Recommended Leagues - an idea, some testing and some observations

Deciding my setup, i.e. what leagues to play, is the worst part of starting a new game for me. I always procrastinate over starting my first game on a new FM because I am always concerned over getting the right mix of realism and ensuring that my game will run at an acceptable speed for years to come. In my opinion, it is essential to get the right database to ensure longevity.

Many people report that the 'recommend' button on the 'setting up a new game' page on FM is not particularly insightful. With this in mind, I have been running some tests and trying to think of a way of coming up with a guide to starting new games. A guide that will offer some kind of standard way of working out how many leagues can be loaded on different systems (whether in playable or view-only) and what database settings are best (including loading players in DDT files). This might put an end to the confusion and the threads where people write out their PC specs in a desperate attempt to get some advice from other gamers as to hoe many leagues they can run.

At the moment, what I have to present is very modest as well as being messy, random and badly worked out but I hope some of it might be slightly useful.

Below I have copied and pasted some of the tests that I have carried out with FM09. Some of you might have seen these tests posted up on other threads. At the bottom of the page, I have indicated what I feel might be a good way of moving forward and making suggestions to confused FMers who don't know what leagues to load.

I started off measuring file sizes with the belief that you could compare the amount of data that has to be processed and that more processing should result in slower progress. After that, I did some timing as well.

My PC spec is Dual Core 2 1.66 GHz with 3GB RAM.

Here we go then.

One league playable: A game with all English divisions loaded as playable. Creates a file size of approx. 40MB. This is speedy as can be on my machine and doesn’t take long to setup at all.

View-only experiment: A game with all English divisions loaded as playable plus 18 other view-only countries (top two divisions of each) creates a file size of approx. 65MB. The game itself seems just slightly slower than the one playable league option above. Took slightly longer to load up. More players appear to have been loaded although it should be noted that some teams in the view-only leagues still have grey players. I also have additional international countries that are manageable, noticeably those related to the countries I have loaded on view-only. Moving through a week definitely took a noticeably longer time than the one league playable option above (not painfully so though). More results were coming up on the screen as I progressed through the week.

Four leagues playable experiment: A game with all English divisions loaded, plus top two divisions of Spain, Italy and all divisions of Scotland – all countries playable. This took ages compared to the other two to load up. Processing through a week actually seemed quite reasonable though and not that much different to the view-only one above. The game size was 53MB.

One playable, three view-only: After the above, I decided to run the same leagues but with only English playable and the rest of view-only. Setup was slightly quicker than above, as you would expect, but not that much quicker. File size was 46MB. Processing through a week seemed barely different to just one league playable to me.

One playable, seven view-only: So double the leagues, added the top two four more countries on view-only. This set up in reasonable time, not an incredible amount of difference to above. Twice the amount of leagues gave a file size of 53MB. This is very interesting because this is the same file size as that of the ‘four leagues playable’ experiment. Processing, again, not really noticeably different to above. More results coming through but not slowed down at all.

Another couple of random tests coming up.

102 leagues from 51 countries (default selections), all playable:

Took years to get setup. Perhaps even decades. I could have reread the entire oeuvre of James Joyce and then listened to an entire Gustav Mahler symphony cycle. Well, maybe not that long, but still. A long time. There is no way I’m ever going to try all leagues. Goodness knows how long that would take.

Saving the game, to start with, took an age. Processing was, quite frankly, too slow for me. It still seemed to go ok but took a lot longer to get through a week due to the extra results being processed. Not disastrously slow but would require some extra patience, for sure. Playing this game would result in a lot of hanging about for me. File size was 200MB.

Now going to try the same but with all but England on view-only.

102 leagues from 51 countries (defaults selections), only England playable, the rest view-only:

A lot quicker in loading up but not incredibly so. I could probably have read one Joyce novel and listened one Mahler symphony this time. Maybe A Portrait of the Artist as a Young Man and Symphony No. 1 - two of their shorter works respectively. Actually, it was quicker than that. Not bad at all considering.

Processing decent speed, slower than most of the settings I have tried, but vastly superior to above. Size of file only 75MB.

Got loads of great obscure national teams that are manageable: Tonga, Samoa, Cook Islands, Fiji – just a few of the highlights.

So, I can conclude from this that view-only can add more players from countries that are not usually considered ‘core’ or important enough to load their famous players.

I have found that the file sizes are hugely increased when loading players from all countries. The difference in file size between a large database with no additional players retained, and a large database with all players loaded from England, Scotland, N.Ireland, Ireland, Wales is 15MB. This is to be expected with all of the extra players involved. Loading extra players also has a significant influence on how quickly the game sets itself up.

One way of avoiding loading all players from a nation is to make a DDT file and load players from specific clubs. This can add realism without having to run extra leagues.

I have experimented with setting up some DDTs to retain players from the top clubs in Europe and also from a few of the major top flight leagues in Europe. For instance, I have retained all Serie A players and all clubs that have made the Champions League group stages.

I have tried loading DDT files for clubs from certain leagues and then running those leagues on view-only to see if this is superior, in terms of speed, to loading the league as playable. The results indicate that this saves a small amount of processing time. I concluded that by loading the league on playable I got a larger database of players which had an effect on file size and setting up times but it didn't make a really significant difference to processing speeds.

My current plan is as follows:

DDT for non-league steps 3&4

DDT for European Champions League sides

DDT for UEFA CUp sides

DDT for top divisions of Italian, French, Spanish, German, Scottish

Then, load all England playable with top two of Italian, French, Spanish, Belgian on view-only, and entire Scottish league structure also on view-only.

Some final tests now measuring processing times for holidaying the first month of game time (from July to August):

- All English divisions on their own (5 mins).

- All English divisions with a collection DDT files configured to retain top european leagues etc. (6 minutes).

- All English divisions playable with top two of Italian, French, Spanish, Belgian on view-only and all of Scotland on view-only plus DDTs as above (7 minutes).

- All English divisions, top two of Italian, French, Spanish and all of Scotland on playable, with Belgium on view-only plus DDTs (8 minutes).

(Going a bit off-topic for a minute, I have found that making the DDTs is not as time consuming as some people report. I have found an excellent way of making up DDTs very quickly. You can use the short names of the club, so what I do is copy and paste the clubs from each league from Wikipedia. Then I copy and paste the 'retain players' command in front of each one. Then I quickly check the FM editor by selecting the competition and viewing the teams and make sure that I have the right team names. Finally, I add the quotation marks to each team. Doesn't take very long at all using the method above and means that I don't have to get involved with the dreaded ID numbers!)

Finally, I can certainly conclude from all this that setting up DDTs for leagues you don't want to load but would want to have all the players from that league is a superior method than loading additional leagues. For those with poor specs, having DDT files and just playing with one league will be far superior to loading extra playable leagues or view-only leagues which will add strain to their computer. I cannot really confirm whether having DDTs for view-only leagues is superior to just loading those leagues as playable. It seems to save on file size and, I believe, makes a small amount of difference in processing times.

Right, well that's all for now in terms of experiments. The DDT files is certainly an interesting find and something I recommend to those who are restricted by their computer specifications as to how many leagues they can play.

So, to conclude this rather long post, I was thinking that it would be really good if anyone could come up with a solid and standard way of testing a setup on any given system. My leading idea for this is to get people to holiday with their given setup for the first month of game time and time it. Then they could compare this time to a given time for 'acceptable' processing time.

The file size of a save game increases dramatically during the first two seasons and then slows down afterwards, presumably because it reaches a level where it begins overwriting some of the data it stores. Going along with this, processing times seem to increase after the first few seasons too, in my experience.

So my idea is to suggest a standard 'holidaying time' for the first month of a save game that is decided in relation to processing speeds occurring later on in that save game. At the moment, I have tested settings that take up to 15 minutes for the first month without having any 'slowing down' problems that occur later in the save games life so this is my initial suggestion, i.e. find a configuration that takes 15 minutes to holiday through the first month on your system and you should be fine well into the future.

It would be wonderful to hear from others who have something to contribute, especially with different PC specs. Like I say, the idea is to come up with a quick and solid way of testing potential setups according to the experiences of others and in order to maximise the realism and potential of the game on every system.

C.