article that Juicey posted re 3rd set lays on 3 setters, I've been running some live code, thresholds based on a little back testing, since the 8th June. No rank\surface filtering as yet, I'd have to pull some external data in to achieve that


3rd_set.jpg

Only 106 entries so far, so it's a small sample but it's not doing too badly.
I can get scores from ITF & Challenger in my code that aren't available in BA, those are actually doing better than ATP\WTA, with WTA performing quite badly to date. Running this in BA in the same period would have you at a slight loss, assuming WTA\ATP only

It's hands free so will leave it running, but looks like it might be an easy slow burner to leave on in the background.

That's very cool. I wonder if once you move past grass season and get a decent sample size for hard courts it will turn around for the main tour. Clay feels like it has been very sticky this year compared to past years.

[sniffer66]

Gong back to the article that Juicey posted re 3rd set lays on 3 setters, I've been running some live code, thresholds based on a little back testing, since the 8th June. No rank\surface filtering as yet, I'd have to pull some external data in to achieve that


3rd_set.jpg

Only 106 entries so far, so it's a small sample but it's not doing too badly.
I can get scores from ITF & Challenger in my code that aren't available in BA, those are actually doing better than ATP\WTA, with WTA performing quite badly to date. Running this in BA in the same period would have you at a slight loss, assuming WTA\ATP only

It's hands free so will leave it running, but looks like it might be an easy slow burner to leave on in the background.

That's very cool. I wonder if once you move past grass season and get a decent sample size for hard courts it will turn around for the main tour. Clay feels like it has been very sticky this year compared to past years.

That's the hope. I have the tournament names in the data so can quite easily create a surface lookup for each entry. Will need a lot more data to know if it's going to make a difference but will analyse it once I get a few months in

[sniffer66]

Bit more data in, looks like filtering out Challenger is the way to go. WTA back in the green

tennis_thirdset_cumulative_points_5_lines_latest_copy4_no_mens_wimbledon.png

tennis_thirdset_cumulative_points_no_challenger_latest_copy4.png

[jamesedwards]

Thought this was an interesting read:

https://medium.com/@temple.daniel/i-fou ... ba4445a3ea

I built a Bet Angel automation for this for a laugh.

8 straight losers so far.


z158.jpg

This looked promising for a while, then turned to mush. Thought it sounded too easy.

z287.jpg

[sniffer66]

Thought this was an interesting read:

https://medium.com/@temple.daniel/i-fou ... ba4445a3ea

I built a Bet Angel automation for this for a laugh.

8 straight losers so far.


z158.jpg

This looked promising for a while, then turned to mush. Thought it sounded too easy.


z287.jpg

Mines done the same the last week or so, even with the selective filters.

I've got a coded paper version running so will leave that on in the background for interests sake

[Euler]

You may want to look at surface and tournament types.

Grand slams have different incentives from 125 tour events.