Worth checking out on DB?

[jackad]

I happened to come across a particular configuration of four factors that showed the following results over 51 races in which it was found:
Bet to win: 47% won; ROI 30.3%
Bet to place: 72.5% won; ROI 19.2%

Is someone willing to check this out with a database? Here's the one rule: Must be Rank 1 for VEL, (K), and EFF, but NOT Rank 1 for PER. That's it.

Thanks.
Jack

[Carl]

Doesn't look like I have EFF ranked on my download Jack, so I can't.

Mike or Hurrikane may be able to though, they build their own rankers I think.

later,
Carl

[MikeDee]

had some stuff I was playing with so gave your play a try. fast dirt races with any races with first time starters excluded. rk=1 and rvel=1 and reff=1 and rper not = 1

Trk nWin nBet Win% aMut aBet aWon aP/L ROI
PHA 42 128 32.81% $4.86 $256 $204 ($52) -20%
TDN 38 115 33.04% $5.37 $230 $204 ($26) -11%
WO 33 108 30.56% $4.67 $216 $154 ($62) -29%
PEN 25 66 37.88% $4.80 $132 $120 ($12) -9%
DEL 25 78 32.05% $5.52 $156 $138 ($18) -12%
FE 25 91 27.47% $5.48 $182 $137 ($45) -25%
GLD 24 84 28.57% $5.00 $168 $120 ($48) -29%
RD 21 60 35.00% $4.52 $120 $95 ($25) -21%
RKM 17 65 26.15% $5.47 $130 $93 ($37) -28%
MTH 15 46 32.61% $4.93 $92 $74 ($18) -20%
PRM 12 44 27.27% $5.25 $88 $63 ($25) -28%
LAD 10 40 25.00% $5.60 $80 $56 ($24) -30%
MNR 10 50 20.00% $5.10 $100 $51 ($49) -49%
EVD 9 54 16.67% $7.67 $108 $69 ($39) -36%
SAR 4 14 28.57% $5.00 $28 $20 ($8) -29%
310 1043 30% $2,086 $1,598 ($488) -23%

[jackad]

Thanks MikeDee. Now I won't spend any more time tediously checking this combination one race at a time. It's obviously a no-go.
Jack

[hurrikane]

can't remember and my *%&$* computer with my database is having a cow right now. (good thing we all back up our data, isn't it). Seems to me there were great differences in EFF results in relation to distance. Perhaps running this seperately for spr and rtes. Also...the 2 or 3 tracks that show a low neg could probably be tweeked by odds or field size to show a nice profit.

[pizzaman]

there are many combinations of category rankings that give consistent flat bet positive roi, but have the disadvantage of occuring on average only about once a day.

for example, htr=1, f1=1, pscan = bottom third (that is: 5 or 6 in a 6 horse field, 7-9 in a nine horse field) produces a +.36 roi and 24% win rate (335 times in last year or so). Eliminating post time short prices less than 5:2 reduces the win rate to 22% but increases the roi to +.61 (237 plays). This combination works for both sprints and routes and dirt and turf.

[Glen]

Pizzaman appears out of lurk mode and delivers another quality post. Good to see you out and about!

[Glen]

BTW, the only factor I have no idea how to use is the EFF factor. Kinda reminds me of Hush Puppies at Red Lobster. You go to Red Lobster for the fine seafood selection. Then as a filler side item they give you hush puppies to surround the juicy fish product specimen. EFF is HTR version of Red Lobster Hush Puppies...Gimmie one vote to nuke factor EFF. :)

Actually I'm sure Eff has some use, just haven't figured it out yet.

[Carl]

I use EFF in one spot play as ">54". I am quite happy with it here, as it out performs the "<=54" horseys continually in this particular context.

And no I don't know what it "means". But then again, I don't know what a lot of HTR factors mean. And I don't know why my car starts when I turn the key.......

later,
Carl

[Glen]

Carl,

Your thinking of %EE aka Early Energy. EFF is Effort or Consistancy of the animal.

[Carl]

Well then, as Gilda used to say,

Never Mind!

later,
Carl

[hurrikane]

Carl,
the reason your car starts when you turn the key is because you put gas in it yesterday. :D

[HarryH]

I ran stats on Pizzaman's spot play (small dataset, about 6 weeks) with the addition
of nJTW% >20. got only 23 bets, but win% about 30% and ROI% about 95% I probably
would not be able to stand the boredom, but the numbers look good.

[Ron]

HarryH;

Just curious;.. about how many cards and tracks did you have to look at
during your 6 week study to find those 23 plays?


Ron

[HarryH]

I examined 8627 cards over 53 days; his original specs for the spot play came up with

49 bets 20.4%win 10.3%ROI Avewin $10.80
adding Caveat JTW%>24:
18 bets 33.3%win 128%ROI Avewin $13.68
or adding Caveat JTW%>20:
23 bets 30.4%win 95%ROI Avewin $12.84
using Caveat JTW%>20 AND MLOdds>3:
16 bets 31.3%win 145%ROI Avewin $15.70
using Caveat JTW%>28 AND MLOdds>3:
7 bets 28.6%win 187%ROI Avewin $20.15
Maybe that's going too far.....
The spot play seemed to do better with sprints:

using Caveat JTW%>20 AND MLOdds>3 AND nDIST<8F:
11 bets 36.4%win 195%ROI Avewin $16.27 !!!

One of the problems with this type of spot play, is that a few big hits can run the ROI way up; also, you are cranking maybe 10-30 cards per day to find a bet maybe once or twice a week. These stats were run after the fact (after scratches). Even if the bet you select pre-race is not scratched, the criteria may be altered so that you selection no longer passes the filter; and, vice-versa: a selection may occur which wasn't there pre-race.

This is a very limited dataset; it would be interesting to see some of these stats on the larger set: anybody???

The stats are intriguing, though. Thanks Pizzaman

[MikeDee]

ran the pizza man play on my db and results are attached. I use PL5, fast dirt and exclude any race with a first time starter. I found only 130 plays in over 26000 races, or about 1 in every 185 races. If you included turf, wet and races with FTS'ers the play would increase but my db not set up that way. Since hits were so low did not look at the JT% filter.

Summary results are 26% win 42% ROI As the filed size is increased the number of plays drops as would be expected as one would expect fewer and low end ppscan and htr 1 fr1-1 hits.

Question pizzaman what is the logic behind the combination of low htr/fr1 and high ppr rank?

[pizzaman]

Combinations of categories with all high rankings have high win% but generally show negative roi. It appears that the only way show a long term profit on these horses is to only play them when their odds creep up to a respectable level. I have found many combinations of such categories that are terrific bets when they go off above 5:2. An example is dirt sprints at major tracks with PER=1,R(K)=1,JKY=1 and HTR=1. Flat bets on this combination won 41% of 743 samples for a -.02 roi. The win% drops to 29% for horses between $7 and $24 but the roi is +.33 (141 samples). So this play only works in the long run if you wait til post time to make the wager -- with the problem that most of the time there will be no play and these may be significant losing streaks.

Some combinations of several categories with high rankings and other categories with low rankings certainly have lower win% but can show positive roi as a flat bet. The explanation may be that the negatives about the entry turn off the public and result in overlays. These combinations can be played without watching the odds.