Tennis is the only sport which is played on different surfaces throughout the year. So this is a special feature for this sport. In this paper four surfaces are distinguished, namely clay, grass, indoor hard-court and outdoor hard-court. Each of these surfaces have their own characteristics. Clay is a slow surface where the ball has a high bounce, therefore suitable for quick and defending players. While grass has the characteristic to be very fast with a low ball bounce which is better suited for players who like to attack. This while hard-court is characterized by uniform ball bounces. Ideal for players who like to hit the ball flat and hard. Most of the players have a surface they prefer and some of them are real specialists. Now the most difficult part of all these different surfaces is the switch between them. As a tennis player you have to get used to the different ball bounces and different way of movement between different surfaces. The most difficult switch is from the clay to the grass, because there is almost no time to practice on grass. This lack of time is because there are almost no grass courts in the world. So most of the surprises occur on grass. Table 4 shows if the bookies take all this into account.
Firstly when looking at the mean returns of all the matches played on a surface, playing on grass is the least negative category. Playing on grass generates on average a 50% less negative mean return than playing on clay, hard-court indoor or outdoor. Also when taking the 0.0645 take of the bookies in mind a mean return of -0.049 can be positive if the take would not exist.
Secondly the surfaces clay, indoor and outdoor hard-court generate a favorite longshot bias pattern. On all these three surfaces betting the huge favorite generates the best mean return of all categories, while vice versa the heavy underdog generates the worst mean return possible.
So the bookmakers set the odds for the heavy favorites too high and for the heavy underdogs too low. This could be caused by the bookies being biased or by the bettors being influenced by the odds and therefore over betting the underdogs and under betting the favorites. These results are in line with the previous literature, for example [Win06] and [RMG49].
|
grass
|
clay
|
Indoor hard-court
|
Outdoor hard-court
|
odds category
|
N
|
mean return
|
St. dev.
|
t-test
|
N
|
mean return
|
St. dev.
|
t-test
|
N
|
mean return
|
St. dev.
|
t-test
|
N
|
mean return
|
St. dev.
|
t-test
|
0,0-0,1
|
72
|
0,125
|
4,460
|
0,238
|
154
|
-0,649
|
2,188
|
-3,683
|
89
|
-0,427
|
3,821
|
-1,054
|
227
|
-0,577
|
2,354
|
-3,693
|
0,1-0,2
|
129
|
0,066
|
2,539
|
0,295
|
305
|
-0,241
|
2,201
|
-1,912
|
176
|
-0,509
|
1,683
|
-4,010
|
448
|
-0,182
|
2,239
|
-1,720
|
0,2-0,3
|
213
|
0,090
|
1,822
|
0,722
|
576
|
-0,204
|
1,636
|
-2,998
|
375
|
-0,110
|
1,690
|
-1,261
|
705
|
-0,134
|
1,654
|
-2,151
|
0,3-0,4
|
221
|
-0,216
|
1,274
|
-2,521
|
596
|
-0,087
|
1,341
|
-1,577
|
357
|
-0,055
|
1,360
|
-0,760
|
652
|
-0,107
|
1,324
|
-2,056
|
0,4-0,5
|
200
|
-0,112
|
1,114
|
-1,420
|
628
|
-0,083
|
1,120
|
-1,851
|
355
|
0,011
|
1,140
|
0,190
|
781
|
-0,085
|
1,125
|
-2,119
|
0,5-0,6
|
137
|
-0,081
|
0,929
|
-1,016
|
435
|
-0,079
|
0,922
|
-1,790
|
277
|
-0,084
|
0,929
|
-1,505
|
554
|
-0,069
|
0,928
|
-1,747
|
0,6-0,7
|
258
|
-0,044
|
0,751
|
-0,932
|
770
|
-0,057
|
0,754
|
-2,091
|
431
|
-0,091
|
0,761
|
-2,493
|
940
|
-0,068
|
0,760
|
-2,728
|
0,7-0,8
|
206
|
0,003
|
0,591
|
0,068
|
598
|
-0,039
|
0,609
|
-1,561
|
394
|
-0,083
|
0,626
|
-2,642
|
692
|
-0,039
|
0,608
|
-1,673
|
0,8-0,9
|
199
|
-0,109
|
0,511
|
-3,012
|
515
|
-0,037
|
0,463
|
-1,806
|
317
|
-0,041
|
0,470
|
-1,568
|
625
|
-0,032
|
0,459
|
-1,759
|
0,9-1,0
|
171
|
-0,070
|
0,340
|
-2,705
|
369
|
-0,015
|
0,262
|
-1,101
|
160
|
0,010
|
0,248
|
0,499
|
566
|
-0,035
|
0,292
|
-2,866
|
total
|
1806
|
-0,049
|
|
|
4946
|
-0,105
|
|
|
2931
|
-0,099
|
|
|
6190
|
-0,099
|
|
|
Table 4: All ATP matches played in the period 2010-2012 divided by surface played
Thirdly, in contrast to the other surfaces, grass does not generate the favorite longshot bias. Table 4 shows that the underdogs with a prior match winning probability of 30% or lower generated a positive mean return. This suggests that underdogs playing a match on grass are under betted by the bettors or/and the odds set by the bookies are too high. When looking at the favorites there is no under or over betting. A mean return of -0.07 is almost equal to the bookies take. So the matches played on grass exhibit a reverse longshot bias.
Something confirmed by [Lin94] for the baseball betting market but contradicted by the results found overall for tennis, at the Grand Slams and on surfaces clay, indoor hard-court, outdoor hard-court and previous literature as [Muk77].
Fourthly all the positive mean returns found are not significantly different from zero. Also there is no other literature studying tennis matches on grass surfaces3.
Share with your friends: |