Eric Bowers narrowing the field

[JennyK]

Leodis

Perhaps I should have put up the full sentence to which I referred:

"Bowers did not claim 80% strike rate.His minimum acceptable odds were 2/1 . His average odds were 4/1 and never backed greater than 10/1."

I've no real knowledge of Bower's work and don't know whether the above is an accurate encapsulation or not, but my point, and it applies equally to gerry's suggestion, is that both, if applied automatically, are arbitrary. Any decision to eliminate part of a field because one has an arbitrary screening rule is bound, sometimes, to exclude the winner before one really gets into studying the race. We are all looking to exclude most runners as we try to work out whether there is one or more worth backing, but there is a difference between doing so through the automatic application of a rule rather than as a result of appraisals based on all the factors one believes to be pertinent in finding winners.

 

[Chesham]

I think somethings are taken out of context and unless one has actually read the Bowers Book, perhaps that causes the Confusion.

It was part of the chapter on using computers and back in 1990 the only readily available was RSB

A 5 year study that Bowers carried out using a computer

87% of winners= Odds less than the number of Runners

ie: 10 runners = 9/1 or less, 15 runners = 14 /1 or less, 20 runners = 19/1 or less

 

[Chesham]

Rob Morris who demonstrated his odds lines to a small Yahoo group that I was fortunate to belong to. Over a week and posting up daily he was well in profit. Now Rob would only back up to his tissue of 9/2 if VFM was on offer. ie: Rob knew that even if his 9/2 tissue was accurate he would still have losing runs.

Of course Rob had a Bank that and the Psychological Mindset to withstand losing runs.

I don’t think Bowers was backing in races like some of those recently covered on this thread. In fact t yesterday’s York was not a balanced handicap and avoided. The Trends analysis did find the eventual winner.

The main thing is that this Bowers thread has enabled members to join in and try something a little different. Unlike VDW, 40 year old form books are not required to explore races that were all aftertimed, no having to find the “KEY”

 

[Michael231]

JennyK
You Say Jakajaro was cut off, but if you added the pace factor it would have been a top pick.
I got AI to go through the race and it picked it as a top pick on bowers as when I asked it to go through the race it gave so a few and mentioned Jakajaro then I asked it to go through the race using the pace as well and give me who it thought was the best bet and it gave me Jakajaro. May be able to find it and put it up later. The thing I was looking at was how Bowers managed to get so many winners and I think you will find he went through his own way of picking his choice of the winner and like VDW he had other things taken into consideration, eg pace and head to head races. I did back it but only small wager as not used AI for this method before, it got that one but the next one I asked about was the 5:15York and the pick ended up last think it was the joint fav so shows not 100% just letting you know that there is more than one way to get the winner even with a draw bias.
I'm waffling now so enjoy the rest of your day.

 

[Michael231]

Chesham looks like I will have to get the book now a bit dear at the cheapest I've seen is £18 plus postage used.
AI tells you how to but not why I'm sure the book will.

 

[Monty]

Chesham has put most of the book’s salient points on here, Michael. I doubt that you’ll find more inspiration from the book. Just research all of Chesham’s posts on this subject. Like me, I’m sure you will be educated and enthralled.

 

[JennyK]

"You Say Jakajaro was cut off"

Yes, Michael231, but only if, for the race, one made a single-figure draw an arbitrary screening rule. As I pointed out:

".... in the twenty-three 19 and above runner York 5f handicaps that comprise the population in my base, seven were won by horses from double-figure stalls, so a single-figure stall winner, though likely, is by no means certain."

 

[Michael231]

My apologies for that I hadn't read it all

 

[Michael231]

Hi M Monty that's how I have been doing it to get this far, tell a lie he put up the Smartsig mags and I have them so that helped as well.

 

[Chesham]

These are the stall numbers for The Trend for Yesterdays Race. Yesterdays winner is included

 

[Leodis]

Testing a few algo's with a Bowers race. The first one is trying to get it to read and understand form. The second one is from a Lee perspective.

19.50 Hml

RPR trajectory vs today’s OR and weight impact​

• Obito: Latest three RPRs (90 on 30Sep25, 87 on 16Sep25, 81 on 04Nov25). Latest peak 90 ≥ Today's OR 82 + 3 → meets "Improved-on-drop" criteria; TS rose to 62 then fell to 50 (mixed recency). Add +2 score for improved-on-drop but flag partial TS drop; no jockey claim; weight today 9-11 vs past carried similar — no weight advantage. Net score: +2 (needs confirmation due TS dip)

• Austrian Theory: RPRs 74 (16Apr26), 80, 86 (18Sep25). Latest RPR 74 is below OR79 so no immediate uplift; earlier 86 > OR by 7 but not recent—treat as isolated uplift → +1 and "needs confirmation". Weight today 9-8 similar to past; no claim

• Flying Fletcher: RPRs 78,78,83 (most recent 26Apr26 RPR78). Peak 83 > OR75 by 8 but dated 27Mar26 (earlier). TS profile 71→54→71 indicates bounce; treat as supported uplift but partly inconsistent — +1 (or +2 if assuming TS support recent). Jockey claims none except 27Mar26 had 3lb claim then; today no claim

• Qitaal: RPRs 74,80,82 with TS 58→68→61. Peak 82 > OR77 by 5 but some older; TS reasonably strong — +1 to +2 depending recency. Jockeys varied; today Connor Beasley (no claim) is a plus

• Quiet Resolve: Latest RPR 76 > OR73+3? 76 = OR73+3 → qualifies; TS 61 on most recent win supports uplift; weight 9-2 vs claimed 5lb in that win? Today jockey Rhys Elliott 5lb claim reduces carried weight by 5lb relative to standard — gives additional +1 per rule. So treat as Improved-on-drop +2 plus +1 for claim = +3 (high confidence)

• Draupnir: RPRs 75,55,76 (recent 75) vs OR72 → latest 75 ≥ OR+3 → qualifies; TS 69 supports; weight similar — slight positive

• Sanafi Zabeel & Naturalia & Tee Aitch Aye show weaker RPRs below OR; no uplift.

TS backing for form​

• Strong TS support: Flying Fletcher (TS 71/67/71), Qitaal (68/61), Draupnir (69) — supports recent RPRs.
• Mixed/weak TS: Obito TS dropped to 50 recently despite high RPR earlier — reduces confidence.
• Weak TS: Salam Dubawi and Tee Aitch Aye — low TS or missing.

Jockey impact and recent booking changes​

• Obito retains David Nolan (consistent) — neutral/stable.
• Austrian Theory Oisin Orr (top rider) — positive booking for a 79-rated horse.
• Qitaal Connor Beasley (strong on sprinters/milers) — positive.
• Quiet Resolve switches to Rhys Elliott with 5lb claim today — significant positive (extra +1 per rules).
• Flying Fletcher jockeys variable; no today’s claim.
• No other notable late booking jumps.

Pace/draw influence​

• 9f at Hamilton on Good: race likely run at steady pace; horses with form at 10f (Obito, Austrian Theory, Qitaal, Draupnir) should handle 9f. Seas of front-runners: Flying Fletcher often prominent — may set honest pace and suit closers (Obito, Qitaal). Draw not supplied; assume neutral impact.

Betting suggestions & stake guidance​

• Win: Quiet Resolve (each-way preference) — confidence raised by recent win, TS support and 5lb claim. Suggested stake: small win bet (2 units) + place (4 units) or 1.5u win + 3u place.
• Win: Obito — interesting each-way given high past RPRs but TS drop means lower confidence. Suggested stake: 1.5u win + 3u place (or 1u win + 2u place).
• Place/Dutch: Dutch between Quiet Resolve and Obito for cover (total risk similar to single win stakes). If you prefer a trio for places add Qitaal (each-way 1u).
• Each-way percentages: follow standard 1/4 odds 3 places if available.
• Market watch: If market shortens strongly to Obito or Quiet Resolve, increase stake by 25–50% (raise confidence); if they drift, reduce stake by 25%.

Confidence summary:

• Quiet Resolve — High (recent win, TS support, jockey claim).
• Obito — Medium (high past RPRs but TS dip; needs confirmation).
• Austrian Theory, Qitaal, Flying Fletcher — Place prospects/each-way fodder.

Summary comparison — Form analysis vs L33 Model​

Horse	Form model verdict (brief)	L33 score (Steps1–9)	L33 verdict
Quiet Resolve (OR73)	Top selection — recent win, TS support, 5 lb jockey claim; high confidence (bet)	9/9	BET — primary (win + EW)
Obito (OR82)	Top selection (ability clear from past high RPRs) but TS dip → medium confidence; EW suggested	7/9	FOLLOW — each-way (needs confirmation)
Austrian Theory (OR79)	Place/each-way; earlier high RPR (86) but not recent → lower confidence	8/9	FOLLOW — each-way/place candidate
Qitaal (OR77)	Place contender; TS and RPRs supportive → EW option	8/9	FOLLOW — each-way/place candidate
Flying Fletcher (OR75)	Useful recent RPRs and TS; each-way/possible win value	9/9	BET — value (win + EW)
Draupnir (OR72)	Live for places; RPR/TS supported but trip step-up concern	7/9	FOLLOW — place chance, trip query
Naturalia (OR62)	Weak — below OR, minor each-way chance	6/9	FOLLOW — minor each-way chance
Sanafi Zabeel (OR64)	Outclassed relative to leaders; place at best	1/9	FOLLOW/No bet
Salam Dubawi (OR81)	No form support; avoid	0/9	No bet
Tee Aitch Aye (OR66)	Poor recent form; avoid	0/9	No bet

Notes on interpretation:

• Form model used RPR vs OR, TS backing, jockey claims and market guidance — produced Quiet Resolve and Obito as top two with Flying Fletcher/Qitaal as each-way threats.
• L33 applied a strict 11-step checklist focussing on TS trajectory, credibility, TS→RPR relationship, exposure and distance/weight logic. Horses passing all/most steps get stronger support; failures reduce bet recommendation.

 

[JennyK]

Too late to be of practical help today, Leodis, but re the draw, in 9f handicaps at Hamilton with 8 to 11 runners, my data shows proportionately more winners came from stalls 7 to 11 than from 1 to 5. But with only twenty races comprising the population, the apparent difference may well be a product of small numbers (as, I strongly suspect, is the lack of any winner among the twenty from stall 6).

 

[Leodis]

No problem, JennyK. The top pick came 2nd @17/2 so happy with that. A lesson learned, however, not to get involved with races that include runners with unexposed, UK form (see winner).

 

[Chesham]

Too late to be of practical help today, Leodis, but re the draw, in 9f handicaps at Hamilton with 8 to 11 runners, my data shows proportionately more winners came from stalls 7 to 11 than from 1 to 5. But with only twenty races comprising the population, the apparent difference may well be a product of small numbers (as, I strongly suspect, is the lack of any winner among the twenty from stall 6).