The Forward Testing Thread
-
Post your Strategies that you want to put forward for Forward Testing here. I(or Martin) can then add them to the first post for everyone to monitor/tinker with. All strategies will assume that they are being tested from the date of posting unless you say other wise. Just post your strategy, a brief description and the .JSON file exported from your strategy bunker.
John Folan
Zero Paced Lays. Consigned to the bin after four losing months.Turf and AW Short Odd Lays. Laying short priced horses on the Turf and AW. No Colts and some course filters.
Start Date for this strategy is 30/07/2023
Nothing made and nothing lost on these. Binned on the basis I am looking at a different angle. The original strategy apart from June this year looks solid on the megathread.@Greg-Mitchell
Lay 3 Year Old Horses. Laying three year old horses subject to certain criteria. Start Date for this strategy is 01/08/2023
1690879857623-lay-3yo.json
Keep these going for a further three months. 21 points profit. BUT look at them excluding Mondays and Sundays and the profit is much better.@Karl-Pick
Back Pace 8-10 Horses. Excludes 7f races. 2-10 odds range. Start Date for this Strategy is 07/08/2023
1691418557480-rules_export-5.json
Keep these on forward testing. 26 points profit since inception.@martin forward testing strategies:
Lay 3rd Favourite Would suggest binning these, Loss since posting
Irish Dobs (28th July 2023) - rules_export (8).json
Irish Dobs 1.0.gz Continue testing. 11 points profit since posting.Ones to avoid for now but still testing -
Back to lay dobs Binned as 3 points loss since posting
Lay Beaten Fav The bane of everyones existence. 20 point loss since august. Shame as these were amazing last year and this. -
@simon-topper said in The Forward Testing Thread:
@John-Folan This one looks interesting, set it up myself as well.
Bad downturn in September but only 10% of the years profit
-
@John-Folan This one looks interesting, set it up myself as well.
-
Created a bot for the last post. Will run it as a test and see what happens. It's fully automated as well
-
I've been looking into each race type. I've started with the AW. I am predominantly going to be looking for Lay Strategies. This is the first one. I am looking at 2023 onwards for my results and in theory this one is a cracker.
It's laying all horses in a race between a specific odds range at certain aw courses. The only other filters are sex, class and runner numbers. Forward testing starts today.
-
@Dan-MacKinnon tricky month normally with different ground everyday and every year on the turf......Mine is AW related so not a lot of variance with ground! Just cold! 3 today fingers crossed...
16:45 Dundalk Moral Turpitude
19:45 Dundalk Theriverrunsdeep
20:15 Dundalk John Alexander -
@Karl-Pick I've looked at all the strategies I've forward testing and they all go south in December/ January
-
@John-Folan solid stuff so far then! well done to those who contributed so far
and great work John!
-
I’ve updated the OP. Will go through the thread and check any that aren’t in the OP and add them with updates.
-
@Martin Dobby was 28th Oct. Remembered thanks to specific date rule which I should probably remove now!
Lay 3rd fav was last week. Back testing looks better now, a bit of forward testing to do though. Its one that really annoys me, but will be awesome if i get it right...
-
@Tony-Hastie when did you tweak these? date wise?
-
@Tony-Hastie looks interesting will be interesting to see how it performs moving forward!
-
Also did a deep dive on Lay the third fav and am seeing better results...
-
I combined the B2L and Irish DOBing, removed a few courses....added a few other things and its going alright at present if anyone is interested in taking a look...
-
@John-Folan that will be great!
-
I’m going to go through these and do a “how it’s going” snapshot over the next few days. A few of these have been running a while now.
-
@Andrew-Gray That's my strike rate since 1st August.
the back testing says it's profitable so can only forward test now and see if it holds true, as long as when I look at the back testing it's close to my results so far then I think it should be ok.
Looking forward to the extra years of data Adam is going to put in soon to see if it still holds true.
Here is a google sheet with all the results (hope that works):
https://docs.google.com/spreadsheets/d/1MEgvc5dOn3T91dfgV4MmHg15cH8VUD0Odmuil6FfxHI/edit?usp=sharing
-
@simon-topper Interesting stuff Simon.
I think I got lucky and missed the down turn. Since 26th October I have been running your strat and it's got a 95% strike rate (is that what you've also seen?). Can't argue with that
What is your theory on why this works/will work going forward?
-
So first off, apologies to anybody using my bot for LDW. Let this be a lesson to triple check everything.
My forward testing results weren't matching the back testing but while I was making money wasn't too concerned. It took a bad dive for a couple of weeks so started looking at what was going wrong, I had sometime to watch a few races so had BFBM open and Betfair. I noticed some bets weren't being placed even though it should, so knew something was wrong with the bot. At some point I had excluded 1st & 2nd favourites but hadn't reverted back!
Also I now have had over 900 selections and noticed a couple of things.- 5f races are the worst performing distance -26 points during testing, specifically 5f handicaps. Anything over 27f also performed poorly.
- no. of runners also had a couple of outliers so have trimmed that down as well.
Have updated the bot and the strategy to reflect this.
-
Oh and seeing as I’m a mission today do yourselves a favour and run lay the 3rd favourite looking at the race type breakdown. You are welcome .