Simple way to find Value

[Mulkis]

Hi everyone! This is my approach to finding "true odds" and finding "value bids" (tennis in this case): 1) I take decent amount of historical data with fixed odds indicated (2 years, over 5000 matches); 2) DCOUNT how many games within a specific range of odds are played (+-0.1 decimal odds value); 3) MatchesWon_at_Odds / TotalMatches_at_Odds = Probability of Favorit Win; 4) 100/Probability of Favorit Win = TRUE ODDS (0% profit). Next step is to place a back bid with better then TrueOdds in the exchange and wait it to be matched (and its almost 100% it will be matched due to in-play volatility). Hm... Looks too simple to be true, right? Then what's wrong with this approach? Your comments will be much appreciated!

[Lardonio]

Re: Simple way to find Value Toss it in the oven and see if it bakes...

[Drawsandmore]

Re: Simple way to find Value Hi Mulkis, If I've understood you correctly, the "problem" is that your data relates to the s/p BEFORE the match is started, but (sometimes at least) you need the match to start before the odds will be acceptable value. This will almost invariably mean that your selection is losing the match...Effectively your selection is handicapped and starting the match a couple of games or a set behind. Your value is based on both players starting level if you see what I mean?

[Mulkis]

Re: Simple way to find Value

Hi Mulkis, If I've understood you correctly, the "problem" is that your data relates to the s/p BEFORE the match is started, but (sometimes at least) you need the match to start before the odds will be acceptable value. This will almost invariably mean that your selection is losing the match...Effectively your selection is handicapped and starting the match a couple of games or a set behind. Your value is based on both players starting level if you see what I mean?

Yes, this is one of the main issues with this "value system". And it would be perfect "deal breaker" in soccer or some other sports, when for example one goal changes situation dramaticaly or 15min no goal is major factor for match outcome. But in such game as tennis in-play volatility is often overdone by simple favorit lose of one point in the beginning of the first game in the first set. Odds react 10-20% virtualy for no reason. So, it is quite easy to catch "value odds" with no real "handicap". At least this is my observation.

[Drawsandmore]

Re: Simple way to find Value

Yes' date=' this is one of the main issues with this "value system". And it would be perfect "deal breaker" in soccer or some other sports, when for example one goal changes situation dramaticaly or 15min no goal is major factor for match outcome. But in such game as tennis in-play volatility is often overdone by simple favorit lose of one point in the beginning of the first game in the first set. Odds react 10-20% virtualy for no reason. So, it is quite easy to catch "value odds" with no real "handicap". At least this is my observation.[/quote'] You've obviously seen the potential problem, but if you've observed the odds changing that dramatically so quickly, it certainly does appear to be a "Simple way to find value" in tennis at least. Good luck with this Mulkis, and I will keep checking to see how you're doing.

[slapdash]

Re: Simple way to find Value I suspect you're underestimating the effect of the times you're not matched, even if they are rare. They will happen: for example, if you get an early goal in football or early break of serve in tennis and hang on to the lead then you probably won't get matched. The point is that the times you don't get matched will almost all be times that your bet would have won. This means that you need a very high rate of being matched just to break even. Look at the numbers in an example. Suppose the true odds are 3.00 and you ask for 3.30. Then if the times you are not matched are all times that you would have won then you need to be matched over 95% of the time to break even, and that's not taking commission into account. If you're paying 5% commission then you would need to be matched over 97% of the time to break even.

[Mulkis]

Re: Simple way to find Value Valid point, slapdash, but you are not completely correct. The tricky thing is numbers: when you take such high odds like 3.00-3.30 in your example you are right: one would need to be mached 97% to break even. But, in shorter odds such as 1.30-1.43 one would need to be mached only 77% to break even. At 1.50-1.65 one needs 85% "match rate", but at 1.20-1.35 65% "match rate" - lower the odds lower match requirement.

[slapdash]

Re: Simple way to find Value True. Though I'd expect that when you're backing a very short-odds favourite it's more likely that your bet is not matched: those are precisely the matches where it's most likely that your team/player will get an early material advantage and hang on to it.

[Mulkis]

Re: Simple way to find Value

True. Though I'd expect that when you're backing a very short-odds favourite it's more likely that your bet is not matched: those are precisely the matches where it's most likely that your team/player will get an early material advantage and hang on to it.

Absolutely not! Just look at now in-play Federer-Monfils. Fav starting odds (Federer) was at 1.20. Now it is over 2.00. And believe me this is a very common situation in tennis.

[slapdash]

Re: Simple way to find Value

Absolutely not! Just look at now in-play Federer-Monfils. Fav starting odds (Federer) was at 1.20. Now it is over 2.00. And believe me this is a very common situation in tennis.

I'm not saying that it won't get matched the majority of the time, just that I expect it will get matched less often when you're backing a short-odds favourite. Usually when Federer starts at 1.20, he doesn't go a set down.

[Mulkis]

Re: Simple way to find Value Agree. Shorter the odds - more risk to miss the game unmatched. But as we know at 1.20-1.35 break even "match rate" is at comfortable 65%. And what is even more interesting, if one increses "target odds" from 1.35 to say 1.45, "minimum match rate" decreases to 54%. Crazy, isn't it?