Asian Handicap Question

[geoffw123]

I never use Asian handicap markets, so they always puzzle me when I working on my App. Here is another one that has got me pondering.....

What I am unsure of currently is how the commerical apps that implement a Practice mode can calculate a profit or loss amount internally, without getting the actual monetary amount returned by Betfair (Obviously this wont be returned as a practice mode bet is not a real bet and is inviisible to the Betfair Server)

lets say I place £2 bet on Team A +1 & +1.5

So we are effectively having 2 bets on Team A of £1 each with Team A getting a 1 goal and a 1.5 goal advantage before we start

Possible outcomes are

Team A Wins by any score      Both Bets win

Any Draw                      Both Bets win

Team B Wins by 1 goal         Bet +1 is level outcome so bet is void, £1 stake returned
                              Bet +1.5 still wins so we get 1 * (odds-1) profit returned

Team B Wins by 2 goals or more
                              Both Bets Lose we lose 2 quid

As far as I am aware at market close Betfair will just return me a "Win" or "Lose" result for this market row.

That means I cant distinguish between the condition where both my £1 bets win or the condition where 1 bet wins and 1 bet loses. So how can I calculate the profit/loss for this bet ?

Would be great if one of the commercial guys (Hello Mr BFExplorer) could explain this please ?

Thanks
Geoff

[WTPooh]

If status for A +x.25 is 'WINNER' then check status for A +x. If it is 'WINNER' then both bets won.
For A +x.0 you will get 'LOSER' status if team B wins by x goals. You need to check status for A +x.25. If it is 'WINNER' then your stake returned.
For A +x.75 you will get 'LOSER' status if team B wins by x+1 goals. You need to check status for B -(x+1). If it is 'LOSER' then half of your stake returned.

[geoffw123]

Hiya

Thanks for the feedback.

I did wonder if the way to do this was look at the other row results, but I had thought that was so fiddly that Betfair would have provided a cleaner way to do this.
It is kind of misleading even for the single score rows eg Team A +1 that Betfair for certain results returns "LOSER" when in reality the bet is a void outcome.
Anyways I have coded up something now, it looks very ugly and is about 2 pages long but seems to work on a quick test.

Thanks again
Geoff

[WTPooh]

I have another idea. You just need to find a row where both A+x and B-x have status loser. There can be no such row if one team won by more then maximum handicap on the market.

[geoffw123]

Hi

I only wrote this yesterday, so absolutely no guarantee that this is corrent.

if handicap ends in .0 or .5 then its a single item row, logic is

if the Asian HC row status is "Winner", it really is a winner. P/L is just standard simple stake * (odds-1) etc

If he Asian HC row status is "Loser" then add 0.5 goals and if that row is status = "Winner" then this bet is void, no loss
else if its still a "LOSER" then this is really a losing bet, your stake is lost.

For the .25 and .75 double rows, I just split it into 2 bets of stake/2, and follow the same logic as above.
-----------------------------------------------

Hmmm, as I was reading this back, there is an obvious flaw with this. What if the bet is on the Team A +4 goals ?

I cant go and check +4 + 0.5 goals row as it doesnt exist

This might be solvable, as Team A +4 is the same as a bet on Team B -4 goals, I can check row Team B -4 + 0.5 = -3.5 goals as that does exist. Need to ponder this a bit more.

Thanks Betfair for making this so simple and straightforward

[WTPooh]

If you can find a row where both A+x and B-x are losers then team B won exactly by x goals.
Example if there is no such row: A+4 is a loser, B-4 is a winner. Then team B won by more then 4 goals. You can assume that team B won by 5 goals.
Now you know which team won and win margin. You can settle your bet.

[geoffw123]

Hi WTPooh

Thanks again for the feedback, I like that last suggestion much more than my first solution, so I ditched my first attempt and rewrote it to use the above logic.
The code is shorter and looks cleaner, hopefully it works OK ! I can post a bit of code if anyone is interested, give me a shout if you want to see it.

Regards Geoff

[Gox]

Hi WTPooh

Thanks again for the feedback, I like that last suggestion much more than my first solution, so I ditched my first attempt and rewrote it to use the above logic.
The code is shorter and looks cleaner, hopefully it works OK ! I can post a bit of code if anyone is interested, give me a shout if you want to see it.

Regards Geoff

Hey Geoff,
I've been struggling with the same problem for days now. May I ask if you could post a snippet of the code, just to see how your algorithm works?

Thanks,
Goran

[geoffw123]

HIya

I am not really dialled into the details of this now, it was months back when I was struggling with it. I came up with this C# code, its fairly short, but no guarantee it works. Well volunteered you can test it for me. Please let me know if you think it is correct or not.

Regards

     struct resultAsian
      {
         public short margin;
         public string result;
      };

      static resultAsian figureOutAsianResult(CMarketDataESA mktData, out long winnerSelId)
      {
         resultAsian[] asianRow = {new resultAsian { margin= 4, result = "Home" },
                         new resultAsian { margin= 3, result = "Home" },
                         new resultAsian { margin= 2, result = "Home" },
                         new resultAsian { margin= 1, result = "Home" },
                         new resultAsian { margin= 0, result = "Draw" },
                         new resultAsian { margin= -1, result = "Away" },
                         new resultAsian { margin= -2, result = "Away" },
                         new resultAsian { margin= -3, result = "Away" },
                         new resultAsian { margin= -4, result = "Away" },
                        };

         winnerSelId = -1;   // draw

         for (int i = 0; i < asianRow.Length; i++)
         {
            var idx = i * 8;
            if (mktData.runnerData[idx].status == RunnerStatus.LOSER &&
                                 mktData.runnerData[idx + 1].status == RunnerStatus.LOSER)
            {
               if (asianRow[i].result == "HOME")
                  winnerSelId = mktData.runnerData[0].selectionId;
               else if (asianRow[i].result == "AWAY")
                  winnerSelId = mktData.runnerData[1].selectionId;

               return asianRow[i];
            }
         }
         // shouldnt get here
         return new resultAsian { margin = 0, result = "?" };
      }