Querying a database of statistics to get counts of different events - mysql

Question

I'm making a database of a soccer league that has these tables:
+---------------------+
| Tables_in_league484 |
+---------------------+
| player |
| statevent |
+---------------------+
18 rows in set (0.09 sec)
and the player table in question look like this,
mysql> desc player;
+-----------+-------------+------+-----+---------+----------------+
| Field | Type | Null | Key | Default | Extra |
+-----------+-------------+------+-----+---------+----------------+
| pid | int(11) | NO | PRI | NULL | auto_increment |
| lastname | varchar(55) | YES | | NULL | |
| firstname | varchar(85) | YES | | NULL | |
| dob | date | YES | | NULL | |
| posid | int(11) | YES | MUL | NULL | |
| tid | int(11) | YES | MUL | NULL | |
| shirtnum | int(11) | YES | | NULL | |
| email | varchar(85) | YES | | NULL | |
+-----------+-------------+------+-----+---------+----------------+
8 rows in set (0.09 sec)
posid is fk for position table;
tid is fk for team table;
mysql> desc statevent;
+--------+-------------+------+-----+---------+----------------+
| Field | Type | Null | Key | Default | Extra |
+--------+-------------+------+-----+---------+----------------+
| eid | int(11) | NO | PRI | NULL | auto_increment |
| gid | int(11) | YES | MUL | NULL | |
| pid | int(11) | YES | MUL | NULL | |
| minute | int(11) | YES | | NULL | |
| typeid | int(11) | YES | | NULL | |
+--------+-------------+------+-----+---------+----------------+
5 rows in set (0.09 sec)
where the typeids are:
1 for shot
2 for save
3 for goal
4 for assist
how can i structure a mysql query that gives me a result that looks like this
+--------+------+------+-------+---------+----------------+
| Name | Team | Shots| Saves | Goals | Assists |
+--------+------+------+-------+---------+----------------+
| Nick | 1| 8| 0| 4| 1|
| Jeff | 4| 5| 0| 5| 6|
| Jim | 7| 7| 0| 6| 3|
+--------+------+------+-------+---------+----------------+
that ends after the 10th result? (limit 10)
I've been trying for hours and I'm knackered thinking about it. What do I count? What do I group by? Can I order by aliases?
EDIT
I failed to mention in my first edit that, while there are 18 helpful tables in this database, they are all empty (thus entirely useless) as they relate to the stat events.
They would have been wonderfully helpful.
However, I have to structure my query on this one table of statevents using only typeid. Is this possible?

Answer 1

Essentially, you're just trying to construct a simple PIVOT TABLE query. Personally I'd advocate just returning a GROUPed result set and handle the data display at the application level, but if you must do the pivoting in MySQL then it might look something like this - I've changed some column/table names to get you thinking a bit...
SELECT p.firstname
, p.team_id
, COUNT(CASE WHEN event_type_id = 1 THEN 'foo' END) Shots
, COUNT(CASE WHEN event_type_id = 2 THEN 'foo' END) Saves
, COUNT(CASE WHEN event_type_id = 3 THEN 'foo' END) Goals
, COUNT(CASE WHEN event_type_id = 4 THEN 'foo' END) Assists
FROM player p
JOIN stat_event e
ON e.player_id = p.player_id
GROUP
BY p.player_id;

Answer 2

You would have to join the player table with the other tables you need counts from (shots, saves, goals etc).
One you have the join in place, you would need to aggregate on player id, player name and team with the help of a group by clause.
Your final query will look something like this..
SELECT p.firstname, t.team, COUNT(sh.shots), COUNT(sa.saves), COUNT(g.goals),COUNT(a.assists)
FROM player p
INNER JOIN team t
ON p.tid = t.tid
....
GROUP BY p.pid, p.firstname, t.team
LIMIT 10
EDIT:
I am not a DB expert. I have one SUBOPTIMAL way of achieving this.
I would create a temporary table containing information of the form (it would have to contain pid and tid information too):
...
Nick Goals 13
Matt Saves 4
Nick Saves 11
...
This should be simple to achieve.
I would then use a SQL cursor to iterate over all distinct player ids and recover statistics from the temporary table we constructed above.