James Quinn is regarded as one of the “greats” of the American handicapping scene. We have chosen his book The Best of Thoroughbred Handicapping, Leading Ideas and Methods as one of the 10 best racing books of the last 20 years.

In this exclusive extract, Quinn pays tribute to the work of fellow US professional and former PPM columnist Barry Meadow.

Every bet, every day, Barry Meadow plays the races in harmony with his betting lines. In three decades of chasing the money at racetracks, I can think of no other player that resembles Barry on the point. The objective is to find value, and to make money. If the wager should over-pay, Barry may play. If the wager should not overpay, Barry never plays. The trick is to estimate with a pronounced degree of precision what the payoff should be to render good value, the mark of expertise in handicapping. I imagine nobody does that as carefully and effectively as Barry.

Years ago, to cite a classic case, the pick six at Del Mar featured a seductive carryover, and Barry took the Amtrak from his spot in Anaheim to the station at Del Mar. Minutes later he arrived unannounced at the box.

It’s unfailingly fun to spend a day at the races with Barry, because not only is he smart and clever, but also he’s full of good humour. Knowing that Barry was chasing the carryover, I immediately pricked him by saying I had no interest in betting the pick six, but he countered by notifying me that that would not stop him.

Unknown to Barry at the moment, something else would stop him. In the box that day too were a handful of nationally renowned handicappers, each of them eager for the pick-six races to begin.

On the train ride, Barry had isolated the horses he liked in each of the six legs. As his custom, he had arrayed the numbers of the horses he liked in several six-race combinations, e.g. 2-5, 3-6-7, 8, 1-7-9, 3-5, 6-7 (72 combinations) such that his singles and A-horses (top choice in a spread race) would dominate each ticket. He had constructed as many as a dozen potentially playable tickets. Using his betting lines, Barry estimated the chances that each ticket would win. Using the morning lines, size of the pool, and the win probabilities he had attached to each ticket, Barry next estimated the number of winning tickets and the pick-six payoffs he could expect. The pick-six races looked formful to a fault, and following the numerous calculations, Barry announced he had found no bets.

If memory serves, Barry’s cost to play would have amounted to $1,200 or thereabouts, and he had estimated the pick-six payoff at no greater than $7,000 to $8,000, or less than 7-1; no value, no play. His colleagues in the box were not similarly deterred. One of Barry’s colleagues had invested $2,500 and when he hit the pick six, he was thrilled, except in short time the payoff was posted at $8,000, and the thrill was gone.

Barry’s commandment is to make a betting line. The betting line must be a 100 per cent line, which equals the chances of all the horses in the race, and Barry is insistent that the handicapper’s line correspond to 100 per cent, even if that means agonising on whether Contender C should be 4-1 or 9-2. Barry juggles his original line until he has reached between 99 and 100 per cent. He suggests to inexperienced linemakers this may take as long as 15 minutes to begin, while causing tension headaches, but soon enough the process can be completed pain-free in a couple of minutes.

In making a betting line, Barry is merely asking handicappers to quantify their opinions. Instead of saying, “I like 6 the best,” or “It’s a closely matched race among the 4, 7 and 9,” or “The 5 sticks out but he’ll probably be overbet,” or “As many as six horses can win here,” assign each of the horses a fair-value betting line by asking rhetorically how often each would win if the race were repeated a hundred times. Once the percentages have been assigned to each horse, sum them. The percentages should add up to 100. If they do not, juggle the line, adding and subtracting points for each horse, based upon the various handicapping factors that apply.

Once the 100-point line has been established, the next step is easy. The percentage chances are converted to fair-value odds, using the odds-percentage table. If Contender A has 38 points, the fair-value odds will be 8-5. If Contender B has 28 points, the fair-value odds will be 5-2. If Contender C has 18 points, the fair-value odds will be 9-2, and so forth. Barry urges handicappers to memorise the percentages and the corresponding fair-value odds, a sensible and elementary task.

The next step is crucial, and it’s Barry’s signature point. Handicappers should not bet at any time, unless they are getting a 50 per cent bonus. Contender A, with 38 points and a fair-value line of 8-5, must be 5-2. Contender B, with 28 points at 5-2, must be 7-2. Contender C, with 18 points at 9-2, must be 7-1. Fair value is converted to good value. Barry’s chart, “Win Overlays” (on the next page), presents the 50 per cent bonus line for horses having a handicapper’s fair-value line of 2-5 to 6-1. Memorise these adjusted lines now.

The 50 per cent bonus is not pulled from midair, but is subjectively biased. In Money Secrets At The Racetrack (TR Publishing 1987, 1990), his signature book, Barry reports that after studying thousands of his personal bets, he has concluded the 50 per cent bonus will guarantee profits of 8 to 10 per cent. Any smaller bonus percentage may convert profit to loss. This happens because:

(a) handicappers have a tendency to overestimate their horses’ winning chances, especially the low-priced contenders;

(b) late betting often lowers the odds to the player’s disadvantage; and

(c) the race will be susceptible to a relatively large error factor, which is not considered in setting the betting lines.

If handicappers have no access to the odds-percentage table, but have assigned the percentage chances on a 100 per cent betting line, fair-value odds are estimated easily with a calculator by dividing the percentage chance of each horse into 100, and subtracting 1. Contender A above has 38 percentage points, so 100 divided by 38 equals 2.63, minus 1, equals 1.63, or 8-5.

As demanding as Barry can be that handicappers must make a betting line, or they’ve probably been kidding themselves, he compromises to the extent that the line might be set for the authentic contenders only. Eliminate the horses having a nominal chance.

Instead of a 100-point line, now use an 80-point line, allowing the non-contenders a 20 per cent chance. In other words, the handicapper’s contenders should be winning four of five bettable races, a proficiency standard that can be examined, and should be.

Barry’s records show his non-contenders win 15 per cent of the races he bets. Handicappers without records can assume their non-contenders will win 20 per cent of the races. The weakness of the 80 per cent line, however, extends in undesirable ways to exacta wagering, a topic Barry covers exceedingly well, and which we address next.

Reiterating the importance of setting a betting line, Barry summarises his betting procedure succinctly:

  1. Analyse the chances of every horse in the race.
  2. Determine the fair-value odds using the odds-percentage table.
  3. Bet to win on any horse having odds 50 per cent higher than the betting line.

The betting lines accomplish all the important money-management objectives:

  1. They force handicappers to quantify their opinions, such that they can consider intelligently, not just one horse, but several horses, and how each relates to the others – make decisions, not selections.
  2. They allow handicappers to test abilities that can be measured with personal records across the seasons, e.g. do your even-money shots win half the time, 2-1 shots a third of the time, 4-1 shots 20 per cent of the time, 9-1 shots 10 per cent of the time; how much money do you make on horses that go at 5-2 and below; has your handicapping been improving, or is it in decline?
  3. Best of all, the line enables handicappers to spot the true overlays, and the horses and combinations by which bettors can win, not only in the straight pools, but also in double pools, exacta pools, quinella pools, and the pick-three, pick-four, pick-six pools.

Barry accomplishes something else, quite wonderful, in Money Secrets At The Racetrack. He presents an amazing array of charts that tell handicappers exactly how much they should bet in relation to the odds and what the cut-off points are for finding his 50 per cent overlay combinations in doubles, exactas, quinellas, and the serial bets. I counted 32 charts, and eight of them displayed pick-six cards and scenarios that featured various combinations of singles, A horses, and B horses (backup horses), how many combinations must be covered for each scenario, how to fill out the several betting cards exactly, leg by leg and horse by horse, and is easily the best, most accurate material for pick-six shoppers on record.

In his chapter “How Much to Bet,” a brief treatment of the Kelly Criterion, Barry explains the dangers of overly aggressive betting and reminds handicappers the optimal bet size is equal to the edge divided by the odds. If handicappers enjoy a 15 per cent edge when wagering to win (33 per cent winners at average odds of 5-2) and the odds on Contender A will be 3-1, the optimum bet is 5 per cent of the bankroll. To bet more is to overplay your edge, and risk ruin, and to bet less is to win less than you’re capable of winning.

That is, the Kelly Criterion guarantees the maximum growth of a bankroll, as no other money-management method does. But as Barry explains, the problem with the Kelly Criterion in pari-mutuel settings (racetracks) is that the edge is always subjective, not objective. Handicappers believe they have a 20 per cent edge in an exacta, but the handicapping has been less than perfect.

The correction is a fractional Kelly. Barry suggests a half-Kelly, such that a 15 per cent edge to win on a 3-1 shot is no longer an optimal bet of 5 per cent, but half of 5 per cent, or 2.5 per cent. Few handicappers have records that inform them of their betting edge to win, not to mention the edge in doubles, exactas, and the serial bets. An a priori edge when betting to win might be assumed at 15 per cent, if handicappers are confident they can win 33 per cent of their bets at an average odds of 5-2, an attainable goal for competent handicappers. Divide the 15 per cent by the odds to win on selections, and bet a 50 per cent Kelly. A 2-1 shot would warrant 3.75 per cent of the bankroll (15/2 x 0.50). A 5-1 shot would warrant 1.5 per cent of the bankroll (15/5 x 0.50). A 10-1 shot would warrant 0.75 per cent of the bankroll (15/10 x 0.50). This follows Kelly, as the percentage amount should decline as the odds grow greater.

Barry recommends a $2,000 bankroll for each of the several pari-mutuel pools – win, place-show, exactas, doubles, pick-threes – and his charts reveal the optimal bet sizes for the $2,000 bankroll at the various odds levels. Examine the Win-Bet Chart. It shows the optimum bet size where the handicappers line varies from 1-5 to 6-1 and the tote odds vary from 4-5 to 10-1. If handicappers believe a best bet should be 3-2, and the public offers 2-1, the optimal bet for a $2,000 bankroll is $26. To bet more is to risk converting profit to loss. To bet less is to win less than the handicapper is capable of winning. If an overlay is not listed on the chart, play the highest listed number, e.g. your 3-5 standout that is sent away at 8-5 deserves an $80 bet.

Barry’s discussion of place and show wagering is concise and excellent, and will have tremendous appeal to contemporary handicappers finding an unprecedented abundance of underlays in the win pools.

Underlays to win are often overlays to place and show, as the oversupply of win wagering is not reflected in the place and show pools, where the casual customers will be supporting their favourite longshots. As Barry tells, the crowd generally overbets longshots to place and underbets favourites and low-priced contenders to place. An overlay occurs when a horse’s percentage of the place pool is significantly less than its percentage of the win pool. A favourite that has 50 per cent of the win pool, but just 25 per cent of the place pool, and 20 per cent of show pool, is probably an overlay to place and show.

But is it? How seriously must a horse be underbet to place and show to qualify as an overlay? Barry provides the chart that tells. Using the crowd’s well-known efficiency when betting to win, Barry has estimated the maximum percentages of the place and show pools that expose the overlay opportunities. Examine the Place-Show Maximum Percentages Chart. The even-money favourite that has 25 per cent of the place pool can be bet to place, as that percentage is less than the 26.95 maximum percentage the horse can have to place. Similarly, the 1-1 favourite that has just 20 per cent of the show pool can be bet to show, as 20 per cent is less than the 23.99 maximum percentage.

Ignoring the win pools, if handicappers have a calculator, they can calculate the place-show percentages and compare them to Barry’s friendly charts. Overlays are identified quickly. Handicappers leave literally thousands of dollars untouched among their place and show overlays annually, as I discovered on my inaugural day at the races with Barry, at Santa Anita, in 1987. Overlays I covered to win in the first and second races were overlays to place and show as well, the first paying $6 to place and $5 to show, and the second paying $6.20 to place and $4 to show. Their percentages of the place and show pools fell significantly below Barry’s cut-off percentages, but I did not realise that. Handicappers should notice that Barry’s place-show percentages have been estimated for horses at 4-1 and below, and not for the mid-level and longer-priced horses that inevitably underpay to place and show.

NEXT MONTH: We continue our extracts from James Quinn’s marvellous book. The book is available from most Internet booksellers, including Daily Racing Form at www.drf.com.

Click here to read Part 2.

By James Quinn


PRACTICAL PUNTING – MARCH 2007