Cheers mate and yes I have - that calibration chart includes turf races and the algo performs similarly on such
The on thing is that my data from the past couple of years is all weather heavy (probably around 70:30) and there are A LOT more turf courses so my sample sizes for some of these aren’t particularly significant
**FRIDAY 27TH FEBRUARY - AI GENERATED PREDICTIONS**
Busy day with 12 price qualifying value bets across @LingfieldPark and @WolvesRaces tomorrow, with a few extra bets added on the @tipstrr profile too
Lingfield (AW) Bets:
- STAR OF ALBION (26%) @ 4/1
- LORD HARCOURT (23%) @ 5/1
- MERESIDE DIVA (24%) @ 6/1
- BRIDGET'S VIEW (15%) @ 15/2
- FERROUS (29%) @ 7/2
- WYLD BILL (20%) @ 5/1
Wolverhampton (AW) Bets:
- MAXIMISING (30%) @ 4/1
- STENMARK (12%) @ 10/1
- BUCKLAND BELLE (13%) @ 11/1
- BEAUMADIER (12%) @ 9/1
- PRIDE OF DONEGAL (31%) @ 7/2
- GEMMARI (26%) @ 4/1
Hopefully including the algorithm's predicted win %'s sheds some light on the odds range and why we often go some days without a winner - i.e. I am certainly not guaranteeing 'booms' but rather trying to spot overpriced value opportunities.
Bet anything 9/2+ (5.5+) each-way personally and will be including in results from March.
3 additional bets that just missed out on target price can be found on the tipstrr profile here: tipstrr.com/tipster/racebo…
@racebotuk@LingfieldPark@WolvesRaces@tipstrr Damn so it’s as in depth as you can get by the sounds, well done for building that by the way.
Have you tested during the turf season, I think this could help predictions even more so
It analyses up to 400 variables per horse to determine a horses % chance of winning that race and is accurate within a couple of % points on average.
E.g. if a horse is given a 25% chance of winning, in the long run horses given that value will win between 23-27% of the time (it's tighter than that for the most part).
The below chart shows the calibration for the top ranked picks in each race (each dot represents 150 1st ranked horses), with tightness to the actual win % trend line being the main goal.
From there I apply a 1.15-1.3x EROI price target for each horse, with those sitting above such qualifying for a bet.
@LBTrade_@LingfieldPark@WolvesRaces@tipstrr Always reassuring!
February hasn't been the best of months for this system however I had (regrettably) been making a few tweaks throughout and had some very unfortunate variance swings... still in profit (just) but confident the system is improved now - 2 big days to go 📈
**MONDAY 16TH FEBRUARY - AI GENERATED PREDICTIONS**
7 price qualifying value picks across @WolvesRaces tomorrow after a couple days of nice profit on the trot 🏇
Wolverhampton (AW) Bets:
- ALONDRA @ 15/8
- LION'S HOUSE @ 4/1
- BAD HABITS @ 3/1
- NAMMOS @ 13/2
- PAPA COCKTAIL @ 9/2
- MAGIC RUNNER @ 4/1
- HACKNEY DIAMONDS @ 4/1
Accidentally included EL BUFALO as a pick over on tipstrr which can't be deleted so just going to have to roll with 3 picks in 1 race 🤷♂️
tipstrr.com/tipster/racebo…
**SUNDAY 15TH FEBRUARY - AI GENERATED PREDICTIONS**
Profit day today (from what I can tell on my accounts) but will do an official results spec tomorrow. 5 bets at @NewcastleRaces tomorrow..
Newcastle (AW) Bets:
- HIGHFIELD COMRADE @ 11/4
- FREE PIC @ 9/2
- BOSSY PARKER @ 7/2
- FLOWSTATE @ 4/1
- DIGITAL @ 15/4
Slightly increased the TARGET PRICE parameters to slightly reduce picks but that has resulted in DUCHESS just missing out on price by 0.05....
Included on tipstrr here: tipstrr.com/tipster/racebo…
**SATURDAY 14TH FEBRUARY - AI GENERATED PREDICTIONS**
Would like to say tomorrow will be a lovely day but the world can be a funny place... here's tomorrow's value bets across @LingfieldPark and @NewcastleRaces tomorrow anyway 🤷♂️
Lingfield (AW) Bets:
- BOWNDER @ 7/1
- THECOFFEEPODDOTCO @ 10/3
- CAYMAN TAI @ 10/1
- WAY TO DUBAI @ 10/3
- GENNADIUS @ 11/2
- HEATHCLIFF @ 3/1
- LEGAL REFORM @ 5/1
- MART @ 3/1
Newcastle (AW) Bets:
- FOREST CAPER @ 9/2
- GOLSPIE @ 9/1
- PITBOSS @ 6/1
- FRANCESCI @ 8/1
- MYSTICAL LAND @ 7/2
- BRAZILLIAN ROSE @ 5/2
- MERSESIDE DIVA @ 4/1
- POPTY PING @ 9/2
+100pts since Nov26 but the past couple of days have been cruel nevertheless.
The staking system has exaggerated matters the past 2-3 days too due to winners predominantly being made up of the lower stake bets by chance.
A few high stake bets just missing out on target price for tomorrow such as INSPIRED and SILVER STATE but I dare not give any leniency and play as a bet at this point...
All bets and tips verifiable here as always: tipstrr.com/tipster/racebo…
Had the trifecta in the 1st race yesterday n all... 😅
Don't hold your breath though, a couple headbobs the wrong way can make or break a day as you well know!
The white rows (non-price hitting) perform considerably better but ultimately if you bet on them long-term you'll make a loss despite the higher win rate...
Currently +12% ROI for the month for qualifying bets and +0.5% for those which didn't hit price (despite 18.5% and 27% win rates respectively)
**WEDNESDAY 11TH FEBRUARY - AI GENERATED PREDICTIONS**
Another 19 price qualifying picks across @Southwell_Races and @kemptonparkrace tomorrow including 3 in 1 race - getting a bit hectic now...
Southwell (AW) Bets:
- SHALAA ASKER @ 13/8
- PAPA DON'T PREACH @ 8/1
- HIMSELF @ 11/8
- FLEUR DE MER @ 9/1
- IMOLA @ 3/1
- STATE OF MADNESS @ 4/1
- KING OF CHAOS @ 85/40
- WELCOME RETREAT @ 8/1
- GREY HORIZON @ 13/2
Kempton (AW) Bets:
- TRAPPER JOHN @ 9/2
- KARAKULA DANCER @ 10/3
- CHILLICONCARNEIGH @ 15/2
- ALL MOONSHINE @ 7/4
- TRUE COLOURS @ 9/2
- EL BODON @ 4/1
- CARBINE HARVESTER @ 11/1
- RAJINOORA @ 3/1
- AIMING HIGH @ 9/1
- VIZZAVONA LADY @ 6/1
Either there are some massively overpriced bets tomorrow or the algo is having a seizure with the likes of STATE OF MADNESS, WELCOME RETREAT, KARAKULA DANCER and CARBINE HARVESTER currently leagues above their target odds.
The 3 picks in the 4:44pm making me a little nervous considering none of them are the favourite.
Personally bet anything 9/2+ each-way but can result in a lot of outlay, particularly when a 2pt bet hits that price.
2 additional tips (I know) and easy best odds analysis can be found here: tipstrr.com/tipster/racebo…
@LBTrade_@LingfieldPark@NewcastleRaces Fair play, hopefully something in that as LM is currently priced much bigger… I’m going to have to take a proper look at all this later as I say.
My algo currently doesn’t perform great on lightly raced fixtures so something like this could be the missing piece to the puzzle!
**TUESDAY 10TH FEBRUARY - AI GENERATED PREDICTIONS**
Just the 17 value bets across @LingfieldPark and @NewcastleRaces tomorrow with the introduction of improved prediction accuracy and a 3rd ranked pick for each race.
Lingfield (AW) Bets:
- PROFIT STREET @ 13/8
- FIDELIUS @ 13/8
- DIOMED SPIRIT @ 6/1
- WRIST ART @ 6/1
- HEADSHOT @ 8/1
- BEAUZON @ 9/4
- WYLD BILL @ 4/1
- KATE O'RILEY @ 12/1
Newcastle (AW) Bets:
- LORD MAYOR @ 9/2
- MADEMOISELLE BELLE @ 7/2
- FRANCESI @ 4/1
- POWERFUL RESPONSE @ 9/1
- SIR MAXI @ 7/2
- SPARTAN FIBHTER @ 13/2
- ESQUE ELEGANCE @ 2/1
- LADY DELILA @ 5/1
Appreciate this is a lot of bets to follow for most but value is value - I'd rather have a +15% ROI over 1000 bets than +30% over 100
Much easier to follow over on tipstrr here with live best odds and fast bet links for all tips: tipstrr.com/tipster/racebo…
I advise to bet each-way for odds of 9/2+ but P/L recorded to win singles on X
@racebotuk@LingfieldPark@NewcastleRaces Just thought I’d highlight this one from yesterday, lion ring massively clear on SR, didn’t get looked at in terms of betting, wins at 12/1
@racebotuk@LingfieldPark@NewcastleRaces tough one here due to 3 of them not having a SR for most recent logged run, but Lord Mayor comes out on top of the SR’s for most recent runs, just wanted to highlight that I found wickseys most recent win due to the POR of 58 and he ran off 51! Don’t always need highest SR
@racebotuk@LingfieldPark@NewcastleRaces You should check it out just out of interest, racingthroughtime.uk
Can build custom race cards with races from today, shows the speed figures and projected rating etc. I’d probably say 70% of ur model picks today align with the top SR’s
@LBTrade_@LingfieldPark@NewcastleRaces Regrettably it considers just about everything other than speed/pace factors which I know are important (as they aren’t available from my data source)
Uses 400+ other variables per horse to determine predicted win chance however so any confluence could prove promising…