Why Do My Selections Always Lose?

The most common question I receive is…

Why do my system tests always work paper trading but as soon as I bet them, lose?

I’m sure that you’ve experienced this in the past. It’s happened to all of us, and that’s important to remember.

But, the big question is… why is it happening?

Assuming that you’ve been paper trading a system on real live bets, rather than running it on past results which is a separate topic, then I’m going to tell you why in just two words.

Sods Law.

I realise that’s not very helpful, but it’s the truth.

Don’t worry, I’ll explain it better!

Let’s take Steve as an example.

Steve had developed a system that looked to be making excellent profits. He’d spent a long time working on it, and now the time to test it was at hand. But, to be safe, he decided that he was going to start by paper trading the selections. After all, he didn’t want to go and lose his bankroll on something that could still be backfitted, even though he was pretty confident it wasn’t.

At this point Steve was excited, he’d spent a lot of time developing something that works and he couldn’t wait to prove to himself that he was right.

Once he began paper trading he hit a good run. After just three days he was starting to find it harder and harder to continue paper trading. He’d spent a long time building the system and deep down he knew that it worked. He’d even proved to himself that it was working over the last few days where it made consistent profit. Now he was just losing money by not betting on the selections.

But… he left it for another day just to be sure.

And the next day, he made more profit. That was it. He wasn’t going to wait any longer missing out on these profits. So the very next day Steve began to place real bets and…

…he started losing!

He lost for the next week, and by now he was getting worried. The system had been working so well until he started placing real bets on it, then it immediately began to lose. What was happening! All the weeks of work and the successful paper trading, only to be losing when he started placing bets.

There must be a reason.

So, Steve went back to his original system and looked at the results of the last week to see how he could adjust it to make a profit. And he found it.

The very next day he implemented his changes and… he lost again!

But if he’d stuck to the original rules he would have won. So, he switched back the following day and… lost.

No matter what he did he kept losing.

The reason he kept losing was nothing to do with the system that he built. The only problem that Steve had was… himself.

He made just two critical mistakes. These two mistakes resulted in him not only losing his bankroll, but also throwing away a system that worked!

His first mistake was when he stopped paper trading.

It’s very difficult to stop yourself betting from your live account when you’re paper trading through a winning streak. But, if you’re paper trading a system that has a winning streak, the last thing you want to do is to start betting it live because only one thing comes after a winning streak…

A losing streak!

If you’re already seeing the winning streak then you’ve missed it, by the time you start placing your real bets you’re going to be placing them just in time to catch your losing streak. And because this is a new system, you’ve never experienced a losing streak for it before. Which means that you don’t know how long it’s likely to last in a downswing.

Psychologically you’re not yet ready to cope with that downswing, especially if you’ve started betting at the beginning of it.

When you paper trade you should give yourself a set time of one month, two months or however long you’re comfortable paper trading for and stick to it. It doesn’t matter if there are winning streaks or losing streaks. If you’ve built the system properly then it will be profitable. What you’re actually doing, is training yourself to cope with the upswings and downswings of a new system.

The second mistake that Steve made was changing his strategy. Because he didn’t realise that he had started betting on a downswing, he thought he could solve it by changing his selection approach. An approach he had spent weeks developing and testing was changed in one night based on the results of just a week.

What he was doing was backfitting a new system to the results from the last week, rather than relying on an approach that had been carefully developed.

Because he hadn’t been through the complete paper trading process that would have trained him in coping with the upswings and downswings, he didn’t realise that a losing run of a few days was completely normal. In fact, he could have expected the losing streak to go on for much longer.

We can take two very important lessons from Steve’s experience:

  1. We must choose a paper trading timeframe and stick to it. We are training ourselves to understand how the upswings and downswings work and how to cope with them.
  2. You must never change your selection process based on just a few days data. Changing a process you’ve developed carefully should only be done with a lot of results and a large amount of consideration. It is a very last resort.

Michael Wilding

Michael started the Race Advisor in 2009 to help bettors become long-term profitable. After writing hundreds of articles I started to build software that contained my personal ratings. The Race Advisor has more factors for UK horse racing than any other site, and we pride ourselves on creating tools and strategies that are unique, and allow you to make a long-term profit without the need for tipsters. You can also check out my personal blog or my personal Instagram account.


  1. Your reasoning is spot on michael. I watched Andy bell for 3 months. He was having a fantastic run then at the start of feb hit over 100 points in 1 day. I took advantage of a great offer and in just over 4 weeks has lost over 120 points. Doh.

  2. The basis of your arguement is sound however I bet on my systems from the start but only for nominal amounts. Three weeks ago I put ideas on paper and bet with Betfair for £2 win and place together with £1 e/w on 365. The horse I started with won at Betfair SP of 429/1 and with a place bet and e/w with 365 £8 returned £1405 (2.00 Southwell Direct Approach market SP 50/1 – BOG 66/1 – Betfair 429/1)

      1. Thanks Michael, wish it happened more often. How anyone can lay a horse at that sort of price baffles me but bless them

  3. Winners and losers are a bit like buses, there will be another one along in a minute. I once had my own selection process that found me approx 4 horses a day, and there was usually at least one winner in those four, often two. I then ran into a losing run of 20 – and couldn’t believe it. The next day I went armed with 5 possibles to the Bookmakers, and one of them was 20-1, before I left the bookies that day 4 had won, repaying me all that I had lost, and the last one was 20-1, and I almost walked out and left it, but then I thought well it has all the criteria, why not do it. And yes it won. So that gave me my profit of £300 and taught me the lesson Michael that you are teaching us here.

    I used to do experiments by writing winners and losers on little bits of paper and pulling them out of a box, and you would be surprised at the way they come out, they never come out nicely spaced unfortunately, but it teaches you the laws of probablility. They are set in stone.

    1. Thank you for your comment Wendy. Pulling paper out of a box is an interesting experiment to learn how probabilities work and to demonstrate clusters of winners and losers.

  4. Hi:

    What You Have written Has Been expoused Many Times By Many People Over The Years, Yet It Makes It No Less True: You Are Absolutely Correct About Upswings and DownSwings And Both Have To Be Ridden With The Tide Of Good Or Bad Fortune:

    However, If The System Is Sound In Principle And Logical In It’s Application Then Over The Longer Periiod A Profit Should be Made–And Held: The Watch Word For This Priceless Rule Is—–PATIENCE, And A Bank Roll To Sustain It:

    If All These Attributes Are Brought Together And Adhered To Then The Future Will Do The Rest, And A Very Bright Future It Should Be:

  5. Michael, that was another perfectly painted picture of an ever present concern.
    The take-home point is as ever down to the individual mindset and once that is set in balance, then most other problems fade away.
    Have a great few days awa. It is always great to step outside of the habits to see the obvious and I suspect we might have a whole raft of new inspiration upon your return :D.

  6. One way to deal with losing runs is to use a portfolio of systems and my system builder (HRB) tells me what losing runs I might encounter for each system, However in my experience you can still encounter long losing runs (I’ve just had one of 20). Is there any way of calculating the length of losing runs if you are using a portfolio of systems?

    1. Thank you for the comment Mike. You can calculate an estimated losing run length for a portfolio using the same calculation as you would for a single approach by take the strike rate and number of selections figures from the portfolio as a whole rather than just one approach. You should be aware that the biggest danger is not the losing run but the downswing. You may have a single losing run of twenty but it’s when you get them back to back with just one or two winners in between its important to be confident in your approach and that it will come back.

    2. Maybe I am impatient but I dump a system if it has a losing run of more than 5. I will only continue with one that has a SR of better than 50%. This leads to lower prices but a smaller bank. It evens itself out because with a better strike rate you can use higher stakes so such a system can be as or more profitable than
      low SR high prices

      1. I proimise I’ll shut up after this last one but the way of the universe is not to have all good or all bad. From your own personal finances to those of countries there are never unending periods one or the other but a drift up and down over a period of time.

      2. Thank you for the comment Clive. You can expect to have a losing run of at least 10 in a row with a strike rate of 50%, so getting rid of a system when you have had a losing run of more than 5 means you could be throwing away methods that are working.

  7. Hi Michael,
    Your comments regarding systems is a good one. May I throw a spanner into the works by saying they tell us it takes us about 90 days (3 Months) to re-wire our Brains to learn something NEW. The trouble with doing a System is the human factor by this I mean we doubt it will happen as we are always thinking and changing our minds so any System that any person has will Win and Lose what you need to do is keep emotions out of the equation which is very difficult especially if you have 2-5 horses that could win for you at any given time that is why Book Makers are so profitable as they take the emotional part of it out by what they do. The best thing to do if someone creates a System is to try it out on paper then if you are happy try it out with small stakes and try to keep feelings out of the choices you make very hard to do though. Regards Robert.

    1. Thank you for the comment Robert. You are absolutely right, removing emotion is key. For me the way of achieving this is to use automation in both selection and betting processes which removes me from the events themselves. I am only then involved in the development of strategies and monitoring performances of monthly/quarterly/yearly periods. This distance allows me to remove that emotion.

Back to top button