More data from MySql ("complex query")

[eevan79]

Here is my code:

SELECT *, case when (kills = 0) then 0 when (deaths = 0) then 1000 else ((kills*1.0)/(deaths*1.0)) end as killdeathratio, ($scoreFormula) as totalscore 
   FROM ( 
   SELECT gp.name as name, bans.name as banname, avg(dp.courierkills) as courierkills, avg(dp.raxkills) as raxkills,
avg(dp.towerkills) as towerkills, avg(dp.assists) as assists, avg(dp.creepdenies) as creepdenies, avg(dp.creepkills) as creepkills,
avg(dp.neutralkills) as neutralkills, avg(dp.deaths) as deaths, avg(dp.kills) as kills, count(*) as totgames, 
case when (kills = 0) then 0 when (deaths = 0) then 1000 else ((kills*1.0)/(deaths*1.0)) end as killdeathratio,
SUM(case when(((dg.winner = 1 and dp.newcolour < 6) or (dg.winner = 2 and dp.newcolour > 6)) 
     AND gp.`left`/ga.duration >= $minPlayedRatio) then 1 else 0 end) as wins, 
     SUM(case when(((dg.winner = 2 and dp.newcolour < 6) or (dg.winner = 1 and dp.newcolour > 6)) 
     AND gp.`left`/ga.duration >= $minPlayedRatio) then 1 else 0 end) as losses 
     FROM gameplayers as gp 
     LEFT JOIN dotagames as dg ON gp.gameid = dg.gameid 
     LEFT JOIN dotaplayers as dp ON dg.gameid = dp.gameid 
   AND gp.colour = dp.colour 
   AND dp.newcolour <> 12 
   AND dp.newcolour <> 6
   LEFT JOIN games as ga ON dp.gameid = ga.id 
   LEFT JOIN bans on bans.name = gp.name
   WHERE dg.winner <> 0  
   GROUP BY gp.name having totgames >= $games) as i 
   ORDER BY $order $sortdb, name $sortdb LIMIT $offset, $rowsperpage

 

As you can see I am using AVG to get users average data and all in one query. I also need to find when users left the game and count this as disconnects.

 

There is table gameplayers (as gp. in code above)and when user left before the game ends, leftreason filed is"Disconnected" or "ECONNRESET".

 

So it need to add this code, but I always get error.

MAX(...code to select table...WHERE (`leftreason` LIKE('%ECONNRESET%') OR `leftreason` LIKE('%was dropped%') )  )

 

Is this even possible?

[Adam]

This query looks a mess. I haven't wrapped my head around it properly yet, but the DK ratio case statement- aren't you repeating the logic? The sub-query already works out the ratio and returns it as "killdeathratio". Assuming we can remove that, you then have ($scoreFormula) as totalscore - you're injecting a variable into the query so that you can select it? That makes no sense. Assuming we can remove that, you're then just selecting * from a sub-query. Remove the outer query then and you have a simpler query to deal with. Add some indentation and you're left with:

 

SELECT
    gp.name as name
  , bans.name as banname
  , avg(dp.courierkills) as courierkills
  , avg(dp.raxkills) as raxkills
  , avg(dp.towerkills) as towerkills
  , avg(dp.assists) as assists
  , avg(dp.creepdenies) as creepdenies
  , avg(dp.creepkills) as creepkills
  , avg(dp.neutralkills) as neutralkills
  , avg(dp.deaths) as deaths
  , avg(dp.kills) as kills
  , count(*) as totgames
  , case when (kills = 0) then 0 when (deaths = 0) then 1000 else ((kills*1.0)/(deaths*1.0)) end as killdeathratio
  , SUM(case when(((dg.winner = 1 and dp.newcolour < 6) or (dg.winner = 2 and dp.newcolour > 6)) 
        AND gp.`left`/ga.duration >= $minPlayedRatio) then 1 else 0 end) as wins
  , SUM(case when(((dg.winner = 2 and dp.newcolour < 6) or (dg.winner = 1 and dp.newcolour > 6)) 
        AND gp.`left`/ga.duration >= $minPlayedRatio) then 1 else 0 end) as losses
FROM gameplayers as gp
LEFT JOIN dotagames as dg
    ON gp.gameid = dg.gameid 
LEFT JOIN dotaplayers as dp
    ON dg.gameid = dp.gameid
    AND gp.colour = dp.colour
    AND dp.newcolour <> 12
    AND dp.newcolour <> 6
LEFT JOIN games as ga
    ON dp.gameid = ga.id
LEFT JOIN bans
    ON bans.name = gp.name
WHERE dg.winner <> 0
GROUP BY gp.name having totgames >= $games
ORDER BY $order $sortdb, name $sortdb LIMIT $offset, $rowsperpage

 

Does this work? Probably be easier for you to edit in future if so.

 

To your original question, where are you wanting to add that SQL?

[eevan79]

Yes this code is working, but I dont have score formula...and I needed.

Current score formula:

 

$scoreFormula = '((((kills-deaths+assists*0.5+towerkills*0.5+raxkills*0.2+(courierkills+creepdenies)*0.1+neutralkills*0.03+creepkills*0.03) * .2)+(wins-losses)))*10'; 

 

If I change scoreFormula users will get different score.

 

But I can use that formula after fetching data from sql.

 

Regarding to your question...I want to count data from gameplayers (gp) table where users disconnected or dropped  from game (as disconnect)

 

gp.leftreason LIKE('%ECONNRESET%') OR gp.leftreason LIKE('%was dropped%')

 

 

O have tried this:

 

LEFT JOIN gameplayers as disconnects ON (gp.leftreason LIKE('%ECONNRESET%') OR gp.leftreason LIKE('%was dropped%'))

 

But, of course, its not working.

 

Can I add it to this query?

[eevan79]

Correction: your code is not working:

MySQL Error: 1064 ( You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near ' bans.name as banname  , avg(dp.courierkills) as courierkills  , avg(dp.raxk' at line 3)

[kickstart]

Hi

 

Can't see what is causing the error you are getting.

 

However to get the extra data I think we need more info.

 

What is the structure of gp? Are there multiple records there for a single game / player and you want to count the number of disconnects for that game player? Or are you only interested in disconnected games?

 

All the best

 

Keith

[eevan79]

Ok, before I answer you look at this code:

 

SELECT *, 
    case when (kills = 0) then 0 
    when (deaths = 0) then 1000 
    else ((kills*1.0)/(deaths*1.0)) end as killdeathratio, 
    ($scoreFormula) as totalscore 
     FROM ( 
     SELECT 
     gp.name as name, 
     bans.name as banname, 
     avg(dp.courierkills) as courierkills, 
     avg(dp.raxkills) as raxkills,
     avg(dp.towerkills) as towerkills, 
     avg(dp.assists) as assists, 
     avg(dp.creepdenies) as creepdenies, 
     avg(dp.creepkills) as creepkills,
     avg(dp.neutralkills) as neutralkills, 
     avg(dp.deaths) as deaths, 
     avg(dp.kills) as kills, 
     COUNT(*) as totgames, 
     case when (kills = 0) then 0 
     when (deaths = 0) then 1000 
     else ((kills*1.0)/(deaths*1.0)) 
     end as killdeathratio,
     SUM(case when(((dg.winner = 1 and dp.newcolour < 6) 
     or (dg.winner = 2 and dp.newcolour > 6)) 
     AND gp.`left`/ga.duration >= $minPlayedRatio) then 1 
     else 0 end) as wins, 
     SUM(case when(((dg.winner = 2 and dp.newcolour < 6) 
     or (dg.winner = 1 and dp.newcolour > 6)) 
     AND gp.`left`/ga.duration >= $minPlayedRatio) 
     then 1 else 0 end) as losses 
          
     SUM(case 
     when(
     (gp.`leftreason` LIKE ('%ECONNRESET%')) 
     OR 
     (gp.`leftreason` LIKE ('%was dropped%'))
     )
     then 'DISC' ) as disc
     

     FROM gameplayers as gp 
     LEFT JOIN dotagames as dg ON gp.gameid = dg.gameid 
     LEFT JOIN dotaplayers as dp ON dg.gameid = dp.gameid 
     AND gp.colour = dp.colour 
     AND dp.newcolour <> 12 
     AND dp.newcolour <> 6
     LEFT JOIN games as ga ON dp.gameid = ga.id 
     LEFT JOIN bans on bans.name = gp.name 
     WHERE dg.winner <> 0 $_sql
     GROUP BY gp.name having totgames >= $games) as i 
     ORDER BY $order $sortdb, name $sortdb 
     LIMIT $offset, $rowsperpage

 

I separated "wrong" code between empty lines. This error occured:

MySQL Error: 1064 ( You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near 'SUM(case    when(    (gp.`leftreason` LIKE ('%ECONNRESET%'))    OR    (gp.`le' at line 33)

In gp there is data for one single user (who joined the game):

id, gameid, name, left, leftreason, team, color etc...

 

I want to match leftreason with ECONNRESET or was dropped and count it as  disconnets from all games.

[awjudd]

I believe that the syntax error you just provided us with is because you are missing the comma between losses and SUM ... i.e.

SELECT *, 
    case when (kills = 0) then 0 
    when (deaths = 0) then 1000 
    else ((kills*1.0)/(deaths*1.0)) end as killdeathratio, 
    ($scoreFormula) as totalscore 
     FROM ( 
     SELECT 
     gp.name as name, 
     bans.name as banname, 
     avg(dp.courierkills) as courierkills, 
     avg(dp.raxkills) as raxkills,
     avg(dp.towerkills) as towerkills, 
     avg(dp.assists) as assists, 
     avg(dp.creepdenies) as creepdenies, 
     avg(dp.creepkills) as creepkills,
     avg(dp.neutralkills) as neutralkills, 
     avg(dp.deaths) as deaths, 
     avg(dp.kills) as kills, 
     COUNT(*) as totgames, 
     case when (kills = 0) then 0 
     when (deaths = 0) then 1000 
     else ((kills*1.0)/(deaths*1.0)) 
     end as killdeathratio,
     SUM(case when(((dg.winner = 1 and dp.newcolour < 6) 
     or (dg.winner = 2 and dp.newcolour > 6)) 
     AND gp.`left`/ga.duration >= $minPlayedRatio) then 1 
     else 0 end) as wins, 
     SUM(case when(((dg.winner = 2 and dp.newcolour < 6) 
     or (dg.winner = 1 and dp.newcolour > 6)) 
     AND gp.`left`/ga.duration >= $minPlayedRatio) 
     then 1 else 0 end) as losses
          
     , SUM(case 
     when(
     (gp.`leftreason` LIKE ('%ECONNRESET%')) 
     OR 
     (gp.`leftreason` LIKE ('%was dropped%'))
     )
     then 'DISC' ) as disc
     

     FROM gameplayers as gp 
     LEFT JOIN dotagames as dg ON gp.gameid = dg.gameid 
     LEFT JOIN dotaplayers as dp ON dg.gameid = dp.gameid 
     AND gp.colour = dp.colour 
     AND dp.newcolour <> 12 
     AND dp.newcolour <> 6
     LEFT JOIN games as ga ON dp.gameid = ga.id 
     LEFT JOIN bans on bans.name = gp.name 
     WHERE dg.winner <> 0 $_sql
     GROUP BY gp.name having totgames >= $games) as i 
     ORDER BY $order $sortdb, name $sortdb 
     LIMIT $offset, $rowsperpage

[kickstart]

Hi

 

Looks like you are trying to SUM a character result there (rather than converting disconnects to 1 and others to 0 and summing that).

 

All the best

 

Keith

[eevan79]

Ok, I'l separate problematic code:

 

   

then 1 else 0 end) as losses

 

, SUM(case

    when(

    (gp.`leftreason` LIKE ('%ECONNRESET%'))

    OR

    (gp.`leftreason` LIKE ('%was dropped%'))

    )

    then 1 else 0 end) as disc

 

    FROM gameplayers as gp

 

Now I get values 1 or 0 instead of SUM. This large code totally confused me.

[ngreenwood6]

I briefly reviewed your code but couldnt you take some of the processing out of that query and just do the processing via php. Most of the time you can and can save you a big headache of trying to figure out complex queries.

[eevan79]

I can only remove killdeathratio ratio and calcualte it with php. If I remove more coding then I need to prepare aditional queries for that.

 

With this code I have all data about users from 5 different tables. Also, this code is used for highscore list.

[kickstart]

Now I get values 1 or 0 instead of SUM. This large code totally confused me.

 

To answer this I think I need the full table declarations and a few sample lines.

 

However you are looking at the gp table and joining other tables to it. This could result in multiple rows for the same gp row, in which case you SUM would be the sum of the duplicated rows which are disconnected (which doesn't seem useful to me).

 

All the best

 

Keith

[eevan79]

I know that data, in this case, can be populated better in mysql, but that depence of hosting program (which is used to insert data into mysql).

Here is samples and tables:

dotagames (dg)

INSERT INTO `dotagames` 
(`id`, `botid`, `gameid`, `winner`, `min`, `sec`) 
VALUES
(31, 0, 32, 0, 0, 0),
(30, 0, 31, 0, 0, 0),
(29, 0, 30, 2, 53, 48),
(28, 0, 29, 2, 53, 18),
(27, 0, 28, 1, 53, 19); ...etc.

 

dotaplayers (dp)

INSERT INTO `dotaplayers`
(`id`, `botid`, `gameid`, `colour`, `kills`, `deaths`, `creepkills`, `creepdenies`, `assists`, `gold`, `neutralkills`, `item1`, `item2`, `item3`, `item4`, `item5`, `item6`, `hero`, `newcolour`, `towerkills`, `raxkills`, `courierkills`) 
VALUES
(1, 0, 1, 2, 2, 3, 22, 1, 2, 0, 0, '\0\0\0\0', '\0\0\0\0', '\0\0\0\0', '\0\0\0\0', '\0\0\0\0', '\0\0\0\0', 'Hvsh', 3, 0, 0, 0),
(2, 0, 1, 3, 1, 1, 7, 0, 0, 2, 0, '\0\0\0\0', '\0\0\0\0', '\0\0\0\0', '\0\0\0\0', '\0\0\0\0', '\0\0\0\0', 'N016', 9, 0, 0, 0),
(3, 0, 1, 5, 0, 2, 5, 0, 1, 18, 0, '\0\0\0\0', '\0\0\0\0', '\0\0\0\0', '\0\0\0\0', '\0\0\0\0', '\0\0\0\0', 'U000', 5, 0, 0, 0),
(4, 0, 1, 10, 11, 3, 46, 1, 2, 2, 9, '\0\0\0\0', '\0\0\0\0', '\0\0\0\0', '\0\0\0\0', '\0\0\0\0', '\0\0\0\0', 'UC42', 10, 0, 0, 0),
(5, 0, 1, 9, 9, 5, 219, 1, 10, 2, 37, 'I099', 'I06B', 'I0AW', 'I0A9', 'I0A8', 'I0A1', 'U00K', 8, 2, 0, 0); ...etc

gameplayers (gp)

INSERT INTO 
`gameplayers` (`id`, `botid`, `gameid`, `name`, `ip`, `spoofed`, `reserved`, `loadingtime`, `left`, `leftreason`, `team`, `colour`, `spoofedrealm`) 
VALUES
(1, 0, 1, 'sVeTiDjOkA', '91.150.116.97', 1, 0, 12656, 626, 'has left the game voluntarily', 0, 3, 'europe.battle.net'),
(2, 0, 1, 'adida_s', '77.66.157.119', 1, 0, 9453, 1068, 'has left the game voluntarily', 0, 2, 'europe.battle.net'),
(3, 0, 1, 'BuBi.Mr_PoRn', '84.132.243.78', 1, 0, 13078, 718, 'has left the game voluntarily', 0, 5, 'europe.battle.net'),
(4, 0, 1, 'changer4', '194.170.95.210', 1, 0, 34844, 1763, 'has left the game voluntarily', 0, 4, 'europe.battle.net'),
(5, 0, 1, 'DsD)WitE(', '86.149.188.238', 1, 0, 7984, 3655, 'has left the game voluntarily', 1, 8, 'europe.battle.net'),
(6, 0, 1, 'Neubivljiv', '192.168.1.5', 1, 1, 7828, 3977, 'has left the game voluntarily', 0, 1, ''),
(7, 0, 1, 'Dirty-Tool', '91.49.241.117', 1, 0, 25328, 2795, 'has left the game voluntarily', 1, 11, 'europe.battle.net'),
(8, 0, 1, 'Loucien', '79.103.98.165', 1, 0, 15859, 2818, 'has left the game voluntarily', 1, 10, 'europe.battle.net'),
(9, 0, 1, 'shernat', '77.65.107.184', 1, 0, 22500, 3696, 'has left the game voluntarily', 1, 9, 'europe.battle.net'),
(10, 0, 1, 'Xalaropsolis', '79.103.121.80', 1, 0, 7984, 3983, 'has left the game voluntarily', 1, 7, 'europe.battle.net'),
(11, 0, 2, 'Ricky.Boy', '86.124.133.233', 1, 0, 8375, 30, 'has left the game voluntarily', 0, 4, 'europe.battle.net'),
(12, 0, 2, 'Human_From_MARS', '87.126.10.162', 1, 0, 7031, 61, 'has lost the connection (connection error - ECONNRESET)', 1, 10, 'europe.battle.net'),
(13, 0, 2, 'Neubivljiv', '93.87.149.182', 1, 1, 9391, 189, 'has left the game voluntarily', 0, 1, 'europe.battle.net'); ...etc

games

INSERT INTO 
`games` (`id`, `botid`, `server`, `map`, `datetime`, `gamename`, `ownername`, `duration`, `gamestate`, `creatorname`, `creatorserver`) 
VALUES
(1, 0, 'europe.battle.net', 'MapsDownloadDotA Allstars v6.60b.w3x', '2009-06-26 13:27:33', 'dota -sdso!!! ghost!', 'Neubivljiv', 3983, 16, 'Neubivljiv', ''),
(2, 0, 'europe.battle.net', 'MapsDownloadDotA Allstars v6.60b.w3x', '2009-06-26 17:10:19', 'dota -sdso ghost++ !!!', 'Neubivljiv', 189, 16, 'Neubivljiv', 'europe.battle.net'),
(3, 0, 'europe.battle.net', 'MapsDownloadDotA Allstars v6.60b.w3x', '2009-06-26 18:24:43', 'dota sdso! ghost++', 'Neubivljiv', 73, 16, 'Neubivljiv', 'europe.battle.net'),
(4, 0, 'europe.battle.net', 'MapsDownloadDotA Allstars v6.60b.w3x', '2009-06-26 21:53:24', 'dota -sdso ghost!!!', 'Neubivljiv', 2790, 16, 'Neubivljiv', ''),
(5, 0, 'europe.battle.net', 'MapsDownloadDotA Allstars v6.60b.w3x', '2009-06-27 00:10:48', 'dota -sdso!!! ghost!!!', 'Neubivljiv', 3037, 16, 'Neubivljiv', '');

 

And data from code (first post) can be viewed here

[kickstart]

Hi

 

If I understand it, gameplayers has one row for per / game. A player will not have multiple rows for the same game? Your code gets you the details of a players games, and the extra column you want is the number of disconnections that player has had in any game.

 

If so something like this:-

 

SELECT
    gp.name as name
  , bans.name as banname
  , avg(dp.courierkills) as courierkills
  , avg(dp.raxkills) as raxkills
  , avg(dp.towerkills) as towerkills
  , avg(dp.assists) as assists
  , avg(dp.creepdenies) as creepdenies
  , avg(dp.creepkills) as creepkills
  , avg(dp.neutralkills) as neutralkills
  , avg(dp.deaths) as deaths
  , avg(dp.kills) as kills
  , count(*) as totgames
  , case when (kills = 0) then 0 when (deaths = 0) then 1000 else ((kills*1.0)/(deaths*1.0)) end as killdeathratio
  , SUM(case when(((dg.winner = 1 and dp.newcolour < 6) or (dg.winner = 2 and dp.newcolour > 6))
        AND gp.`left`/ga.duration >= $minPlayedRatio) then 1 else 0 end) as wins
  , SUM(case when(((dg.winner = 2 and dp.newcolour < 6) or (dg.winner = 1 and dp.newcolour > 6))
        AND gp.`left`/ga.duration >= $minPlayedRatio) then 1 else 0 end) as losses
, SUM(case when((gp.`leftreason` LIKE ('%ECONNRESET%')) OR (gp.`leftreason` LIKE ('%was dropped%'))) then 1 else 0 end) as disc 
FROM gameplayers as gp
LEFT JOIN dotagames as dg
    ON gp.gameid = dg.gameid
LEFT JOIN dotaplayers as dp
    ON dg.gameid = dp.gameid
    AND gp.colour = dp.colour
    AND dp.newcolour <> 12
    AND dp.newcolour <> 6
LEFT JOIN games as ga
    ON dp.gameid = ga.id
LEFT JOIN bans
    ON bans.name = gp.name
LEFT JOIN gameplayers as gp2
ON gp.name = gp2.name
WHERE dg.winner <> 0
GROUP BY gp.name having totgames >= $games
ORDER BY $order $sortdb, name $sortdb LIMIT $offset, $rowsperpage

 

All the best

 

Keith

[eevan79]

Ok, I hope we are almost done.

 

With your code I get an error: MySQL Error: 1054 ( Unknown column 'totalscore' in 'order clause')

I have added totalscore and then I got one huge query (delay) before top players page displayed. 

 

For now this code is working:

 

SELECT *,

    case when (kills = 0) then 0

    when (deaths = 0) then 1000

    else ((kills*1.0)/(deaths*1.0)) end as killdeathratio,

    ($scoreFormula) as totalscore

    FROM (

    SELECT

    gp.name as name,

    bans.name as banname,

    avg(dp.courierkills) as courierkills,

    avg(dp.raxkills) as raxkills,

    avg(dp.towerkills) as towerkills,

    avg(dp.assists) as assists,

    avg(dp.creepdenies) as creepdenies,

    avg(dp.creepkills) as creepkills,

    avg(dp.neutralkills) as neutralkills,

    avg(dp.deaths) as deaths,

    avg(dp.kills) as kills,

    SUM(dp.kills) as totkills,

    SUM(dp.deaths) as totdeaths,

    COUNT(*) as totgames,

    case when (kills = 0) then 0

    when (deaths = 0) then 1000

    else ((kills*1.0)/(deaths*1.0))

    end as killdeathratio,

    SUM(case when(((dg.winner = 1 and dp.newcolour < 6)

    or (dg.winner = 2 and dp.newcolour > 6))

    AND gp.`left`/ga.duration >= $minPlayedRatio) then 1

    else 0 end) as wins,

    SUM(case when(((dg.winner = 2 and dp.newcolour < 6)

    or (dg.winner = 1 and dp.newcolour > 6))

    AND gp.`left`/ga.duration >= $minPlayedRatio)

    then 1 else 0 end) as losses

   

    , SUM(

    (gp.`leftreason` LIKE ('%has lost the connection%')) 

    OR (gp.`leftreason` LIKE ('%was dropped%'))

    OR (gp.`leftreason` LIKE ('%Lagged out%'))

    OR (gp.`leftreason` LIKE ('%Dropped due to%'))

    ) as disc

   

   

    FROM gameplayers as gp

    LEFT JOIN dotagames as dg ON gp.gameid = dg.gameid

    LEFT JOIN dotaplayers as dp ON dg.gameid = dp.gameid

    AND gp.colour = dp.colour

    AND dp.newcolour <> 12

    AND dp.newcolour <> 6

    LEFT JOIN games as ga ON dp.gameid = ga.id

    LEFT JOIN bans on bans.name = gp.name

    WHERE dg.winner <> 0 $_sql

    GROUP BY gp.name having totgames >= $games) as i

    ORDER BY $order $sortdb, name $sortdb

    LIMIT $offset, $rowsperpage

I have added more maches to see result. And then I got disconnects, but not same number when I click on each user (user details). There is same query for lefreason, just for specific user.

For example on top page for one player I see 1 disc, when I click on user details I can see 4.

 

This is query for single user - user info page (for disconnects)

SELECT COUNT(*)

    FROM `gameplayers`

    WHERE

    (`leftreason` LIKE('%has lost the connection%')

    OR `leftreason` LIKE('%was dropped%')

    OR `leftreason` LIKE('%Lagged out%')

    OR `leftreason` LIKE('%Dropped due to%')

    )

    AND name= '$username' LIMIT 1

 

Here is 2 screenshots (taken from top page, and for user info)

(user disc marked with red color.)

[fenway]

It's impossible to debug sql statements with placeholder variables.;

[eevan79]

Yes...finally I debug it. Problem is

WHERE dg.winner <> 0

So this query is only for wins and losses. 0 means DRAW. And thats why other disc arent counted. I totally forgot that this query ignore unfinished/uncompleted games.

 

So many headaches .... and always turns out to be a simple problem.