Dynamically changing coefficients of sports events
Line change betting odds
Computer program "KoefficientyRU". Perhaps this is exactly what you have always been looking for
Description of the program KoefficientyRU (Analytics)
And so, the main question:"DOES THERE BE A MEANING IN RATES ON MOVING COEFFICIENTS?" Or how to test your strategy of rates on historical data (back-testing).
In the presence there is, I'm not afraid of this word, BIG statistics on the meetings for 2016, developed by the program KoefficientyRU for seven main sports, well, a little for 2015.
||The number recorded by the program KoefficientyRU MARATHONIC bets
||Number of lost, NOT compared meetings*
||Number of mapped meetings*
| 1||2016||January ||11 235|| 786||10 449|
| 2||2016||February ||12 244||1 076||11 168|
| 3||2016||March ||12 913||1 214||11 699|
| 4||2016||April ||12 992|| 932||12 060|
| 5||2016||May || 9 472||1 145|| 8 327|
| 6||2016||June || 6 991||1 246|| 5 745|
| 7||2016||July || 9 737||1 391|| 8 346|
| 8||2016||August ||11 952||1 369||10 583|
| 9||2016||September|| 8 976|| 356|| 8 620|
|10||2016||October ||15 351||1 448||13 903|
|11||2016||November ||12 126|| 960||11 166|
|12||2016||December ||11 335|| 403||10 932|
- * The number of mapped and NOT compared meetings requires some clarification. The MARATHON company before the meeting on its page writes: "Tournament", "Team Name", "Meeting Time", and, of course, "Odds", and after the match ends on another page, writes "Tournament", "Team Name" , "Time of Meeting", and, of course, "Account". So, the "Tournament", "Team Name" and "Meeting Time" on these pages sometimes differ from each other by ochyatki, permutation of words, changing the start time, and simply banal cancellation of matches. So you need to compare them. Approximately 5-10 percent of the program can not be compared, and those lost in the lost meetings are those whose start time has changed by more than an hour. That's because of this and there are these losses. Therefore, we work with the rest ~ 90-95%, and here, I must say, it suffers, mainly, TENNIS. He often hosts meetings, and the names of athletes are distorted and cut very much. To get answers to your questions, 90% should be enough.
For example, we dig in last year's data. There is an archive database of December 2015, it recorded 8,656 meetings. Of these, Soccer - 2 983, Basketball - 2 090, Hockey - 1 250, Handball - 507, Tennis - 568, Volleyball - 1 123 and Baseball - 135. Not the most eventful month (here it is necessary to take into account that there were skips when scanning data ), But there is something in it.
We disassemble only such rates: Ï1 / Ï2. The initial capital is $ 1,000. The size of all bets is the same and is $ 10. First, it does not affect the final result in any way, if we take all the meetings in any sequence (since several meetings start at the same time and there are dilemmas for the variable rates), and secondly, the rate is not the most An important parameter in the game with the bookmaker. The main task of the size of the bet is to sit out, without special moral losses, a bad luck zone that always comes to everyone. Each determines its own allowable amount of losses independently, and choose it according to Kelly, on the percentage of the deposit, at a fixed value - it's all the same. If you want to earn a lot, then be ready to lose a lot, you want to lose a little, then you will not earn much - that, in my opinion, is the only criterion when choosing the rate.
Let's review all possible options for December 2015, to find the most optimal betting strategy.
- There are two betting strategies - the Lower Coefficient (L) And the Greater Coefficient (G).
- There are three variants of the movement of the coefficients: 1st option Smaller Coefficient Drops; 2nd option Smaller Coefficient Grows; 3rd option The Big Coefficient fell so much that it became the Smallest Coefficient, i.e. Coefficients Overwhelmed.
- It is possible to optimize the input parameters by filtering the rates by their value. Input parameters: Limits of the Small Coefficient Ê1, Bounds of the Greater Coefficient Ê2, Limits of the Degree of Modification of the Smaller Coefficient Change.%, Boundaries The number of steps taken Step Down or Step Up , Borders per Percent Bookmaker% bookmaker, Number of measurements Col. Zam. (The maximum number of measurements is 576 - this is for 4 days, I remind you that the scan is every 10 minutes), and, of course - Sports.
Data for optimizing the selection of input parameters abound. Without regard to the division according to the types of sport, there are only 48 Approaches.
Summary table (Approach number)
|Filter \ Move
|Change. % (L)
|Step Down (L)
|Step Up (G)
|% bookmaker (G)
"Approach-1" - we take absolutely all games and bet on a smaller coefficient.
"Approach-2" - we take absolutely all the games and bet on a larger ratio.
The results of the first two approaches are not impressive, one is worse than the other. Here it is possible to reflect a little - as so, I put everything on a smaller coefficient and lose 4 deposits, but I do the opposite, put on a larger coefficient and lose two more times more? Well, firstly, some sports have such an outcome as ANYTHING, and secondly, the percentage of the bookmaker, too, will not allow the reverse action to go into the plus.
The remaining approaches in themselves, also almost (see below) did not give a positive result, but here in a combination of different approaches a positive picture is nevertheless obtained. And here there are always supporters and opponents of such approaches to history. Therefore, I suggest that everyone independently test their strategies on historical events.
If you sit with a pencil in your hand and a sheet of paper in search of your strategy for the current sporting events, then this ANALYTICA module will help you find your strategy and understand the depth of the sports betting market with the help of a very large database of historical events that have already happened. On what you planned to spend months, now you will need to spend only a couple of days and at the same time you will be able to cover a huge period of time. This will definitely save you a lot of money and time, because it will immediately show you the absurdity of your first approach to rates and, possibly, teach others more literate approaches.
ANALYSIS is launched directly from the program "KoefficientyRU" of the registered commercial version.
Immediately proposed months, starting in January 2016 and ending with the last last month. Here you need to click the mouse three times: Year -> Month -> OK.
The main window with a complete list of all the sports events of the selected month will appear. It is possible to sort this list by clicking on the table headings, filter, right mouse button -> "Filter", and also analyze your strategies - right mouse button -> "Strategy Analysis".
With the filter of the general list table everything is clear and there is no need to explain, but the "Analysis of strategies" needs to be explained. In the "Analysis of strategies on historical events" window that appears, you need to specify the selection criteria by filling in fields that are easy to understand and then clicking on the "Analysis" button. If you need to return to the initial state and make all filters passable, then this can be done with a single "Reset Filters" button. There is another possibility for analyzing and optimizing the autopilot's work schedule, for those who have studied the chapter "Àâòîïèëîò", There must be everything clear.
After clicking the "Analysis" button, a window will appear with a detailed list of all games received as a result of the analysis and sorted by time. The top part of the window reflects all the conditions of the Analysis chosen by you. Taking into account this sequence of events, I supply each line with the current values: "Deposit", "Guessed", "Drawdown". The remaining positions in the lines are completely understandable, except that H1_0, H_0 and P2_0 are the very first measurement coefficients, and P1, H and P2 are the last values of the coefficients.
And also, immediately in the upper left corner of the screen, another window appears with a clear graph of the change in the deposit.
Hand-picking all possible options for input parameters is long and tedious, and not using computer capabilities would be silly. Therefore, there is in the module ANALYTICS such a convenient function as optimization. It allows you to optimize each numerical input parameter individually. To do this, use the small buttons located on the right, select the desired parameter and click the "Optimize" button. We make all filters pass-through and optimize only the first input parameter K1, according to our summary table this is the Approach #9 (see above).
Optimization in our understanding is the calculation of absolutely all possible values of the input parameter (in steps of 0.1) and the choice based on the results of calculating the best value of the Deposit, or even better if the large value of the Deposit is supplemented by a small value of Drawdown. And so, before us the table with the received results of calculation on the input parameters K1Min and K1Max, sorted by "Deposit". What do we see? ... BINGO !!! There are options when you get a profit.
Having studied the optimization only on the first input parameter, we make the following conclusion: if in January 2016 we made absolutely all bets on sports events (7 sports), where the smaller coefficient would be in the range from 2.2 to 2.3 (inclusive) Profit $ 314 with the surviving Drawdown of $ 164, making 477 bets. For a deeper understanding of the result obtained, let us use two more indicative characteristics of the analysis of strategies. "Expected Profit" (OP) - this is what the profit will be for each bet, in theory, if we have all the bets guessing, then OP = 10, if all bets are not guessed, then OP = -10, between these two values ??will be found Our real OP and it is quite clear that the closer its value will be to +10, the better will be our chosen system. The second indicative characteristic: "Recovery Coefficient" (KB) - this parameter characterizes the stability of our system, which tells how quickly after the drawdown
obtained Our system will be able to recover. And so, our OP would be $ 0.66 (Profit / Bids) and KV = 1.91 (Profit / Drawdown), a good system is considered to have a KV of more than 3. We get not quite brilliant indicators, but still, it would be profit. Now substitute our optimized values ??(2.2 and 2.3) in the filter (Lower coefficient) and analyze the situation. Yes, here's another thing, to get away from Optimization and return to Analysis, you need to turn Optimization off with the small button [x] located just to the right of the "Reset Filters" button in the value pop-up window. As a result, we get this graph.
Looking at such a graph of profitability, especially on the unevenness of its growth, it is better to beware of such raw strategies and continue its further improvement. Also, I want to draw attention to the most common mistake of emotional players on the example of this chart. The second quarter of this chart shows very good results and here many can "demolish the roof", and they will begin to increase their rates. IN NO CASE!!! You can not do this, I advise everyone to remain calm - this market forces you to spread your blood. Already by the middle of the chart, at best, you will lose all your profits, and in the worst case, the market will shake you all out to a penny. Do not see this trap - adhere to strict rules, "but if you do not care," then increase the maximum from $ 10 to $ 13 - this is 1% of the Deposit and never exceed this value.
When, having exhausted all possible values of all input parameters, finally choose a comfortable strategy for yourself, you know - you are ready to make a decision to trade not with emotions, but with analytics. Another method is called the Mechanical Trading System (MTS). In this case, you are not interested in any insider information, because Everything is already reflected in the graphs, and the schedule takes into account everything. It remains only to catch their games and make the right bets. The first half of this task is easily taken over by the KoefficientyRU program, see section "AutoPilot".