[OUTDATED] Turns to win study

[ratchetspyro94]

Might I suggest that when the 'time to win' study happens that 'clicks per turn' be used to record the time?  As such things as low/high quality, CPU speed, individual clicking speed can affect a strictly time test.  And clicks per turn is more objective...

I see where you're going and this would be an interesting study.  What I'd probably have to do is take 2 separate runs with the deck over 50 games each.  1 based solely on time were I play 50 games straight and record score, electrum and time before and after, and a second were I recorded clicks per turn or game.

Having turns to win, score and electrum per time, and clicks per game would be telling stats about the efficiency of the fastest rush decks.  When I finish updating most of the faster decks for 200 wins, which will take a few weeks, this will be my next project.

*waits for jmdt's wins to hit 20,000*

he play a lot. I still have 3000 wins.

[kobisjeruk]

Might I suggest that when the 'time to win' study happens that 'clicks per turn' be used to record the time?  As such things as low/high quality, CPU speed, individual clicking speed can affect a strictly time test.  And clicks per turn is more objective...

i dont know if you're serious CB! but my mockery* radar is on high alert on this one

* i just insert this to waste another 5 seconds of your life, hah! take that SG

[Retribution]

Anyone feel like testing a golem deck with pendulums?
Or would that be too slow?

[Scaredgirl]

Might I suggest that when the 'time to win' study happens that 'clicks per turn' be used to record the time?  As such things as low/high quality, CPU speed, individual clicking speed can affect a strictly time test.  And clicks per turn is more objective...

That's actually a pretty good idea. It would definitely give a better understanding of the overall speed of the deck. Downside is that it would be a real pain to count those clicks

[smuglapse]

Might I suggest that when the 'time to win' study happens that 'clicks per turn' be used to record the time?  As such things as low/high quality, CPU speed, individual clicking speed can affect a strictly time test.  And clicks per turn is more objective...

That's actually a pretty good idea. It would definitely give a better understanding of the overall speed of the deck. Downside is that it would be a real pain to count those clicks

If that ever happens, I would suggest counting "Done" as a separate click or action.  Because a lot of people could be using the Space bar.

[jmdt]

Might I suggest that when the 'time to win' study happens that 'clicks per turn' be used to record the time?  As such things as low/high quality, CPU speed, individual clicking speed can affect a strictly time test.  And clicks per turn is more objective...

That's actually a pretty good idea. It would definitely give a better understanding of the overall speed of the deck. Downside is that it would be a real pain to count those clicks

If that ever happens, I would suggest counting "Done" as a separate click or action.  Because a lot of people could be using the Space bar.

You guys have fun planning out the next several months of my life don't you.   

[CB!]

Might I suggest that when the 'time to win' study happens that 'clicks per turn' be used to record the time?  As such things as low/high quality, CPU speed, individual clicking speed can affect a strictly time test.  And clicks per turn is more objective...

i dont know if you're serious CB! but my mockery* radar is on high alert on this one

* i just insert this to waste another 5 seconds of your life, hah! take that SG

Actually I was serious.  If we're concerned about 1-2 seconds a turn, a slower computer or a slower clicker could eat that time up.  Clicks per turn would tell more about the deck, rather than the person playing the deck...

and you successfully wasted 5 seconds of my life... 15 if you count this reply

[ratchetspyro94]

Might I suggest that when the 'time to win' study happens that 'clicks per turn' be used to record the time?  As such things as low/high quality, CPU speed, individual clicking speed can affect a strictly time test.  And clicks per turn is more objective...

i dont know if you're serious CB! but my mockery* radar is on high alert on this one

* i just insert this to waste another 5 seconds of your life, hah! take that SG

Actually I was serious.  If we're concerned about 1-2 seconds a turn, a slower computer or a slower clicker could eat that time up.  Clicks per turn would tell more about the deck, rather than the person playing the deck...

and you successfully wasted 5 seconds of my life... 15 if you count this reply

True so...

Count up all the clicks then divide by how many turns you took? Or just count up all the clicks?

[jmdt]

True so...

Count up all the clicks then divide by how many turns you took? Or just count up all the clicks?

Honestly, I think clicks per game is sufficient for comparison purposes.

[CB!]

True so...

Count up all the clicks then divide by how many turns you took? Or just count up all the clicks?

Honestly, I think clicks per game is sufficient for comparison purposes.

I agree.  From there you can calculate the clicks per turn, if you wanted to... So, keep track of the number of turns + the number of clicks, not including clicking done (pressing space)...

[kev]

SG is right, of course, but sometimes you gotta look yourself in the mirror and say "When in Rome."

Counting clicks is borderline insane.  The time it takes to move your finger a number of millimeters is NOT the relevant data to collect.  If I pause to decide whether to immolate a fire spirit, I could have clicked 810 times in that gap.  It should be obvious why it's pointless to measure how many decision points we come to or how many seconds per decision is spent.  A better measure is electrum or score gain per hour but even that is tremendously flawed due to things like processor speed.

ttw is a useful statistic because
1) it's a very good proxy for match length,
b) it's easy to calculate, use, and understand, and
4) it's related to win/loss because giving opponent fewer turns to act is in our best interest.

jmdt had it right from the start and SG has it right now.  We should continue tracking ttw but recognize it's an imperfect measurement of deck speed.

[guolin]

SG is right, of course, but sometimes you gotta look yourself in the mirror and say "When in Rome."

Counting clicks is borderline insane.  The time it takes to move your finger a number of millimeters is NOT the relevant data to collect.  If I pause to decide whether to immolate a fire spirit, I could have clicked 810 times in that gap.  It should be obvious why it's pointless to measure how many decision points we come to or how many seconds per decision is spent.  A better measure is electrum or score gain per hour but even that is tremendously flawed due to things like processor speed.

ttw is a useful statistic because
1) it's a very good proxy for match length,
b) it's easy to calculate, use, and understand, and
4) it's related to win/loss because giving opponent fewer turns to act is in our best interest.

jmdt had it right from the start and SG has it right now.  We should continue tracking ttw but recognize it's an imperfect measurement of deck speed.

I agree with this entire post.