select only last try - sql

Question

Consider the following table structure for an imaginary table named score:
player_name |player_lastname |try |score
primary key: (player_name,player_lastname,try)
(dont discuss the table schema, its just an example)
This table holds the scores of all players - every player should be able to play either one OR two times. Now, how could I fetch data about every player's last try only (i.e. first tries should be ignored for those who played more than once)?
An example of what I'm trying to achieve:
player_name,player_lastname,try,score
=====================================
bart, simpson,1,250
lisa,simpson,1,150
lisa,simpson,2,250
homer,simpson,1,300
homer,simpson,2,350
maggi,simpson,1,50
The result should be:
player_name,player_lastname,try,score
=====================================
bart, simpson,1,250
lisa,simpson,2,250
homer,simpson,2,350
maggi,simpson,1,50

Answer 1

One option is to JOIN the table to itself using a subquery with MAX:
select s.*
from score s
join (
select max(try) maxtry, player_name, player_lastname
from score
group by player_name, player_lastname
) s2 on s.player_name = s2.player_name
and s.player_lastname = s2.player_lastname
and s.try = s2.maxtry
SQL Fiddle Demo
Depending on your database, you may be able to take advantage of analytic functions such as ROW_NUMBER() though which would make this easier. Here is a another fiddle to demonstrate.
Since you are using postgresql, then you should be able to use the analytic ROW_NUMBER() function. This should work as well:
select *
from (
select try, player_name, player_lastname, score,
Row_Number() Over (Partition By player_name, player_lastname order by try desc) rn
from score
) s
where rn = 1
BTW -- I'd consider adding a player_id as a primary key.

Answer 2

This will probably have the best performance
select distinct on (player_name, player_lastname)
player_name, player_lastname, try, score
from score
order by 1, 2, 3 desc

Answer 3

A Rank function can solve this:
SELECT player_name,player_lastname,TRY,score
FROM (SELECT player_name,player_lastname,TRY,score,RANK() OVER (PARTITION BY player_name, Player_Lastname ORDER BY TRY DESC)AS try_rank
FROM score
)sub
WHERE try_rank = 1
I'm assuming 'try' is the number that can be 1/2.
Edit, forgot Partition BY

Answer 4

SELECT player_name,player_lastname,try,score
FROM scores sc
WHERE NOT EXISTS (
SELECT *
FROM scores nx
WHERE nx.player_name = sc.player_name
AND nx.player_lastname = sc.player_lastname
AND nx.try > sc.try
);

Answer 5

Try this out:
Sel player_name,
player_lastname,
try,
score
from score where try = 2 or
try = 1 and
(player_name,player_lastname) not in
(sel player_name,player_lastname from score where try=2);