Horse racing draw advantage bias. There are 2 charts here.
Originally, only the 2yr old chart was shown, but this could mislead as the chart for 3yr olds and older horses differs considerably.

The 2yr old pattern shown on the Top chart, has proved very reliable in the past, but the chart for older horses shown at the bottom has not produced a pattern in the past worth following in the long term.
There is quite a lot of cross-course manouvreing in these races, but not all races split.
First thoughts may be that splits may appear to benefit 2yr olds more than older horses, but splits in the last 2 seasons, started around the 16 runner mark.
As the 2yr old pattern is strongest in races of only 8 - 13 runners, splits in these races may only be affecting races with larger fields, which show a more even pattern.
To sum up, there appears to be an advantage towards the High side of the course in races for older horses, but in 2yr old races, this advantage is less effective.
Most 2yr olds races of 8 - 13 runners have been won by horses drawn towards the Centre of the course.
As races with larger fields tend to split more often, both patterns appear to suffer, with winners starting from a wider range of stall numbers.

NEWCASTLE 7F (2 yr olds)
Stalls usually positioned HIGH ( Centre shown C, Low shown L )

Link to course and distance menu
5 Races were run on the ROUND course due to drainage work in the straight during October 2005.
The chart below shows results in the STRAIGHT course only.

These are the only 2yr old races that I have found produce a consistent pattern.
The stalls were positioned Centre for the first time in 2004.

On good or better going, there have been 43 races.
Above 13 runners, there were 7 wins from the Bottom 1/2, and 8 from the Top 1/2 of the draw.
In races of 8 - 13 runners, stalls positioned High or Centre, stalls 1 - 6 won 21 of 28 races 75%.
With the stalls positioned High or Centre, these winners started from stalls away from the High running rail, that is towards the Centre of the course, or towards the Low side.
During October 2005, due to drainage work, 5 races for 2yr olds were run over 7F on the Round course.
Those results are not shown on the above chart.

2006 results  :- 2-9, 4-9 (C), 2-10 (C), 3-10 (C), 2-13 (C), 11-14 (C),
3-15, 1-16, 13-16, 17-17
All had stalls positioned High except where indicated.

NEWCASTLE  7F (3yr olds and older )
Stalls usually positioned HIGH ( Centre shown C, Low shown L )

Link to course and distance menu

A very weak pattern.
On good or better going, the Bottom half, Top half split is pretty even, 32 wins against 24.
One other winner was drawn dead Centre.
On good or better going, the Top 6 stalls won 32 of 57 races 56%
That is a low percentage for 6 stalls.
Stalls 1 to 6 won 23 of 57 races.
Splits often occur in these races as jockeys search for faster ground towards the Low side of the course.
This tactic may be one reason for the spreads of results in races of 16 or more runners.
The following observation remains unaltered from 2005 :-
On good or better going, in races of 9 to 15 runners, the Top 6 stalls won 21 of 30 races 70%.
In 2006, there were 4 wins in 5 races from the Top 6 stalls under those conditions.
That is very much the opposite of the 2yr old pattern shown above.
You may choose to follow that trend, but 7F on a straight course is a very long distance for a draw advantage to produce a reliable pattern of results in the long term.
Historically, there has not been a pattern worth following in these races.
There is no recommendation.

On good/good to soft in places  or softer going, there is a similar pattern.
In races of 9 to 15 runners, the Top 6 stalls won 12 of 18 races 66%.
Above 15 runners, wins are spread across most stalls.

2006 results :-  6-8, 6-8, 9-14, 5-15 (only 10 ran), 9-15, 11-15, 5-16, 17-17, 17-17
In 2005 and 2006, one race, winners drawn 11-14 (C), 5-15(C), took place on the Round course due to work on the straight course.
Those 2 results are excluded from this analysis, and are not on the chart.