Fast time vs 100%

[gavinl]

Just wondering, at what time do the fast solvers (say, average under 20 seconds) pull out of trying to do the solve in order to maintain their average?

[oddcouple]

Beats me. I'm always content to average the 40-50 second range with a high solve percentage (preferably 100%). But that's just my personality.

[Deanna48]

Gavinl, when going for speed, I try not to send in any that have taken more than 40 seconds. And less than 30 is best.

[nedzeppelin1453]

I guess I'm probably near the bottom of that group of "fast solvers." I just stop doing a puzzle when I run out of patience with it. So, a couple minutes maybe.

[Zachman285]

My time last month was 15.6.

I rarely submit anything over 25. If I don't get a good start in the first 3 to 4 seconds I'm usually out on the puzzle. The 25 number has gone down as I've gotten better over the years. When I used to average 21-22 seconds that number was closer to 40.

My goal on the site is to get the solve time as low as possible while staying above the 75 percent fastest solver threshold. Hope that answers your question!!

[maradnu]

I do 100% consistently and pulled down stats on my last 200 cryptos solved. My average on all 200 was 42.3 while my average for my 75% fastest was 32.4 - basically a 9.9 second difference. Now that would surely vary from one night to the next and I am far from fast, obviously.

[nedzeppelin1453]

maradnu, how do you find those stats?

[pickleball]

yes hi, i would also know how to find these stats, thank you

[maradnu]

I went to recent games, copied the data, pasted it into Excel, then spent some time stripping the numbers out and then getting them usable for calculations. Probably fifteen minutes of work for me and I know Excel pretty well. I pasted it into a grid, then did a @left to pull numbers out of that column. I copied and pasted values then had to edit and enter on each cell to make them usable. I sorted by time than did an average on the whole list and on the 75 fastest.

[Fudi]

Funny, maradnu. I was just thinking, today, about how it would be a neat thing if the Fastest Solver monthly competition considered only top 75% solve times for each player, regardless of what percentage each player actually solved successfully. That way, players could still try for 100% success, but not be penalized for slower solves. I was thinking that this would be a fairer way, really of determining Fastest Solvers. Now, mind you, I don't think I'd make the monthly cut if it were done this way, as I have for the past few months, but I nonetheless think it'd probably be more fair.

[maradnu]

I used to barely make the top 20 on speed, but now can't even manage that. I miss the success rate, which I could nearly always hit the top three.

[gavinl]

Yes Fudi, or even a Fastest Solvers with 100%...though i think Nanrich would win that one every time.

[azhang941]

To the original question, I generally pull out of trying whenever I feel frustrated that I'm not getting anywhere instead of after any set time period.
Also depends on how I'm trying to play, as I have done 100 rate before, and I've also tried doing fast times where I pulled out after not seeing something immediately. The thing I noticed is that after a certain point (around 85-90% solve rate), solving less brings very diminished returns in lowering average solve time.

[nedzeppelin1453]

I also like the idea of the fastest times scoreboard only taking into account each player's fastest 75% of puzzles. That would mean you would no longer have to choose between speed and success rate.

[will0416]

I also stop more at the point of frustration, rather than some personally-set time limit. My patience ceiling is probably at ~3 "resets" (which I guess translates to the high 20 second range).