AD's daily selections

[AustinDillon75]

More for fun these ones as the value shouts, I'm just asking Gemini for the each way returns with these so don't whack me over the head if I'm wrong. I quite like the idea of doing a 1pt win, 1pt each way on these. A couple of them shortened from the morning line odds.

(When will Golden Ace get the credit from the market her performances actually deserve?)

1pt Win Stake on All Selections​

Total Outlay: 19 selections at 1pt Win = 19.00 pts

Here are the returns for the three horses that won their races:

Horse	Result	Odds	Return (1pt Win)
Loud And Proud	WON	12/1	13.00 pts
My Gift To You	WON	9/4	3.25 pts
Arthurs Sixpence	WON	6/1	7.00 pts
Total			23.25 pts

Final Balance (Win Only)​

• Total Staked: 19.00 pts
• Total Returned: 23.25 pts
• Net Outcome: +4.25 pts

---

Combined Totals (1pt Each Way + 1pt Win)​

If you placed the 1pt Each Way bet and an additional 1pt Win bet on every horse (a total stake of 3pts per race):

• Total Staked: 57.00 pts
• Total Returned: 71.13 pts (47.88 pts from E/W + 23.25 pts from Win)
• Overall Net Outcome: +14.13 pts

 

[AustinDillon75]

Total Outlay
10 selections staked at 1pt Win + 1pt Each Way (which equals 2pts to Win and 1pt to Place, totaling 3pts per race) = 30.00 pts

Actual Field Sizes and Place Terms
Here are the official field sizes from May 2, 2026, which lock in the exact place terms for your successful bets:

• 13:55 Uttoxeter (Final Surprise): 7 runners. Standard terms for 5-7 runners are 1/4 odds, 2 places.
• 16:10 Newmarket (Ten Carat Harry): 11 runners (Handicap). Standard handicap terms for 8-11 runners are 1/5 odds, 3 places.
• 16:15 Punchestown (Place De La Nation): 4 runners. Fields of 1-4 runners are strictly Win Only.
• 18:27 Doncaster (Papa Cocktail): 6 runners. Standard terms are 1/4 odds, 2 places.
• 16:40 Hexham (Mr Globalist): 6 runners. Finishing 3rd in a 6-runner field means the place bet loses, as only the first two places are paid.

Returns Breakdown​

Time	Horse	Result	Odds	Terms Applied	Return
13:55	Final Surprise (IRE)	WON	3/1	1/4 odds (7 runners)	9.75 pts
16:10	Ten Carat Harry	WON	28/1	1/5 odds (11 runners)	64.60 pts
16:15	Place De La Nation	WON	10/1	Win Only (4 runners)	33.00 pts
18:27	Papa Cocktail	2nd	12/1	1/4 odds (6 runners)	4.00 pts

Calculation Details:

• Final Surprise: 2pts Win @ 3/1 (8.00 pts) + 1pt Place @ 3/4 (1.75 pts) = 9.75 pts.
• Ten Carat Harry: 2pts Win @ 28/1 (58.00 pts) + 1pt Place @ 28/5 (6.60 pts) = 64.60 pts.
• Place De La Nation: As this was a 4-runner field, the 1pt place stake defaults to the win under standard rules. The total stake is therefore applied as 3pts Win @ 10/1 = 33.00 pts. (Note: If your specific bookmaker voids the place stake instead of rolling it over, your return here drops to 22.00 pts, with the 1pt place stake refunded).
  
  
• Papa Cocktail: The 2pt Win bet loses. 1pt Place @ 12/4 (3/1) = 4.00 pts.

Final Balance​

• Total Staked: 30.00 pts (87pts)
• Total Returned: 111.35 pts ( +182.48 pts)
• Net Outcome: +81.35 pts (+95.48 pts)

 

[AustinDillon75]

The main top rateds absolutely tanked yesterday, just 9 wins from 50 that ran (3 non runners) and a loss on the day of 17.20 pts to BSP. I know it seems an odd thing to say but all we are doing is regressing towards what is expected.

TOP RATEDS AFTER DAY 6
234 RUNS
79 WINS
33.8% STRIKE RATE
51.31 PTS TO BSP
21.9% RETURN ON INVESTMENT.

Will put up today's horses in a bit.

 

[AustinDillon75]

Today's ratings, and the Newmarket card. We are 80.1% correlated to the morning line market.

 

[AustinDillon75]

Value bets

​

Total Outlay​

11 selections staked at 1pt Win + 1pt Each Way (2pts Win, 1pt Place = 3pts per race) = 33.00 pts

Returns Breakdown​

None of the selections won their races, so all Win portions of the bets lost.

For the Place portions, here is how the results apply based on the official field sizes:

• 13:45 Spinning Lizzie (3rd of 6): Loss. Standard terms for 5-7 runners pay 2 places.
• 14:05 You Mystify Me (2nd of 4): Loss. Fields of 1-4 runners are Win Only.
• 16:10 Lazurite (2nd of 8): Return. Standard terms for 8 runners pay 3 places at 1/5 odds.
• 16:45 Noble Raider (3rd of 7): Loss. Standard terms for 5-7 runners pay 2 places.
• All other selections finished outside the standard place parameters.

Time	Horse	Result	Odds	Terms Applied	Return
16:10	Lazurite	2nd	9/1 SP	1/5 odds (8 runners)	2.80 pts

Calculation for Lazurite:

• 2pts Win @ 9/1 = 0.00 pts (Loss)
• 1pt Place @ 9/5 = 1.80 pts + 1.00 pt stake = 2.80 pts

Final Balance​

• Total Staked: 33.00 pts (120 pts)
• Total Returned: 2.80 pts (+ 185.28 pts)
• Net Outcome: -30.20 pts (+ 65.28 pts)

 

[AustinDillon75]

Better day for our top rateds today, 14 wins from 36 races (38.9%) and a BSP profit overall of 4.57 pts. Steadying the ship today.

TOP RATEDS AFTER DAY 7
270 RUNS
93 WINS
34.4% STRIKE RATE
55.88 PTS TO BSP
20.7% RETURN ON INVESTMENT

Back up tomorrow. We will keep seeing if the model can be improved but will only implement changes in the background vs the currently successful one. After 7 days it's still some way ahead of where I would expect it to be on average. An average 7 days is 250 runs, 77 wins (30.9% SR) and a 37.15 pt (14.9% ROI) in 4 years worth of backtesting. This week's strike rate is in the 86th percentile, the profit is in the 73rd percentile, and the return on investment in the 70th percentile.

We can afford plenty of ups and downs as we continue. And this is definitely continuing.

 

[AustinDillon75]

This is a big day for us. I've checked the backtesting, which the live tests have outperformed, and found it has failed to slice the parquet file correctly meaning that where it has looked at 1 and 2 year jockey / trainer performance it has had access to a whole month's performance even if it is testing a race on 1 April, meaning it is distorting. So I am doing a bit of tweaking around to the model and revising the approach to testing. If we do this, it should re-calibrate the models and revise the level of importance to attach.

The issue is that the models do place a fair bit of weight on 1 and 2 year performance, so it's not exactly helpful or confidence inspiring. If the performance does start winding back to negative territory and perhaps if we end up at 100 pts in deficit we'll throw it in, but it's done so well thus far on blind performance that we should continue for now.