A look at perfection : an analysis of Lapsilap's WR 40-line run

Started by caffeine, April 24, 2011, 03:04:13 PM

Previous topic - Next topic

caffeine

Very cool! Is there a way we can download this new mod that has the fixed TPM?
Quote from: belzebubIf you wanna see other metrics of this sprint, let me know.
Something that would be interesting to me would be "effective keys per second." For all tetrominoes that caused a fault, instead of raw keys, only add the keys that would be required for an ideal placement to the running total. Divide this number by the total number of seconds, and you have an adjusted KPS, or maybe "eKPS."

So, if you waste KPS on extra movements and hold and such, it won't do anything to increase your eKPS, which will reflect only your accurate key presses.  This way, we would be able to compare different players actual finger speed.

Edit: Actually, you could have another statistic as well: wasted keys percentage. Subtract the total keys pressed from the total that went to eKPS, divide that by your total keys, and now you have a percentage of keys that went to wasted/suboptimal placement faults.

belzebub

For everybody: I added a large PART II on my Analysis, hope you like it!!

Quote from: caffeine
Very cool! Is there a way we can download this new mod that has the fixed TPM?

Something that would be interesting to me would be "effective keys per second." For all tetrominoes that caused a fault, instead of raw keys, only add the keys that would be required for an ideal placement to the running total. Divide this number by the total number of seconds, and you have an adjusted KPS, or maybe "eKPS."

So, if you waste KPS on extra movements and hold and such, it won't do anything to increase your eKPS, which will reflect only your accurate key presses.  This way, we would be able to compare different players actual finger speed.

Edit: Actually, you could have another statistic as well: wasted keys percentage. Subtract the total keys pressed from the total that went to eKPS, divide that by your total keys, and now you have a percentage of keys that went to wasted/suboptimal placement faults.

I didn't package it yet, but it has some cool features like dig replay with previews too, so I will, i'm currently thinking of what stat's I'm gonna include in the rep file when it saves.

I think eKPS = raw KPS - faults - holds. And I think wasted pieces = wp = finesse faults + holds.

Wasted keys percentage = wkp = wp / total keys = (finesse faults + holds) / total keys.

I can simply include that in the statistics file when I fix the hold thing, but I'd prefer to further discuss this on the other thread :-)

Paradox

thats awesome! so you pretty much showed that finesse can really help in certain situations
[!--ImageUrlBegin--][a href=\\\"http://oi46.tinypic.com/2zqx63k.jpg\\\" target=\\\"_new\\\"][!--ImageUrlEBegin--][img width=\\\"400\\\" class=\\\"attach\\\" src=\\\"http://oi46.tinypic.com/2zqx63k.jpg\\\" border=\\\'0\\\' alt=\\\"IPB Image\\\" /][!--ImageUrlEnd--][/a][!--ImageUrlEEnd--]