So I've been posting the results of the tennis model I use fairly regularly and as noted by reader @Polish_Middle it was time to add a few helpful tracking tools to the presentation....
With this in mind the updated tennis model presentation now looks like this with two added columns of data and symbols:
As always we have the tourney info for tracking purposes, the player, country and opening price from MarathonBet which is regularly the first shop to hang a money line for a tennis match. The calculated implied probability of the opening price is presented next, followed by the "Current" price which is simply the price at 5Dimes (where I bet most of my tennis) at the time that I'm working on my handicap. Finally the Model win % is based on a simple algorithm that combines past performance in general, past performance on given surface, past head-2-head performance, tiredness, recent retirement, success at the given tournament and a current form score to come up with the expected likelihood of the player winning the match-up.
Now at this point I generally stop, create the graphics and post the results before continuing on my handicap. As I've discussed on past Deep Dives, for me the model is just a starting point to assess which sides present expected value and from there I make a series of decisions on the most effective way to back or fade a certain player, recognizing that in a lot of cases when the underdog presents value, there may be a better way to cash a ticket than backing the ML. This is a long way of saying I do not blindly tail the model and it's a discussion for a different post, so onto the new stuff.
The last two columns now show the difference between the model win %tage and the current price with a handy colorful data bar that scales so that you can see the bar clearly when it is over 5% and very clearly at 10%. The final column indicates sides where there is more than 5% perceived edge on a side and if the line is moving in the direction of the model valued player, a green check appears. For players with a perceived edge of more than 5% where the line is moving against the player with value, a yellow exclamation point appears. The reasoning behind this is the model can't capture everything and it's worth knowing if the market is moving in the direction of the model or against it which should help inform the handicap.
Hope you enjoyed this soliloquy and make intelligent decisions with this information. BOL!