picking contenders -- how many?

[andicap]

I'm in a bit of a tussle over best way to pick contenders..not so much the method, of which I have several, but mainly, how many contenders to get down to. I know some authors say, you have as many contenders as you have. But I've had races where only 3 horses qualified under my standards, but the 4th best contender won. (average and larger size fields.)
Obviously it means my standards might be too tight, they have served me well at other times, eliminating lots of horses that lost.

I think Dave Schwarz might use half of field size plus 1 (or minus 1?)


Here are the methods, using different tpes of figures (speed, some pace figures in certain races or when the track is biased, or there's an extreme pace match-up. Or on the turf I usually won't use speed figs.)


1. Take 3rd best figure in last 90 days, subject 2 ticks for flexibility, errors, etc. and use that as a race "par." Any horse who in his last 3 races at the
distance/surface meets the par (preferably within 6 months and within 5 races overall) is a contender. Ex., top 3 Beyer figs in last 90 days are 93, 89 and 83. Horse must have 79 in his recent record to qualify.

2. Take 10 best figures (or in smaller field, half the qualified lines plus 1) from today's distance/surface over last 6 months. Horse can qualify up to 3 times. (prevents races in which one horse has 6 of top lines, but has other flaws.)

Like I said earlier, the problem I have is that sometimes I get only say 3 horses as contenders and my next highest horse -- the 4th best -- will win, in a field size of say, 8 because the "figure" horses have flaws on pace, form cycle, a bad trainer, terrible post position (outside post with

Sartin always recommended five contenders. Of course in shorter field sizes like 6 or 7, five contenders would be too many. More like 4.

Any opinions?

[IRISHLADSTABLE]

Hi Andicap,
I allways use Four,somethimes I'll get a scr and it becomes three. I've been burnt too many times over the yrs
by throwing out that 4th selection.In ex races or Tri races
I'll play a Part Wheel with my top selections and box the other 3.

Jimmy

[shots]

Originally posted by andicap
....the 4th best -- will win, in a field size of say, 8 because the "figure" horses have flaws on pace, form cycle, a bad trainer, terrible post position...

Any opinions? [/B]

Looks to me that you have the reasons why the 4th best should make the contender list.

[BillW]

Andi,

The question (before you examine your contender selection process) is: Is your 4th contender winning such that it is producing a profit? If not your contender selection process is working as is.

Bill

[lousycapperII]

Originally posted by andicap
I'm in a bit of a tussle over best way to pick contenders..

If the winner is in your top 4 contenders at least 80% of the time you should be at the optimal point. Then start your selection process.

-LCII

[Binder]

With Val 2
I have always "hid" down to the top 5
Total Enery and Primary Line Score horses
Meaning after picking my pace lines I eliminate
any hores thats not in the top 5 of those corrallaries

They are my contenders
I use the odds and the E/L screen to eliminate
down to the 2 horses I will bet

This works for me

[lousycapperII]

Originally posted by Binder
With Val 2

I use the odds and the E/L screen to eliminate
down to the 2 horses I will bet

This works for me

Are you saying you bet 2 horses in every race or is just one of them your final choice depending upon odds, et cetera?

-LCII

[dav4463]

I get burned more often by the fringe horses in exotics if I limit my trifectas to only the top 4 choices. Some races I may have one contender, sometimes two, three, or four, but I always rank the top 7 horses, using the 7th as a scratch alternate, and consider the 5th and 6th rated at high odds on the bottom half of exotics tickets.

[alysheba88]

80% of the time I use 4, other 20% is 3.

Am talking win betting only.

When I do play trifectas, I often use horses that I dont have on my win contender list (underneat of course)

[cj]

Andi,

I just pick as many horses as I think are true contenders, but with a catch. I use the amount of contenders I've selected heavily when I make an odds line for the race.

As an example, Steve Fierro's templates use 80% as a guideline for his contenders, only leaving 20% for the non-contenders. I do something different. I figure the random winning percentage for my contenders. Say I have 3 contenders in a 10 horse field, the percentage is 30. I add that to 100, then divide by two, and that is the percentage I assign to the contenders (65% in this case). I don't know why it works, but I tried it 2 years ago, and still use it today to lead me to overlays without overestimating the percentages of my contenders.

My biggest problem was when I'd only have like 2 contenders in a 12 horse field. In the past, I would have assigned them 80% of the odds, but now it would be only 58%, much more realistic in my opinion. With the former, I was almost guarenteed to be betting one of the two, and was not winning in those type races in the long run. I am now.

[andicap]

CJ
In other words, you now assign them 58% using the NEW method after assigning them 80% under your OLD methods, right?

thanks. very informative and helpful

[cj]

Exactly...I had to admit (not always easy) that I was kidding myself if I thought I could pick the winner 80% of the time with 2 or 3 contenders in big fields...sounds good in theory, but in practice, I was getting killed!

[Binder]

lousycapperII

Yes I only bet races when I.......

Have 5 contenders
I can bet 2 horses
That are in my top 3 or 4 horses on VAL 2
That fit my model on the "sticks" on the Early/Late Screen

I pass almost all the races that don't have
these factors

The exception is
I will bet other horses in races
that I follow from stable mail

I only really handicap races from Philly Park
so I bet some horses that I know and follow
:)

[MV McKee]

I think it should vary from race to race.
I use an exclusion method as a foundational step when I handicap. Over the long run this process eliminates between 31 and 32% of the runners. But I have had specific races in which no horse was eliminated in 11 or 12 horse fields, as well a races in which 8 or 9 horses were eliminated. Don't think there are any hard and fast rules regarding how may or what percentage of entrants to eliminate in a given race, or even a given type of race.
IMO, the key is to balance the percentage or number of entrants you want to eliminate and the percentage of potential winners you are willing to sacrifice.
In my case, I started with a process that eliminated <1% of the winners. The problem was that I was only eliminating @ 5% of the horses. As you work forwards (or backwards as the case may be) from that point you eventually reach a point of diminishing returns. By that I mean for every horse you eliminate as a contender, you start eliminating a greater percentage of winners. In my case, this began to occur around 32%. I probably could have pushed the contender elimination process to 35% or so, but then I would have been approaching a 10% "missed winners" rate. That may sound good on the surface, but the problem is that there is a strong tendancy to eliminate a greater percentage of your higher priced winners as you move forward.

[kitts]

Andi-

Played for many years with 5 contenders and now use only 4. Win bet only. I have been doing four long enough that it makes me a small profit so I am sticking with the Four Plan.

[freeneasy]

throw no contender out and with that keep track of the cost and and payoffs using all contenders in a trifecta box.

[Shacopate]

I think CJ has nailed it.

Especially in the day we're in, where logical first choices are overbet. Chances over-estimated.