Lower league??!?

Okay. Here comes a little rant about the qualifying system.

In the Voller EFA I was one spot from making it to the upper tier. Which means I'm stuck in the lower tier. And it really annoys me, because I lose out a lot of money and opportunities to play against the best teams. And what really gets me is that it takes soooooo long to climb back up to the top. The first theoretical chance I have of winning the premiership is now in season 5.

The thing is. After I changed to my weighted approach to players I've done really well. So well in fact that I truly believe that with my current team I'll walk over both the lower league(s) this season and the middle league(s) next season.

So really, from a financial perspective, why would I bother buying players and paying more wages when I know the current squad will do perfectly well for at least two seasons?

I'm driven by two things in life. Business and impatience. Luckily, impatience outweighs my business sense, so I'm already on the ball looking for new good players to be found!

On the topic of my Golden Retrievers team, I signed a DM ranked 20th and an AM ranked 27th. They cost me a small fortune, but hey, we're only using Monopoly money anyway.

I've set a goal to break into the top 10 ranked teams in Voller. It might be tough since I miss out quite a bit in terms of price money as a result of being in the lower league. On the other hand, I've got a list of attributes for over 3500 players printed on my wall. Surely that must count for something?

And in case you wondered, I'm 112th ranked right now.. :)

Preferred foot

Simple test. Low data sample.

Does two-footedness improve GpM rate for strikers?

432 strikers in Voller appearances 7+.

# observations - Preferred foot - Average GpM
13 - Either - 0.22
35 - Left - 0.25
199 - Right Only - 0.28
166 - Right - 0.28
19 - Left Only - 0.30

Conclusion is that two-footedness does not help. But the disclaimer is that the sample size is low and the variation is high due to the low number of appearances.

Golden Retrievers: A Flying Start

Following from my last blog entry The weighted approach to glory!

First two days of new season in Voller has passed, and I've got some good signings based on my weighted attributes list.

I have played 10 FA games with the following results: 7-3-0
Scored 17 and only conceded 5.
Remember though, this is the lower league of the EFA, so there are better teams to play out there :)

My team consists of the following players now. I'll list their placement on my weighted list for each position. N/A means these are my old players not even close to matching the best attributes for their position!

Position (Natural position) - AF - Weighted rank
GK (GK) - £28k - N/A
DL (DL) - £50k - 239th
DC (DM) - £220k - 17th
DC (DC) - £35k -8th
DR (DR) - £20k - N/A
DML (AMC) - £16k - N/A
DMC (AMC) - £230k - 7th
DMR (DMC) - £28k - N/A
FCL (DMC) - £16k - 178th
FCC (AMC) - £24k - 68th
FCR (DMC) £40k - 123th

All in all, I'm really happy with my defence. Those two DCs which score high for defenders on my weighted list are really spot on. Assisted with the brilliant DMC, I have a really strong central defence.

Looking at the attacking players, I'm not sure if I use the correct parameters. Right now my list is based on maximising Goals per Match (GpM), but I'm thinking for the team as a whole that it would be better to base the list on attributes needed for scoring team goals. However, I'll leave it for now.  My FCR has a 0.7 GpM ratio, the FCC has 0.5 and the FCL 0.1.

Next for my team, I think I'll secure my defence by getting a new GK, DL and DR.

On a side note, I'm also looking at a project of creating a team in another GW which will be based on maximising the average rating. Obviously the players would all need jumping 20, but it could be an interesting project to see how much average rating actually corrolates with team performance.

The weighted approach to glory!

So I've decided to put my Voller team Golden Retrievers up as my test team.

The background is that by now I've gathered enough information for identifying most important attributes for DCs, DLs and GKs for not conceding team goals. I also have a fairly good overview of which attributes are needed for FCs to score goals (GPG).

The second thing I've done is to sort all players in Voller based on these attributes, screen grabbed from FML, and used OCR to get all the data into a SQL database I set up.

So there we are. On one side I know the weighting of attributes, on the other side I have the actual player data from Voller. Perfect.

By combining the data, I've got a weighted list of best players in Voller for each respective position. I've now started purchasing some of the top (cheap) guys based on my weighted list. So far I've bought two DCs and three FCs.

Well, my FCs are technically not FCs. They are 2xDMs and 1xAM. They still came quite high up on my FC GPG-test though, so I think they will do quite well even though they play in a bit of an akward position.

The results so far, including where the players rank in my weighted GPG system:
My target man FCC (AM - £26k AF - 63rd) has scored 1 in 7 matches
My poacher FCL (DM - £35k AF - 170th) has scored 4 in 6 matches
My poacher FCR (DM - £16k AF - 120th) has scored 3 in 6 matches

More to follow on my progress. I don't expect to build a top team by playing DMs as strikers and there are still many players out there better than mine. But I do think my team will end up stronger than many others, plus I'll get the bonus of a much lower wage bill..

Come have a look at my team in Voller. It's called Golden Retrievers.

And now for something else: Side backs.

First, I'll define side backs as players with their primarily function to avoid conceding team goals. That makes things a bit easier.

Remember when I said that jumping was the most important attribute for side backs? Well, I was wrong. Partly. Jumping is the most important attribute for improving a side back's average rating.

Jumping is the 7th most important attribute if you don't want to concede team goals!

Looking at the most important attributes for DL/DRs if you want to score team goals however, jumping is still the most important one. I guess this comes down to the now-fixed corner exploit where many of the cornerbangers were actually side backs.

So what is the main attribute you should look for if you want your side backs to concede less team goals?
Well, it's not a technical attribute, neither is it a mental one.. ;)

JUMP!

It's been a while since my last post. Partly because my trial version of SPSS ran out, partly because I've been busy with other stuff, but most importantly because this is my blog and I can update it whenever I want!

I'll follow up the discussion from this thread. The question asked:
Does the 6.2 rating of a bad jumping DC (good otherwise) or the 7.2 rating of a good jumping DC (bad otherwise) have the equivelant effect on the final result?
Can a team with the 6.2 DC win more games than the team with the 7.2 DC?

Well. To find the result, I simply removed average rating from the equation.

I did a regression analysis on team goals conceded per game for DCs to find the most important attributes for not conceding goals. I've OCRd 1196 DCs from Voller, so the numbers should be significant enough to provide quite a clear result.

And the result?

*drumroll*

*more drums*

*drumrollllllll*

Get a DC that can jump! ;)

Feed the ball..

A quick look in Valderrama today.

View: Position; Assists Per Match; Crosses; Appearances
Filter: Corners < 10; Appearances > 20

By simply scrolling down the list, it looks like most of the players with high number of assists per match either play AMC or Striker. Their crossing/appearances ratio isn't too impressive either.

The conclusion?
Either most people prefer to play a narrow formation (4-3-3 anyone?), or ball distribution from the centre is most effective for scoring goals.

I say a bit of both! :)

Skin Magic!

Here are three cool skin tricks I've messed around with.

#1: Improve the font size
You can do this by changing the variables font_x_sz and font_y_sz in skins.cfg and default.cfg in your skins folder. I prefer font size 14.

#2: Change the font
This one is cool. Grab some free fonts for instance from here. Replace (and rename) the downloaded fonts with the ones existing in your fonts folder.

#3: Mess about with your image files
Just when you thought you couldn't have more fun!
In this example I changed the grass to purple, why not let the pitch match my team's ugly shirt colors?

Launch Photoshop, navigate to the image files in your client/app/data/images folder and start getting creative!