引用:

原帖由 lung 於 20-3-2006 04:28 PM 發表
玩馬最緊要細水長流，
有入有出最開心，
孤注一擇最不要得。
就算比你贏一百幾十萬如何？
若繼續賭，又會好快又比翻馬會，
跟著愈賭愈大，跟著愈賭愈狼，
問你有幾多身家？
若果無唔賭 ...

You are correct. That's why we need applied Math. Suggest to use Kelly's formula.

引用:

原帖由 KK 於 21-3-2006 02:03 PM 發表
我覺得Kelly's formula始終要依賴每隻馬的勝機來計算,若然不準,結果都是只得一個...輸.

Of course la. If we are bad at estimating the 勝機, can we win by any methods? Kelly's formula only max the profit if we have better estimation than the public. It does not mean it is a MUST WIN formula.

Also agree it is difficult to estimate p of all horses. But we can leave the public odds unchanged, and then adjust the p of the horses we have different estimates from the public. And then, let the program normalise the p. It is not the best way but it saves time.

Use Kelly's formula may cause capital drop to 50% easily if unluckily. Can do in a conversative way by betting the fraction of suggested amount.

Welcome for discussion :)