hooray, to the solving

price loses?

The downers are proving much more tricky - here I go trying to get the stats on Reg's list:

Sommeil = net xfers -9500, teams selected 19% approx 6.3%
Henry = net xfers -21800, teams selected 35% approx 8% PRICE DOWN
Butt = net xfers -870, teams selected 2% approx 5.5%
Carragher = net xfers -12040*, teams selected 18% approx 8.5% (price stays cos injured??)
Giannokopolous = net xfers -2420, teams selected 4% approx 7.6% PRICE DOWN
King = net xfers -4180, teams selected 6% approx 9% (price stays cos injured??)
McCarthy = net xfers -1350, teams selected 3% approx 5.7%
Rooney = net xfers -16290, teams selected 17% approx 12% (price stays cos suspended??)
Zokora = net xfers -1170, teams selected 3% approx 4.9%

Bentley = net xfers -360, teams selected 1% approx 4.6%
Chimbonda = net xfers -17430, teams selected 17% approx 13% PRICE DOWN
Djourou = net xfers = -170, teams selected 1% approx 2.2%
MGP = net xfers -2960, teams selected 7% approx 5.4%

*carra's net xfers = since monday's change

teams selected % stat is off the website, following approx% stat is approximate percentage of teams that have him that have dropped him (forgive errors, it's late).

None of the above went down except those stated. But loads of others did, too many to get stats on.

It looks like the 'down rule' may be greater than 5% of teams that have a player needing to drop him - maybe 7%ish?

Grouped by percentage selected: Bold is net% drop (ignore the T/Out, it doesn't seem relevant)

Looks like we're getting a curve of % drop req'd for change vs % ownership

YoungCHA net -5321 9% DOWN 0.1 T/Out 5626 -7.7%
Defoe net - 5160 9% NO CHANGE T/Out 6931 -7.5%

Gray net -4864 8% DOWN 0.1 T/Out 5268 -7.9%

Rosicky net -8910 5% DOWN 0.1 T/Out 9076 - INJURED -23.1%
KingWAT net -2289 5% NO CHANGE T/Out 3142 -5.9%

Stelios net -2416 4% DOWN 0.1 T/Out 2701 -7.8%

Scholes net -3716 2% DOWN 0.1 T/Out 4463 - SUSPENDED -24.1%
Arca net -2886 2% DOWN 0.1 T/Out 2930 - INJURED -18.7%
Cole net -2486 2% DOWN 0.1 T/Out 2514 - INJURED -16.1%
Makelele net -2464 2% NO CHANGE T/Out 2574 - INJURED -16%
Li Tie net -1638 2% DOWN 0.1 T/Out 1978 -10.6%
Ambrose net -1378 2% DOWN 0.1 T/Out 1476 -8.9%
Long net -1079 2% DOWN 0.1 T/Out 1357 -7%
Bent M net -784 2% NO CHANGE T/Out 1162 -5.1%
Skoko net -615 2% NO CHANGE T/Out 1195 -4%

Faye net -2090 1% DOWN 0.1 T/Out 2108 - INJURED -27.1%
Lauren net -1388 1% DOWN 0.1 T/Out 1400 - INJURED -18%
Baros net -1347 1% DOWN 0.1 T/Out 1386 - INJURED -17.5%
Vidic net -1128 1% DOWN 0.1 T/Out 1140 -14.6%
Reyes net -1022 1% DOWN 0.1 T/Out 1049 - INJURED -13.3%
Borgetti net -859 1% DOWN 0.1 T/Out 951 -11%
Quinn net -799 1% DOWN 0.1 T/Out 931 -10.4%
Radzinski net -792 DOWN 0.1 T/Out 1016 -10.3%
Pollitt net -772 1% DOWN 0.1 T/Out 815 -10%
Stewart net -741 1% DOWN 0.1 T/Out 861 -9.6%
Murphy net -688 1% DOWN 0.1 T/Out 732 -8.9%
Kilbane net -539 1% DOWN 0.1 T/Out 611 -7%
Cox net -491 1% DOWN 0.1 T/Out 708 -6.4%
Anichebe net -477 1% DOWN 0.1 T/Out 805 -6.2%
Lisbie net -473 1% DOWN 0.1 T/Out 771 -6.1%
Richardson net -473 1% DOWN 0.1 T/Out 645 -6.1%
Landzaat net -457 1% DOWN 0.1 T/Out 570 -5.9%
Legwinski net -415 1% DOWN 0.1 T/Out 481 -5.4
Johansson net -390 1% DOWN 0.1 T/Out 543 -5.1
Sodje net -390 1% DOWN 0.1 T/Out 425 -5.1
Khiz etc net -343 1% DOWN 0.1 t/Out 375 -4.6%
Euell net -290 1% DOWN 0.1 T/Out 343 -3.8%
Elrich net -183 1% NO CHANGE T/Out 315 -2.4%
Rochemback net -124 1% NO CHANGE T/Out 514 -1.6%

Parlour net -220 0% NO CHANGE T/Out 260
Dabo net -171 0% DOWN 0.1 T/Out 229
Routledge net -144 0% DOWN 0.1 T/Out 161
Pongolle net -144 0% DOWN 0.1 T/Out 188
Adebayor net -132 0% NO CHANGE T/Out 362
Zakuani net -123 0% DOWN 0.1 T/Out 134
Connolly net -97 0% NO CHANGE T/Out 132
Mellor net -87 0% DOWN 0.1 T/Out 118

Looks like it would be wise to ignore 0% players in any calcs

OK, so i wioke up through the night, popped the PC on to see if there had been any changes, and im still here. May as well stay up now i reckon

Well done to all for gathering the info (especially Reds for gathering the data fom me ), the picture for transfers in is certainly clear i think. 1.5% of total no. of players leads to a rise. However, is it the no. of players as at start of GW 1? (i.e. 737233 --> 11058), or when the system was switched on (770000 ish --> 11550). Ive gone for start of GW1 --> 11058.

The price drops are indeed more tricky and i believe Slarty has uncovered something with his 'curve' theory. But before i go onto that, have a look at this....(figures taken at 4.00 am - minimal difference to those taken immediately after the price rises)



In the spreadsheet above i have put in the 1.5% rise equation and 0.7% fall equation (as suggested by Reds). Where these values should have causd a rise/fall, i have indicated this in the 'transfers' column by shading the cell yellow. The text is red to indicate an expected fall, blue to indicate an expected rise. As you can see the formula was 100% successful (Rooney doesnt count due to injury/suspension)

Howver before we all start celebrating, some players are requiring a less than 7% drop to cause a change, see Slartys earlier figures and malefice's from yesterday.

More data is needed to prove the difefrent thresholds depending on the teams selected by %. A very (very) rough guide at the minute is ....

teams selected by % -----> %Net transfers required for change

0 ------------------------------------------> ??
1 ------------------------------------------> between 2.4 and 3.8 Figures for Bentley disprove this???
2 ------------------------------------------> between 5.1 and 7.0?
4 ------------------------------------------> above 7.0?
5 ------------------------------------------> above 7.0?
9 ------------------------------------------> above 7.5?

More work to do! Back to bed !!

That's a thing of beauty, good work sir.

excellent work - I shall use my watch list to test the theory....

the early work here is already paying dividends as I used my transfer this week to bring in Kenny. This was prediced very early yesterday morning....(along with Carvalho)

who tonight? Bridge, Sheva, David James

Bridge must have been on the very edge of the fresh hold for going up tonight as he had about 11,286 net transfers in since his rise.

Bentleys stats are...

teams selected by = 1%

This would indicate a change of over -2.4 -->-3.8 % would cause a drop.

His actual % diff = -4.88 %, but he didnt change??

I cant work it out, any ideas anyone?

could it be to do with rounding? maybe he was selected by 1.6 or so[/quote]

This is top notch stuff. I feel slightly guilty that my inability to add 2+2 won't allow me to contribute - yet I will benefit all the same... but only slightly guilty. I think we will soon have the FPL people pulling their hair out with one hand and clapping with the other - if you can clap one handed.

You lot are the best. Keep up the good work.

I had considered that actually JG and have done some further analysis..

Assume for the 1% players the % required to cause a drop is -3.5% (as opposed to the -2.4 to -3.8 listed above).

If Bentley were to have a 'teams selected by' value of 1.4% it would have been enough to prevent a price change.

1.4% of 737233 = 10,321

-3.5% of 10,321 = -361

Actual transfers out = -360, thus no price drop.

Obviously 1.4% would be rounded down to 1% when displayed on FPL.

Maybe the formula uses the exact teams selected by figure rather than the rounded %.

Does that make sense?

Excellent work last night, you must be shattered reg.

Cheers Whuff, still a long way to go though i reckon.

Could you do me a favour please and check your figures for me as i think there may be a few mistakes in there.

Some of the transfers in figures are higher on your list than they are currently, even though yours were taken 10 hours beforehand.

e.g.

Stewart - transfers in

Your figure = 362
current figure = 127

Macnamee - transfers in

Your figure = 119
current figure = 78

Murphy - transfers in

Your figure = 107
current figure = 45

Lie Tie - transfers in

Your figure = 673
current figure = 369

or am i suffering from sleep deprivation

I haven't had time to follow through your logic completely so excuse me if this is a dumb suggestion but could it be possible there are special rules in place for players who are selected by only 1% of players?

Yes we believe there are Nash, and that special rule was that drops occur somewhere between a -2.4 and -3.8% change. However only Bentley breaks that rule at the moment, thats why im confused.

It could be me, I was a bit bleary eyed last night as well. I still have the data so I will give them a quick check.

My Mistake. I thought that the way they formatted the teams was uniformly the same across each field. So I merged the tables together to save on space. This however isn't the case. These are the correct three tables for sheff Utd including %, in, out.
If you need the stats for the other three teams then let me know and I will upload them. Apologies for the mistake.

Cheers for clearing that up Whuuf, had my head spinning for a while.

It would be a real bonus if the game allowed you to sort by players names rather than by the no. of transfers in/out, it requires a lot of editing in excel to get it into a mangeable state.

If you have the time, could you post the stats for...

Bangura, Blizzard, Stewart, Macnamee, Sheringham, Murhphy, Routledge, Lie Tie and Quinn?

Actually, The images will do if easier.

These players should give an indication of the price change thresholds.

yes it's a nightmare getting the data into editable form which is why I opted for screen grabs. I will post those bits up ASAP.
If you need any data from arsenal, chelsea, blackburn and man utd I have those stats from last night as well.

The more the merier, thanks.

Blackburn:

Chelsea:

man utd:

tottenham:

watford:

west ham:

Cheers Whuuf, i shall plug some of these figures into the machine, wind the handle and see what happens!

Edit: Whuuf, not Whuff

Right, i cant take it any more, i give up.

I think ive established that its 7% for players who have a teams selectd by % of 2 or more.

However for those with a teams selected by % of 1, there is something not quite right and the pattern isnt always followed.

Look at the values for Emerton, Park and Kishanishvillimajig...



Maybe it is a rounding issue but i dont have enough data to tell for sure

I agree I think it is a rounding issue. I would say forget the 2.4/3.8% idea, they wouldn't have different rules specifically for the 1% players.

Any player that the site claims is selected by 1% could in fact be anywhere between 0.5% and 1.5%, we can simply never tell.

This could account for a huge 50% error in the calculations Slarty did, hence why some have dropped in price when they only appear to have been sold by 4% of their owners (rather than the 7% that now looks close).