Chalk

[gtapp]

With so many races at Northfield having short payouts I was curious on how you betting experts handle chalk. Do you skip races that have a 1 to 5 favorite? Bet them only in exotics?

I read* that you shouldn't bet anything that paid less than 5 to 2. (I assume that would be a straight bet)

Do you have a rule of thumb you use?

*Tom Ainslies complete guide to Harness Racing. A good book but old. I think the last edition was 1980

[cliff]

When faced with a "super chalk", if I can't play against it, I pass. The exception would be singling such a horse in a pick exotic (pick4, 5, or 6), and then only when I really like a value horse or two in the other legs.

As far as what price to play a horse, I'll play chalk when I consider it an overlay. For example, if I think a horse should be 3-5 and he is sitting at 8-5, then I will pull the trigger. If I think he should be even money, and he sits at 3-1, then it's a go. If I think there are only two horses in the race, and one is 4-5, and the other 5-2, I'll take my chances with the higher priced horse.

[thegiss]

I pretty much agree with Cliff.
My column this week in Harness Eye (out Friday) deals with identifying and beating false favorites.
Here is an excerpt:

We saw the problem of overbet favorites during my month-long Betting on Haiti To Win benefit. In 22 cards, I turned a profit (betting $2 per horse per race) just three times, despite some good handicapping.

So how do we avoid this trap? First and foremost, as always, limit and or vary your plays. If you don’t like a horse...or a race… or a price…our security officers are not going to drag you kicking and screaming to a window to force you to bet. Go get a coffee or study the next race a bit closer. Skip the race if there is something about it you do not like.

Second, we need to find favorites that are vulnerable. Now, by definition every favorite is vulnerable. That is why the actually run the race. Otherwise, we would just cut a check for 50% of the purse and giver it to the connections of the favored (rather than the winning) horse. Favorites win a lot of races, sure. But, even at Northfield, where over 42% of favorites win, that means 58% lose. So, if we find the favorites with a good chance to lose, we have found a good chance at a value play. And finding them is easier to do than you might think. Sort of. We simply deconstruct our handicapping—we take the same things we look for when we like a horse and see if any of those things give us a reason to make us a dislike that horse. Why is a particular horse the likely favorite? Is it his form? His post draw? His driver? Is it a change (an angle) from his last race or is it an expectation that he will race the same way as he did in his last start?

[letitrideman]

Excellent advice and thread. I am grappling with this a bit, as a neophyte, trying to form a strategy myself.

I currently won't bet anything less than 1-2 consciously. Of course one problem is that a lot of money comes in right before the gate closes. So there are times that a reasonable price fav, that I like a lot, drops to 1-5 or 1-9 on some huge bet made at the end and what had value now doesn't.

But forgetting about that "issue", a 1-5 fav has to score/win 5/6 ( 83.3% ) of the time just to break even on them overall. That's a pretty daunting number. Even 2-5 has to hit 5/7(71.4%).

So, as Keith said, you can just passola. Or look to find value in others as the overlay on the fav typically means "better than they should be" odds on any other contender.

Or, what I have started doing is looking for opps in the Show pool on your 2nd and 3rd favs - particularly if someone has dumped a lot on the favorite there as well. This only works IF there are other strong horses tho. Still having two shots to win - even if the chalk expectedly runs away with the race - is key here IMO.

I also have started doing what cliff ( and trackrat also suggested it to me last year ) does and key the huge chalk with the others I like in exotics. IMO, if there are only 2-3 quality horses in a race, and one has such short odds, that's a good opp to score some decent priced exacta, tri, pick3+ hits. Better to cash a $12 exacta ticket than risk 20 to win 1 on some overbet chalk.

I still get surprised how often "bargain overlays" can be found in the Show pool tho. I know many "vets" think its for old ladies and people playing "names", but in a way that is good as it creates opps for those not there just to throw darts at the board.

Besides waiting on my Win bets until almost gate, the other thing that I'm concentrating on is watching for opps in the Show pools.

[trackrat]

I have had some success lately using the favorite on top in my trifecta key wagers, with four or five horses underneath. A $1 trifecta key wager with the favorite on top and four horses underneath is $12, while five underneath is $20. To break even with these two examples, the tri would only have to pay $24 or $40 for the base $2 trifecta wager.

Lately, I've swallowed my pride and used Beau Blevins on the underneath but only from an inside post. He aint much of a versatile driver, but he follows the leader fairly well. Cashed with him driving second in the fifth race last Friday night in a trifecta that paid $254. With this strategy, you need to pick your spots.

[cliff]

Aug 11, 2010 17:34:50 GMT -5 letitrideman said:

Excellent advice and thread. I am grappling with this a bit, as a neophyte, trying to form a strategy myself.

Strategies are good. Handicapping is only part of the battle (sometimes the easiest part), wagering and bankroll management are the keys to beating the game.

I currently won't bet anything less than 1-2 consciously. Of course one problem is that a lot of money comes in right before the gate closes. So there are times that a reasonable price fav, that I like a lot, drops to 1-5 or 1-9 on some huge bet made at the end and what had value now doesn't.

But forgetting about that "issue", a 1-5 fav has to score/win 5/6 ( 83.3% ) of the time just to break even on them overall. That's a pretty daunting number. Even 2-5 has to hit 5/7(71.4%).

I proved to myself many years ago (statistically) that prohibitive favorites win at about the same rate as all favorites (generally about 40%).

So, as Keith said, you can just passola. Or look to find value in others as the overlay on the fav typically means "better than they should be" odds on any other contender.

Or, what I have started doing is looking for opps in the Show pool on your 2nd and 3rd favs - particularly if someone has dumped a lot on the favorite there as well. This only works IF there are other strong horses tho. Still having two shots to win - even if the chalk expectedly runs away with the race - is key here IMO.

I'll differ with you here. I haven't wagered the show pool in probably 20 years, and I can't think of a single scenario that would entice me to wager that way in the future. The reason is very basic...the payout is entirely dependent on what other horses come in, and thus how the pool is shared. If the heavy chalk hits the board (again from my statistical work, about 75% of the time), then the payouts on the other two will be affected adversely. Sure, if the chalk falls off the ticket as the example you cited in another thread, then the payouts are generous, but seriously, wouldn't you rather have the winner in that race?

The better strategy, imo, is to find the vulnerable favorites and play against them. There are many ways to do this, such as straight win bets if there is a horse you really like, or wheeling the other contenders over the favorite in an exacta or tri, boxing the other contenders, and so forth.

I also have started doing what cliff ( and trackrat also suggested it to me last year ) does and key the huge chalk with the others I like in exotics. IMO, if there are only 2-3 quality horses in a race, and one has such short odds, that's a good opp to score some decent priced exacta, tri, pick3+ hits. Better to cash a $12 exacta ticket than risk 20 to win 1 on some overbet chalk.

Indeed.

I still get surprised how often "bargain overlays" can be found in the Show pool tho. I know many "vets" think its for old ladies and people playing "names", but in a way that is good as it creates opps for those not there just to throw darts at the board.

Besides waiting on my Win bets until almost gate, the other thing that I'm concentrating on is watching for opps in the Show pools.

I already spoke to this, but let me put it another way...one cannot judge overlays from the show pool, because one doesn't know the payouts until after the race has been run. If you think the chalk is beatable, then by all means wager against him, but use the win pool as your guide.

[cliff]

Aug 11, 2010 20:35:25 GMT -5 trackrat said:

I have had some success lately using the favorite on top in my trifecta key wagers, with four or five horses underneath. A $1 trifecta key wager with the favorite on top and four horses underneath is $12, while five underneath is $20. To break even with these two examples, the tri would only have to pay $24 or $40 for the base $2 trifecta wager.

Lately, I've swallowed my pride and used Beau Blevins on the underneath but only from an inside post. He aint much of a versatile driver, but he follows the leader fairly well. Cashed with him driving second in the fifth race last Friday night in a trifecta that paid $254. With this strategy, you need to pick your spots.

I'm primarily a win, and pick (3,4,5,6) player, but when I do play trifectas and supers, I always wheel the best other five horses behind my pick. I consider myself a fair handicapper, but for the life of me I haven't found a reliable way of telling in what order the other contenders might complete the ticket.

Example, if you figure there are only two possible winners in a race, you might play something like ab/ab/cdef. What happens in that case is that either a or b come first over, tires and falls off the ticket. The same thing happens when you exacta box a and b. The way I play is a/bcdef, then if I'm in the mood to hedge I'll play a small win ticket on b, enough to break even on the race.

If I think I have the winner and he happens to be a value, then I'll key him in the exotic and also place a win wager. Too many times I've picked a nice $12 winner only to have a 40-1 suck along the rail the whole race and blow up my tri or super ticket. Always have something on him to win.

[thegiss]

Cliff- with net pool pricing, the show pool is no longer "entirely" dependent on what other horses come in. I can't release excerpts yet, becauise it is current in Harness Eye, but my August Column, Show Me The money, identifies several scenarios where a show wager does make sense.

[letitrideman]

Some interesting comments Cliff.

I'm a big stats guy typically, so I'm curious on your statement that short favs don't win any more than all favs ( 40% of the time ). That sounds rather implausible to me. While I doubt a 1-9 fav wins enough to make blindly backing them a long-term sound strategy, I'd be shocked if they didn't win well over 40% of the time they run. Maybe not the >90% they'd need to ( to stay in the black ) but it would really be surprising to me if they didn't win over say 80%, at least, long-term.


While probably not a perfect corrolation by any means, I'd expect that the odds do correspond loosely to the actual winning %'s a great deal, over time. A 1-9 fav winning more than a 1-5 than a 2-5 more than a 1-2, 3-5, etc. down the line. If it was 40% regardless, then any horse at above 3-2 odds would have value and anything under that would not. I doubt you are proposing it is that simple! If the fav wins at a 40% clip regardless of its odds it would be a surprisingly easy prop to beat long-term.


My "good guess" is that the actual win %'s by odds largely correspond to the odds minus the vig/take of the track. Say the odds are even-money on the fav at the bell. If the vig is 25%, then the odds should be 5-4( if their take/vig were not being subtracted, so my guess is the actual winnig % on an even money fav is closer to 44.4%. Not the 50% needed to break-even, but better than 40%. For a 1-2 horse the odds should be like 5-8, so I'd guess they win in the 61.5% range or so. A 2-1 fav should be 5-2, so in the 28.6% range I'd guess.


As for your eschewing the Show pool because it is dependent on what horses come in, I'm not sure I follow you. While I know that a disproportionately-backed fav can create negative-pool conditions for the track, that isn't an every-race condition. Most races you can certainly judge how much a horse will pay by the relative size of their portion of the pool. I grant you it isn't always easy to "approximate" in your head, but it isn't magic either. I'm curous to see Keith's article on it, but I am a newbie and I can tell when a horse is underbet or over bet in the Show pool.


Once or twice a racenight I'll ID an opp where a fav or 2nd fav will pay close to as much in the Show Pool, with the cushion of 3 times the chances of cashing, than they will in the Win pool. IMO, when they are relatively close, there is FAR MORE VALUE in the Show Bet. So many focus entirely on the win pool ( for which the track tells you the current proportions by the odds so you don't have to figure it out ) - tho it is easy as you take the total pool minus vig and then divide by the $2 tickets sold on that horse. Say the Win handle is 4000, 25% vig and 1500 bet on a horse. 3000/750 =4.00 so even money. the Show pool works the exact same way except you divide the handle minus vig by 3 ( unless there is a negative pool situation ). If that same horse has say only 276 bet on it out of 2000 in the Show pool. It would pay 1500/3/138=3.60. Doesn't it make more sense to have 3 shots at a 3.60 payout than 1 at a 4.00? And if you look at the payout tables, this kind of scenario does happen here and there. It is even more likely when looking at the 2nd fav. They are sometimes disproportionally lowly backed in the Show Pool compared to the Win Pool. These opps offer true value, IMO. Backing your 2nd fav ( or first if u feel the huge fav is vulnerable ) for a bridge-jump in the Show pool is certainly a good bet ( even if the fav does hit the board you had a shot at huge prices for minimal risk ). But that isn't the only opp out there! I don't frequent the track that often and I've scored a half dozen higher paying Show tickets than win tickets would have on the same horse.

[letitrideman]

Cliff peaked my curiosity, so I took an hour to manually calculate the fav stats for last month ( August ) at NFD.

I know it is far from a thorough and statistically solid sample set, but I just looked at the last month's results to glean some info about favs and hit rates. I'm going to speculate that what I found falls more in line with what I'd expect than what Cliff has stated.

In 234 races from August, I have favs winning 128 times for a 54.7% overall mark, so decently above the 40% mark cliff had projected.

For those favs below 1-2 ( .05 - .4 ) they were 25-6 or 80.6%
For those favs from 1-2 to below even money ( .5 to .9 ) they were 56-31 or 64.4%
For those favs even money to below 3-2 ( 1 to 1.4 ) they were 31-31 or 50%
For those favs from 3-2 to below 2-1 ( 1.5 - 1.9 ) they were 12-26 or 31.6%
For those favs above 2-1 ( 2 - 3 ) they were 4-12 or 25%

Pretty much stratified as I'd expect!
What I didn't quite expect is that playing all favs blindly would have actually netted you 109.05 units won - 106 lost for a tiny profit on the month, I'd think that is slightly unusual. I didn't account completely for dead heats tho ( and I know there were at least a couple on days I was there ) so it is probably a little under break-even in reality.

Also interesting that in the sweetspot of ( my 1-2 low limit and 1.3-1 ) you'd be 85-57 60% with 71 units won minus the 57 = 14 units profit. 14/142 is just under 10% ROI which is not bad. So betting favs above 1.3-1 is definitely negative EV blindly, but in that sweetspot it is a positive EV - for this small sample it was at least!

CONCLUSION: Your capping may vary. LOL But overall, it was not a bad idea to play the chalk in August at least.

While it may have been a little higher for favs than avg, if I get some time I may pick another month to analyze - maybe something earlier in the year randomly, for comparison sake.

[letitrideman]

In 221 races from April, I have favs winning 93 times for a 42.1% overall mark, more in line with Cliff's prediction. I note though that there are a lot more > even money favs in that month than in August. So it is probable that August was a low-odds-favs month and thus why more of them hit the winner's circle %-wise.

For APRIL:
For those favs below 1-2 ( .05 - .4 ) they were 15-3 or 83.3%
For those favs from 1-2 to below even money ( .5 to .9 ) they were 45-38 or 54.2% (10% worse than Aug )
For those favs even money to below 3-2 ( 1 to 1.4 ) they were 25-36 or 41% (9% worse than Aug)
For those favs from 3-2 to below 2-1 ( 1.5 - 1.9 ) they were 7-42 or 14.3%(a whopping 17% worse)
For those favs above 1.9-1 ( 2 - 3 ) they were 1-9 or 10%

So still clearly stratified and I feel safe in saying that a 40% ATB hit rate just doesn't pass statistical analysis. But OVERALL it is probably close as August appears to have been a particularly hot month for favs.

CONCLUSION:
April's numbers are more in line with my predictions of the favs hitting pretty much 25% below their odds overall, and stratified down from lower to higher proportionally.

One thing this has me thinking about is "watching it" when I like a fav that is higher than 1.3-1 in odds. They are an abysmal 30-104 combined for those two months and seem to be an "avoid" situation. So when WIN betting on favs I'll be more conscious of staying within the "sweet spot" range of 1-2 through 6-5. Below that is too little reward for the risk, IMO, and above that is bucking some bad historical hit rates, it appears!

[cliff]

Sept 1, 2010 18:43:50 GMT -5 letitrideman said:

As for your eschewing the Show pool because it is dependent on what horses come in, I'm not sure I follow you. While I know that a disproportionately-backed fav can create negative-pool conditions for the track, that isn't an every-race condition. Most races you can certainly judge how much a horse will pay by the relative size of their portion of the pool.

Not exactly correct. While the portion of the pool that is bet on your horse is known, you cannot tell how much will be distributed to your horse until the race is over and the three top finishers are known. You cannot calculate the return the same way as the win pool because the winnings are split three ways after the betters' original wagers are returned.

Redo your calculations with this method and you will see how the payout changes depending on which horses finish in the money with your horse.

I'll grant you that there are times when a favorite is "under valued" in the show pool relative to the win pool. So, if you are quick with numbers, and the relationship doesn't change much as the horses go off (not a good assumption these days with all the simulcast money that is wagered in the final seconds), you can as you say "have 3 shots at $3.60 rather than 1 at $4.00". I don't believe that is typical, but rather the reverse is true, i.e., the win pool is under valued relative to the place and show pools.

[jimhorseman]

Letitrideman:

Love all the time and effort you took to do the August and April analysis. I never really thought about rating probability of winning in relation to how heavily favored the horse was. As Cliff just pointed out a high percentage of the wagers are not recorded till the final seconds before a race goes off. Too many times I thought I made a good win wager on a horse at maybe 2-1 only to see the odds change to 6-5 at the quarter pole. Sadly it seems that favorites now dominate Northfield due to the declining quality of stock and drivers. Thus finding playable favorites (as win bets or heading exotic bets) will be crucial to any near-term Northfield wagering success.

[letitrideman]

Those factors ( late $ and negative-price-pools ) are just as relevant in win pool betting as well. So I see no valid arguments being made for why show pool betting is inherently bad, or should be ignored when a value bet can be found. It's cool if you don't care to bet any bet type, but that doesn't mean value can't be found.

In all sports betting there are two main factors, handicapping and betting. Some are good at one and not the other. The betting side can make a huge difference tho. Money management, ROI, % of bankroll, etc. are often overlooked aspects. I know a fair amount about the betting side in other sports. But horses and parimutual are new to me. Lots to learn!

Good Luck!

[wire2wire]

I appreciate the effort in the sta calculating as well but I like to think of it this way...show betting is for cowards (please don't take offense) and not a way to make a decent profit worth a trip to the track or the power of turning on your PC. I will say If I really like a horse to win I will typically place $100 W and $100 to place, and back myself up with a tri ticket that would look like 1234/1234/the horse I like. sometimes a $32 Tri back up can pay off quite nicely especially if the horse you like is chalk.

If you bet $20 show on a favorite that pays $2.40 to show you only make $4.00, the risk/reward factor isn't there. And from what I hear $4.00 couldn't buy you a beer at Northfield ;D