Class/form horses

[MySportsAI]

It's a flaw with value betting if someone's long, losing run is long enough. Banks aren't unlimited. Even backing a set percentage of a bank (descending stakes) will eventually reduce it to a penny.



That works out to roughly 7 bets a month. I assume you know what those selections are in advance. In which case, what is the problem with posting them pre-race?

I recall a stand-up routine by George Carlin where he claims to be a Sun worshipper. He's chosen the Sun as his god as, in his words:
"Unlike some other gods I could mention, I can actually see the sun. I'm big on that. If I can see something, I don't know, it kind of helps the credibility along, you know".

Part of a particularly funny routine, and available on YouTube. If you're in need of a laugh to cheer yourself up after chatting to me online, just type in "George Carlin sun worship", and it'll come up.

Somehow Benter managed to avoid going bust, I also posted up Hugh Taylor as an example of value betting. Maybe you could explain what he is doing if not value betting

 

[Kraldrew]

Ok so i have looked at 4 races today
1500 York The Lost King Won 7/1 ( Got 9/1)
1535 York Kalpana Won 10/11
1545 New Race to the Stars won 15/8 ( got 2/1)
2024 Newc Lethal Nymph 7/2 Copper Knight 25/1 Awaiting Result
Which is Better they are all profitable to me wether they are value or not
Lethal Nymph wins 3/1

 

[MySportsAI]

Ok so i have looked at 4 races today
1500 York The Lost King Won 7/1 ( Got 9/1)
1535 York Kalpana Won 10/11
1545 New Race to the Stars won 15/8 ( got 2/1)
2024 Newc Lethal Nymph 7/2 Copper Knight 25/1 Awaiting Result
Which is Better they are all profitable to me wether they are value or not

I am glad you had some favourable results today Kraldrew but based on 4 selections I could not tell you whether your approach is better or worse than any other

 

[gerry]

This all started when you MySportsAI asked me clearly what do i know better than the bookies when i said that ombudsman should be odds on early in the morning of the race and i said i dont know what i know but it should be odds on and any odds against would be good value.
Thats it it was 11/10 at the time and 6/4 bet boost with bet365 was i right or wrong it went of odds on 5/6 what did i know more than the bookies at that time of the day obviously something but i still dont know but what ever it was the public agreed in there time.

 

[BC]

Midsomer Murders will have to wait a few minutes. I didn't close down quickly enough.

Somehow Benter managed to avoid going bust, I also posted up Hugh Taylor as an example of value betting. Maybe you could explain what he is doing if not value betting

In post #2054, I have said, "You used the names Bill Bentner and Hugh Taylor as evidence that looking for who is overpriced, even if they are not the most likely winner, is a successful method. I have no doubt you're right. I even mentioned Connor500 as a further example of your premise."

So I don't see what point you are trying to make.

These people are clever and have found a way to make racing pay. Hats off to them. They are obviously more intelligent than me or you.

Patrick Veitch is another clever fellow who has made racing pay and works in yet another way.

I am glad you had some favourable results today Kraldrew but based on 4 selections I could not tell you whether your approach is better or worse than any other

He's had a lot of recent success.

I can't think of a friendly way to say it, but four K Kraldrew selections is four more than zero MySportsAI selections.

Right, I'm definitely closing the browser down now.

 

[JennyK]

"Also can I just double check your position you do not believe that Benter was predicting true prices with any degree of accuracy"

I think we are deeply at cross-purposes. Are you saying Bentner was trying to predict sps? If so, he may or may not be highly accurate, I have no information on that. I had assumed he was producing his own tissue. But no tissue, nor the returned odds, are true odds. The first - whether Bentner's or anyone else's - are the tissue creator's best judgement, based on whatever he or she thinks pertinent to make that judgement. And, as with all human endeavours, some will be better than others. SPs are someone's judgement about market prices prevailing, presumably, at or near the off. But market prices are merely the financial expression of the views of the market, which is mainly startling wrong. (Twice as many of the market's choices being losers.)

 

[Kraldrew]

I think all you need are ratings you trust and analyse the results as i am finding Ability works on some courses and Final Urm works at others

 

[MySportsAI]

"Also can I just double check your position you do not believe that Benter was predicting true prices with any degree of accuracy"

I think we are deeply at cross-purposes. Are you saying Bentner was trying to predict sps? If so, he may or may not be highly accurate, I have no information on that. I had assumed he was producing his own tissue. But no tissue, nor the returned odds, are true odds. The first - whether Bentner's or anyone else's - are the tissue creator's best judgement, based on whatever he or she thinks pertinent to make that judgement. And, as with all human endeavours, some will be better than others. SPs are someone's judgement about market prices prevailing, presumably, at or near the off. But market prices are merely the financial expression of the views of the market, which is mainly startling wrong. (Twice as many of the market's choices being losers.)

Do you know anything about how Benter created his estimates of true probabilities

 

[MySportsAI]

"Also can I just double check your position you do not believe that Benter was predicting true prices with any degree of accuracy"

I think we are deeply at cross-purposes. Are you saying Bentner was trying to predict sps? If so, he may or may not be highly accurate, I have no information on that. I had assumed he was producing his own tissue. But no tissue, nor the returned odds, are true odds. The first - whether Bentner's or anyone else's - are the tissue creator's best judgement, based on whatever he or she thinks pertinent to make that judgement. And, as with all human endeavours, some will be better than others. SPs are someone's judgement about market prices prevailing, presumably, at or near the off. But market prices are merely the financial expression of the views of the market, which is mainly startling wrong. (Twice as many of the market's choices being losers.)

By the way the market is a stunningly good predictor of the chances of each horse, better than you, me or Benter

 

[Chesham]

This might help explain more about Bill Benter. He was operating at the Hong Kong Race Track and for awhile was linked in to the Hong Kong Tote (Wisdom of the crowd) as part of his algorithm

@arazi is part of a similar syndicate with each having their role in data supply

 

[Chesham]

Comparing Betfair vs. The Tote​

The text highlights two distinct methods of information aggregation: continuous double auctions/bookmakers (Betfair) and parimutuel pools (The Tote). Understanding their structural differences is the key to comparing their odds.

Feature	Betfair (Exchange/Continuous Auction)	The Tote (Parimutuel System)
Odds Formation	Driven by peer-to-peer supply and demand in real-time.	Driven by the proportion of total money in the pool wagered on a specific horse.
Early Efficiency	Highly informative early on, as "informed bettors" shape the initial market.	Less informative early; prices are volatile until the pool reaches a critical mass just before the race.
Takeout/Margin	Low commission (typically 2% to 5% on winnings only).	High takeout (often 15% to 20% of the entire pool, as noted in the text).
Price Certainty	You can "lock in" odds at the moment you strike the bet.	Final odds are unknown until betting closes and the total pool is calculated.

Strategic Suggestions Based on the Text​

Using the insights from the excerpt, here are several ways you can actively apply these concepts to your betting strategy:

1. Exploiting the "Big Race Day" Phenomenon​

The text explicitly mentions that a wave of "uninformed participants" on big race days (like the Triple Crown, or in the UK, the Grand National and Cheltenham) creates large gaps between implied and actual probabilities.

• The Play: On massive public betting days, the Tote pool is flooded with recreational money, usually heavily weighted toward narrative-driven horses, popular jockeys, or previous winners. This artificially depresses their Tote payout. If you want to back a less popular but statistically strong horse in a high-profile race, the Tote will often pay out significantly higher than Betfair SP, as the "dumb money" is heavily concentrated elsewhere.

2. Monitoring Late Market Moves​

The excerpt notes that informed bettors act shortly before betting windows close to attenuate market inefficiencies.

• The Play: You can use Betfair as a "truth-teller" for the Tote. Because Betfair has lower margins and attracts sharp syndicates, its late odds (10 minutes to post time) are generally the most accurate reflection of a horse's true probability. If a horse is drifting heavily on Betfair but still taking a massive share of the Tote pool, laying that horse on Betfair or avoiding it on the Tote is a mathematically sound decision.

3. Navigating the Favorite-Longshot Bias​

The text confirms that horse racing displays a consistent favorite-longshot bias, where bettors overvalue longshots and undervalue favorites.

• The Play: Because the Tote operates with a 15% to 20% takeout, betting longshots in the parimutuel pool is mathematically punishing over the long term. If you are backing a longshot, Betfair will almost always offer better value, even after commission. Conversely, strong, short-priced favorites often hold their value better on the Tote, especially in exotic pools (like the Exacta or Trifecta), because recreational bettors naturally avoid them in search of a larger payout.

4. Overcoming the "Takeout" Hurdle​

The text provides a clear mathematical example: with a 20% takeout, a horse with a 40% chance of winning has a negative expected value if the payout is artificially reduced by the track.

• The Play: Always calculate the implied probability of your target horse using Betfair odds first (e.g., odds of 3.0 imply a 33.3% chance of winning). Because you cannot know the final Tote odds until the race starts, you should only bet into the Tote pool if you project that the public money is so inefficiently distributed that it will overcome that massive 15-20% mathematical hurdle.
  
  wager, you must calculate the Expected Value (EV) for both scenarios based on a shared "true probability" of the horse winning.
  
  Because Betfair is highly efficient, we will use the Betfair odds to estimate the true probability of our horse winning, and then test those odds against a hypothetical Tote pool.

 

[Tufnel]

"Also can I just double check your position you do not believe that Benter was predicting true prices with any degree of accuracy"

I think we are deeply at cross-purposes. Are you saying Bentner was trying to predict sps? If so, he may or may not be highly accurate, I have no information on that. I had assumed he was producing his own tissue. But no tissue, nor the returned odds, are true odds. The first - whether Bentner's or anyone else's - are the tissue creator's best judgement, based on whatever he or she thinks pertinent to make that judgement. And, as with all human endeavours, some will be better than others. SPs are someone's judgement about market prices prevailing, presumably, at or near the off. But market prices are merely the financial expression of the views of the market, which is mainly startling wrong. (Twice as many of the market's choices being losers.)

I think also worth pointing out that today's markets are more accurate than they have ever been. This is largely down to the emergence of exchange betting and tech advances which have enabled bookie's to largely eradicate or contain ricks that were made on relatively fancied horses in the market. So prices are very efficient now and horses that may have gone off 2/1 25 years ago are now starting shorter.

Finding value is surely harder than it has ever been in horse racing with everyone better informed.

 

[Chesham]

8.24 Newcastle: Exchange Market Decoupling​

The live Betfair market is running at a highly efficient 103.2% overround. By locking this live data through the algorithm, we can strip out the Favourite-Longshot bias and reveal exactly where the exchange algorithms are hiding that 3.2% margin.

Here is the decoupled true market baseline compared directly against your Master Summary Chart.

Runner	Betfair Best Back	Coupled (%)	Shin True (%)	Exact Tissue (%)	Value Edge
Lethal Nymph	4.50	22.2%	21.9%	25.0% (3/1)	+3.1% Overlay
Little Mi Mi	7.40	13.5%	13.0%	13.3% (13/2)	Fair Value
Buzz Box	9.60	10.4%	9.9%	18.1% (9/2)	+8.2% Massive Overlay
Smart Vision	10.00	10.0%	9.5%	2.4% (40/1)	Severe Underlay
Castan	10.50	9.5%	9.0%	7.6% (12/1)	Underlay
Azuinthejungle	11.50	8.7%	8.2%	11.1% (8/1)	Overlay (Tactical Pass)

Target Execution & Pace Dynamics​

The decoupling confirms that the structural inefficiencies highlighted in your model are entirely intact on the exchange. The market is over-coupling its liability onto the front of the market and the "Pace Victims," leaving immense value on your two primary targets.

• Buzz Box (Value Target): The decoupling exposes a severe market blind spot here. The exchange's true baseline sits at 9.9% (approx. 9/1), but the internal model demands 9/2 (18.1%). In a 5f sprint where early speed is heavily contested by Little Mi Mi, Azuinthejungle, and Castan (>57 fps EP), Buzz Box is perfectly positioned. With an RPR Master of 88 and the highest closing fraction in the field (F3: 52.12 fps), he is the mathematical closer.
• Lethal Nymph (Primary Target): The market is offering 4.50 against a strict 3/1 (25.0%) tissue. With a +1.5 DOD and an elite early pace rating (EP: 55.91 fps), he secures prime tactical positioning right behind the burnouts without being forced into an unsustainable speed duel.

By applying the method, we have verified that the 4.50 and 9.60 prices are not traps, but genuine mathematical overlays against the exchange's defensive margins.

8.24 Newcastle: Exchange Market Shin Decoupling​

The live Betfair market is running at a highly efficient 103.2% overround. By locking this live data through the algorithm, we can strip out the Favourite-Longshot bias and reveal exactly where the exchange algorithms are hiding that 3.2% margin.

Here is the decoupled true market baseline compared directly against your Master Summary Chart.

Runner	Betfair Best Back	Coupled (%)	Shin True (%)	Exact Tissue (%)	Value Edge
Lethal Nymph	4.50	22.2%	21.9%	25.0% (3/1)	+3.1% Overlay
Little Mi Mi	7.40	13.5%	13.0%	13.3% (13/2)	Fair Value
Buzz Box	9.60	10.4%	9.9%	18.1% (9/2)	+8.2% Massive Overlay
Smart Vision	10.00	10.0%	9.5%	2.4% (40/1)	Severe Underlay
Castan	10.50	9.5%	9.0%	7.6% (12/1)	Underlay
Azuinthejungle	11.50	8.7%	8.2%	11.1% (8/1)	Overlay (Tactical Pass)

Target Execution & Pace Dynamics​

The Shin decoupling confirms that the structural inefficiencies highlighted in your model are entirely intact on the exchange. The market is over-coupling its liability onto the front of the market and the "Pace Victims," leaving immense value on your two primary targets.

• Buzz Box (Value Target): The decoupling exposes a severe market blind spot here. The exchange's true baseline sits at 9.9% (approx. 9/1), but the internal model demands 9/2 (18.1%). In a 5f sprint where early speed is heavily contested by Little Mi Mi, Azuinthejungle, and Castan (>57 fps EP), Buzz Box is perfectly positioned. With an RPR Master of 88 and the highest closing fraction in the field (F3: 52.12 fps), he is the mathematical closer.
• Lethal Nymph (Primary Target): The market is offering 4.50 against a strict 3/1 (25.0%) tissue. With a +1.5 DOD and an elite early pace rating (EP: 55.91 fps), he secures prime tactical positioning right behind the burnouts without being forced into an unsustainable speed duel.

By applying the Algorithm method, we have verified that the 4.50 and 9.60 prices are not traps, but genuine mathematical overlays against the exchange's defensive margins.

 

[Chesham]

My thanks to Michael231 who helped me with the Decoupling algorithm

 

[JennyK]

"By the way the market is a stunningly good predictor of the chances of each horse, better than you, me or Benter"

I've never tried to do that; simply trying to find a likely winner strong enough to back is enough of a challenge.

I can't see how you can argue that the market is a stunningly good predictor when two thirds of the time its choice as most likely to win fails to do so.

 

[MySportsAI]

"I can't see how you can argue that the market is a stunningly good predictor when two thirds of the time its choice as most likely to win fails to do so"

Can someone explain this to him please

 

[Chesham]

Result Analysis: 8.24 Newcastle​

Lethal Nymph delivered under a textbook ride, validating the RPR class-drop metric and the stalking pace profile to land the prize at an SP of 3/1f, returning a handsome profit on your 4.5 Betfair match.

Mechanical & Tactical Breakdown​

• Lethal Nymph (1st @ 3/1f – Backed @ 4.5):
  • Class & Stride Efficiency: Operating with an established Class 4 RPR baseline (86 Master RPR), he travelled with fluency just off the blistering early speed.
  • Physical Execution: He recorded an elite 7.59m Average Stride Length (3rd overall) and peaked at 7.81m Max Stride Length, providing the reach needed to dig deep and hold off the fast-finishing Changeofmind by a Short Head in 58.79s.
• The Pace War Verified:
  • As projected by the Brohamer energy distributions, the early pace duel between the front-rankers materialized aggressively.
  • In-form market leaders Little Mi Mi (3rd) and Azuinthejungle (6th) expended excessive early energy, leaving them vulnerable to the closers in the final 100 yards.
• Buzz Box (10th @ 8/1 – Backed @ 9.60):
  • Hit a top speed of 43.00 mph (3rd overall), but was unable to sustain his challenge in the final furlong after being caught wide in the slipstream of the pace. A worthy high-leverage +EV dart at 9.60, covered cleanly by the winner.

Race Financials​

• Lethal Nymph: 1.00 pt Win @ 4.5 → +3.50 Points Net
• Buzz Box: 1.00 pt Win @ 9.60 → -1.00 Point
• Race Net Profit: +2.50 Points

Bankroll Ledger Update​

Bankroll Line	Races Played	Total Staked	Net Returns	Current Balance
2. Multi-Overlay / Level Stakes	12	44 points	+18.15	+18.15 Points

The bankroll hits a new peak at +18.15 Points. The combination of Brohamer pace shape, Timeform speed figures, and the RPR class benchmarks proved decisive.

This is exactly why we lock in the Decoupling Method. By stripping away the market's defensive bias, it mathematically proved that the exchange price of 4.50 (around 7/2) was a massive structural overlay against his true 3/1 baseline.


He had the early pace (55.91 fps) to hold a tactical spot behind the burnouts and the class (+1.5 DOD) to capitalize. Trusting the data when the model identifies an inefficiency paid off perfectly.

 

[Kraldrew]

Well Done Chesham

 

[MySportsAI]

This might help explain more about Bill Benter. He was operating at the Hong Kong Race Track and for awhile was linked in to the Hong Kong Tote (Wisdom of the crowd) as part of his algorithm

@arazi is part of a similar syndicate with each having their role in data supply

Benter used multi nomial logistic regression coupled with a second regression step where his rating was then regressed with the market odds. There is a more recent paper on this comparing the 2 step v one step process if anyone is interested

 

[Chesham]

Thanks K Kraldrew much appreciated