Throughout my tennis tipping and tennis betting career (4/5 years), I have noticed an against-the-grain trend, mostly in challengers but also definitely existing on the tour-level. Many times, I have noticed that a player who has had a long, successful tournament in the prior week (a finalist or champion), has lost or at least struggled against the betting line or expectations if they play the very next week (specifically in their first round). I am posting this now because my goal is to compile statistics over the last 5ish years or so on players who have made the finals in a prior week and their performance if/when they play the immediate following week (specifically their first round performance). I don't just want to look at a simple W-L record because of the varying quality of opponent, but perhaps analyze performance based on an average ranking of that opponent and compare it to the expected performance when a player did not compete or reach the finals in the week before.
For example, here are the results and lines of last week's (16.2.2015) challenger finalists and champion's results in these weeks current tournaments...
Last week, New Delhi CH
This week, Koyata CH
R1: Devvarman (-130) vs. Ramanathan
(+310) 64 36 46
Bhambri wins R1 vs. Lopez-Perez, loses R2 vs. Albot as favorite
Last week, Wroclaw CH
Winner: Dustov, loses R1 this week in Cherbourg as -300 favorite to Setkic, 26 36
Finalist: Basic, loses R1 this week in Cherbourg as -160 favorite to Couacaud (+110), 26 46
Last week, Morelos CH
Winner: Estrella Burgos, loses R1 in Acapulco 250 to Lu, straight sets (Lu was favored)
Finalist: Dzumhur, did not play this week
Yes, this is just one week, but it's another week of many where this trend has lived up to my observations.
I hope to compile these statistics and find a trend that fits my hypothesis that past week's winners struggle and underperform against expectation and the betting line in their first round match the following week. I would love to possibly publish an article using this data for school. If any of you all has insight on a good way to compile this data, please let me know. Thanks!