Internal fractions for 5f & 5 1/2 f?

[GameTheory]

Another question for my parser:

How are you supposed to calculate internal fractions for a horse from a 5f or 5 1/2f running line when the position/beaten length calls are at 3/16 & 3/8 but the leader times are still at the 1/4 & 1/2?

I guess I haven't noticed or have forgotten about this discrepancy since I gave up doing those kind of calculations in my own handicapping years ago. I looked in a couple of books where they talk about calculating these things (using the standard add 1/5 sec for each beaten length), but no mention of the fact that the calls are different below six furlongs, while the leader times remain the same. Is there a standard solution?

[Tom]

The infor you get is probably not correct to begin with, and even if it was, there is no "correct" way to adjust for beaten lengths anyway, so I alway just ignore the fact that it is different and treat 5-5.5 furlong race the same as 6 furlong races. I figure the errors will all even out in the long run.

Tom

[GameTheory]

While I agree that any kind of beaten lengths = ticks is flawed, at least it's consistenly done the same way.

But to take a time from one part of the race, and apply beaten lengths from another part seems pretty screwy. It seems it should be the time that is adjusted somehow. There must be some sort of procedure people use for that, right? 4 1/2 furlong races I'm not too worried about, but 5 & 5 1/2 are pretty common...

[Tom]

I am sure it will get fixed someday, but then again it only took a hundred years to get most tracks to use the same color saddle cloths. I think they call it tradition.
:rolleyes: :rolleyes: Tom:rolleyes:

[Rick]

GameTheory,

You could probably interpolate to get a reasonable approximation but, as others have pointed out, the accuracy of the the data is questionable anyway. My approach would be not to use a single race for anything but maybe something like a median of last 5 races where the extremes are eliminated.

[Big Bill]

Game Theory,

I use BRIS Pace Ratings in my handicapping and several years ago I shared your concern as they, (BRIS), state that their E2 ratings are at the 1/2 mile in sprints. Here is what I asked them in an e-mail:

"In a 5.5 furlong, the points of call are at the 1/4, 3/8, strech and finish. Since your description of the pace ratings state that the E2 rating is at the 1/2 mile call in sprints, is the E2 pace rating really accurate, i.e., is the rating calculated on the time to the half or to the 3/8 call? Guess this question would apply to 5 furlong races as well."

This was the BRIS reply:

"This is a complicated situation because although we get a 3/8 call we only get a 4f fractional time, so we incorporate an extrapolation process to determine the beaten lengths. So, it is based on the 4f time but we use a process to determine an "accurate" E2 (3f) number. I hope this makes sense."

I've never been certain if that made sense to me or not! Sounds like they use an extrapolation algorithm to apply to their pace rating algorithms. <g>

The accuracy of data will, IMHO, always be suspect. I just move on hoping that in the long run, if you use the data the same way, it will even out. What do you think?

Big Bill

[NorCalGreg]

I am sure it will get fixed someday, but then again it only took a hundred years to get most tracks to use the same color saddle cloths. I think they call it tradition.
:rolleyes: :rolleyes: Tom:rolleyes:

Was doing a search on computing internal fractions, and as happens many of my handicapping-related google searches--led me once again to the P.A. site.

Hate to break it to you, Tom--15 years later---ALL TRACKS STILL AREN'T USING THE SAME COLOR SADDLE CLOTHS.

[MJC922]

The way I always dealt with it is if the first quarter (2.0f) falls between the two points of call 1.5 and 3, and as it's 2/3rds closer to 1.5 the estimated beaten lengths where the time was taken would be first BL multiplied by .66 and second BL multiplied by .34.

[NorCalGreg]

The way I always dealt with it is if the first quarter (2.0f) falls between the two points of call 1.5 and 3, and as it's 2/3rds closer to 1.5 the estimated beaten lengths where the time was taken would be first BL multiplied by .66 and second BL multiplied by .34.

MJ, you DO realize everything above my post---was from 2002? LOL...and BRIS is still using the "although we get a 3/8 call we only get a 4f fractional time, so we incorporate an extrapolation process to determine the beaten lengths" method.

CJ will be along shortly and explain how it's done in the modern world.

CJ...you have the floor.......

[cj]

MJ, you DO realize everything above my post---was from 2002? LOL...and BRIS is still using the "although we get a 3/8 call we only get a 4f fractional time, so we incorporate an extrapolation process to determine the beaten lengths" method.

CJ will be along shortly and explain how it's done in the modern world.

CJ...you have the floor.......

I can only speak for TimeformUS, I doubt anyone else is doing it like us. Most just pretend the calls match up. It is funny you are bringing this up now because we are changing a few things I wasn't happy with in the display of these races.

The calls listed are the "official calls" ... things like 1 by 2 and 2 by 1.5, etc. We don't tinker with those at all.

All of our figures are based on time and are for 2f and 4f regardless of the distance of the call. What we do is put the pace figure underneath the call that matches most closely. At all distances less than 6f, this is the stretch call, not the 2nd call.

At 4.5 furlongs, there is no second call, so the figure is placed under the stretch call which comes at 3.5 furlongs.

At 5 furlongs, the second call comes at 3f and the stretch call comes at 4f, so obviously the stretch call is a perfect match for our pace figure.

At 5.5 furlongs, the second call comes at 3f and the stretch call at 4.5 furlongs, so we still put pace figures under the stretch call.

As for how we calculate the pace figures for 4f at these distances, we do the best we can with the information we have. If a horse is 4 lengths back at 3f of a 5.5f race and 1 length back after 4.5 furlongs, we assume that he was 2 lengths back at the 4f point. He made up three lengths in 1.5 furlongs, so that equates to two lengths in one furlong.

We make similar adjustments for the other calls that don't line up, like the first 5f call and the stretch call at 4.5 furlongs. It isn't perfect but it is the best we can do with the data given.

[MJC922]

MJ, you DO realize everything above my post---was from 2002? LOL...and BRIS is still using the "although we get a 3/8 call we only get a 4f fractional time, so we incorporate an extrapolation process to determine the beaten lengths" method.

CJ will be along shortly and explain how it's done in the modern world.

CJ...you have the floor.......

Sorry, didn't notice you were bumping a necro-thread.

[highnote]

Brohammer wrote (and I paraphrase) that races under 6 furlongs are best handicap using final time speed figures because pace is less of a factor.

I do not use pace analysis in short sprints. It's never caused any problems -- in fact, makes things simpler.