FIND_IN_SET too slow with GROUP_CONCAT (Dense Rank in MySQL) - mysql

Question

I have a query that calculates dense ranks based on the value of a column :
SELECT id,
score1,
FIND_IN_SET
(
score1,
(
SELECT GROUP_CONCAT(score1 ORDER BY score1 DESC) FROM scores
)
) as rank
FROM score_news;
This is what the query results look like:
+----+--------+------+
| id | score1 | rank |
+----+--------+------+
| 1 | 15 | 1 |
| 2 | 15 | 1 |
| 3 | 14 | 3 |
| 4 | 13 | 4 |
+----+--------+------+
The query takes Nx longer time when number of scores increases by N times. Is there any way I can optimize this ? My table size in the order of 106
NOTE: I have already tried a technique using mysql user variables but I get inconsistent results when I run it on a large set. On investigation I found this in the MySQL docs:
The order of evaluation for user variables is undefined and may change
based on the elements contained within a given query. In SELECT #a, #a
:= #a+1 ..., you might think that MySQL will evaluate #a first and
then do an assignment second, but changing the query (for example, by
adding a GROUP BY, HAVING, or ORDER BY clause) may change the order of
evaluation...The general rule is never to assign a value to a user
variable in one part of a statement and use the same variable in some
other part of the same statement. You might get the results you
expect, but this is not guaranteed.
My attempt with user variables :
SELECT
a.id,
#prev := #curr as prev,
#curr := a.score1 as curr,
#rank := IF(#rank = 0, #rank + 1, IF(#prev > #curr, #rank+#ties, #rank)) AS rank,
#ties := IF(#prev = #curr, #ties+1, 1) AS ties
FROM
scores a,
(
SELECT
#curr := null,
#prev := null,
#rank := 0,
#ties := 1,
#total := count(*)
FROM scores
WHERE score1 is not null
) b
WHERE
score1 is not null
ORDER BY
score1 DESC
)

Answer 1

The solution with variables could work, but you need to first order the result set, and only then work with the variable assignments:
SELECT a.id,
#rank := IF(#curr = a.score1, #rank, #rank + #ties) AS rank,
#ties := IF(#curr = a.score1, #ties + 1, 1) AS ties,
#curr := a.score1 AS curr
FROM (SELECT * FROM scores WHERE score1 is NOT NULL ORDER BY score1 DESC) a,
(SELECT #curr := null, #rank := 0, #ties := 1) b
NB: I placed the curr column last in the select clause to save one variable.

Answer 2

You can also use following query to get your dense rank without using user defined variables
SELECT
a.*,
(SELECT
COUNT(DISTINCT score1)
FROM
scores b
WHERE a.`score1` < b.score1) + 1 rank
FROM
scores a
ORDER BY score1 DESC
Demo
Demo using your data set
An index on score1 column might help you