Re: Today's Plays/Races and Summer End (522 Views)
Posted by:
TGJB (IP Logged)
Date: September 07, 2004 01:33PM
1-- The analysis is certainly not a good guide for deciding whether TG has value, even if you did a study over a representative period of time, at a lot of tracks (say, 100 days at 5 tracks). We have never claimed the analysis would show a positive ROI-- it is done the day before racing in order to have it available for those who want to download at night, without access to odds, fluctuations, track conditions, scratches (which at Sar are done off the program-- I was forced to pass a lot of races because I had no way of knowing whether a fast AE would draw in). Late decisions have a big effect-- one late decision I made was to throw in that $60 Fout horse and that caused me to hit a $1,400 pick 3, which took me from red to black for the meet, and would have done the same thing for the analysis.
2-- Over 36 days the results will be pretty random, subject to a couple of big hits and winning or losing a high percentage of photos. If we had done more than 10% profit or 30% loss I would have been surprised-- you are slightly lower than I would think we would be over a longer sample. My guess would be about minus 10%-- doing better than most and beating the takeout, but not showing a profit. Too many disadvantages doing it in advance.
Having said that, I'm curious how you measured it. Did you set up a unit for each race, and adjust it for the strength of the bet (unit, double unit for B+)? Did you use seperate units for doubles and pick 3's? If we had a win bet and 3 horse exacta box in the same race, how did you split it up? Did you only count ones that surpassed the odds minimum, or did you upgrade/downgrade the unit by odds?
3-- Part of my play on the Pa Derby was a 5 horse tri/super box which included the first 2 finishers, but not the third. It should have been obvious from my ROTW comments that I would use the Dutrow horse at the price-- there was no way to know whether he would handle the distance, but a repeat of his last at the weights saving ground made him almost a sure thing to finish in the money. Sure looks like he went forward.
I said in my comments that MSB was only about 50% to run a big one off the two tops. I read PP to run well off the pattern-- to not bounce off an effort that he has bounced from (depending on how you view the off races) twice before. He did not fire. Considering the price, if I'm right half the time in that situation I'll do okay (and would have here given the first two finishers). Beav's nonsense notwithstanding, those who have been following ROTW over the years know I have indeed done okay.
3-- My request for you to post examples in advance where another handicapping theory would cause opinions that are contrary to ours (even assuming TG players agree) has nothing to do with Ragozin-- those guys are FAR more dogmatic than I am. Please do so, so the lines can be drawn clearly in advance, and the issues discussed. My interest in this is as a teaching tool and to show how TG works in conjunction with or sometimes as opposed to other methods. It has nothing to do with Ragozin at all.
4-- By the way, did you use the Haskell in your ROI? What was your opinion going in, and would you have hit it?
5-- There is probably some super hot pace at which the Dutrow would have spit it out. But if you are asking me whether the pace played an important part in the outcome, I doubt it.
6-- There are ways to tell whether TG is a powerful tool, and it's much easier than with any other data I know of. We have gone on the record in advance with an analysis of several hundred ROTW's now-- take a look at them, look at the odds for the race, and work out a play based on our comments. It's a lot of work, but it can be done. Also, as you say, the RBR is up every day-- anyone can look at the data and the results and draw their own conclusions, as they can with all other data and theories, except we make it available for exactly that purpose. Finally, I have a pretty good record buying and managing racehorses using the data-- see if you can get anyone else to make a complete record available for any lage sampling of doing that.
TGJB