This thing is a breeze to use! +karma since I haven't done that yet.
Oh wow, as Kirch said, it's a breeze to use.
I love it.
Thanks guys ... glad you like it.
@kirch: What happened to the display of your win-rate?
It is set to format to "00,00%" so I am kind of puzzled how you got "020%" .
A couple of thoughts...
1. The enter deck code cell on the first page seems to be protected? I couldn't enter mine (fairly unimportant)
2. (We've already discussed this) It would be great to get uber accurate win results for HP/dependant non-EM wins. Also an electrum tracker would be great!
To show the simplicity... this post took me less than 60 seconds to show great stats!
1. Damn, the "bug" is back. I thought I had dealt with this but apparently not.
For some reason, my Excel seems to reset the cell-formatting for this particular field,
hence joining 3 cells to make a large one keeps failing. Whyever the error-message claims
it has something to do with sheet-protection.
In the next update, this field will simply be 1 cell to avoid complications.
2. Hehe, yes. For our study-purpose, I will create the Statmasta-"Pro":
Über-accurate results based on HP-left for any match that is not a straight loss or EM.
However, the regular Statmasta will just get an update featuring an approximation
to score- and electrum-gain (as is already done with electrum won/game right now) because I hear
too many people complaining that entering the times for ones games is waaay too much work already.
I don't think anybody would bother to mark down HPs considering how little impact it has, especially on
electrum-gain.
Here is the math behind HPs left and elec-/score-gain:
x -> HPs left
y (top) -> score-gain dependent on HPs left
y (bottom) -> electrum-gain dependent on HPs left
HP=0 (loss) and HP=100 (EM) are not in this graph
So elec- and score-gain is a linear function.
If you finish with 1HP you get 35elec and 65score.
If you finish with 99HP you get 59elec and 89score.
The exact middle-ground is at 50HPs with 47elec and 77score.
47elec and 77score are the values used by the regular Statmasta in the next update.
Speaking for electrum won (outside of card wins) the worst possible case of
miscalculation would be something like this:
"CCYB", 50% win-rate, 6min games, 20% EM-rate
100 games -> 50won, 10/50 EM, 50 lost, 10hours played
Now if the player were to win every single match that is not an EM (=40matches) with 99HPs,
the Statmasta would cheat him out of 12 electrum per match, so:
40x12=480electrum ... over the course of 100 games played in 10 hours!
With the above assumptions about the decks speed, that would amount to only 48 elec/hour (FGei).
Now considering nobody will ever win every match with either 99HPs or EM and considering that most
likely some of those unrecorded gains are balanced by some matches barely won with e.g. 20HPs,
that is really not worth worrying about.
Now for score-gain the case is a little different since there are no card-wins that make
the actual per-game score-gain just a sideshow ... there's just one way to increase score:
65-89 (120 for EM) per game and that's it.
If I were to play a deck that consistently finishes above the 50HPs-line and yet only gets
77 score per match accounted for, the difference would be worth noting:
"CCYB", 50% win-rate, 6min games, 20% EM-rate
100 games -> 50won, 10/50 EM, 50 lost, 10hours played
all non-EM games won = 72HPs left on average
72HPs left mean a score-gain of 83 for the match, so the player actually won:
40x83=3320score (wins)
10x120=1200score (EMs)
50x-30=-1500score (losses)
Sum: 3020score net-profit
The Statmaste, assuming HP50 for each game won, awards:
40x77=3080score
10x120=1200score (EMs)
50x-30=-1500score (losses)
Sum: 2780score net-profit
Basically, the HP-oriented deck got cheated out of 240score over the course of 100 games here.
That is 8% of the actual score-gain total ... not really acceptable for testing-purposes even when
considering that other speed-based contenders wind up with a score-gain of 0 or even a slight loss.
(Meaning the tendency of a decks score-gain should be obvious either way ...)
Either way, the score-track in the regular Statmasta will really just be an approximation whereas
electrum-gain can be considered accurate in both future versions.
