Tennis Analysis 2009 – Part 6
When looking at the ATP analysis for the favourites, the data is shown below with a brief discussion following:
| Overlay Prob 50-60% | Overlay | #Bets | #Wins | %Wins | $Bet | $Profit | %ROI |
|---|---|---|---|---|---|---|---|
| 0 | 0.05 | 310 | 168 | 54.2% | $8,354.41 | -$116.16 | -1.4% |
| 0.05 | 0.1 | 236 | 114 | 48.3% | $18,503.13 | -$789.72 | -4.3% |
| 0.1 | 0.15 | 209 | 118 | 56.5% | $24,715.32 | $3,785.48 | 15.3% |
| 0.15 | 0.2 | 182 | 84 | 46.2% | $27,730.66 | -$257.90 | -0.9% |
| 0.2 | 0.25 | 121 | 50 | 41.3% | $21,556.05 | -$1,496.39 | -6.9% |
| 0.25 | 0.3 | 92 | 41 | 44.6% | $18,652.49 | $955.28 | 5.1% |
| 0.3 | 0.4 | 155 | 57 | 36.8% | $36,699.97 | -$3,866.22 | -10.5% |
| 0.4 | 0.6 | 142 | 76 | 53.5% | $40,041.90 | $17,305.16 | 43.2% |
| 0.6 | 1 | 64 | 19 | 29.7% | $21,659.77 | -$995.08 | -4.6% |
| 1 | 100 | 8 | 2 | 25.0% | $3,162.71 | -$227.10 | -7.2% |
| TOTAL | 1519 | 729 | 48.0% | $221,076.40 | $14,297.35 | 6.5% | |
| Overlay Prob 60-70% | Overlay | #Bets | #Wins | %Wins | $Bet | $Profit | %ROI |
| 0 | 0.05 | 390 | 246 | 63.1% | $16,425.86 | -$3.46 | 0.0% |
| 0.05 | 0.1 | 334 | 194 | 58.1% | $37,185.94 | -$1,298.69 | -3.5% |
| 0.1 | 0.15 | 203 | 111 | 54.7% | $34,549.59 | -$2,014.82 | -5.8% |
| 0.15 | 0.2 | 124 | 88 | 71.0% | $26,857.46 | $7,696.79 | 28.7% |
| 0.2 | 0.25 | 77 | 37 | 48.1% | $19,193.79 | -$1,599.41 | -8.3% |
| 0.25 | 0.3 | 47 | 26 | 55.3% | $12,921.87 | $1,428.25 | 11.1% |
| 0.3 | 0.4 | 77 | 39 | 50.6% | $24,102.28 | $1,531.27 | 6.4% |
| 0.4 | 0.6 | 79 | 46 | 58.2% | $29,149.85 | $10,079.84 | 34.6% |
| 0.6 | 1 | 26 | 11 | 42.3% | $11,457.57 | $1,702.99 | 14.9% |
| 1 | 100 | 4 | 0 | 0.0% | $2,020.76 | -$2,020.76 | -100.0% |
| TOTAL | 1361 | 798 | 58.6% | $213,864.96 | $15,502.00 | 7.2% | |
| Overlay Prob 70-80% | Overlay | #Bets | #Wins | %Wins | $Bet | $Profit | %ROI |
| 0 | 0.05 | 397 | 287 | 72.3% | $25,340.54 | -$704.02 | -2.8% |
| 0.05 | 0.1 | 239 | 183 | 76.6% | $40,745.16 | $4,501.75 | 11.0% |
| 0.1 | 0.15 | 160 | 113 | 70.6% | $39,582.71 | $2,358.64 | 6.0% |
| 0.15 | 0.2 | 90 | 56 | 62.2% | $27,059.35 | -$443.73 | -1.6% |
| 0.2 | 0.25 | 59 | 40 | 67.8% | $20,527.97 | $2,126.82 | 10.4% |
| 0.25 | 0.3 | 30 | 17 | 56.7% | $11,524.19 | -$404.54 | -3.5% |
| 0.3 | 0.4 | 32 | 22 | 68.8% | $13,355.16 | $3,451.87 | 25.8% |
| 0.4 | 0.6 | 19 | 7 | 36.8% | $9,001.15 | -$2,301.09 | -25.6% |
| 0.6 | 1 | 17 | 6 | 35.3% | $9,372.89 | -$1,989.66 | -21.2% |
| 1 | 100 | 1 | 0 | 0.0% | $658.26 | -$658.26 | -100.0% |
| TOTAL | 1044 | 731 | 70.0% | $197,167.38 | $5,937.79 | 3.0% | |
| Overlay Prob 80-90% | Overlay | #Bets | #Wins | %Wins | $Bet | $Profit | %ROI |
| 0 | 0.05 | 290 | 242 | 83.4% | $31,962.03 | $843.43 | 2.6% |
| 0.05 | 0.1 | 160 | 130 | 81.3% | $43,877.81 | $2,154.22 | 4.9% |
| 0.1 | 0.15 | 85 | 72 | 84.7% | $31,825.82 | $3,961.21 | 12.4% |
| 0.15 | 0.2 | 68 | 55 | 80.9% | $29,340.91 | $4,022.53 | 13.7% |
| 0.2 | 0.25 | 26 | 20 | 76.9% | $12,844.48 | $1,650.55 | 12.9% |
| 0.25 | 0.3 | 14 | 9 | 64.3% | $7,416.68 | -$254.63 | -3.4% |
| 0.3 | 0.4 | 10 | 6 | 60.0% | $6,049.96 | -$230.79 | -3.8% |
| 0.4 | 0.6 | 8 | 4 | 50.0% | $4,853.90 | -$588.57 | -12.1% |
| 0.6 | 1 | 4 | 4 | 100.0% | $2,600.35 | $2,788.48 | 107.2% |
| 1 | 100 | 1 | 0 | 0.0% | $707.49 | -$707.49 | -100.0% |
| TOTAL | 666 | 542 | 81.4% | $171,479.46 | $13,638.93 | 8.0% | |
| Overlay Prob 90-100% | Overlay | #Bets | #Wins | %Wins | $Bet | $Profit | %ROI |
| 0 | 0.05 | 178 | 171 | 96.1% | $47,492.24 | $1,209.08 | 2.5% |
| 0.05 | 0.1 | 54 | 47 | 87.0% | $25,576.47 | $248.17 | 1.0% |
| 0.1 | 0.15 | 20 | 16 | 80.0% | $12,661.32 | -$170.12 | -1.3% |
| 0.15 | 0.2 | 14 | 12 | 85.7% | $8,747.44 | $902.11 | 10.3% |
| 0.2 | 0.25 | 6 | 6 | 100.0% | $4,002.88 | $1,373.17 | 34.3% |
| 0.25 | 0.3 | 0 | 0 | - | $- | $- | - |
| 0.3 | 0.4 | 1 | 0 | 0.0% | $700.42 | -$700.42 | -100.0% |
| 0.4 | 0.6 | 1 | 0 | 0.0% | $733.27 | -$733.27 | -100.0% |
| 0.6 | 1 | 0 | 0 | - | $- | $- | - |
| 1 | 100 | 0 | 0 | - | $- | $- | - |
| TOTAL | 274 | 252 | 92.0% | $99,914.04 | $2,128.72 | 2.1% |
Here one sees good profits amongst all groups. From 2.1% ROI in the 90%-100% range, to 8% ROI in the 80-90% range. There is little to be concerned of here. One may note that when the probability is under 70%, then the model has not been profitable when the overlay is under 10%. Punters may want to restrict their betting in these cases. However with massive profits obtained in the 10-15% overlay range when the probability is between 50 and 60%, and again between 70 and 80%, this could well be just a issue of random variation due to a small sample size.
Interestingly, when the overlay is greater than 60% and the probability greater than 50%, a 4.1% ROI loss was made. It is hard to look to closely at this, because just the next step down – when the overlay is between 40 and 60% – for the same probabilities, a 28.3% ROI profit was made.
It is hard to get any conclusive analysis except for the fact that the model is strongly positive for ATP for favourites. A similar analysis for WTA will follow in the next post.
Let us know what you think in the comments below.
Sign up to the Sportpunter Tenns Model by 31/12/09 to receive a free Sportpunter Polo shirt and free entry, along with myself, into day 1 of the Australian Tennis Open.
Thanks. I think its very interesting to see some numbers for the favourites as well. I am surprised by the 50-60% probs. It looks like all the real profit there is when the overlay is greater than 40%. I think I will change my betting strategy for that segment.
When you talk about small sample size Jon, how many bets do you think are needed before a sample size is good and variation is not a factor?
its really hard to say bovina, probably around 500-1000 bets, which is of course very high. We can only approximate
If we look at the 50-60% probs then. Theres about a 1000 bets combined between 5 and 40% overlay with no real profit. So isn’t that a good sample size and a fair call to say that we should only bet overlays greater than 40% in the 50-60% probs? Or am I missing something there?
yes for sure you make a good point, and it would probably be a godo wise decision to only bet 40% overlays plus in this braket. However it may not pan out that way. Look at the massive profits from 40-60% overlays, compared to 30-40% overlays. That is how much natural variation occurs.
That said, I agree with you, that it might be better in the long term, to bet 40% min overlay for this bracket. At least you wont be wasting time betting bets that are break even or only marginally positive
Also another thing. considering the federer factor, it makes little sence to me, why the minimum overlay for the 50-60% range should be greater than the 40-50% range. Hence the random variation between ranges
Thanks for the input Jon. I will use a more defensive strategy in ATP this season in an effort to get back on the winning track. Only time will tell if its a wise decision or if I am missing out on some nice profits in the lower overlays.