ROTW??

[fkach]

Jimbo,

>It isn\'t \"obvious\" to me that the difference in performance was more than dirt and synthetic<

I haven\'t seen the figure that Jerry assigned to the Ruffian yet, but on the surface it looks like it will probably be a very fast number because of the margin. The difference between the synthetic figures she earned out in CA and the figure I suspect she earned in the Ruffian looks very large. It almost can\'t be just the syth to dirt issue. The reason I say that is that there is almost no evidence at all of other horses moving forward that much on the switch. IMO, this particular horse is a great one for study specifically because she has a prior dirt record.  

As a matter of general discussion, I threw out the other possible contributions to her superior win, but I have no rock solid view about which factors made the primary contribution or even whether this was just a random new peak.  

If you put a gun to my head, I\'d guess that Baffert knew she was doing extremely well and that\'s why he shipped in to begin with. I think it\'s also possible that some very mild improvement was masked by her racing on synth in CA.

It\'s also not unusual to see extremely large gaps between horses on very off tracks because some horses love it, some hate it, mud is getting kicked in horse\'s faces, horses get eased etc... I agree that there was nothing in her record that suggested that she liked the mud. But I did notice that the last 3 races that day (which is when the rains really came and the track became a mess) had winning  margins of 10, 12 1/4, and 5. THey were all logical horses, but those are extraordinary margins. Two went wire to wire and TTS stalked a mediocre horse. So I through it out there.

These are all extraneous to the issue we would like to understand, but they complicate the analysis. I\'ve pretty much said all I can say on this.

[miff]

Not that its completely irrelevant to the string, but Jerry could legitimately give TTS something like -5 or more. Just did the day and the race is completely off the charts. TTS topped by app 7-8 points without getting too creative.

Even if you regress Copper State(which I don\'t feel is too justified), the number is ginormous.


Mike

[TGJB]

fkach-- \"... there is almost no evidence at all of other horses moving forward that much off the switch\".

Alan Greenspan said that everyone is entitled to their own opnion, but not their own set of facts. There is a TON of evidence of horses jumping foward a lot synth to dirt-- that\'s what prompted my original post a few months ago. Go find that string-- it has a partial list. You can also add Monterey Jazz, who jumped out of his mind when he shipped to Lone Star.

Here\'s the key point-- the fact that horses going East to West run their figures in California (I gave a bunch of examples, specifically all the ones I have sent there), means IT\'S NOT THE FIGURES-- those are right. So it\'s something else.

[fkach]

TGJB,

>Alan Greenspan said that everyone is entitled to their own opnion, but not their own set of facts. There is a TON of evidence of horses jumping foward a lot synth to dirt-- that\'s what prompted my original post a few months ago. Go find that string-- it has a partial list. You can also add Monterey Jazz, who jumped out of his mind when he shipped to Lone Star. <

Sorry for long post.

There are examples of just about everything. The problem was that many of the big move ups I noticed had other very logical reasons for their improved figures \"in addition\" to the potential synth/dirt relationship question. That makes it hard to seperate which factor contributed how much. I thought the Ruffian was another example of that. That\'s why I took the conversation off topic. There were other reasons that horse could have moved forward too.  

We can discuss Montery Jazz, but I\'m sure you don\'t want to go there. You would just get pissed off at me for taking the conversation in another direction and tell me \"we don\'t handicap that way here\". MJ running a huge figure when he shipped to Lonestar was no shock to me or some others. I\'ll leave it at that.

Few may care what I think (especially you), but I offer this to those that do.

These are the things I am 100% certain about that I believe most would agree on.

1. The top horses on turf earn slower figures than the top horses on dirt.

2. Turf paces tend to be slower than dirt paces.

3. It\'s at least \"reasonable\" to suspect that the slower paces of turf races have something to do with the tightness of the finishes (I think that part is certain) and the flatter figures at the top (that\'s a maybe).

These are the things that I am already 100% certain about synthetic tracks because I have the data.

1. The paces at all distances for every synthetic track are running several fifths slower than they did for the same final times on dirt (all adjusted for track speed etc...)

2. The data compiled by a friend of mine suggests that the top of the scale is slower by a few fifths of a second - somewhat like turf. I didn\'t do that study, but I trust his data because he has everything in a database and is brilliant. He gave me a chart estimating it. It\'s a sliding scale chart. At the top it\'s a few fifths and it slowly moves to zero.

These are the things I suspect are true based on limited empirical evidence.

1. Every time I have seen a very fast speed figure earned on a synthetic track by a very high quality horse, I took a close look at the pace. There really aren\'t all that many. However, I\'ve already seen several that were accompanied by a very fast pace.

All of this suggests several things to me that I can\'t prove.

What I or anyone else should do with these insights is up to them.

Personally, I use my friend\'s chart to upgrade the synth speed figures for versatile horses moving from synth to dirt \"but only at the very top of the scale\" and not if the figure was earned by a closer coming off a very fast pace (and vice versa).

[HP]

These are the things I am 100% certain about that I believe most would agree on.

1. The top horses on turf earn slower figures than the top horses on dirt.

2. Turf paces tend to be slower than dirt paces.

3. It\'s at least \"reasonable\" to suspect that the slower paces of turf races have something to do with the tightness of the finishes (I think that part is certain) and the flatter figures at the top (that\'s a maybe).

I\'m with you on one and two, but I don\'t agree on three.  I don\'t think the slower pace of turf races has anything to do with it.  Nor does the tightness of the finishes.  The final track record times on turf and dirt are pretty close, despite the generally slower paces of turf racing (in fact, the turf track records over a mile are generally faster raw times than the dirt track records over a mile - so even if the pace is slow, the TG figure is based on the FINAL time, and the final times of the best turf races are equal to or faster than the best dirt races).  Other variables are probably very close (the weight, wind, ground loss, etc.).  

So in my opinion, the \"flatter figures\" at the top are probably attributable to...something in the track variant/adjustment that Jerry makes.  Maybe grass surfaces don\'t change as much during the day.  Maybe grass surfaces vary less from track to track than dirt surfaces.  

I could be wrong about this, but I don\'t think what you are referring (flatter figures at the top of the scale) to has anything to do with slow pace or tight finishes.  

HP

[fkach]

HP,

I wasn\'t actually saying the paces themselves are causing horses to run slower final times (though they may be). I was saying that the figures may look slower at the top because of the different paces. It could also be a process issue.  In terms of tight finishes, I think it\'s a certainty that slow paced races produce tighter finishes.

Beyer had some interesting comments on this subject. I believe it has been discussed here a little too. Beyer eventually changed his beaten length charts for turf racing only, but as far as I am concerned he hasn\'t solved the problem. Not even close.

I would not argue that Curlin is as good a horse on turf as dirt, but if he was (theoretically), I would have bet large sums of money that he would not run the same figures. They would suggest he was slower.

No time to discuss this today. Gotta go. ;-)

[HP]

My point was...the slower paces are NOT causing the horses to run slower final times.  The final times of the best turf races are just as fast the best final times on dirt at comparable distances.  This assumes you are talking about the BEST races at each distance (there are probably a lot of turf races where the slow pace does indeed yield a slower final time).  Since we are talking about the PEAKS here, the FASTEST turf horses are running final times that are just as fast as the FASTEST dirt horses, so the pace is entirely irrelevant in these cases.  Slower pace does not account for why Ghostzapper gets a minus six and the fastest turf horses can\'t do better than minus one or two.  

As for the \"tightness of the finish,\" I would say this is also completely irrelevant, but of course JB would have the last word on this.

[miff]

HP,

You may be forgetting that turf courses are much faster inherently than dirt, therefore a horse running a mile in 1.34 on turf is NOT running nearly as fast(fig wise) as a horse running the same time on a dirt surface, all else being equal.


Mike

[HP]

I agree with you miff and I\'m not forgetting that fact -- I was just focused on the fact that \"slower pace\" probably has nothing to do with it.  There are probably plenty of top level turf races where there was a fast pace and a fast final time.  Your point probably has more to do with why the top turf horses don\'t get a minus six than anything about pace or \"tight finishes.\"

[miff]

HP,

I agree. Also must consider,aside from pure breeding, that most horses go to the turf because they are too slow to compete on dirt.

Years ago,I asked about slow paces on the turf and a few NY jocks generally offered that most turf races are won by horses that can settle and make a big late run. Speed horses going slow  on turf still seem to get run over more often than not, not so much on dirt. Remember a great dirt runner may shade 24 in the last quarter, a turfer 22-ish, thats app ten lengths difference,raw.


Mike

[fkach]

HP,

Thanks for clarifying. I understand what you are saying now.

I would simply question whether very fast raw final times on turf are so fast because turf surfaces in general are faster. That may not tell us about figures.

The tightness of the finishes and this pace issue are actually very significant to synth figures because it relates to a theory I have about the process of making turf figures (and now also synthetic figures) and some of the problems figure makers like Beyer run into.

WARNING WARNING WARNING - If you\'ve never made speed figures, have no interest in how they are made, and/or have no interest in my silly theories skip to the next message.



One of the first Beyer books had a interesting way of teaching novices why one second is more significant at a shorter distance than a longer distance.

Basically, if you are 1 second slower than the best 100 Meter runner, you are a mediocre sprinter. If you are 1 second slower than the best miler, you are still a world class miler. This phenomenon is built into all sophisticated figures. Jerry has touched on the issue from time to time.

Now, let me propose a possibility.

When the pace of a route race is very slow (let\'s call it an 8F race), it may effectively reduce the distance by some percentage because the horses are only running hard for part of the race. That allows inferior horses to finish closer in time and distance to their superior rivals than the true difference in their abilities. Yet when figure makers make their figures and use their beaten length charts, they use the 8F charts and formulas because that\'s how long the race actually was.

I have never seen the TG charts and formulas because they are proprietary information. So forgive me for illustrating this in Beyer terms.

Each 1/5 of a second is worth a hair over 2 Beyer points at one mile (8F).

A two turn mile in 137 is equal to a Beyer figure of 101 (source Beyer on Speed).

A two turn mile in 136 is equal to a Beyer figure of 112 (source Beyer on Speed).

But suppose my theory is correct and we should really be using the 6F chart or the 7F chart instead because many turf/synth races have such slow paces and the effective distance is actually shorter?

Then, each 1/5 of second would be worth more than a hair over 2 points and instead of a one turn mile in 136 being equal to a 112, it might be equal to a 113 or 115.

Using my new adjusted formulas would have the effect of suggesting that the very best turf horses in the world are as fast as the very best dirt horses in the world. The beaten length charts would also reflect the fact that beating a horse by a length or two on turf is more significant than on dirt. (all very relevant to the questions we are facing with synthetic to dirt figures)

All that said, you simply can\'t look at a few horses that ran massive new tops on the switch from CA synth to dirt and conclude that the entire figure move was related to that issue. Not when most of the horses were lightly raced 3YOs in the spring that can and often do explode due to normal development, one lightly raced potential champion filly on the improve, a horse from the Baffert barn shipping from CA to NY (since he and other trainers have been awesome with that move for decades), and a speed crazed unrateable horse that dropped in class, shortened up in distance, and got away with an easier pace.

Doing that will cause a massive misunderstanding of the issue and the type of move that can be expected \"on average\". That\'s what makes all my other extraneous comments on this issue ABSOLUTELY VITAL.

[BitPlayer]

HP –

Perhaps I\'m misunderstanding what you are saying, but in my mind tightness of finishes has to be relevant.  Imagine every race as represented by a stick, the length of which represents the range of TG figures earned by the field in a given race.  Since turf races produce tighter finishes (and thus a narrower range of TG figures) than dirt races, the sticks representing turf races will generally be shorter than the sticks representing dirt races.

Now pile up all the sticks for a given period of time to create a structure.  That structure represents the TG scale for that type of races.  The rule for creating the structure is that sticks for races featuring horses who ran relatively well in both races have to overlap, so that those horses can be given relatively consistent figures.  The rule applies because (oversimplifying a bit) TG is making figures \"using all the horses\" to judge the speed of the track (be it turf, dirt, or synthetic).  Because the sticks representing turf races are shorter than those representing dirt races and there is lots of overlap between sticks (because turf horses run in a narrower range), the resulting structure (scale) for turf races has to be more compact than the structure (scale) for dirt races.   That\'s the figure compaction to which Jimbo was referring (for synthetic figures) when he started this thread.

Now, because TG has only a single scale for both turf and dirt races, you have to line up the turf and dirt scales in some fashion.  In theory, you could line them up any way you choose (e.g., so that the fastest horses on the two scales get the same figures).  In fact, TG seems (sensibly, I think) to have lined them up so that the more compact turf scale is somewhere in the middle of the dirt scale.  Thus, the dirt scale sticks out at the top end and the fastest dirt horses earn better figures than the fastest turf horses.

[HP]

Why wouldn\'t there be at least one race where all the sticks bunched up at minus six instead of zero?  Also, there are actually races on grass where they are not all bunched up and the winner wins by a few lengths and breaks a track record.  And the winner still does not get a minus six or even close.  

I\'m not saying the tight finishes are entirely irrelevant to why the turf figures come out the way they do in general, I\'m saying it does not account for the lack of peak races on grass that are comparable with peak races on dirt.

Miff\'s point about grass being inherently faster than dirt is arguable, but I think this may have more to do with it than the other points made about pace and tight finishes.  Just my opinion...  HP

[fkach]

>Why wouldn\'t there be at least one race where all the sticks bunched up at minus six instead of zero? <

Because if you start your scale in the middle and each 1/5 of second is actually more significant for a turf/synth race than for a dirt race (as I explained in one of my prior posts on speed figure methodology and distance), you won\'t get to -6 (or an extremely slow number either) because you are using an inappropriate unit per 1/5th.  

http://www.thorograph.com/phorum/read.php?1,46631,46731#msg-46731

>Also, there are actually races on grass where they are not all bunched up and the winner wins by a few lengths and breaks a track record. And the winner still does not get a minus six or even close.<

Typically, the margins are smaller, but as I suggested in prior posts, empirical evidence suggests that the margins can actually expand a bit for turf/synth races when the pace is a little more like a dirt race (faster). That\'s what you are often seeing.

Regardless, it would be incredibly rare to get to a -6 if my theory is correct.  It would require a truly great turf/synth horse, on his best day, in a race where the pace was more like a dirt race and he wasn\'t used up in it (to make up for the issue described above and previously).  In other words, it doesn\'t happen, but it could if God intervened. ;-)

[BitPlayer]

HP -

I don\'t think I can explain my opinion better than I already have, and I hate it when threads like this drag on with posters repeatedly talking past each other, so I\'ll address your question and drop it here.

Your question:  \"Why wouldn\'t there be at least one race where all the sticks bunched up at minus six instead of zero?\"  In my scenario, one stick represents one race, so I think your question is why there isn\'t a short turf \"stick\", representing a race run in very fast time, with a bunch of numbers near minus 6, instead of near zero.  I think the answer is that TGJB would sooner believe that the turf was very fast for that race than that several horses in the same race ran very fast significant new tops.  Where one horse runs a new top and wins by a lot, TGJB has to tie the race the to previous races run by the also-rans.  If top turf figures are generally slower than top dirt figures, he\'d need an awfully long stick (a huge winning margin or lots of weight and ground loss) to stretch all the way out to minus 6.

The key is to think about making races fit neatly together in TG\'s database.