Staking plan, robustness and yield in excel???

I´ve got a spreadsheet where I keep a record of my bets. I also got functions for yield and robustness (from joes site).

My current yield is 6,89%

The problem I have is that I vary my stakes according to my estimate of the bets value using a scale from 1-10. If I hadn´t used this stakingplan my yield would have been -14.47% (using my average stake for all bets).

From this I´ve concluded that my stakingplan is working pretty well, nothing peculiar about that. My question is if there is any good way to "rate" my stakingplan statisticaly and if it should have any effect on the robustness of my bets since it effects my winnings....

[Joe]

Actually, if your staking plan deviates significantly from fixed stakes, then my current method of calculating a record's profit significance (i.e. its robustness) is not technically valid. My robustness test assumes that each bet has the same weight, i.e. fixed stake. Since you vary your bet weights (stakes), you may need to think up another way to test your record and profit significance.

The difference between +6.89 and -14.47% is a lot. How many bets have you had? My conclusion from those values would be that your high stake bets (ones you're more confident in) are working well, whilst the low stake ones are not. Am I right here? I would also assume that your confident bets have lower odds. If this is all so then it would suggest to me that the higher odds/low confidence bets aren't worth taking at all. Instead, just bet on the high confidence stuff using a staking strategy closer to a fixed stakes method. For example, dump all your 1s to 4s and just bet with 5s to 10s.

If most of your odds fall in a narrow range, say between 1.5 and 3, then this staking strategy would seem suitable to me.

Hmm...

Then the robustness only tells me how good I am at picking matches with value (the chance in percent that my correct picks aren´t just a result of the laws of probability), right?

Your conclusions are correct. My high stake bets generally have lower odds (1,7-2,3) and my low stake ones have higher odds.

I checked the yield I would have had if I´d only used my >4-bets- 37%. The <5-bets gave a yield of -26%.

This suggests that I should abandon my <5-bets and focus on the high stake/low odds ones. The problem is that the bulk of my bets (70%) are <5 units. It´s easier to find value with the underdog than with the favorite...

The number of bets are also quite low, only 31 so far. Since my <5-bets often have odds around 3,5-6 it´s really hard to say anything from the results of these bets yet. They have a long turnover, and maybe I´ve hit a losing streak...

The same reasoning can be applied to the >4-bets I guess.

How big a material would you recommend to be able to draw a good conclusion from the data??

Thanks for the help, I really appreciate it!!

[Joe]

Aye, 31 games is not enough to draw any concrete conclusions from. I normally say at least 100 as a working minimum, but really you want a season's worth which should be anything from 150 to 500, depending on how many bets you have each week. The more the better. You can never have enough data with which to test statisical significance.

Quote:

---

This suggests that I should abandon my <5-bets and focus on the high stake/low odds ones. The problem is that the bulk of my bets (70%) are <5 units. It´s easier to find value with the underdog than with the favorite...

---

You say it's easier to find value with the high odds underdog, but the fact is that these bets are losing, so perhaps it's not as easy as you think.

I believe that the the higher the odds (which usually means long priced away wins), the higher the bookies overround. You may think you have found value, but at this end of the market there is a much steeper hurdle to overcome in the first place.

Everything I've ever studied regarding football betting points to the best way being to concentrate on shorter priced home win singles, in the region of 1.5 to 2.25. First off, profits are statistically more significant than those achieved with few wins/high odds because the greater strike rate means that the element of chance is reduced. Secondly, and perhaps more importantly, you are winning more often. OK, you're winning less each time, but you will be less likely to go through a long losing run which can be very psychologically damaging in terms of the faith you place in your system.

Quote:

---

You say it's easier to find value with the high odds underdog, but the fact is that these bets are losing, so perhaps it's not as easy as you think.

---

It seems so...

About the overround, if the bookies increase their overround on long priced aways they should attract more money to the lower odds with the lower overround. This should mean that they are losing money compared to an equal overround in the long run. All of this in a perfect world of course.

So why are they doing it? Is it because people generally likes to bet on the underdog? I can't see any other reason.

[Joe]

Yes, I think that is part of it. People always like to follow an underdog, but perhops more importantly when novice punter sees 6s and 7s offered, they think "Christ, alll I have to do is bet 1 to get 6. How easy can it be?" Large numbers can be deceiving.

Another point worth considering is that a betting strategy concentrating on lower odds will suffer less bank fluctuation in the long run (assuming fairly fixed odds).

Your 'average' person doesn't bet on underdogs, they bet on favorites...

[Joe]

This depends on the sport and the event. Granted, a lot of money is taken on Man U each weekend, but look at the ridiculous levels of money (£100 million was it?) put on the National. Only one bugger wins out of about 40 and most don't even finish anyway. If ever that was underdog betting gone mad, that's it.

Of course more people bet on favourites, that's why they're favorites with lower odds. But the point is that there is still a fair proportion of punters who take long shots because there's the thrill of the bigger win. I'm sure this must explain some of the reason why it's harder to make money from longer prices (i.e. the bookies are proportionally meaner on them because they have to lower the odds to take account of the proportionally higher level of interest). Just a thought.

horse racing and sport are different things

I think in horse racing you may get less return betting on underdogs, you won't in soccer though, for example...

(and I don't know much about horse racing betting/bookmaking anyway)