After the first wave of the coronavirus, most of the leagues have now resumed their games and Soccerwidow started a public experiment to see whether old statistics can still be used and what can be observed after this unexpectedly long break.
Since the 1st of July, we have been running an HDAFU Tables experiment on Soccerwidow, and a parallel Over/Under Goal betting experiment on our German-language sister site Fussballwitwe.de.
Whilst it is too early to say whether the HDAFU Tables will perform to expectations, the Over/Under picks are doing outstandingly well. The original starting bank of 3,000 increased by over 50% in 25 betting days.
Slideshow Picks
The picks for the respective day appeared here around 1 p.m. (sometimes earlier) as well as the results of the previous day.
Please click on the arrows to scroll through the entire history of the picks.
Below are all the picks that were published during the July 2020 Corona experiment (the 2019-20 Winter League seasons finished now). The bank grew from a starting point of 3,000.00 to impressive 4,617.56 during just one month. It was very pleasing to see that the statistics taught in the coursebook in combination with the Cluster Tables did so reliably well despite this Corona outbreak and very long breaks of the leagues.
[slideshow_deploy id=’6755′]
The expected probability and zero odds are calculated exactly as described in the coursebook using the Cluster Tables. The selection criteria is:
- if it has a minimum probability of 60%, and
- if it has a positive value, and if not,
- the bet with the lowest negative value in the 60% – 80% cluster is selected
- only 1 bet per match is selected
The basis for calculating the stakes is the following risk adjustment
- Odds up to 1.1: 5% from the bank
- Odds between 1.1 – 1.16: 4% from the bank
- Odds between 1.16 – 1.39: 3.8% from the bank
- Odds between 1.4 – 2.25: 2.5% from the bank
- Odds between 2.25 – 7.50: 1.5% from the bank
- Odds over 7.50: 0.5% from the bank
Stakes are always rounded to the nearest whole number.
However, not only are the stakes calculated according to the risk but a ratchet system is also applied. This means that the stakes increase with each round in accordance with the highest bank total achieved and remain at the same level even if the bank then decreases again. The stakes are only reduced if the bank erodes to 60% of the starting bank (i.e. starting bank loses 40%).
Starting Bank (at the start of the experiment on July 1, 2020): 3,000
Highest Bank (25th July 2020): 4,729.44
Bank will increase each day if there are winnings; bank for calculating stakes will only reduce when it drops below 2837.66 (60% of starting bank).
Duration of the experiment
We all know that the coronavirus interrupted/paused the leagues for different lengths of time.
The EPL broke on March 9th and, after a 100-day break, started playing again on June 17th.
Italy also stopped on March 9th and started playing after a 103-day break on June 20th.
Poland suspended on March 13th; their break was only 81-days and they started playing again on May 29th. The league concluded on 19 July 2020 and all matches of 31–37 round have been played with “no more than 25 percent of the number of seats allocated to the public”.
Spain suspended on March 10th and started playing again since June 11th after a 93-day break. There were matches played nearly every day for 39 days – concluding on Sunday 19 July.
Each league will make up the lost time differently, however, the last game of this winter season is scheduled to be played on August 2nd. This will end our experiment. In summary, we are expecting from the 1st July until the close a total of 85 matches for cluster table betting.
Important information about the current risk!
Even if we trust our own coursebook and statistics and are actually pretty sure that the published picks will lead to a profit, we are currently playing safe by not risking real money on this experiment.
Just like everyone else at the moment, we can only guess what effect playing in empty stadiums will have on match results. Will home advantage be affected?
How do psychological factors affect results? Like all of us, the players were locked up in their houses for months and subjected to strict curfews.
Did everyone continue to train equally? What effect has this break on the fitness of the players?
There are currently so many questions and unknown factors that could potentially affect game results. Therefore, be careful! Should you follow our picks with real money, then please stake only as much as you can afford to lose and please adhere strictly to the staking plan!!!
Fingers crossed that things go our way! 🙂
Enjoy & share, Your Soccerwidow
Over Under Betting as of 15 July 2020 ~ 11 days Picks: 42 Games
Since the 1st of July, 42 matches have been evaluated and ‘live’ betting recommendations ahead of the games were published.
After 14-days into this trial what can be said is that, at the moment, it is debatable whether one can take past statistics and select bets based purely on mathematical formulas and calculations.
Here are our observations so far:-
People who bought the coursebook know about the recommended use of the Profitability/Yield quotient. Unfortunately, the quotient currently proves to be very volatile and using it for choosing bets may lead to losses.
Selecting by ‘value’ only is also backfiring at the moment. There is a clear trend of more goals than usual in the matches and bookmakers are adjusting their odds to reduce their payout risks. Hence, bets that look on paper like they contain ‘value’ are probably ‘valueless’.
Nevertheless, every cloud has a silver lining and, although the probabilities seem to have shifted a little, it seems that the 60% to 80% probability cluster has an especially higher hit rate than actually expected (i.e. mathematically speaking, using past statistics). If the expected Zero-odds are calculated using the Cluster Tables, it can be clearly observed that bookmakers are reacting to this current change by lowering their prices (betting odds).
Therefore, the current course of action suggested is to consciously search in this probability cluster (60% to 80%) and to include bets in the portfolio within this range that have a low or even negative ‘value’.
As you’ve seen in the above graph, with these conditions in place, the bank grew from 3,000 units to a respectable amount of 4,308 units in just 14-days…
Fingers crossed that these observations and conclusions are correct. We are only halfway through this experiment so time will tell.
Report II as per 24th July 2020 ~ 19 days Picks: 67 Matches
The Coronavirus experiment is coming to its end and it can definitely be said that it is going very well indeed… So far, in just 19 days of betting, the bank has increased from a starting point of 3,000.00 to 4,642.44 units (54.7%).
I have been asked by some in the comments section below why I have been including not only positive values but also negative ones in the published picks.
The reason was that I wanted to give everyone the opportunity to see how and if statistics (and my coursebook) are still applicable both during the pandemic itself and when taking into consideration some pretty long lockdown suspensions/breaks of various leagues.
Below is a graph showing the profit curve applied to the Profitability/Yield quotients:
As you can see on the red curve the point 2.0 is the transition point (Profitability/Yield Quotient: 2.0). The profit up to this point is 1,421.78 (84.4% of a total of 1,684.04), achieved with 36 (of a total of 67) bets (53.7%).
The lesson therefore is… Past statistics are certainly still applicable and so are the teachings in my coursebook. Should you be using the Cluster Tables then it is prudent to choose the bet selections by applying the Profitability/Yield quotient (do not choose any bets below a P/Y quotient of 2.0!).
Nevertheless, for the public, I will continue to publish the picks until the end of this experiment using the same criteria (positive as well as a negative value), but from now on I will also publish the P/Y quotient with the picks.
Final Report as per 2nd August 2020 ~ 25 days Picks: 77 Matches
The bank grew from an initial figure of 3,000.00 units to an impressive total of 4,617.56 using ratcheted stakes (from a starting point of 100 units per bet) during the course of just one month.
It was very pleasing to see that the Cluster Tables performed so reliably well despite the coronavirus outbreak and the consequent very long pauses in our featured leagues.
Read the full reports and its findings here: Over Under Betting Experiment July 2020 ~ Final Report & Further Findings
Did you ever try backing under 0.5? Looks a large difference in zero odds and market odds, but obviously will be very high variance.
Hello again Andrew,
Yes, we’ve looked into the Under 0.5 market and it is always worth checking the 0-0 correct score odds if you fancy a dabble at this. The correct score market is in my opinion a more liquid market than Under 0.5, so you should find higher prices there.
As you say, 0-0 FT results account for just over 7% of all FT results, so you will find an awful lot of losers along the way, but if you can get the zero odds substantially bettered with every bet you make, and can stand the long losing streaks, then there is definitely mileage in riding the 0-0 wagon. Not for me though – you need a big bank and a thick skin if you want to turn a long-term profit here.
Well i’m used to high variance gambling in casinos using their bonuses, and from experience high variance = high long term profits if you keep going, but not sure if it’s going to be the same with this type of thing in terms of higher profits.