>> FSGP RESULTS WEEK 12
>> 01. John(3-3) ... 14.50 [174]
>> 02. Plaidmoon(4-2) ... 12.08 [145]
>> 03. Sean(4-2) ... 11.83 [142]
>> 04. Peter(3-2) ... 10.80 [108]
>> 05. Bob(4-2) ... 10.58 [127]
>> 06. Eric(4-2) ... 10.33 [124]
>> 07. Patrick(4-2) ... 9.75 [117]
>> ASCP(3-3) ... 6.33 [76]
>
> There's something not right with these weekly scores. Since conference
> game scores count double, ie multiplied by 2, all the totals should be
> even numbers. But 3 of the 7 weekly totals above are odd numbers.
Found it. The box indicating an "Intra Game" for Utah-Arizona wasn't
ticked. (well it was, but the reference on the spreadsheet to indicate that
referenced a different cell, why IHNFC), so it wasn't tabulated as such. 4
of us (me included) got a zero for that game anyways, so it only really
affected you, Peter, and Patrick. Plaidmoon went from 5-10 points.
Luckily no major shake ups.
01. John(3-3) ... 14.50 [174]
02. Plaidmoon(4-2) ... 12.50 [150]
03. Peter(3-2) ... 12.40 [124]
04. Bob(4-2) ... 12.00 [144]
05. Sean(4-2) ... 11.83 [142]
06. Patrick(4-2) ... 10.67 [128]
07. Eric(4-2) ... 10.33 [124]
SEASON
01. Patrick(64-19) ... 16.01 [2113]
02. Eric(67-17) ... 15.88 [2112]
03. Peter(60-17) ... 15.88 [1953]
04. Plaidmoon(68-16) ... 15.80 [2102]
05. Bob(67-17) ... 15.04 [2000]
06. John(60-23) ... 14.76 [1933]
07. Sean(56-22) ... 14.63 [1843]
08. Warren(41-11) ... 13.83 [1300]
> Also a couple of weeks ago some of us had a couple of losses dropped
> from our season won-loss records shown in the results. But the formula
> calculations of points divided by number of games seems to be done
> correctly with the missing game counts included. Example: I'm 67-17
> but the results show i'm 67-15. The season average calculation was
> done using 67+17+49 conference games =133 total.
Damn, good catch. So, the W-L record is not what is used to calculate the
scores, I use a seperate "adjusted picks" total to do that. (to account for
the intra games being doubled). In another one of those "how the frick did
this happen" moments, the cells totalling losses were one letter off for
Nov 4. Meaning, row E was referencing D, F referencing E, etc etc. Records
have been adjusted.
Always appreciate the help finding this stuff, your skill and eye with
numbers is very impressive.