Showing posts with label Steve Davidowitz. Show all posts
Showing posts with label Steve Davidowitz. Show all posts

  • The Truth About Progression Betting

    POSTED Jul 11, 2014 By Derek Simon
    Before the more scholarly handicapping authors like Tom Ainslie, Andrew Beyer and Steve Davidowitz came along in the 1970s and 1980s, most racing literature featured two prevailing qualities:

    1) It was poorly written.
    2) It was gobbledygook, often espousing theories only slightly crazier than those proffered by the Flat Earth Society.

    Take, for example, a recent gem I came across entitled “Picking the Winners with Systology.” Published in 1933, “Picking Winners with Systology” is a compilation of selection and betting methods that, according to the (wisely) unnamed author or authors, “fills and completes the cycle of information which has long been a pressing want of turf players.”

    “Using any one of the eight workable and tested systems contained herein, players may be sure that only misfortune and the failure of horseflesh and blood to do that which could be confidently expected and figured, will result in a loss,” the book claims.

    So, in other words, if you lose, blame the horse.

    Still, although “Picking Winners with Systology” is rife with crazy notions that the more sophisticated players of today could easily spot — how about a parallel speed chart that equates five furlongs in 56 seconds with six furlongs in 1:09? — the book also contains some ideas that aren’t as easily dismissed (although they should be).

    I fear that far too many bettors still believe that progression methods actually work… because, theoretically, they do.

    For instance, the Quadru Methods detailed in “Picking Winners with Systology” combine several fairly simple and straightforward selection approaches — none of which are profitable, by the way — with a well-known progression betting system.

    “Play is based upon the sum necessary to recoup losses and show profits. If the choices in three races have lost, the sum lost, plus profits sought, are added together and the player wagers enough to bring back that total. Thus, if the player had lost $12 on three previous races and wishes to show a profit of $10, he would wager $6, presuming that his choice in the fourth race was quoted at odds of 4 to 1,” the book notes.

    On the subject of odds, the player is warned that he “should never back a horse at less than even money.” (Presumably, this is to keep the amount wagered in check.)

    So, with all this in mind, I decided to run a simple test.

    Using my database of more than 14,000 races run during 2012-13, I first looked at the stats on sole favorites (no entries or ties) that went to post at even odds or greater. I chose race favorites for my test for two reasons: 1) They win more often than most — if not all — one-factor methods on the planet, thereby providing the consistency that progression betting systems need; and 2) They produce an ROI similar to, if not in excess of, what the methods outlined in the book can achieve.

    SOLE FAVORITES AT EVEN ODDS OR GREATER

    Number: 10,451
    Winners: 3,303
    Win Rate: 31.6%
    $2 Net Return: $1.66
    ROI: -17.01%

    Now comes the fun part: Using the same horses listed above, I sought to make a profit of $10 per race. If I lost, the $10 was added to the amount wagered, along with another $10 in desired profits. All bets were rounded down to the nearest dollar and, of course, the $2 minimum bet was always observed.

    The overall results were, well, fantastic. As I mentioned previously, in theory, progression betting works. After 10,451 bets, profits stood at $99,808.70 — a sizable chunk of change.

    However, before you rush to cash out your 401(k) and start betting progressively on favorites, take a stab at what the maximum wager amount in this sequence of 10,451 bets was?

    If you guessed $403,080, give yourself a gold star and move to the head of the class.

    And it gets worse.

    After accumulating a little over $39K in profits in 4,265 races, the wheels came off, as 23 straight losses led to a capital deficit of — are you sitting down? — $726,242.65. Add that to the subsequent $403,080 bet to recoup profit and losses and it becomes apparent (at least to those with a calculator) that one would have needed a bankroll in excess of $1.1 million to proceed with this progression.

    What’s more, my test doesn’t even begin to address what a $403K bet would do to the pools at most racetracks in the country. In fact, in a comical twist, the horse that broke the 23-race skid in my test was Gin Shot and he won the 7th race at Mt. Pleasant Meadows, a track that no longer exists.

    There was $456 in the win, place and show pools combined.

    So, the next time you read about a progression betting system guaranteed to show profits, go search for Bigfoot or the Loch Ness monster instead. It’ll be about as fruitful — not to mention a whole lot less expensive.

  • The ‘Key Race’ and the Kentucky Derby

    POSTED Apr 20, 2014 By Derek Simon
    Many moons ago, in his master work “Betting Thoroughbreds,” Steve Davidowitz coined the term “key race” to describe a race that featured an inordinate number of next-out winners. According to Davidowitz, these winners are generally not coincidental. 

    “Either [the key race] was superior to the designated class or else it contained an unusually fit group of horses. In either case, that’s important information,” the author wrote.

    Apparently the Daily Racing Form agreed, as that esteemed publication soon began italicizing the names of next-out winners in its result charts and past performances.

    Yet, by its very nature, the Key Race Method suffers from one very large and significant drawback: typically, by the time an event can confidently be deemed a key race it has lost its value as a predictive tool. After all, what good is it to discover a particularly strong race after half a dozen horses have already won their next start? Not only that, but given how infrequently horses compete today, it can take many weeks or even months to determine whether a past contest qualifies as a key race.

    Hence, I decided to come up with a method of assessing key races that doesn’t require validation after the fact. Like Davidowitz’s initial technique, however, I wanted to keep it simple, so that even novice horseplayers could judge the merits of a particular race with just a modicum of time and effort.

    Here’s how it works:

    A)  Using the result chart from a horse’s last race, find the median finishing position for all the entrants in their prior race. This information can be found in the leftmost column following the (abbreviated) track name. For example, by examining the chart below, one will discover that On Lockdown finished fifth in his last race, which was run at Oaklawn Park (OP) on March 15, 2012.

    Note: For those who don’t have kids in school, the median is simply the middle value of an ordered array of numbers. If the array is even, it is the average of the two values closest to the middle.  

    B)  Divide the number of entrants, or the field size, by the figure obtained above to get the key race Rating.

    (Click on image to enlarge)

    That’s all there is to it; the higher the rating, the better the race was for the class. Notice I said “for the class.” Keep in mind that these key race ratings need to be viewed in light of the overall level of the race. A high rating in a $15,000 claiming event does not make a horse a contender against a rival that earned a much lower figure in a Grade I affair — it’s just common sense.  However, in races featuring horses that last raced against similar competition, the ratings can be invaluable.

    For the Kentucky Derby, I came up with a grading system (below) to address this problem:


    The last horse to win the Kentucky Derby after exiting a race with an “F” grade was Strike the Gold in 1991.

    Here are the grades of the major preps in 2014:

    (Click on image to enlarge)

  • Bogus Bias

    POSTED Jan 23, 2014 By Derek Simon
    John F. Kennedy once said that too often we “enjoy the comfort of opinion without the discomfort of thought.”

    Clearly, President Kennedy was talking about track biases. For few other subjects evoke as much hyperbole and outright nonsense than the notion of a track bias, a term popularized by Steve Davidowitz in “Betting Thoroughbreds.”

    According to Davidowitz, “Every racetrack has its peculiarities. Some are small in circumference, some have pasteboard-hard running surfaces, some card races that place a premium on early speed or post position.

    “And nothing can help to change or create a bias as effectively as a shift in weather conditions. A sudden rainstorm is odds-on to force a premium on early speed. On the other hand, a few days of rain, a sudden frost, or extreme heat can have totally unpredictable effects.”

    In order to spot a track bias, Davidowitz advises players to do four things:

    1. Watch the turns.
    2. Watch the break from the gate.
    3. Watch the run to the first turn (especially in route races).
    4. Watch the top jockeys.

    If this is not possible, Davidowitz adds that a “careful reading” of the results charts can also lead to the discovery of a track bias.

    Given this ambiguity, it is easy to see why “track bias” has become the modern-day equivalent of “cry wolf.” Tell people to search for something, especially something hard to define, and, by golly, they’ll find it!

    In fact, I got a chuckle out of the reactions to an article written by Daily Racing Form columnist Mike Watchmaker entitled “Talking Track Bias,” in which Watchmaker dared to say there was no speed bias on Belmont Stakes Day this past June.

    Those commenting on the piece couldn’t seem to agree on what a speed bias is, much less whether or not one existed:

    “Mr. Watchmaker fails to mention days when speed wins almost every race and other days where a front runner can't [find] the winner’s circle. This holds true for turf courses as well and, in both cases, certain racetracks are conformed and constituted in such a way as to favor a particular type of runner. … One of the reasons this wonderful sport has not caught on as well with later generations is the repetitive sight of wire-to-wire winners.”

    “… Go watch where all the winning closers came from, at least 5 to 6 lanes out. Watch how many horses two lanes in were getting caught.”

    “Yes, there was a bias. Orb's trainer said as much when he compared the surfaces between Churchill and Belmont. One was like a springboard, the other was looser and sandier.”

    “Bias is such a big and important topic that it may be worth writing a book about. Right off the bat most people confuse rail bias with speed bias. A hot rail is caused by track maintenance moving dirt away from the rail in anticipation of bad weather (because they can't grade a muddy track, and the rain, due to the slope of the track, will gradually move dirt from the outside to the inside). The thinner surface at the rail makes it fast. Usually people call this a speed bias, because the horses up front have the inside path, but a closer can win on it just as easily.”

    “Most of [the] horses that closed were horses with speed except, I believe, one race — if my memory is correct.”

    “You have to know if horses are holding on much deeper into races than you thought they had a right to if everything was equal, or that horses aren’t making the late moves you figured they should have.”

    “There was something a little different about Belmont on Saturday. I initially picked Golden Soul on top, but after watching three dirt races, you can see that front runners and stalkers were holding on all the way to the end. Closers weren't even plodding up for show money.”

    Well, that’s as clear as mud. So maybe the track was biased on Belmont Stakes Day, maybe it wasn’t — kind of like how I “may have won” a million dollars from Publisher’s Clearing House?

    Of course, I've heard the same gobbledygook from those who believed that Princess of Sylmar was done in by a “speed-favoring” track at Santa Anita in the Breeders’ Cup Distaff.


    (Click on image to enlarge)

    The fact that Princess of Sylmar was beaten by a whopping 16 ¼ lengths in the Distaff is overlooked by the bias boosters. The fact that Royal Delta, the champion mare “Princess” edged in the Beldame, was actually closer to the pace than Beholder in the Distaff? Summarily dismissed.

    Instead, those who cry bias point to Goldencents, the impressive winner of the Breeders’ Cup Dirt Mile.

    A Facebook friend of mine summed up the feelings I know many share regarding the speedy colt when he noted that Goldencents’ triumph on Breeders’ Cup Day “made perfect sense, but not after being hung out wide and running a savage pace.

    “Yet he kept going. The track carried him. Even if he tired a bit despite the bias it would have [been] understandable after that pace and trip. When he came back on a honest track, he could not duplicate that huge race.”

    True, in the Cigar Mile 29 days later, Goldencents finished a well-beaten seventh. However, that’s not the whole story. What my friend didn’t say is that Goldencents was bumped hard at the break of that race and, consequently, was third at the opening call.

    (Click on image to enlarge)

    This is of no small consequence, as Goldencents has never won when he’s been worse than second at the first call. In fact, the son of Into Mischief is 0-for-4 in such instances. He is 2-for-4 when running second at the opening call and 3-for-5 when leading.

    Bias? I think not.

    Something tells me JFK would agree with me.