The Apprentice – Betting System Task

Guest post by David Renham

The fading sound of Big Ben in the background meant it was just after 6am. The phone rang and a half-asleep Frankie rushed down the stairs.

‘Hello this is Lord Sugar’s office. Lord Sugar wants you to meet him at Epsom racecourse. The cars will pick you up in 30 minutes.’

Frankie put the phone down and while sprinting back up the stairs barked out,

’30 minutes everyone. We’re off to a racecourse!’

There was a mass of movement – frantic ironing, quick showers, hairdryers bursting into life, shoes buffed, etc, etc. At 6.30 calm had descended and all 6 candidates made their way to the front of the building where the cars had just pulled up. 40 minutes later, they were standing in the racecourse paddock facing Lord Sugar’s trusted aids Karen and Nick. That famous number plate ‘AMS 1’ came into view and parked up. Lord Sugar got out of the car and made his way over to where everyone was standing.

‘Morning.’

‘Good morning Lord Sugar.’

‘You may well be wondering why I’ve bought you to one of the most prestigious racecourses in the country. Well the gambling industry is now massive business across the world. In 2009 the legal gambling market totalled an estimated £210 billion globally. That is a serious amount of money. Horse racing has always had a strong association with gambling and making money from horse racing is the object of today’s task…..

Your task today is to develop a horse racing system for turf flat racing that has made the biggest profit over a 3 year period from 2008 to 2010. You are allowed to calculate your profits using Befair Starting Price rather than the traditional bookmaker’s SP. However, there is catch. The software that I will be giving you to help you with the task can only test data from 2008 and 2009 – you will have to be confident in your system that it will not have ‘bombed out’ in 2010 as the 2010 profits and losses will be added to the 2008-09 ones. The software will take 5% commission into account when calculating profits and losses, as commission is paid on all winning bets on Betfair.

OK, we are going to mix the teams up and will start as we started the process with girls versus boys. So Kirsty you come across and join Haley and Amy; Frankie you come across and join Jamie and Kieran. Nick will be following the girls; Karen will be keeping an eye on the boys. So there it is. The team making the most theoretical profit will win the task. You have until 6pm and at 7pm I will see you all back in the boardroom, when one of the members of losing team will get fired.’

Each team went into a separate room to be greeted by three high speed computers with horse racing software loaded. Numerous newspapers, books and periodicals were also on hand to help the teams. The boys went about the task of choosing a team leader. Kieran got in first,

‘Well we need to pick a team leader and as I used to work in a betting shop I feel I have the best credentials of the three. What do you reckon?’

Jamie nodded and Frankie agreed saying,

‘Yup, you are the obvious choice and clearly we need all the time possible to produce the biggest profits, so let’s not waste anymore time.’

Over in the girls room they had been equally quick choosing Haley as the team leader. She had worked on the marketing side at Newmarket racecourse for 3 years in her early twenties and as with the boys it was a fairly straight-forward choice. Haley looked at her two colleagues and said,

‘OK. What knowledge do you two have when it comes to gambling and making money?’

‘Well, I have been racing a few times,’ commented Kirsty. ‘My ex was heavily into horses and gambling in general so I feel I have a pretty good idea. His uncle used to use this super duper system in the 70s, but it did not seem to work when we tried it!’

‘I’m afraid’ said Amy, ‘I have little idea about it if I’m honest, but maybe my computer skills will coming in handy in terms of churning out and analysing the results data. I think we make a pretty good team here.’

The others nodded before team leader Haley spoke once again,

‘OK. My understanding of the Betfair market is that there are very good prices to be gained when backing outsiders. For example I know that some winners paid as high as 200 or 300/1. I think we need to go for an outsider system. Remember we are simply concerned with making a profit here, and the biggest profit at that. It is not about number of winners, although the more we get the better of course! Hence we need to look at bigger prices as there must some scope for some decent profits, if we can find the right system.’

The girls spent the next half an hour familiarising themselves with the software, the variables within the software and various betting terms. After this initial fact-finding mission they had another quick discussion. There were still 9 hours of the task to go and it was agreed that they should look at bigger priced runners to begin with. If no real profitable avenues were achieved in the first 3 hours, they decided they must re-convene and plan an alternative strategy.

Back to the boys and they were still discussing ideas. They had gone around in circles, but eventually team leader Kieran put his foot down.

‘Clearly we cannot agree on a decent strategy here as a team so I will take it upon myself to make the decision on what path we will take. I think we need to look for a decent volume of bets, and I think we need to find a method that has a high strike rate. The higher the strike rate, the more likely it will repeat the performance of the previous two seasons.’

‘I not sure that …’

‘Look Frankie, I am team leader and let’s just follow my lead right?’

Frankie looked at Jamie and rolled his eyes. Kieran continued,

‘So high strike rate; probably looking at favourites then. Let’s start with that and see what we find.’

The boys plugged in the data – favourites in turf flat races from 2008 to 2009. That gave then 2390 winners from 7532 bets but that made a loss of £132.70. However, a loss of that amount equates to an ROI (return on investment) of -1.7% so not a bad starting point. The best news being that there were over 7500 bets to work with.

‘I see that horses that are running again with a week have made around 50 quid profit’ mused Jamie.

‘Yeah, but the number of bets are already down to under 1500 and I don’t think 50 odd quid is going to be enough?’ replied Frankie. Team leader Kieran agreed and starting looking at the numerous variables he had at his disposal. He decided that he would combine horses that finished in the first 4 LTO, focusing only on the months from June to September. He did this as these two filters forced the system into profit; albeit by £17.59.

‘This is more like it!’ cried Kieran.

‘Hmm not convinced myself,’ replied Jamie. ‘This smells of back-fitting to me.’

‘Back-fitting phaaarrr!’ Kieran squeaked. ‘These are two sensible filters using the time of the year when the going is likely to be less variable and hence favourites are likely to perform better…… AND it is focusing on in-form horses. Can’t see your problem!’

‘It’s all very well dressing it up like that, but despite both filters having some logic to them, you have only added them because they sneak us into profit. Remember this has got to do OK in 2010 as well.’

Kieran ignored the lecture and five minutes later the boys were back number crunching to see if they could improve upon their small profit. They still had only 3800 bets in the system so plenty of scope for further filters to be added. After 45 minutes, Kieran piped up again,

‘Hey, look at the price stats….. if we focus on SPs of 2/1 to 9/2 we increase profits to £113.57’.

The two others tapped they keys on their computers and lo and behold the same figures were plastered across the screen.

‘Great’ said team leader Kieran. ‘I really feel we are getting somewhere’.

The other two looked at each other exchanged weird facial expressions which to you and me meant ‘no we aren’t really getting somewhere, but if we lose surely he is for the chop not us.’

More filters were looked at – restricting the races to class 4 or lower improved profits to £200.25, although the bets had dropped to only 1072. It seemed that all that mattered were profits with no thought of whether this system was in fact any good, or what was likely to happen to it when tested on the 2010 data.

Time was running out for the boys, when Kieran perked up. He chirped in a very excited manner,

‘guys, the price filter can be extended from 2/1 down to 7/4, and up to 11/2 from 4/1. This gives us more bets and more profits. £214.14 to be precise.’

Jamie put his head in his hands and muttered,

‘You can’t change the price filter again just because it gives us an extra £14. This is definitely back-fitting of the worst order. How do you know that this will improve the 2010 figures?’

Kieran ignored his question and just continued on his relentless march towards more profit. He added one last filter which restricted the system to male jockeys only – funny that considering it was a boys against girls task! The final figures for the boys saw

378 winners from 1236 bets for a profit of £221.70.

So what had the girls been doing? Were they able to conjure up a system with profits in excess of £221.70? If you remember their approach was going to be totally different – they were looking at bigger priced horses. They looked at different price brackets but the most sensible price filter seemed to be 20/1 or bigger. This was 20/1 industry price (traditional SP) not Betfair SP. This gave them plenty of selections to work with despite being the best part of £850 down when focusing on all such runners. However, with around 23,000 selections from which to work on, there seemed a decent chance that this could be swung into profit.

Haley decided it was time to discuss their next step.

‘My feeling about all horse racing systems is that they need to be fairly simple. Too many rules will lead almost certainly lead to back-fitting. Also I think we need to think of sensible filters but not test them one at a time because again this will essentially be back-fitting as I’m sure we will ignore the poor filters and simply add the good ones.’

The other two nodded in agreement.

‘Ok. What other filters do you think we should be looking at? Let’s choose one each to begin with. Hopefully those will combine to make a decent profit and we can leave it at that. As I said, the simpler the better.’

‘I reckon only handicap races,’ said Kirsty. ‘These are more competitive and bigger priced runners probably win more often.’

The other two girls agreed. Handicaps are designed to be more competitive so this all seemed a logical first step.

‘I think we should focus on the less competitive races,’ said Amy. ‘By this I mean steering clear of the top tracks and focusing on the poorer courses. I would guess these handicaps at these venues are less competitive giving outsiders a bit more of a chance?’

‘Good plan,’ said team-leader Haley. ‘We have 4 grades of track – go for the bottom 2?

‘Sure. Happy with that.’

‘OK. That gives us which courses?’

Amy looked at her database.

‘Bath, Beverley, Brighton, Carlisle, Catterick, Chepstow, Ffos Las, Folkestone, Hamilton, Leicester, Lingfield (turf), Musselburgh, Newcastle, Nottingham, Redcar, Ripon, Salisbury, Southwell (turf), Thirsk, Warwick, Windsor and Yarmouth.’

‘Great. So that leaves me to give my ten-penneth,’ said Haley. ‘I am going to focus on an area where I feel punters over-bet. I think multiple distance winners are over-bet and there is value ignoring them. Hence, my filter will be simple – only select runners that have either never won over the distance, or won just once or twice. Let’s plug these three filters into the software. Fingers crossed!’

So this simple looking system of horses priced 20/1 (SP) or bigger; in handicap races; at grade 3 or 4 courses; with no more than 2 wins at the distance gave the following results over 2008-09:

151 wins from 5979 races for a profit of £975.73.

The girls smiled at each other. Haley noted,

‘Even if 2010 is a shocker we still should have done ok overall. Of course we only need a couple of real big priced winners in 2010 and this system will make a profit again. I am happy with this – it is simple and targeting value’.

The other two girls smiled and winked at each other. They had done all they could…………………………………………..

An hour later, the two teams were sitting in the reception area outsider Alan Sugar’s office. The phone rang.

‘You can go in now,’ the secretary said. The six rose to their feet and entered the boardroom. Nick and Karen sat expressionless, as they all waited for the man who did the hiring and firing. Two minutes later, Lord Sugar appeared.

‘Evening’.

‘Good Evening Lord Sugar’.

‘So let’s start with the girls. Who was the team leader?’

‘I was’ replied Haley.

‘OK. Makes sense from your background. Good team leader?’

The other two girls nodded in approval.

‘So what was your approach?’

‘Having looked at, and read about Betfair prices, it seemed logical to focus on the bigger prices selections. We felt that was where we could gain an edge.’

‘A logical strategy. What that your only strategy?’

Amy responded,

‘No, we did however want to keep the rules simple as generally the simpler the system the better. According to the literature anyway! We read some Smartsig magazines and they had several KISS systems.’

‘KISS?’

‘Keep It Simple Stupid!’

‘Right!’ Lords Sugar said. ‘I assume you are not calling me stupid!’

‘No, no! That is what the letters stood for!’

Lord Sugar turned his attention to the boys.

‘OK to the boys. So who was your team leader?’

‘I was Lord Sugar’ replied Kieran.

‘Good team leader?’

There was a stony silence.

‘Right’ Lord Sugar said. ‘That answers my question for me! OK so what was the approach?’

Kieran started confidently,

‘I decided that it was necessary to focus on winners and strike rate. The more winners the better. We found that backing all favourites only made a small loss at Betfair SP so it was felt we should continue upon that path.’

‘Hmmm!’ Lord Sugar pondered. ‘That is not quite how Karen relayed it to me. It was more like the Kieran show and bugger anyone else on the team’.

‘Hang on a …’

Lord Sugar stopped Kieran dead in his tracks. ‘Look Kieran. It is clear that you should have been team leader looking at your past experiences, but it seems to be that none of the other two were allowed a look in.’

‘That’s not ….’

‘I don’t want to hear it…….. Frankie, tell me what happened next.’

‘Well Kieran was obsessed with just getting the results data and effectively just trying to manipulate it to create bigger profits. We tried to tell him that he was simply back-fitting the data but he would have none of it. He just ploughed only claiming that any filters he used were logical ones. We tried to tell him that back-fitting might look all well and good superficially, but the chances of similar profits being replicated in other years is slim to say the least.’

‘OK’ Lord Sugar muttered. Time to get the figures. Karen. The boys first.’

Karen turned over her sheet of paper and began,

‘The system in 2008 and 2009 made a profit of £221.70. In 2010 it did not perform as well and although it still made a profit – it was a paltry £15.33. Giving the boys a grand overall profit of £237.03.’

Kieran smiled and looked at the girls. He thought he had it in the bag. But the girls were trying hard not to contain their excitement. They would need a disastrous 2010 to turn this one against them.

‘Nick. Can you give me the girls’ totals please.’

‘I certainly can! In 2008 and 2009 the profit stood at a healthy £975.73.’

Kieran suddenly looked as if he had been hit by Mike Tyson in the solar plexus. Nick continued,

‘In 2010 the profit was £397.43 giving a grand total of £1373.16.’

Lord Sugar turned towards the girls.

‘A very good effort ladies. Not much of a contest eh? You clearly understood the need to keep the system simple and showed that there is value going against the crowd. How many punters would have been happy with a system that produced winners just 2.4% of the time? Very few! But you have shown that value can be found in certain bigger priced selections. Not only that, you managed to nail down an excellent profit – both in the two years you tested and the year you did not. Well done, your treat for winning is perhaps not surprisingly, a day at Royal Ascot. I have hired out a box so you will hopefully have a wonderful time.’

The girls skipped out the room hugging each other; they were extremely happy with their days work.

Lord Sugar turned to the boys and scowled,

‘Right, could you step outside for a minute as I need to speak to Nick and Karen……. and upon your return, one of you will get fired!’

The boys trooped out into the waiting room heads bowed. There was complete silence, but a sense of knowing what was going to come next. Five long minutes passed until the phone rang.

‘Lord Sugar will see you now.’

The boys filed back into the boardroom to face the music. Lord Sugar looked up and focused his attention on Kieran.

‘So Kieran, everything seems to point to the fact that you as the team leader were responsible for the failure of this task. What do you have to say to that?’

Kieran without really thinking out his response babbled,

‘How can you say that it was my fault we failed to win this task? I was the one who came up with all the ideas. I was the one who produced a profit and a decent profit at that. If Frankie and Jamie felt that I rail-roaded them, then that simply shows how weak they are and how devoid of ideas they are. I passionately believe that the methods I used were the right ones but unfortunately the girls simply found something better.’

Lord Sugar choked,

‘Better? £1136.13 better to be precise! OK I hear what you are saying about weak support from your team; Jamie what do you say to that?’

Jamie moved uncomfortably in his seat and responded,

‘Lord Sugar; he simply did not give us a chance. He said “I am team leader and let’s just follow my lead right?”

Kieran started to protest before Karen cut him dead.

‘Kieran; I was there,’ she said. ‘You did say that. You did not listen. They tried to tell you several times that your methods were not the way to go when producing a successful horse racing system.’

‘Thank you Karen’ Lord Sugar interjected. ‘Frankie what have you got to say on the debacle? Why should I not fire you?’

Frankie looked at Kieran, then back at Lord Sugar.

‘I am not responsible for the failure of this task Lord Sugar ………. and if you look back at the whole process I have consistently been one of the stronger candidates. I have been on the losing team just three times and this is my first time in the bottom three.’

‘OK! I think I have heard enough!’ interrupted Lord Sugar.

‘In my opinion, there are two main reasons why this task failed. The most important reason is that you committed the cardinal sin of all system cardinal sins, which is that you back-fitted the system. Not just back-fitted it once, but back-fitted it several times. For God sake you introduced a price filter and then a few hours later you adapted this price filter simply to improve your figures further. That is not the way to do it. And then the crowning “glory” – adding the male jockey filter at the end just to increase your profit by a few quid.

When developing a racing system, you cannot simply manipulate the data to give you better results. That simply does not work long term. These systems have to produce a profit in the future; the past results are simply that; reporting back on the past. Also, Karen told me that you babbled on about how important strike rate is. I am not saying strike rate is not important, but let me give you an example. If you have system that produces 40% of winners but their prices are all odds-on, the strike rate becomes totally irrelevant because you simply lose money.

The other reason for the failure of this task was that one person in particular did not listen and simply thought he could do everything on his own. This is probably the easiest decision I have ever had to make.’

Lord Sugar turned to Kieran and the candidate saw that dreadful arm being raised and the finger pointing in his direction.

‘Kieran YOU’RE FIRED!’