Feeling better

P&L Since last update

Horse Racing £22.35
Football -£0.90
Casino Bonus £71.50


Horse Racing
Thursday 18th -£4.61
Friday 19th £11.68 (includes a loss of £5 on a lay bet that didn't come off)
Saturday 20th £15.28

Trading has been at reduced stakes - max loss 5 ticks/£1.50 instead of 5ticks/£3.00

So back to break even for the month even though I needed a casino bonus to do this - hey it all counts. Also been happy with discipline and resolve to let bot trade uninterrupted. I am happier with reduced stakes as it appears that psychologically I can handle £10 daily losses but suffer when daily losses hit £25

I have another filter I hope to test in the next couple of days. Throughout yesterday I have been building a betfair tracker which can monitor markets long term. My plan is to collect horse racing data for two weeks and see if some ideas I have work in the longer term markets.

I have two laying systems which are easy to implement but find only 2 or 3 bets a week and I am still trying to decide if it is worth it. Also why I have dropped football betting as I feel there are easier ways to make money. My reasons
Firstly you have to be glued to the pc/tele to react if goals are scored
Secondly I was trading the correct score markets and I just wasn't happy with the gaps between best back and best lay prices. I have been spoilt by the efficiency and liquidity of the UK horse racing win markets. I actually tried the Irish and place markets and gave up for the same reasons.

Record loss

Daily P&L -£90.71

Horse Racing -£14.24
Football £1.83
Bot Error -£78.30

Mind is not clear enough to write now so will update tomorrow. Once again after changes to bot program suffered big loss

So far I have lost -£70, +£180, -£500, -£70, -£40 and now -£78 to Bot errors but this is the worst so far because

1) I was watching
2) It wasn't really a programming error but I was running two bots from 2 pcs using betting assistant and also manually interfering with trades and the whole thing got confused. About 3 minutes to the end of a race I manually closed all position (I thought) and went of to do some DIY before the next race. 10 minutes later I saw my betfair balance was 163 instead of about 240. I was so shocked I initially thought someone had hacked into my account

Throughout the day I was constantly messing about manually exiting trades. At 15 loss I implemented stop loss, promptly won the next two and started trading full stakes and lost the next two

If I had let my bot run I would have been down £5 for the day

I am nearly £1k down to Bot errors and wonder if/when I will overcome this. I guess the very expensive lesson is weekly reviews not daily. Also I just have to have faith in the system.

Making decisions during the trading session has to stop

No more tweaks to the bot until March. I will just let it run at 50% stakes no stop loss.

Funny thing all my versions of my bot if I had let them run undisturbed would see me up for the month. As it is I have given away all my profits and really wasted my time

I have been home all week and all I have done is trade and program/backtest the bot

Tomorrow will be a complete break from bot analysing and coding. In the evening I will trade the Fulham and Roma games if I have nothing better to do

Still believe I can end the month up but now will be depending on matched betting

Slow progress

PL since last update £0.32

Yesterday £4.95
Today Horse racing -£1.83
Football -£2.75

Got back briefly into profit for the month yesterday but lost a small amount on today's poor quality racing after Folkestone was abandoned. I haven't done well historically at Southwell so traded carefully and actually made money. Unfortunately lost every trade at Newcastle to be down for the day overall. This is actually my first live trading day at Newcastle so will watch out for this course

Spent the last 3 days coding and back testing my "killer" filter which turned out to be a false dawn as it actually worsened results. I had to learn some serious coding to implement this which I will use in future. It would be an understatement however even if I say I am totally gutted by this as I had high expectations reviewing this indicator on my charts.

So now I have reviewed by Course and by Odds range. Today I will look at results by race type and will leave it at that. I will now move on to collect data from 8am for horse racing and do some match betting and casino bagging

On the positive side my bot has been very stable and I feel I can trust it now.

I also traded a correct score strategy today I copied from another blog. Unfortunately the first day was a loser but I am happy with the theory and possibilities of the system. I traded the AC Milan v Man United game which I and the spreads expected to be low scoring. I risked about £8.50 but traded out with a loss of £2.75 at 75mins when the score was 1-3 to united. Game eventually finished 2-3 to united and I say I am happy as even though the game was totally against expectations the loss was quite small. I wll trade 10 more games to decide if I will continue. This can't really be automated fully and involves a lot of judgement before the game for match selection and also in play. I am also using this as a hobby and to collect inplay football data so will not be dissapointed to make £200 a month on this. This actually helps with the Premium charge

Fighting back

P&L since last update £18.29

Fri -£8.62
Sat £13.00
Sun £13.91

Friday was the third loss in a row so I did some major analysis

Actually Wednesday, Thursday and Friday, if I hadn't implemented stop loss would have done much better however the advantage of the stop loss is after just 2 winning days I have recovered from 3 losing days. I will keep the stop loss but increase it to x 5 of trade stake (currently £3).

From my analysis I have implemented some filters. I was reluctant to do this as I feel I need to get much more data and avoid the dangers of back fitting however I got frustrated with the whole 2 days up 2 days down thing (mostly my fault as if I just let the bot run I would have been profitable).

Filters for now are implement odds range filters for back and lay trades, avoid Wolverhampton and Exeter (after today), only back at Southwell and trade all odds at Kempton my best track so far. When I say avoid I mean trade at 10% of normal stake to gather data.

I am at home now for the next 7 days for half term and I have a few things I need to implement.

Then move on to matched betting and looking at trading the horses from 8am. Also I think information I gather here will be useful in my last 10 minute trading.

So much to do so little time. Now that I have green I am enjoying and hopeful. I know I can triple my stakes and the market will absorb easily however I will stay at £3 risked per trade for the next one week at least.

Dissapointing

Day's P&L -£1.81

Very disappointing after an early high of £18 but could have been worse as hit a low of -£10

Painful but correct decision

Day's P&L -£9.79

Day didn't go according to plan. Set up my bot at 12 then went for meetings until 3pm. Came back and found out because Carlisle was abandoned Gruss's betting assistant got stuck after a few races. Came back and saw had lost about £2.50. Started the bot and then lost a further £7.50 so implemented my stop loss i.e. reduced stakes to 10% of normal.

I say painful because I went on to win £1.99 which means I would have been up £7 for the day

There is no logical reason why I have the daily stop loss and this will seriously test my resolve to keep it. The reasons why I have the daily stop loss is to avoid volatility as I build the bank and road test the system. I am still learning everyday as even though I know what the system can do in back testing, live trading is different.

I have discarded at least two systems that performed very well in backtesting but failed live

So will keep the stop loss but may make an exception for "feature bets". One thing I have noticed is my Bot seems very successful at some courses like Kempton today while it performs poorly at Wolves. I don't want to implement these filters until I have about 300 bets at least for each course

Another good day

Day's P&L £14.79

Another good day. I am beginning to feel good as I still have a filter to apply which looks very good in initial testing. Next is back testing then live testing. If it passes both stages then I will implement hopefully this weekend.

Nearly break even for February now

Good day

Day's P&L £32.68

I am not sure why but things went swimmingly.

Today was a bad day but lessons learnt

Day's P&L -£20.10

Gave up 100% exactly of yesterday's winnings. Lost the first 6 trades and was down £10 so panicked and implemented stop loss mode meaning I was trading at 10% of normal stakes. I was watching, noticed I wasn't getting matched on winning trades and feared my bot was too slow again. Won the next 2 and went back to normal stakes. Won another 2 and was almost at break even then lost 8 of the next 11, panicked again and went to stop loss stakes.

In the end if I wasn't watching I would have lost about £6 instead of £20 as everytime I went to stop loss stakes I won. I have to conclude this was a bad day at the office as I checked and about 26% of my bets didn't get matched. The normal range is between 19% and 28% so far

This really calls into question my daily stop loss principle. If however I had stuck with the stop loss after 6 bets would only have lost £10

I even checked to see if Sunday was a bad day but it was actually OK. Also checked the courses and there may be something here. Southwell is not great historically but I just don't have enough data on Muss and Ling yet.

Should I stop the daily stop loss ? I just don't know. My preference would be to scrap the stop loss and gather as much data as I can. The problem is if there is a fundamental change in the markets I could do my bank. Also large daily losses would affect my confidence.

So I guess my answer is keep the stop loss. I still gather data by trading at 10% of normal. Also today would have been -£10 instead of -£20.

I also reviewed history by day and found out historically Mondays & Tuesdays are not great. I won't rely on this as I have only about 6 weeks data.

The chart below tells the story. Red is actual trades & blue is if I had not messed about with my stakes

I need to stop watching and changing the rules as I go along. One reason I watch is because my homemade excel bot has let me down before. Also you learn when you watch.

May also implement a target where once reached go down to 10% stakes. Hmmm

Fighting back

Day's P&L £20.10

I was fairly confident I knew speed was the problem with my BOT and was proved right. I still need to make changes to improve win/profitability ratio as although I am profitable I am simply losing too many trades. However I will thoroughly test before going live. It seems my first settings are almost unbeatable but we'll have to see. I used reduced £3 stakes instead of £5 when losing so this should/could have been about £35. Anyway won't complain about a win.

3 steps forward 2 steps back

P&L Since last update -£60.32

Twice I changed settings after back testing and both times very costly. The first I simply increased my target by one tick. Luckily I figured out after 1 day's massive loss.
Clawed some of that back changed settings again this time to filter out odds where historically I wasn't profitable. Problem was this slowed down my bot in placing bets meaning I was missing the profitable trades & only picking up the loss making ones. Didn't figure this out for 3 days and finally I changed today, still lost but it was clear that was the difference. Hoping for a better day tomorrow. Will try to speed up Bot tomorrow

So that's 6% of my bank and much of my December profits gone. The big lesson here is always test changes on very small stakes. Also always use a stop loss without this loss would have been about £30 more.

Also hope to start match betting as hoping to make at least £1000 this month