Numpty >> thursday february 23 - 12:43, Edited thursday february 23 - 17:44

An update to add some extra details. 

Secondary Positions

Some playing positions give position xp to 2 different positions. I believe that these are RWB, LWB, DM, AM, RAM, LAM, RW and LW.  The secondary positions have a maximum amount of 500 instead of 1000. Whether the player loses or gains positional xp depends on his starting value. If it's already over 500 then the player loses some xp instead.

Secondary position xp can be calculated in exactly the same way. Simply use the value of 500 instead of 1000 in all of the calculations. 

Example

A defender with Mid xp 100 and Def xp 990 plays a match as a DM. In this case the primary position is in midfield and the seconday position is defence. 

He gets the following: 

• Mid xp increases by (1000 - 100) / 100 = +9
• Mid xp = 109
• Def xp increases by (500 - 990) / 100 = -4.9 (negative)
• Def xp = 985.1

Losing Positional xp

Positional xp only changes when playing a match. So it never goes down by not playing.  When playing as a forward then the player loses Gk xp, Def xp and Mid xp. And the same for other positions, he always loses xp from the positions in which he does not play, at a rate of 1% of the current value. (Secondary positions work as above.)

• New Pos xp = Current * 0.99

Total position xp after playing n matches in a different position

• Total (T) = Current * 0.99^n

Playing in the same position

Naturally when calculating the total xp for a number of matches then the formula only works if the player always plays in the same position. A forward plays as a forward and a defender as a defender etc. 

Similarly for the secondary positions, when calculating the total xp for a number of matches then the formula only works if the player always plays in a position which gives secondary position xp. 

If the player plays out of position, then any calculations have to be done one match at a time. Match order is important as you'll get a slightly different amount if a player plays RAM, RW instead of RW, RAM. 

One match at a time

Quick method if you want to calculate one match at a time:

• Primary Position xp = Current * 0.99 +10
• Secondary Position xp = Current * 0.99 + 5
• Other Position xp =  Current * 0.99 

Starting Amounts

The recorded amounts on each player's proflie are all rounded values. However, based on a hunch I've looked at a few players from when they first joined the game. After tracking their pos xp for a short while I am reasonably confident that they started with integer values - these are not rounded and are therefore accurate. Not important, of couse, but allows you to track their exact values if you wish to do so.

Numpty >> sunday february 18 - 12:25, Edited saturday april 13 - 10:46

Here's an update on how position experience affects star value that I originally posted in another thread. 

I have an estimate for keeper's positional experience. With various caveats, 1000 pos xp (px) increases keeper's star rating by around 60%.  For instance a 10.00* keeper with 0 px would become 16.00* with 1000 px.

• My suggested multiplier is 1.60 ^ (px/1000) 

This makes 100 pos xp worth around 4.8% and 10 pos xp worth around 0.47% on star rating. 

My latest update is that I have now narrowed this down to between 57.93% and 59.66%. So in fact it's a little under 60%. 

Further refinement gives 1000 keeper px as being in the range 58.05-59.21%.  

Other positions may be exactly the same or would at least be similar. Please let me know if anyone wishes to help with collecting data for the research on other positions.

Numpty >> sunday february 18 - 13:00, Edited sunday february 18 - 13:06

I will explain what players and data that I need in the event of any volunteers.

The important thing is to collect data on players where nothing changes other than their px, so this restricts the players that can be used.

• they can't be trained during the research period
• their age factor can't change so they need to be under 21 (100%)
• fairly high rated players are best as this increases accuracy

Ideally I need to collect data on fairly well trained 19-20 year olds who will not be further trained during the research period. Any youngster will do but older ones will obviously have a higher rating which improves accuracy. 

You can choose how long to collect data and not to train them for, but the longer the better. One week is useful, but a whole season or longer would be great. The maximum age is 21y18w as that's the point that the age factor starts to decline. 

Data to be collected on each player. 

The essential data that needs to be recorded at the start and then after each match that they play in: 

• star rating (ignoring the youth bonus)
• position experience (px) for their position

Accuracy is obviously important. It doesn't matter too much if you sometimes miss a match. I also need a link to the players being used.