Forums
Welcome to Live View – Take the tour to learn more
Start Tour
There is currently 1 person viewing this thread.
Wessex
13 Jun 17 11:54
Joined:
Date Joined: 10 Nov 01
| Topic/replies: 249 | Blogger: Wessex's blog
Can someone explain to me why golf win and top 5 prices seem out of kilter?

I have had this question for a while but some of US Open prices illustrate the point starkly.

As an example, Spieth is trading in the win market at 16.5-17.0.  Taking this market as the closest proxy you can find to a true price (there are 62 golfers backable at 1000 which takes out 6% and the total overround is 108%), let’s say his ‘true’ win odds are the mid-point, i.e. 16.75, an implied probability of 5.97%.  If Spieth does not win, and all dynamics of the contest remain the same, surely his chance of coming second can be no less than 5.97%, as he his price could only be shorter in a ‘without the winner’ market.  Applying this logic to 5 places, his chance of finishing in the top 5 would be no less than 5 x 5.97%, i.e. 29.85%.  This would imply a top 5 price of 3.35, yet he is backable in the top 5 market at 5.1.  This applies to numerous golfers in the current market, implying that the top 5 prices (and I assume the other top X prices) for these golfers are over what might be considered the ‘true’ price.
Applying a different approach, if you multiply the implied chance of Spieth not winning (94%) by itself 5 times you get 73.5%.  This would therefore be the probability of Spieth not being in the top 5, which implies a chance of 26.5% chance of him being in the top 5, equivalent to a price of 3.77, still well under the current top 5 price of 5.1.

I appreciate that golfers are human and they might have a different approach if they know they can’t win, making their ‘true’ prices bigger for the other places, but surely this should apply less for a major when there is very serious money, prestige and lots of ranking points available.

I would greatly welcome any comments on this, in particular reasons why my application of maths might be flawed.
Pause Switch to Standard View Inconsistencies in win and top 5 prices
Show More
Loading...
Report therhino June 13, 2017 11:21 AM BST
From a maths perspective, you haven't calculated for the likelihood of a dead heat. In the winner market you are guaranteed an outright result, if Spieth or whoever wins you win 100% of your return no matter what. In the top 5/10/20 markets those results can be split with x amount of other winners in the absence of an outright result, with every tie diluting your return to the point a win can actually be a loss. Prices need extra juice to account for that.

Then there is the humanity factor which you have alluded to - i.e taking on the Matt Kuchar's of the world in the winner market is a far different beast than taking him on in a top 10 market. Completely separate markets, should be used to calculate how you perceive value in either (or at worst use very loosely).

HTH.
Report donny osmond June 13, 2017 11:59 AM BST
i think its more related to their expected score, and where that places them in the order
Report tedstriker June 13, 2017 2:57 PM BST
I don't think the dead heat effect is important. Yes there can be reduced payouts, but when there are more players get a payout, the two factors exactly cancel in financial terms, though arguably there may be a slight advantage to the lesser players over the stronger players.

The reason for the apparent value is because quite simply the elite players win far more when in contention. Using the OWGR site look at number of wins and compare it to number of times 4th - 10th for any player. If top 10 places were distributed evenly you should expect the ratio to be about 1:7. Picking 3 random players the numbers are as follows;  Spieth 12, 26  Moore  5, 46  Na 3, 51. So basically when Spieth gets into contention he converts wins far more often than the lesser players. This is what is being reflected in the odds.

Having said that, when the field is very strong as in a Major, you might expect 3-4 elite players in contention come Sunday, so perhaps this effect ought to be less than for regular events as they cannot all win, and therefore the Top 5 price may indeed represent better value than the win price.
Report El Bandido June 13, 2017 4:03 PM BST
If Spieth does not win, and all dynamics of the contest remain the same, surely his chance of coming second can be no less than 5.97%.

I see no reason why. Suppose a player had a win probability of 25%, by this logic, his top 5 probability would be at least 125%, which is clearly nonsense.

Applying a different approach, if you multiply the implied chance of Spieth not winning (94%) by itself 5 times you get 73.5%. This would therefore be the probability of Spieth not being in the top 5

LOL. This is just plain nonsense! (For a start, if you did this with all players in the field, it would add up to way more than 500%)


There just isn't a simple way to convert win probabilities to top 5 probabilities. There might be some better ways, and some worse ways. But in the end, it's 2 different things you are trying to price up. Some players are more streaky while others are steady eddies, and that's going to affect the relative win/top 5 probabilities.
Report mr meagi June 14, 2017 9:45 AM BST
Thanks for the question Wessex, if only for getting El Bandido to post(I could read EB all day)

Here's my tuppence worth. It's a case of 'Wellies and Icecream'....Confused.....ie no sensible mathematical relevence here.

Let's say Rose is an Ice Cream and Lowry is a Wellie. PP and BS will not be the places to buy wellies. Do not go to BF or WH to buy icecream. Do the opposite. Now if Dustin is a pair of shoes.......the price runs standard....ie try and rip people off on the fav. That is what  bookmakers do........They sell you what you want for the least value that you will tolerate. They make a book for themselves. They have highly calibrated 'instincts' for this task that defy rudimentary mathematical slide rules. They work off a perceived consensus where the litmus test is the flow rate of turnover on any one participant. They respond to 'demand' and to 'disinterest' There is no way that Dustin should be less than 10/1  but why give that when you can sell him at 8/1. Don't be silly.

Is Spieth really 16/1 here.....I think that seems reasonable........Is he 1/5 eight places at that price ....definitely yes.....Is he a 4/1 top five without the chance of the big cash....definitely not...........Top 5 prices are a 'stay away ' every week...........Look at Horschel last week.......he didn't finish top 5 ...............he fell into it by accident, ... Good Luck to All!

(Can't wait for my Pure Luck Picks this week, but no peeping until 12:00)
Report fish in chips June 14, 2017 5:05 PM BST
Looking at career stats for some of the top players (DJ, Speith, Rory, Mickelson, Day) they all seem to have won a significantly large % of tournaments when finishing in the top 10.  Between 18%-22% of their top 10 finishes are wins. Their numbers of 2nd place finishes are mostly lower, and 3rd place finishes lower still. Doesn't look like their odds of finishing in each of the 2nd-5th places should be as low as their odds on winning. I guess when they're in contention, they're often quite far ahead or better at closing out, so they don't finish just short too often.
Post Your Reply
<CTRL+Enter> to submit
Please login to post a reply.

Wonder

Instance ID: 13539
www.betfair.com