regex - Extracting Values from Array in Redshift SQL -

i have arrays stored in redshift table "transactions" in following format:

id, total, breakdown 1, 100, [50,50] 2, 200, [150,50] 3, 125, [15, 110] ... n, 10000, [100,900] 

since format useless me, need processing on values out. i've tried using regex extract it.

select regexp_substr(breakdown, '\[([0-9]+),([0-9]+)\]') transactions 

but error returned says

unmatched ( or \( detail:  ----------------------------------------------- error:  unmatched ( or \( code:      8002 context:   t_regexp_init query:     8946413 location:  funcs_expr.cpp:130 process:   query3_40 [pid=17533] -------------------------------------------- 

ideally x , y own columns can appropriate math. know can easy in python or php or like, i'm interested in pure sql solution - partially because i'm using online sql editor (mode analytics) plot dashboard.

thanks help!

if breakdown array can this:

select id, total, breakdown[1] x, breakdown[2] y transactions; 

if breakdown not array e.g. varchar column, can cast array if replace square brackets curly braces:

select id, total,         (translate(breakdown, '[]', '{}')::integer[])[1] x,        (translate(breakdown, '[]', '{}')::integer[])[2] y transactions;