Sunday, December 3, 2017

Day In The Life

Following on from my last post I thought it might interest people how the nuts and bolts of my new approach works these days.
Preparation for the next day's racing begins the day before as soon as possible after declarations, work permitting. Firstly I go through every card for "stats" qualifiers mainly using Horseracebase.com. This is a catch all term for various systems I am using based on statistics to systematically pinpoint overpriced horses. It could be a sire on a particular surface, a jockey/trainer combination or a trainer with a high strike rate with debutantes. To use a stat it must have: 1) At least 10 actual winners 2) A strike rate of 20% or very close to it 3) Profit. I don't get very interested unless there's a 50% ROI or close to it. People use stats with lower strike rates successfully but for me it increases the risk that the profit is simply chance and it takes too long to find out. Ideas for systems come from everywhere- sometimes a pundit on TV will quote a stat which sounds promising but usually when I run it through Horseracebase it doesn't stack up. If it satisfies the criteria above it goes in. It's a constant process of adding, pruning and removing systems as time goes on- what worked last year won't neccessarily work this.
 At any time I'll be running 20-30 systems on Horseracebase and they'll spit out anything up to 20 qualifiers on busy days. That's where the work starts, as I don't back them blind. I'll then look at each qualifier individually and ask the question-"has the market priced this horse correctly?" Usually if the horse is 5/1 or bigger the answer is "no" and it's a bet. I leave most of the shorter priced ones alone. I won't back all the bigger priced ones either as there may be a logical reason for eliminating them because of trip or ground concerns, or multiple qualifiers in a race.
If it's a bet the last decision is how to back it. In 90% of cases that's done by placing an SP bet on the exchange with a minimum price stipulation. That usually will be placed at 5/1 or more, depending on the nature of the race and the number of runners with chances. The more automatic this process is the better it works, so I try not to overthink it. I back qualifiers to level stakes regardless of price. If I decide the horse is likely to be well backed I may add it to the list to take an early price on the next morning.The above takes anything from 20 minutes on the average Monday to maybe an hour on a busy racing day.
The next phase of prep is on days where there's Irish racing. I'll go through each Irish race in the traditional "form" sense and make an attempt to price them if I'm interested enough in each particular contest.This analysis is different- I have notes on hundreds of Irish horses based on observing their races. I have a fair idea of their individual quirks and requirements which is my edge over the crowd.  I'll start to get an idea of horses I might be interested in the following day when prices are available to me and will write down possible bets and prices to beat. I may also get ideas about in-running bets like back to lay in running or lay to back. I automate this because with the ATR delay, live in-running betting is suicide. I might back a horse at 10 SP then put in progressive lays at 5, 4, 3  and 2 or even lower if I think it might trade short and get beat. Nora Batt was my highlight for this approach this summer!
In the evening I'll see how they've priced up the overnights- I can't have a bet at that stage but I'll have  a fair idea if my target price is likely to be available the next morning.
On raceday I get the shop bets done for me soon after opening time. I like to do multiples on days when I'm backing a fair few juicy priced horses as the Lucky 15 concession (3 times the odds on a single winner) is actually quite valuable. Most of punting feels like chiselling away for small returns at times so there's always the chance of a year transforming lift if 2 or 3 click. Multiples have a bad name but they're actually a great way to get around stake limits- a trick I learned from a really good UK punter I correspond with is that if I want to back a horse ante post, I'll keep sticking it in doubles or even trebles with my daily bets in the hope that one clicks- that way I can build a really nice ante post position relatively cheaply and under the radar.
Before racing I'll have a look at the exchange and see how the markets are shaping both in Ireland and in the races with stats qualifiers. If I have a solid stats bet that's trading much bigger than my price to beat (50%+) I'll add another half stake with that price as a minimum.
If I have time to watch racing live I'll keep an eye on the live market to see if anything I gave a chance to is drifting to a price I didn't expect and I'll stick in a bet at SP with a minimum stipulation.
For me, taking live prices doesn't work- I've found my sort of horse tends to drift and return better at SP. The danger of watching markets live is that you chase prices down and accept lower than you should- I have been that soldier. It's hard to see a horse start at 6.8 and win when you wanted 7, but over time it's good business. Sometimes if there's been a non-runner or a ground change I can make a logical argument for accepting  a shorter price but mostly you're better off letting the shorteners win and backing the drifters. I have very few bets at under 4/1 these days and I've practically given up laying horses above evens- I just can't stomach losing 4 or 5 times my stake in one hit and wiping out a good day. (The real truth is that I'm terrible at it and the stats don't lie)
The decision whether to take an early price or leave it to SP on the exchange is often not easy and is something I get wrong almost every day. I try not to beat myself up about this nowadays and accept that markets are chaotic and difficult to predict- I find myself leaving more and more to SP as the upside to your expectations can be huge. Knowledge of the habits of the trainers or owners involved can help predict if the cash might arrive.
It should be obvious that my turnover is high- it's routine for me to turn over 20-30 times a normal stake in a busy day. That's not for everyone but it suits me, as it smooths returns. You need to set your stakes at a level where a blank day doesn't get to you too much.
Every bet I make goes in a spreadsheet and with the help of someone more tech savvy than myself I hope to be able to do more analysis of this going forward.
My aim on a day with an Irish card is to be able to look at the results and say that I didn't miss a findable, significantly overpriced winner. If favs go through the card, I'm happy to take my medicine.
If there were one piece of advice I could impart to do this well, it's to stop trying to back winners. That may sound perverse, but there's a powerful herd impulse to back the winner regardless of price and shout about it to your mates. It's harder to accept a lower strike rate and accept that you will have losing days, weeks and months even if you're doing everything right. You simply need to work hard to try and ensure that you make good bets, that is bets with positive expected value (look it up) and allow time to prove you right.



Saturday, October 21, 2017

Beating The Machine

Most punters who make a genuine effort to win at the game go through various phases: mug punting, managing to limit losses for a while then eventually getting to the point where winning can reasonably be expected at the start of most years. At the latter stage the difficulties begin- accounts which show profit over time are restricted or closed and you become forced back to the shops. Phone calls to the traders, limited stakes and knockbacks- it grinds you down over time and takes the fun out of it.
At that point you can either spend your energy moaning about your lot on Twitter or try and find another way.
I freely admit that over my punting life, I have struggled to win on Betfair- give me access to 3 or 4 firms' early prices and betting value horses is relatively easy, beating the "machine" is a different matter. Something changed for me in 2017 though and I thought sharing what I think are the reasons might be helpful to some in a similar position.
My shop woes came to a head this year and I made a conscious decision to migrate more to Betfair and have a proper crack at it. The approach is totally different to the early price one however. Spotting overpriced horses the evening before and trying to get on before the value goes doesn't work- by the time there is enough liquidity the price you thought was value the night before may not be there any more.
The first change I implemented was a change of mindset. You need to think of Betfair as a Tote- you are playing against the crowd and it should be just as possible to beat them there as it is to win in other ways.
Secondly the method changes. I now look at the races I'm interested in for the next day and come up with minimum prices for anything from one horse to half a dozen. Then it's a case of either leaving SP bets with this minimum proviso or watching the market live and reacting to the value that's thrown up. This means that I actually have no idea what I'll end up backing until near the off and this takes a little more mental flexibility. The upside is that often horses you didn't really anticipate as bets stick out a mile as value when you've done the requisite prep.
My personal method involves either a full or partial tissue betting to 100% then adding a 50% margin in odds to build in a profit- for example a 2/1 chance would need to be 3/1+ for a bet, and 8/1 would have to be 12/1+ etc. It's important to look at reserves in Ireland as they are often not assessed as accurately. The shape of the race both in how it's run and how they might bet can change a lot if a reserve or reserves get a run.
It's also important to anticipate how the markets and the tissue might change if conditions change before or during the meeting- softening ground, a draw bias on the Flat and so on. Punters without fixed fancies are at a major advantage if they can change tack as races are run.
From taking or leaving Dundalk, for instance, I now wouldn't want to miss a meeting as there are horses with chances trading at insane prices almost every week. Staking is fairly level, but I'm inclined to have another half to full stake on horses which are double my value price or more.
All I can say is that I'm now 80% Betfair and it's paying. I hope this will remain the case- maybe I'll do a progress report at year's end.