As part of my research into a rating system, im interested in discussing how people factor in barrier draws into their respective ratings/assessments.

A few nuggets that ive heard from people are:

* Certain tracks at certain distances (eg: Caulfield, 1400m) - disadvantage to be drawn out wide, but not all.
* Obvious advantages at Flemington or other courses with long straights when drawn on the grandstand rail and the trend of racing is heading there.
* At an advantage to be drawn between 4 and 10 in an 18 horse field
* never an advantage to be drawn #1 unless an on-pacer
* Disadvantage to be drawn out wide if an on-pacer, as will need to sprint from the start to get position, can get caught out wide if they dont make it.

Anyone else got any rules of thumb?

I have been thinking about this to lately. And the only one I think is of concern (and could quite possibly I could be entirely wrong) is your fourth point. A horse that like to lead is going to be disadvantaged from a outside barrier because they have to make sure they get across and dont get caught wide. On the other hand, I dont believe a barrier draw really matter to a horse that sits at the back of the pack, and swoops on the leaders in the last turn and onto the straight. I guess though to be able to decide that, you need to have reasonably accurate speed maps, to determine the horses running position.

I guess by the same token, if you are using speed maps, and there is only one horse that likes to lead, and looks as though no one will contest it, then I guess a wide barrier is not so much of a concern.

Barrier position: Depends on the horse's running style, the distance, rail position, the track and it's condition, the race pace and sometimes none of the above.

So the key to barriers and any ratings that use them is where a horse is most likely to settle:

* as horse that get back and swoop are neither advantaged nor disadvantages by a wide draw
* On pacers and handy horses need to be able to kick early and thus the more distance they have to travel (both in terms of width of the track and number of runners it must out run) will determine how easy it was for them to settle.

Therefore, how many horses in a field does it take for barriers to matter?

eg: There is no need in a race with 8 horses, IMHO, to bring barriers into the equation.

Barriers no doubt do matter, but in any given race and apart from all of the above generally sound reasons mentioned in posts, a combination of luck in the running, fitness and split second Jockey decisions for all a race's runners [not just one] and how they effect each other, will make fools of any attempt to find 'consistent' simple answers to what will or won't be good or bad barriers for any given race. 'Perceived' bad barriers [better odds too] often turn out to be good barriers and vis-a-vis.

Like any aspect of horse racing, there are guidelines for just about anything that a punter wants to consider, but in the end they are so often wrong that they will never put a long term profit in our pocket.

Agree with you crash, however I was intending to run some test cases through SPSS and see what correlation barriers with particular conditions effect the finishing place and speed of the winner. If the correlation isnt strong for any of my test conditions then i'll throw it away but ive got a feeling that it does have a slight correlation, enough to at least consider.

(for example: One test case could be Caulfield 1400m on good going - in my database that gives me a sample size of ~1500 separate events.)